# JudeAI | EDGE ID The one place. Contract fingerprint `1c80b0fa621e444a539ff47dd98c889b3c870d5f700ff7012ee134f079acd919`, contract version `0.1.0`, repository commit `d594aa71f06592d467dc430d2362b37d8ff0b96c`. Statement classes: the contract, kernel, specification, decisions and briefs are AGREEMENTS; research, measurements and the catalogue are OBSERVATIONS; vendor and competitor statements are CLAIMS until observed. Nothing here is a projection of implementation. ## Contents 1. Contract 2. Ontology (TOK) 3. Specification 4. Research 5. Documentation 6. Reusable assets and design system 7. Environments, tools, settings and connectors 8. Engineering philosophy 9. Prompts and intentions: legacy and live 10. Metrics 11. Full specification per capability (index; files in the zip) 12. Files included, with sha256 ## 1. Contract ### contract/estate-contract.json `contract/estate-contract.json` · sha256 `62484e1e78dc7c2fed76f7ed84966bfeb216ec8ba16ce2150c99ea4744ad7b10` ```json { "schema": "tok.estate-contract.v1", "version": "0.1.0", "statement_class": "AGREEMENT", "product": "JudeAI | EDGE ID", "repository": "judeai-edgeid-governed-intelligence-autonomous-agent-estate", "agreed_by": [ "X-X.TECH" ], "dissented_by": [], "authority": "X-X.TECH, the organisation that governs the estate", "applies_where": "the whole estate and every organisation that onboards to it", "applies_from": "adoption of this version", "applies_until": "adoption of a version that supersedes it", "basis": "The estate's governing intention, stated as what the product is and does.", "what_this_document_is": "The agreement that defines what the estate is. It authorises construction as specified. It verifies nothing, counts nothing and awards no readiness; every status in the estate is set from observation and reproduction, never from this document.", "proposition": "Your data. Your intelligence. Your authority. Our responsibility: work done right, with independently verified outcomes.", "purpose": "The living source of truth for all human domain expertise: regulator requirements, live facts and laws, and everything that can be researched, each statement classed as observation, agreement or claim with its sources, dissent and evidence, used by organisations, their people and their agents together to produce verified beneficial outcomes.", "founding_principles": { "el_roi": "To see is to protect, to know is to understand, to have power is to be accountable. Observation protects dignity and privacy; to be seen is not to be watched.", "let_there_be_light": "Truth is distinguishable from falsehood, knowledge from uncertainty and reality from aspiration. No falsehood is accepted as truth; no aspiration is confused with achievement.", "what_god_cannot_do_does_not_exist": "With God all things are possible (Mark 10:27). The limits of existing tools are not the limits of what can be built.", "truth_and_agreement": "Truth is complete and already exists; it can only be observed. Agreement creates laws, rules, standards and plans; agreements can be disagreed with. Every statement is an observation, an agreement or a claim, with its sources, its dissent and its evidence. Actions obey agreements; knowledge follows observation. Where an agreement departs from the evidence, the estate shows both and breaks neither." }, "outcomes": { "dimensions": [ "time", "money", "lives", "planet", "resources", "value" ], "directions": { "SAVED": "preservation: loss, waste, harm, unnecessary expenditure and missed entitlements prevented", "MORE": "abundance: more time, money, life, planet, resources and value made available" }, "units_per_verified_outcome": { "time": "seconds", "money": "GBP", "lives": "lives or life-years, clinically defined per capability", "planet": "gCO2e", "resources": "kWh, litres, tonnes of material", "value": "GBP of value preserved or created" }, "states": [ "OPPORTUNITY_IDENTIFIED", "WORK_ATTEMPTED", "RESULT_ACHIEVED", "OUTCOME_VERIFIED" ], "rule": "Every capability serves at least one outcome. An outcome is claimed only in OUTCOME_VERIFIED with its measurement; attempts and results are reported as what they are." }, "intentions": [ { "id": "source-of-truth", "statement": "The estate is the living source of truth for all human domain expertise: regulator requirements, live facts and laws, and everything that can be researched, refreshed continuously." }, { "id": "truth-separated-from-agreement", "statement": "Every statement is classed as observation, agreement or claim; for every rule, law and fact the estate shows how it came to be, who agrees, who disagrees and on what evidence." }, { "id": "glass-box", "statement": "A glass box, not a black box: light, not darkness; possibility, not limitation; to be seen, not to control. Authorised people and regulators can inspect knowledge, reasoning, authority, execution, verification, recovery and efficiency." }, { "id": "collaboration-and-shared-data", "statement": "Collaboration, community, cooperation and teamwork are invited. Data is shared for good under lawful authorised access, not hoarded, so that organisations create new capabilities, industries, jobs and solutions." }, { "id": "solve-the-crises", "statement": "Capabilities directed at the energy crisis; cyber attack made unprofitable; people given the capability to generate wealth, meaning the power to be; more and better housing; raw materials used better; zero waste; everything monitored and tracked." }, { "id": "agents-native", "statement": "Agents are native to the estate as much as people: always on, inside the estate so data never leaves it, each with its own portal and racetrack so agents cannot crash or collide. The estate is built for agents, which makes it better and safer for people." }, { "id": "clear-screens-deep-help", "statement": "Screens show the essential result first; depth and context live behind a help control as markdown, served to people and agents on demand, so neither people nor agent context windows are overwhelmed." }, { "id": "exhaustive-deterministic-scope", "statement": "Everything is defined and scoped before it is built: every screen, feature, workflow, capability, control, intention, outcome, endpoint and every whole-of-life, customer, agent, regulator and engineer journey. Nothing missing, nothing broken." }, { "id": "match-and-exceed", "statement": "Every competitor capability is matched, exceeded, improved and made easier and faster; benchmarks, baselines, competitors, features, behaviours, workflows and controls are known from research." }, { "id": "compress-time", "statement": "Slow, clunky processes and industries are compressed in time. Human intention is converted into computational instructions; labs automate prompt engineering, creation and solutions; engineering work that took a day takes minutes." }, { "id": "agents-build-agents", "statement": "Always-on agents build further agents that are advanced, fine-tunable, transferable and reproducible." }, { "id": "always-on-monitoring", "statement": "Everything is reproducible, retraceable, auditable and monitored always on; this is how sovereignty and control of every organisation's data are maintained." }, { "id": "lean-repository", "statement": "Clean source code; one instruction markdown; the repository contains exactly what it needs; every file is named and placed by TOK." }, { "id": "compute-light-runtime", "statement": "AI is the factory, not the product: neural models are used at build time for speed; what runs is deterministic, compute-light, more useful and more accurate, and reduces the need for AI over time." }, { "id": "only-true-things", "statement": "The runtime says only what a verified procedure produces inside its certified scope, with its error bound attached, using error-rate control and Bayesian reasoning; otherwise it says it does not know." }, { "id": "same-rules-everywhere", "statement": "One method applies to every domain: identity, relations, processes, logic, probabilities and patterns. Every domain is added by extension and the same rules hold." }, { "id": "better-routes", "statement": "Where a licence or partner is not needed, a better route is engineered: consented data, simulators, open sources, own networks. Where the licence is the network, the legal permission or the clinical safety, the licence or partner is obtained as well." }, { "id": "sight", "statement": "Working backwards from preventing avoidable illness and blindness: the time from first meaningful signal to a verified beneficial outcome is shortened, starting at the front door before the GP, with own capture technology and consented patient data." }, { "id": "approved-design", "statement": "Interfaces use approved UX Pilot, Figma and Untitled UI structure and the estate's original logos, used as-is and never recreated: accessible, readable, complete navigation, real functionality." }, { "id": "every-intention-represented", "statement": "Every intention of the governing organisation and of every onboarded organisation is represented, including capabilities implied and never spelled out, each placed in an application or foundation." }, { "id": "verified-outcomes-fast", "statement": "The governing objective is the shortest independently justified time from legitimate intention to verified beneficial outcome, at the greatest feasible speed and autonomy, with no reduction in scope or acceptance." }, { "id": "education-is-the-source", "statement": "Education is where all capabilities come from; research and development is worth investing in; universities are made prestigious again by being the home from which public, defence, government, procurement, logistics and cyber capabilities grow." }, { "id": "agents-as-real-customers", "statement": "Agents are real users and real customers in every application: lenders, borrowers, buyers, suppliers, patients' representatives, applicants and operators. They act as real customers on their racetracks so they learn what works and what does not, under their own identity and authority." }, { "id": "confidence-reporting", "statement": "Confidence reporting sets the standard beyond credit scoring: a confidence score that is more informationally complete than a credit score, showing not only whether obligations were met but how a person or organisation acted when things went wrong (whether they did all they could, how they communicated, what caused the situation), so a single turn of events does not define them. It strengthens KYC, is explainable and contestable, and is proven by research and measured business and economic outcomes, including wider access to resources." }, { "id": "ai-assurance-standard", "statement": "The estate sets the global standard in autonomous agent assurance, as the first fully agent-native bank: agents as customers and staff with their own identity, certificates, mandates and term, with no gaps in assurance." }, { "id": "solution-banks", "statement": "From think tanks to solution banks: agents generate solutions, and solution banks empower and fund people to enact them in the real world, because having the answer is not enough. Better lending structures and better ways to move money through the economy create jobs and access to resources." }, { "id": "physical-presence", "statement": "EDGE is a real bank and clinic in the physical world, not only online: EDGE points and devices that come to the person, verifying identity by proximity (scanning or tapping an identity document or device), and edge hardware and wearables used for good, including early detection in the clinic." } ], "applications": [ { "id": "finance", "name": "EDGE Finance", "formal_name": "JudeAI | EDGE Finance", "interface_name": "Finance", "intelligence_engine": "JudeFI, Jude Financial Intelligence", "is": "Banking and financial intelligence where people, organisations and agents are all real customers: accounts, payments, cards through a processor, lending and borrowing (agents as lenders and borrowers), KYC, KYB, KYA and know-your-domain in one journey, sanctions and PEP screening with a guaranteed miss rate, transaction monitoring on measured rates, a fraud-sharing network, credit reference on consented Open Banking data, agent payments and mandates, and a sponsor bank route before own permissions. Floors: Starling, Monzo, Revolut and the leading crypto platforms." }, { "id": "clinic", "name": "EDGE Clinic", "formal_name": "JudeAI | EDGE Clinic", "interface_name": "Clinic", "intelligence_engine": "Jude Clinical Intelligence", "is": "Social prescribing with early detection: the front door before the GP, where anyone can come about anything and find out what the problem is. It brings together detection (including retinal screening and other early signals), validated questionnaires, triage including online A&E, an open clinical-grade drug-interaction source, prescribing safety on real terminology and referral to the right service, aimed at the costliest burden on the health service: the time and capacity it takes to be seen. Every clinical capability carries its safety and device obligations as OPEN until cleared." }, { "id": "public-consortium", "name": "EDGE Public Consortium", "formal_name": "JudeAI | EDGE Public Consortium", "interface_name": "Consortium", "is": "The public and institutional home of the estate, founded on the declaration that education is where all capabilities come from and that all research and development is worth investing in. It houses cyber governance (attack made unprofitable), procurement intelligence, vetting and verification, logistics for defence and government, regulator and evidence portals, research consortia with universities, the fraud-sharing and data-sharing networks, and public knowledge from the Research Commons. Capabilities expand here without breaking the form." }, { "id": "judiverse", "name": "Judiverse", "formal_name": "Judiverse", "interface_name": "Judiverse", "contains": [ "JudeAI", "intenJude", "JudeBox" ], "logo": "the intenJude logo", "is": "Where people and organisations work with their agents: Jude as assistant, researcher and operator; documents, mail, calendar, meetings, appointment and interview scheduling, tasks, files, forms and workflows generated from each organisation's versioned workflows; the intenJude lab where intention becomes computational instruction, agents build agents, models are trained, scheduled and compiled into certified procedures; and JudeBox for artists' own music with direct royalties. Floors: every Microsoft product and service, the leading AI assistants, GitHub and GitLab." } ], "application_rules": { "names": "Formal name in formal presentations, introductions and external communications; the short name in the application rail.", "logos": "The estate's logos are admitted from the repository as design-asset objects under design/brand/, used as-is and never recreated: estate-mark.svg; edge-finance.svg (lion alternate); edge-clinic.svg; judiverse-intenjude.svg (Judiverse carries the intenJude mark); EDGE Public Consortium's mark is an open decision.", "every_top_competitor_is_a_floor": true, "oracle_wherever_it_competes": true, "trade_mark_clearance": "REQUIRED_BEFORE_PUBLIC_LAUNCH", "structure": "Four applications. New capabilities are placed in one of the four, or in a shared foundation; a new application is created only by amendment." }, "shared_foundations": [ { "id": "edge-id", "name": "EDGE ID", "is": "Identity and authority for people, organisations and agents: who may do what, issued by a legitimate issuer, with purpose, resource, expiry and revocation." }, { "id": "research-commons", "name": "Research Commons", "is": "Continuously refreshed, open, source-linked knowledge for all human domains, with current state, reconstructible change, corrections and change intelligence." }, { "id": "tok", "name": "Tree of Knowledge (TOK)", "is": "The one ontology; everything resolves from it and every change identifies what it affects." }, { "id": "evidence-portal", "name": "Evidence portal", "is": "Where evidence, proofs and reproductions are inspected and admission is requested from the actual entitlement issuer." }, { "id": "regulator-portal", "name": "Regulator portal", "is": "Where assessors inspect authority, evidence, proof and recovery without the builder's help." }, { "id": "organisation-workflows", "name": "Versioned organisation workflows", "is": "Each organisation's workflows, editable and versioned, from which screens, forms and agent tools are generated." }, { "id": "agent-apis-mcps", "name": "Agent APIs and MCPs", "is": "Every operation callable by an agent under its own EDGE ID with the same authority, preview, audit and undo as a person." }, { "id": "governed-release", "name": "Governed release and adoption", "is": "One release in which everything is correct and working, with no dead controls, rehearsed on the racetrack and accepted on live." } ], "cross_cutting": { "sovereign_intelligence": "Each organisation's data, embeddings, indexes, prompts, memory, logs and agents stay under its own control and never leave the estate or go to model providers.", "glass_box": { "inspection_dimensions": [ "knowledge", "reasoning", "authority", "execution", "verification", "recovery", "efficiency" ], "objectives": [ "Zero silent failures", "Zero fabricated assurance", "Zero unauthorised actions", "Zero avoidable harm" ] }, "accountability": { "principle": "Worth doing. Done right. Answerable to humanity.", "roles": [ "Requester", "Purpose approver", "Deployment authoriser", "Operator", "Independent verifier", "Remedy owner" ], "rule": "An agent is never the answerer of last resort." }, "agent_workspace": [ "portal", "mailbox", "racetrack", "dress rehearsal", "action preview", "audit and evidence", "resource-to-outcome ledger" ], "governing_sequence": [ "See", "Understand", "Imagine", "Discover", "Rehearse", "Challenge", "Authorise", "Execute", "Verify", "Recover", "Learn", "Improve" ] }, "principles": [ { "id": "one-source", "text": "TOK is the only source. Everything is an object in it; every document, view, index, diagram, count and test is generated from it; a hand edit of generated output is a failure.", "enforcement_type": "GATE", "check": "G01 schema, G03 closure, G05 generation, G16 live counts" }, { "id": "truth-agreement-claim", "text": "Every statement is exactly one of observation, agreement or claim, carrying that class's required fields; none is rendered as another.", "enforcement_type": "GATE", "check": "G04" }, { "id": "honest-status", "text": "Every realisation carries a truth status set from evidence, never above it. An attempt is not a completion; a rehearsal is not an outcome; an intention is not an observation. Nobody fabricates evidence, authority, benchmarks, permissions or readiness; a falsehood is rejected and never rewarded.", "enforcement_type": "GATE", "check": "G04 status rules; breaches are incidents" }, { "id": "live-numbers", "text": "Every number is produced from the objects at the moment it is stated, with its method and completeness. No count is written by hand.", "enforcement_type": "GATE", "check": "G16" }, { "id": "everything-identified", "text": "Everything that exists has a permanent identity, a kind, its relations and a working monitor; nothing is undefined, dead, unaccounted for or without an identity.", "enforcement_type": "GATE", "check": "G02, G03" }, { "id": "complete-contract-first", "text": "Every capability, journey, screen, control, operation, endpoint and agent resolves to a complete executable contract before it is built: intention, requirement, steps, executor, authority, design, acceptance, evidence, resources, recovery and answerer. A missing part is an OPEN obligation with an owner, counted and visible.", "enforcement_type": "GATE", "check": "G03 required relations, G10, G11, G13" }, { "id": "research-defines-scope", "text": "Scope comes from research: stated intentions, current primary sources, regulations, standards and every top competitor's screens, flows, capabilities and controls. Every competitor is a floor, matched then exceeded.", "enforcement_type": "GATE", "check": "G15; parity measured and published" }, { "id": "agents-native", "text": "Every operation has a human path and an agent path to the same outcome with equal capability, under the agent's own EDGE ID, with the same authority, preview, audit and undo. Every agent has a portal, a mailbox, a racetrack and a purpose-limited authority from a legitimate issuer; consequential actions are rehearsed with the other agents involved before they commit; people and agents share each workflow with stated hand-off, approval, pause, take-over and undo, conflicts resolved by lease and fence.", "enforcement_type": "GATE", "check": "G12; agent fields in G01" }, { "id": "accountable-authority", "text": "Every effectful actor has an authenticated identity, an accountable owner and a current purpose-limited permission from the competent issuer. Every effectful object names the human or organisational role that answers if it goes wrong; an agent is never the answerer of last resort.", "enforcement_type": "GATE", "check": "G13" }, { "id": "evidence-with-the-effect", "text": "Every action binds its inputs, prior state, release, contract, predicate and receiving effect to durable evidence; the receiving system's own readback establishes the outcome. Rendering or request acceptance alone establishes nothing.", "enforcement_type": "GATE", "check": "evidence fields in G01; G09" }, { "id": "verification-by-reproduction", "text": "Correctness comes from reproducing the exact contract against the actual evidence with the declared checker. VERIFIED records whether the reproduction was by the same lineage or an independent one; where an applicable authority requires independence, nothing less is accepted for that claim.", "enforcement_type": "GATE", "check": "G04 VERIFIED rules" }, { "id": "preventive-enforcement", "text": "Unauthorised, unsafe or conflicting effects are refused at the boundary before execution; emerging risks are predicted and corrected before they become incidents; every discovered defect becomes a correction to the mechanism and a permanent regression test. Zero silent failures, zero fabricated assurance, zero unauthorised actions, zero avoidable harm.", "enforcement_type": "CONDUCT", "check": "incident objects carry cause, correction and regression test (G01); monitors are live" }, { "id": "authorised-recovery", "text": "Every operation has a verified recovery or compensation; irreversible consequences are explicitly controlled; an ambiguous effect is queried, never replayed; recovery preserves the intention and the acceptance predicate.", "enforcement_type": "GATE", "check": "operation recovers_by in G03" }, { "id": "glass-box", "text": "Knowledge, reasoning, authority, execution, verification, recovery and efficiency are inspectable by authorised people, including regulators, without the builder's help. Observation protects the dignity and privacy of the person; to be seen is not to be watched. Knowledge about a person confers no authority over that person.", "enforcement_type": "GATE", "check": "REGULATOR journey in G11; inspection surfaces on every journey" }, { "id": "sovereign-data", "text": "Each organisation's data and agents stay under its own control inside the estate and never go to model providers; agents run inside the estate.", "enforcement_type": "CONDUCT", "check": "access_class on every object (G01); data-flow review on every endpoint" }, { "id": "clear-screens", "text": "Every screen shows the essential result first and keeps depth behind a help control as markdown served to people and agents alike; every screen is accessible and completes for a person with no sight.", "enforcement_type": "GATE", "check": "G14" }, { "id": "approved-design", "text": "Only approved UX Pilot, Figma and Untitled UI structure and the estate's original logos, used as-is; no hand-drawing and no logo recreation.", "enforcement_type": "GATE", "check": "designed_by in G14; design-asset approval field" }, { "id": "outcomes-first", "text": "Every capability serves at least one outcome: time, money, lives, planet, resources or value, saved or more. Outcomes are claimed only when verified with their measurement; resources consumed per verified outcome (tokens, GBP, kWh, gCO2e, litres, seconds) are published for every agent and capability.", "enforcement_type": "GATE", "check": "G10; ledger measurements" }, { "id": "compiled-runtime", "text": "Neural models are the factory, used at build time; what runs is deterministic and compute-light: TOK facts, compiled procedures with certified scope and error bound, answers rendered from proofs with sources. Outside a certified scope the answer is unknown or a question. The AI-free rate is published and rises.", "enforcement_type": "GATE", "check": "G17; AI-free rate, cost, latency and reproducibility measured" }, { "id": "lean-estate", "text": "One instruction markdown; every file accounted for by an object; no duplicates; raw material, datasets and large generated data in storage by digest; nothing is stored without justification.", "enforcement_type": "GATE", "check": "G03, G08" }, { "id": "always-new", "text": "Every build starts from the current sources and the current agreement. Nothing is inherited: research is admitted as observation or claim, never as authority, and every rule, count, permission and status is produced fresh.", "enforcement_type": "GATE", "check": "G04 authority rules; G16" }, { "id": "reconciled-change", "text": "Every change to a principle or decision names what it supersedes or narrows and retires it in the same commit; the principal's wording stays verbatim; a check no program can evaluate is never a gate.", "enforcement_type": "GATE", "check": "G06" }, { "id": "always-on-monitoring", "text": "Everything is reproducible, retraceable, auditable and monitored always on; monitoring is part of assurance and never sufficient alone.", "enforcement_type": "MEASURE", "check": "monitor coverage and freshness published from the objects" }, { "id": "latest-approval-governs", "text": "Only the most recent explicit approval of an exact subject by its competent authority is active.", "enforcement_type": "CONDUCT", "check": "decision objects carry decided_at; the latest governs" }, { "id": "act-on-intention", "text": "Standing intention is direction, not a step approval; agents act on it without questions. A question is asked only where intention and authority do not cover the exact action, or where a legal or irreversible-safety condition requires a recorded confirmation.", "enforcement_type": "CONDUCT", "check": "none mechanical" }, { "id": "one-clock", "text": "Every claimed run is measured on one external monotonic clock including research, provisioning, build, verification, review, deployment, operation, admission and recording; a timeout or incomplete outcome is never a success.", "enforcement_type": "MEASURE", "check": "whole elapsed time published with every claimed run" }, { "id": "extension-not-redesign", "text": "Every human domain is added by extension of TOK with crosswalks to its established ontologies and standards; the kernel does not change to admit a domain.", "enforcement_type": "GATE", "check": "G18" } ], "competitive_position": { "named_rival": "Palantir", "lane": "intelligence", "contrast": { "black_box": "glass box", "darkness": "light", "limitation": "the possibilities of what God can do", "control": "to be seen", "hoarding": "sharing for good", "built_around_agents": "built for agents, and so better and safer for people" } }, "runtime_thesis": { "name": "Compiled governed intelligence", "statement": "Neural models are the factory: at build time they read sources, write procedures, data, tests and proofs, and verify them against known answers. What runs for people and agents is deterministic: TOK facts, compiled procedures with certified error bounds, and answers rendered from proofs. The same input gives byte-identical output; every answer carries its sources and bound; outside a certified scope the answer is UNKNOWN or a question. Small certified models may perceive messy inputs; no large general model is required at runtime.", "measures": [ "AI-free rate", "cost and latency per request against a large model doing the same job", "certified versus measured error rate on expert-graded questions", "reproducibility: identical output on rerun" ], "first_proofs": [ "sanctions and PEP name screening with a guaranteed miss rate on official lists", "an open drug-interaction source with a guaranteed false-assertion rate on a pharmacist-graded held-out set" ] }, "research": { "rule": "Research is pinned input held in storage and referenced by digest; it never enters the repository and is never authority. The factory produces new objects from it, each with its provenance.", "kinds": [ "statements of intention from authorised principals (secrets replaced by keyed-hash markers; human-written text separated from agent-written text)", "regulations and standards", "competitor documentation and screens", "baselines and benchmarks", "source code", "diagrams", "research documents", "datasets and models with their licences" ] }, "open_decisions": [ "the host that runs the self-hosted CI", "the storage location for research and datasets", "whether screens mirror a competitor's exact layout or only its behaviour and features", "which mark is EDGE Public Consortium's: the-edge-consortium.svg, public-edge.svg or cyber-governance.svg (all admitted under design/brand/)" ], "change_control": { "amendment": "tools/amend.py; every change names what it supersedes or narrows and retires it in the same commit", "views": "generated/ is build output; a hand edit is a gate failure", "fingerprint": "tools/gate.py snapshot; a stale fingerprint authorises nothing" }, "what_never_enters": [ "credentials", "machine probes", "host access maps", "personal contact data", "raw research, datasets, scrapes and large generated data", "duplicate or stale generated output", "any rule, budget, permission, approval, count, identity or readiness level not produced in this estate" ] } ``` ## 2. Ontology (TOK) ### tok/KERNEL.md `tok/KERNEL.md` · sha256 `221a53d23963baa81fefcd7a1baaea3ab0431d58fe5c6609c08224dff2276c61` #### TOK kernel, version 0 TOK is the one ontology of the estate and the realisation of the Tree of Knowledge. The machine-readable kernel is `tok/kernel.json`; this file explains it. Where the two differ, `kernel.json` governs. ##### 1. Why one ontology The same machinery describes every domain: identity, relations, processes, time, provenance, logic, Bayesian updating and error-rate control. A sanction designation, a drug interaction, a court ruling and a payment are each *an entity, related to others, asserted by a source, at a time, with some confidence, updated by evidence*. One kernel therefore serves every application and every human domain, and domains are added by extension, never by redesigning the kernel. Gate G18 measures exactly that: adding a domain changes nothing in the kernel. ##### 2. Truth, agreement, claim Truth is complete and already exists; it can only be observed. Agreement creates laws, rules, standards and plans, and agreements can be disagreed with. A claim is what someone asserts. | Class | Records | Example | |---|---|---| | OBSERVATION | what was observed, how, by whom, when, with what reliability and evidence | a regulator's register listing a firm as authorised on a date | | AGREEMENT | who agreed and who dissented, with what authority, where and from when it applies, on what basis | a rule of this estate; a law; a standard | | CLAIM | who asserts it, on what evidence, who supports and who disputes it | a vendor's published benchmark score | Two consequences shape the estate: 1. **Actions obey agreements; knowledge follows observation.** The estate acts under law, permission and policy. What it believes is governed by evidence. Where an agreement departs from the evidence, the estate shows both and breaks neither. That an agreement exists is itself an observable fact. 2. **Confidence comes from evidence, not headcount.** Independent sources with evidence move confidence; a thousand copies of one source count as one; agreement without evidence stays an agreement. The estate is the source of the strongest evidence, not of the most agreement. Gate G04 refuses any object that renders one class as another. ##### 3. Truth statuses and confidence Every object's realisation carries one status: VERIFIED, OBSERVED, HYPOTHESIS, PREDICTION, SIMULATION, UNKNOWN, REFUTED, or ASPIRATION on a claim explicitly about the future. A requirement is an agreement and is never an aspiration; its *realisation* has a status, UNKNOWN until observed. VERIFIED records its independence level: REPRODUCED_SAME_LINEAGE (reproduced by another agent of the same lineage or host) or INDEPENDENT_LINEAGE (different lineage, credentials, host and evidence store). Both are true statements; neither pretends to be the other. A claimed error rate states its calibration set, bound, confidence level and last recalibration. Outside a certified scope the answer is UNKNOWN or a question, never a guess. ##### 4. Identity `tok::`. Identity is permanent, never reused, and separate from label, revision, digest and key. The object's file path is `tok/objects//.json`: the path is the identity, the label lives inside. Retirement leaves a retired record; nothing silently disappears. A generated index maps identities to labels for people. ##### 5. The ten answers Every object answers, for people in plain language and for agents in machine fields: why it exists, what it is, what it must do, what it looks like, how it works, what depends on it, what authority permits it, how it is verified, what happens when it fails, what resources it consumes. An answer not yet known is an explicit `OPEN` obligation with an owner and a date. Open obligations are counted live by the build and reported; they are never hidden and never dressed as aspirations. ##### 6. Kinds and relations The estate layer declares its kinds in `kernel.json`, each with its statement class and its required fields and relations. Required relations that cannot yet be satisfied are `OPEN` obligations, so an object can be admitted early and completed honestly. Relations are typed and directional. Disagreement is a first-class relation: two agreements or claims that disagree are both kept, with their sources, and the disagreement is visible. ##### 7. Admission Nothing enters by copying. `tools/admit.py` mints the identity, classifies, records provenance (origin, digest, admitter, time, licence, access class), scans for what never enters, relates, answers, sets status from evidence, and places the file. Large or restricted bytes go to storage as an `external-artifact` referenced by digest. Secrets found in a source that must be kept are replaced by a keyed-hash marker, so equal secrets remain recognisable as equal without being recoverable. Research is admitted as observation or claim and is never authority: no rule, budget, permission, approval, identity, count or readiness level is admitted as authority unless it is produced in this estate. ##### 8. Amendment `tools/amend.py` changes principles. Every change names what it supersedes or narrows, or states that it supersedes nothing, and the superseded principle is retired in the same commit. The principal's wording stays verbatim on the object; the mechanical check is written separately and typed GATE, MEASURE, CONDUCT or DECISION. A check no program can evaluate is never typed GATE. This is the rule that stops contradictions accumulating. ##### 9. Gates Twenty-two gates run on every commit in the self-hosted CI, and `main` requires them: schema, identity, closure, statement class, byte-identical generation, amendment discipline, tests, hygiene, external artifacts, outcome links, the full journey set, agent parity, answerability, screen density and design source, competitor floors, live counts, procedure certification, extension without redesign, definitions, the behaviour model, screen blueprints and the agent lifecycle. An empty ontology must pass all of them before any content is admitted. ##### 10. Naming and files - Objects: `tok/objects//.json`. - Domain extensions: `tok/domains//`, passing G18. - Sources: `sources/.` with a `source` object of the same identity. - Code: `src//` with a `code-module` object; tests in `tests/` with `test` objects. - Generated: `generated/`, only ever written by `tools/build.py`. Every tracked file is accounted for by exactly one object or by the kernel's infrastructure list (G03). A file nothing accounts for is a gate failure, which is how the repository stays exactly what it needs to be. ##### 11. Runtime The runtime unit of the estate is the `procedure`: a compiled deterministic artifact with a certified scope, an error bound, a calibration set and a deterministic rerun (G17). Neural models are the factory, used in the lab to discover, write and verify procedures; what runs for people and agents is the procedure, on ordinary CPUs, with the same output every time and its bound attached. Perception of messy inputs (images, speech) may use small certified models; no large general model is required at runtime. The AI-free rate, the share of requests served with no model call, is a published measurement that rises over time. ##### 12. One behaviour model for every action Every action in the estate, whether a person presses a button, an agent calls a tool, a schedule fires or a batch runs, is described the same way before it is built. Generators can only produce an action from a complete description, which is what makes every screen and agent route predictable. | Part | What it says | |---|---| | Actor | who does it (person, agent, organisation, system), their role, and who they act for | | Mandate | for an agent or delegate: what it may do, for whom, limits, start, expiry, how it is revoked | | Paths | the human route (a screen) and the agent route (a tool) to the same outcome, with the same authority, preview, audit and undo | | Permission | the authority that allows it, checked before anything happens | | Inputs | every input typed, validated and sourced; nothing guessed | | Guards | every precondition, its data source, and the exact refusal if it fails | | Rehearsal | a dry run on the actor's racetrack showing what would happen, recorded as a simulation | | Approval | who must approve, at what threshold | | Operation | the effect, with an idempotency key (so repeating it never doubles it), a timeout and the receiving system | | Receiving confirmation | the receiving system's own confirmation that the effect happened | | Success | the checks after confirmation, and what the actor sees | | Failure | every failure as a typed code, a plain message for people, a structured one for agents, and whether anything changed | | Unknown | when the outcome is unclear: ask the receiving system and reconcile; never blindly repeat | | Recovery | how to undo or compensate, within what window, who owns it; irreversible effects marked and approved | | Evidence | what is recorded with the effect, bound by digest | | Batch | for many items: per-item idempotency, order, concurrency, rate limits, checkpoint and resume, what happens when some fail, and a summary that accounts for every item | | Output verification | each output checked against its acceptance test, by a named checker, with its error bound | | Monitoring | live time to evidence, prompt to outcome, resource to outcome, success, failure and error rates | Every action moves through the same states: DRAFTED, REHEARSED, AUTHORISED, EXECUTING, CONFIRMED, VERIFIED, or REFUSED, FAILED, UNKNOWN, RECONCILING, RECOVERING, RECOVERED. Gate G20 refuses any action whose description is incomplete. ##### 13. Agents are born, live and die on the record Agents are real customers and staff, so each has a life on the record: - **Birth certificate.** Identity (EDGE ID), when and by whom it was created, the accountable owner (always a person or organisation, never an agent), purpose, a fixed term, model and version lineage, capabilities, mandate, mailbox, portal, racetrack, and keys generated and held by the agent's own host. - **Life events.** Mandates granted, changed or revoked; version upgrades (a new version gets a new certificate linked to its lineage); suspensions; incidents; handovers; embodiments bound or released. - **Death certificate.** When and why it ended, where every open obligation went, keys revoked, what happened to its data, and what records are kept. - **Embodiment.** An agent may act through a device, a robot or an avatar. The embodiment has its own certificate, acts only within the agent's mandate, carries physical-safety controls, and its owner can always stop it. No agent acts without a valid birth certificate and an unexpired mandate; an agent with a death certificate never acts again, and its identity is never reused (gate G22). ##### 14. Complete screen blueprints A screen is generated only from a complete blueprint: its purpose, actors, journey step, entry and exit, essential-first content, approved components, every value bound to its source, every action bound to one operation, every state (empty, loading, success, partial, error, refused, unauthorised, offline, unknown), accessibility, help depth, the agent equivalent, plain-language copy, competitor floor and design source. Every generated screen is checked against its blueprint (gate G21). ##### 15. Every word defined Every term used anywhere is a definition object with a plain-language meaning a brand-new engineer can understand. An undefined term fails gate G19, and the glossary is generated from the definitions. ### tok/kernel.json `tok/kernel.json` · sha256 `358eee947460932cd0c6aa542f77b2f63d4a8567f6813006fe1781e420f03a0d` ```json { "schema": "tok.kernel.v0", "version": "0.1.0", "statement_class_of_this_file": "AGREEMENT", "name": "TOK", "realises": "Tree of Knowledge", "purpose": "The one ontology of the estate. Every object the estate builds, runs, measures or decides is a TOK object. Every generated document, screen, test and proof resolves from TOK. Domains are added by extension, never by redesigning the kernel.", "identity": { "scheme": "tok::", "ulid": "26-character Crockford base32, lower case, minted once by admit.py", "rules": [ "An identity is permanent and never reused.", "Identity is distinct from label, revision, content digest and key.", "Retiring an object leaves a retired record with the identity; identities never silently disappear.", "The object's file path is tok/objects//.json; the path is the identity, the label lives inside the file.", "Aliases and external identifiers are recorded on the object, never used as its identity." ] }, "statement_classes": { "OBSERVATION": { "meaning": "What was observed about reality.", "required": [ "observed_what", "observed_how", "observed_by", "observed_at", "reliability", "evidence" ] }, "AGREEMENT": { "meaning": "What people agreed: law, rule, standard, policy, plan, requirement, decision. Can be disagreed with.", "required": [ "agreed_by", "dissented_by", "authority", "applies_where", "applies_from", "applies_until", "basis" ], "note": "That an agreement exists is itself an observable fact and may be recorded as an OBSERVATION about the agreement." }, "CLAIM": { "meaning": "What someone asserts.", "required": [ "claimed_by", "claimed_at", "evidence", "supported_by", "disputed_by" ] }, "invariant": "A statement is exactly one class. Rendering a class as another is a gate failure." }, "truth_statuses": { "values": [ "VERIFIED", "OBSERVED", "HYPOTHESIS", "ASPIRATION", "PREDICTION", "SIMULATION", "UNKNOWN", "REFUTED" ], "rules": [ "VERIFIED requires evidence reproduced by a party that did not produce it; the independence level is recorded (REPRODUCED_SAME_LINEAGE or INDEPENDENT_LINEAGE).", "OBSERVED requires a direct observation record.", "A requirement in the specification is an AGREEMENT; its realisation carries a truth status; it is never labelled ASPIRATION.", "ASPIRATION is permitted only on a claim object explicitly about the future, never on a requirement, test, readiness level or result.", "UNKNOWN is never zero and never a total.", "No one raises a status above its evidence." ] }, "confidence": { "method": "Bayesian update from evidence; conformal or risk-controlled bounds wherever an error rate is claimed.", "rules": [ "Independent sources with evidence move confidence; copies of one source count as one.", "Agreement without evidence stays an agreement; it never becomes an observation.", "A claimed error rate states its calibration set, bound, confidence level and the date it was last recalibrated.", "Outside a certified scope the answer is UNKNOWN or a question, never a guess." ] }, "ten_answers": [ "why it exists", "what it is", "what it must do", "what it looks like", "how it works", "what depends on it", "what authority permits it", "how it is verified", "what happens when it fails", "what resources it consumes" ], "object_file_format": { "required": [ "id", "kind", "label", "aliases", "statement_class", "statement", "status", "answers", "relations", "open_obligations", "provenance", "fields" ], "optional": [ "accounts_for", "retired" ], "answers_keys": [ "why_it_exists", "what_it_is", "what_it_must_do", "what_it_looks_like", "how_it_works", "what_depends_on_it", "what_authority_permits_it", "how_it_is_verified", "what_happens_when_it_fails", "what_resources_it_consumes" ], "answer_value": "text plus machine fields, or {\"OPEN\": {\"owner\": , \"since\": }}", "relation": { "type": "", "to": "" }, "open_obligation": { "what": "", "owner": "", "since": "" }, "provenance": { "source": "", "digest_sha256": "", "admitted_by": "", "admitted_at": "", "licence": "", "access_class": "PUBLIC|ESTATE|ORGANISATION|PRIVATE|RESTRICTED" }, "retired": { "by": "", "at": "" } }, "infrastructure_paths": { "note": "Files the kernel itself accounts for (gate G03). Every other tracked file must be an object file or be listed in some object's accounts_for.", "paths": [ "README.md", "AGENTS.md", ".gitignore", "pyproject.toml", "tok/kernel.json", "tok/KERNEL.md", "contract/", "spec/", "research/", "tools/", "tests/", "ci/", "generated/", "design/components.json", "tok/ids.lock.json" ] }, "kinds": { "intention": { "class": "AGREEMENT", "requires": { "fields": [ "statement", "principal" ], "relations": [ "measured_by:outcome" ] } }, "outcome": { "class": "AGREEMENT", "requires": { "fields": [ "dimension", "direction", "unit" ], "relations": [ "measured_by:measurement" ] } }, "principle": { "class": "AGREEMENT", "requires": { "fields": [ "enforcement_type", "check" ], "relations": [ "supersedes|narrows|none" ] }, "enforcement_type": [ "GATE", "MEASURE", "CONDUCT", "DECISION" ] }, "decision": { "class": "AGREEMENT", "requires": { "fields": [ "decided_by", "decided_at", "options_considered" ], "relations": [ "supersedes|narrows|none" ] } }, "application": { "class": "AGREEMENT", "requires": { "fields": [ "formal_name", "interface_name" ], "relations": [ "serves:intention", "has_journey:journey", "designed_by:design-asset", "floors:competitor|none_found" ] } }, "foundation": { "class": "AGREEMENT", "requires": { "relations": [ "serves:intention", "used_by:application" ] } }, "capability": { "class": "AGREEMENT", "requires": { "relations": [ "serves:intention", "serves:outcome", "belongs_to:application|foundation", "realised_by:journey", "verified_by:test", "answerable:role", "floors:competitor-feature|none_found" ] } }, "journey": { "class": "AGREEMENT", "requires": { "fields": [ "journey_type", "actor_kind", "steps" ], "relations": [ "realises:capability|application", "passes_through:screen|operation", "rehearsed_by:test" ] }, "journey_type": [ "HUMAN", "AGENT", "REGULATOR", "ENGINEER", "OPERATOR", "WHOLE_OF_LIFE" ] }, "screen": { "class": "AGREEMENT", "requires": { "fields": [ "blueprint" ], "relations": [ "shows:journey", "designed_by:design-asset", "mirrors:competitor-feature|none_found" ] }, "note": "fields.blueprint holds every part of screen_blueprint; gate G21 checks it is complete." }, "state": { "class": "AGREEMENT", "requires": { "relations": [ "transitions_to:state", "guarded_by:control" ] } }, "control": { "class": "AGREEMENT", "requires": { "fields": [ "prevents" ], "relations": [ "enforces:principle", "authorised_by:role" ] } }, "operation": { "class": "AGREEMENT", "requires": { "fields": [ "behaviour" ], "relations": [ "authorised_by:role", "recovers_by:operation", "answerable:role", "exposed_at:endpoint" ] }, "note": "fields.behaviour holds every part of behaviour_model; gate G20 checks it is complete." }, "endpoint": { "class": "AGREEMENT", "requires": { "fields": [ "protocol", "auth" ], "relations": [ "serves:operation" ] } }, "agent": { "class": "AGREEMENT", "requires": { "fields": [ "edge_id", "mailbox", "portal", "racetrack", "purpose", "term", "lineage", "keys_held_by" ], "relations": [ "owned_by:organisation|role", "answerable:role", "mandated_by:mandate", "certified_by:lifecycle-event" ] } }, "role": { "class": "AGREEMENT", "requires": { "fields": [ "answerability" ] } }, "organisation": { "class": "OBSERVATION", "requires": { "fields": [ "legal_identity" ] } }, "person": { "class": "OBSERVATION", "requires": { "fields": [ "role_only_no_contact_data" ] } }, "source": { "class": "OBSERVATION", "requires": { "fields": [ "source_kind", "dated", "licence" ] }, "source_kind": [ "INTENT", "REGULATION", "STANDARD", "RESEARCH", "COMPETITOR", "DATASET", "BLUEPRINT", "RECORDING", "OTHER" ] }, "design-asset": { "class": "OBSERVATION", "requires": { "fields": [ "approval", "origin_tool" ] }, "origin_tool": [ "ORIGINAL_BRAND", "FIGMA", "UX_PILOT", "UNTITLED_UI", "BLUEPRINT" ] }, "code-module": { "class": "OBSERVATION", "requires": { "fields": [ "language", "dependencies_declared", "runs" ], "relations": [ "tested_by:test", "implements:operation|capability" ] } }, "test": { "class": "OBSERVATION", "requires": { "fields": [ "oracle", "reproduction_command", "last_result" ] } }, "evidence": { "class": "OBSERVATION", "requires": { "fields": [ "producer", "receiver_readback", "digest_sha256" ] } }, "measurement": { "class": "OBSERVATION", "requires": { "fields": [ "value", "unit", "method", "observed_at", "counting_method", "completeness" ] } }, "benchmark": { "class": "AGREEMENT", "requires": { "fields": [ "definition", "version_pinned", "scoring" ], "relations": [ "measured_by:measurement" ] } }, "competitor": { "class": "OBSERVATION", "requires": { "fields": [ "name", "observed_at" ] } }, "competitor-feature": { "class": "OBSERVATION", "requires": { "fields": [ "observed_at", "source_url" ], "relations": [ "belongs_to:competitor" ] } }, "external-artifact": { "class": "OBSERVATION", "requires": { "fields": [ "location", "digest_sha256", "bytes", "licence", "access_class" ] } }, "incident": { "class": "OBSERVATION", "requires": { "fields": [ "cause", "correction", "regression_test" ] } }, "procedure": { "class": "OBSERVATION", "requires": { "fields": [ "compiled_from", "certified_scope", "error_bound", "calibration_set", "deterministic" ] }, "note": "A compiled deterministic procedure produced by the lab: the runtime unit of the estate." }, "claim": { "class": "CLAIM", "requires": { "fields": [ "about", "about_future" ] }, "note": "The only kind on which ASPIRATION may appear, and only when about_future is true." }, "definition": { "class": "AGREEMENT", "requires": { "fields": [ "term", "meaning", "plain_language" ], "relations": [ "derived_from:source|none_found" ] } }, "mandate": { "class": "AGREEMENT", "requires": { "fields": [ "scope", "limits", "starts", "expires", "revocation" ], "relations": [ "granted_by:role|organisation", "granted_to:agent" ] } }, "lifecycle-event": { "class": "OBSERVATION", "requires": { "fields": [ "event", "at", "by" ], "relations": [ "concerns:agent|embodiment" ] }, "event": [ "BIRTH", "MANDATE_GRANTED", "MANDATE_CHANGED", "MANDATE_REVOKED", "VERSION_UPGRADE", "SUSPENDED", "REINSTATED", "INCIDENT", "HANDOVER", "EMBODIMENT_BOUND", "EMBODIMENT_RELEASED", "DEATH" ] }, "embodiment": { "class": "AGREEMENT", "requires": { "fields": [ "form", "safety_controls", "stop_control" ], "relations": [ "embodies:agent", "answerable:role" ] }, "form": [ "DEVICE", "ROBOT", "AVATAR", "OTHER" ] } }, "relations": { "serves": "object -> intention | outcome", "belongs_to": "capability | screen | journey | competitor-feature -> application | foundation | competitor", "realises": "journey -> capability | application", "realised_by": "capability -> journey", "has_journey": "application -> journey", "shows": "screen -> journey", "shown_by": "journey -> screen", "passes_through": "journey -> screen | operation", "transitions_to": "state -> state", "guarded_by": "state | operation -> control", "enforces": "control -> principle", "governed_by": "any -> principle", "authorised_by": "operation | control | agent -> role", "answerable": "any effectful object -> role (never an agent)", "exposed_at": "operation -> endpoint", "implements": "code-module -> operation | capability", "tested_by": "code-module | capability -> test", "verified_by": "any -> evidence", "measured_by": "intention | outcome | benchmark -> outcome | measurement", "derived_from": "any -> source", "designed_by": "application | screen -> design-asset", "floors": "application | capability -> competitor | competitor-feature", "mirrors": "screen -> competitor-feature", "depends_on": "any -> any", "supersedes": "principle | decision -> principle | decision", "narrows": "principle | decision -> principle | decision", "agrees_with": "claim | agreement -> claim | agreement", "disagrees_with": "claim | agreement -> claim | agreement", "rehearsed_by": "journey | operation -> test (SIMULATION)", "recovers_by": "operation -> operation", "owned_by": "agent -> organisation | role", "used_by": "foundation -> application", "compiled_from": "procedure -> source | code-module | test", "retired_by": "any -> decision", "granted_by": "mandate -> role | organisation", "granted_to": "mandate -> agent", "mandated_by": "agent -> mandate", "certified_by": "agent -> lifecycle-event (BIRTH)", "concerns": "lifecycle-event -> agent | embodiment", "embodies": "embodiment -> agent", "defines": "definition -> any" }, "admission": { "tool": "tools/admit.py", "steps": [ "Identify: mint the identity; record label, aliases and external identifiers.", "Classify: assign kind and statement class.", "Source: record provenance (origin, digest, admitter, time, licence, access class).", "Scan: refuse credentials, personal contact data, host access details and machine paths; where a source must be kept, replace each secret by a keyed-hash marker so equal secrets stay recognisable without being recoverable.", "Relate: record required relations or OPEN obligations with an owner.", "Answer: fill the ten answers or OPEN obligations.", "Status: set the realisation status from evidence, never above it.", "Place: write tok/objects//.json; large or restricted bytes go to storage as an external-artifact." ], "research_is_never_authority": "Research is admitted as observation or claim. No rule, budget, permission, approval, actor identity, count or readiness level is admitted as authority unless it is produced in this estate." }, "amendment": { "tool": "tools/amend.py", "rules": [ "A new or changed principle names what it supersedes or narrows, or states explicitly that it supersedes nothing.", "Superseded principles are retired in the same commit.", "The principal's wording is kept verbatim on the object; the enforcement type and the mechanical check are written separately.", "A principle whose check no program can evaluate is typed CONDUCT or DECISION, never GATE." ] }, "gates": [ { "id": "G01", "name": "schema", "check": "Every object validates against this kernel." }, { "id": "G02", "name": "identity", "check": "Identities unique, never reused; no identity on main disappears without a retired record." }, { "id": "G03", "name": "closure", "check": "Every relation target exists; every tracked file is accounted for by exactly one object or by infrastructure_paths; every kind's required relations are present or OPEN with an owner." }, { "id": "G04", "name": "statement-class", "check": "Every object has exactly one class matching its kind; the class's required statement fields are present; no status exceeds its evidence; ASPIRATION appears only on a claim about the future." }, { "id": "G05", "name": "generation", "check": "generated/ equals a fresh build byte for byte." }, { "id": "G06", "name": "amendment", "check": "Every principle and decision carries supersedes or narrows relations or supersedes_nothing; every superseded object is retired." }, { "id": "G07", "name": "tests", "check": "All tests pass with dependencies installed from the declared manifest." }, { "id": "G08", "name": "hygiene", "check": "No secrets; no personal contact data; no machine paths, private addresses or host access commands; no file over 1 MB unless a declared data object; no duplicate content." }, { "id": "G09", "name": "external-artifacts", "check": "Each has digest, location, bytes, licence and access class; availability checked when the storage host is reachable." }, { "id": "G10", "name": "outcome-link", "check": "Every capability serves at least one intention and one outcome." }, { "id": "G11", "name": "journey-set", "check": "Every capability has HUMAN, AGENT, REGULATOR, ENGINEER and OPERATOR journeys, or OPEN obligations with owners; every application has a WHOLE_OF_LIFE journey." }, { "id": "G12", "name": "agent-parity", "check": "Every operation's behaviour names a human path and an agent path (MCP) to the same outcome, with the same authority, preview, audit and undo." }, { "id": "G13", "name": "answerability", "check": "Every effectful object names an answerer role; no answerer is an agent." }, { "id": "G14", "name": "screen-density", "check": "Every screen's blueprint declares essential_first, help_markdown, accessibility and design_source." }, { "id": "G15", "name": "competitor-floor", "check": "Every application, capability and screen links competitor features or records none_found with its search evidence." }, { "id": "G16", "name": "live-counts", "check": "Every number in generated/ is produced by the build from the objects, with method and completeness; no count is written by hand anywhere." }, { "id": "G17", "name": "procedure-certification", "check": "Every runtime procedure has a certified scope, error bound, calibration set and deterministic rerun." }, { "id": "G18", "name": "extension-without-redesign", "check": "Adding a domain module changes no file under tok/kernel.json or tok/KERNEL.md; recorded per domain as a measurement." }, { "id": "G19", "name": "definitions", "check": "Every term in the README glossary, and every kind, relation, statement class and truth status that an object uses, has a definition object; the glossary is generated." }, { "id": "G20", "name": "behaviour-model", "check": "Every operation's fields.behaviour contains every part of behaviour_model (or not-applicable with a reason) and only the declared states and transitions." }, { "id": "G21", "name": "screen-blueprint", "check": "Every screen's blueprint contains every screen_blueprint part and every state; every control's action and every action's operation exist; every component is on the approved component list." }, { "id": "G22", "name": "agent-lifecycle", "check": "Every agent has a BIRTH lifecycle event with every birth-certificate field, an unexpired mandate to act, and no actions after a DEATH event." } ], "layers": [ "estate (applications, foundations, capabilities, journeys, screens, controls, operations, agents, procedures, evidence): built first", "domain (every human domain of expertise, by crosswalk to its established ontologies and standards): added by extension", "upper (universal categories): added by extension, with crosswalks", "instance (live population with provenance)" ], "crosswalk_targets": [ "ISO/IEC 21838-2 (BFO)", "DOLCE", "W3C PROV-O", "Schema.org", "Wikidata", "SNOMED CT", "FIBO", "ISO 15926", "Dublin Core", "SKOS", "GS1", "ISO 20022", "HL7 FHIR" ], "extension_rule": "A domain module adds kinds, relations and crosswalks under tok/domains// and must pass G18.", "behaviour_model": { "applies_to": "every operation, whoever performs it (person, agent, organisation, system) and however it is triggered (screen, agent tool, schedule, batch, event)", "parts": { "actor": "who performs it: kind (person, agent, organisation, system), role, and the principal it acts for", "mandate": "for an agent or delegate: what it may do, for whom, spending and volume limits, start, expiry, revocation route", "paths": "the human path (screen) and the agent path (MCP tool) to the same outcome, with the same authority, preview, audit and undo", "permission": "the authority that allows it: issuer, principal, purpose, resource, expiry, revocation; checked at the boundary before anything happens", "inputs": "every input typed, validated and sourced; nothing guessed", "guards": "every precondition as a named predicate with its data source and the exact refusal returned when it fails; evaluated before the effect", "rehearsal": "a dry run on the actor's racetrack showing predicted effects, cost, risk and affected parties before commit; the rehearsal result is recorded as SIMULATION", "approval": "who must approve, at what threshold, and how (none, actor's own confirmation, principal, second person, competent authority)", "operation": "the effect itself, with an idempotency key, a timeout on the one honest clock, and the receiving system named", "receiving_confirmation": "the receiving system's own readback that the effect happened; rendering or request acceptance alone is not confirmation", "success": "the postconditions checked after confirmation, and what the actor sees and receives", "failure": "every failure as a typed code with a plain-language message for people, a structured message for agents, and whether anything changed", "unknown": "when the outcome is ambiguous (timeout, lost reply): query the receiving system and reconcile; never replay blindly", "recovery": "the compensating operation, the undo window, escalation route and owner; irreversible effects are explicitly marked and need approval", "evidence": "what is produced with the effect: inputs, prior state, decision, confirmation and outcome, bound by digest", "batch": "for many items: per-item idempotency, ordering, concurrency limits with lease and fence, rate limits, checkpoint and resume, partial-failure rule (stop, skip-and-report, or compensate), and a batch summary that reconciles every item", "output_verification": "each output checked against its acceptance predicate; the checker and its version named; sampling or independent re-check where declared; the error bound stated", "monitoring": "live measures: time to evidence, prompt to outcome, resource to outcome, success and failure rates, error rate against its bound" }, "states": [ "DRAFTED", "REHEARSED", "AUTHORISED", "EXECUTING", "CONFIRMED", "VERIFIED", "FAILED", "UNKNOWN", "RECONCILING", "RECOVERING", "RECOVERED", "REFUSED" ], "transitions": [ [ "DRAFTED", "REHEARSED" ], [ "DRAFTED", "REFUSED" ], [ "REHEARSED", "AUTHORISED" ], [ "REHEARSED", "REFUSED" ], [ "AUTHORISED", "EXECUTING" ], [ "EXECUTING", "CONFIRMED" ], [ "EXECUTING", "FAILED" ], [ "EXECUTING", "UNKNOWN" ], [ "UNKNOWN", "RECONCILING" ], [ "RECONCILING", "CONFIRMED" ], [ "RECONCILING", "FAILED" ], [ "CONFIRMED", "VERIFIED" ], [ "CONFIRMED", "FAILED" ], [ "FAILED", "RECOVERING" ], [ "RECOVERING", "RECOVERED" ], [ "RECOVERING", "FAILED" ] ], "rule": "An operation without every part of the behaviour model is not specified and cannot be generated, built or executed. Parts that genuinely do not apply are stated as not applicable with the reason." }, "agent_lifecycle": { "birth_certificate": [ "edge_id", "created_at", "created_by", "accountable_owner (person or organisation, never an agent)", "purpose", "term (start and end; fixed-term by default)", "model and version lineage", "capabilities", "mandate", "mailbox", "portal", "racetrack", "keys generated and held by the agent's own host" ], "life_events": [ "mandate granted, changed or revoked", "version upgrade (a new version is a new certificate linked to its lineage)", "suspension and reinstatement", "incident", "handover of obligations", "embodiment bound or released" ], "death_certificate": [ "ended_at", "reason (term ended, retired, revoked, replaced, incident)", "final state of every open obligation and where it went", "keys revoked", "data disposition", "records retained and for how long" ], "embodiment": "An agent may act through an embodiment (a device, a robot or an avatar). The embodiment is its own object with its own certificate; it acts only within the agent's mandate, carries physical-safety controls, and can be stopped by its accountable owner at any time.", "rule": "No agent acts without a valid birth certificate and an unexpired mandate; an agent with a death certificate can never act again; its identity is never reused." }, "screen_blueprint": { "parts": { "purpose": "the job this screen does, in one sentence", "actors": "which actors use it (ids from the unit's actors)", "journey_steps": "which journey steps it serves", "entry": "how actors arrive (screens, links, notifications, agent hand-offs)", "exits": "where actors can go next", "essential_first": "what is shown first: the one result or decision the actor needs", "fields": "every field: id, label, type, required, validation, binding to a data field, help text", "controls": "every control: id, component, label, the action it triggers, the condition that enables it", "actions": "every action: id, the operation it calls, where success and failure go", "transitions": "every move between screens and states: from, on (action result or event), to, condition", "conditions": "every named condition used by controls and transitions, with its expression", "components": "every component on the screen, from the approved component list, with what it binds to", "states": "copy and behaviour for each state: empty, loading, success, partial, error, refused, unauthorised, offline, unknown", "copy": "every piece of text: title, subtitle, labels, hints, confirmations, errors; plain language, defined terms only", "accessibility": "WCAG 2.2 AA notes: reading order, labels, focus, keyboard, contrast, reduced motion, screen-reader announcements", "help_markdown": "the depth behind the help control, served to people and agents", "agent_equivalent": "the MCP tool and structured output that reach the same outcome without the screen", "competitor_floor": "the competitor screens or flows it matches and exceeds, with source links", "design_source": "the approved design (Figma frame or design asset), or OPEN until produced" }, "states": [ "empty", "loading", "success", "partial", "error", "refused", "unauthorised", "offline", "unknown" ], "rule": "A screen missing any part cannot be compiled or generated. Every control triggers a declared action, every action calls a declared operation, every field binds to a declared data field, every component comes from the approved component list, and every state has copy." }, "definitions_rule": { "rule": "Every term used in any object, screen, message or document is a definition object (term, meaning in plain language, source, statement class). An undefined term is a gate failure. The glossary is generated from definitions.", "plain_language": "Definitions are written so a brand-new engineer with no history can understand them." } } ``` ### Objects by kind (compiled, this commit) | Kind | Objects | |---|---| | application | 4 | | capability | 50 | | code-module | 4 | | competitor | 237 | | competitor-feature | 468 | | decision | 21 | | definition | 648 | | design-asset | 13 | | endpoint | 876 | | foundation | 8 | | intention | 27 | | journey | 322 | | operation | 438 | | outcome | 12 | | principle | 27 | | role | 141 | | screen | 346 | ### Decisions - `tok:decision:01m68c87qmz54aqnbm8pqzmgt4` EDGE Public Consortium uses the Public EDGE mark. - `tok:decision:01m696xzzpcbnsexpcdtz7ak9n` Agents hold accounts in their own name. - `tok:decision:01m69b9nqqnxj7bh7p0mner5rf` TRL 9 and CRL 9 mean every intention carried out exactly. - `tok:decision:01m69gty59gnpc17g811317htn` Agent high-risk outcomes are provisional until the owner signs off in person. - `tok:decision:01m69jxxj5cazjefrrj1fvb962` EDGE Public Consortium becomes The Edge Consortium (education and research); Public EDGE is a separate application for government. - `tok:decision:01m69jxxqzrhz7r36ht4pvv5td` No irreversible outcome without a real, analog, in-person human act by the owner. - `tok:decision:01m69k8c61zde0er6j94yb4ekz` Judiverse comes with every app licence, is sold standalone to consumers, and is included in Edge Consortium tuition; students do paid real work offset against tuition. - `tok:decision:01m69m36psxrwhsa148z746ppv` London South Bank University is the first partner institution; SITHub teaches networking in the degree; The Edge Consortium registers with the Office for Students, shares LSBU's campus, and partners with LSBU for R&D and international technical students. - `tok:decision:01m69p22005ent1mresdjz6sh8` Scope frozen on 31 October 2026; every later change enters through scope change intake. - `tok:decision:01m69thft6cmgyfgkjrt6bdrws` The GitHub branch contract (fingerprint d1233bfd) governs the build, untouched. - `tok:decision:01m69thg13xhssvqbxygtr2zp8` Three domains, one site: judiverse.tech canonical, judeai.tech and x-x.tech mirror it; nothing is named estate. - `tok:decision:01m69thg7kks8fwaj0266pf1jy` TRL 9 and CRL 9 are reached when the closure gate passes on the deployed commit with the plug set empty. - `tok:decision:01m69xj6emef4jkw7kgv63bsdz` Jude carries out the whole plan: builds, reproduces, runs, deploys and maintains the system itself. - `tok:decision:01m6a4t0fxra5avfxf0pr2p692` Money provenance: trust from origin and activity, face value at par, four assessments kept separate. - `tok:decision:01m6a7p9c8566pe659w2hpgbjq` The plug set is post-closure: closure is computed on everything the estate itself executes; external receivers bind by one-tap plug or an agent racetrack afterwards. - `tok:decision:01m6a8mtbg94twxzcdgjp4yeac` Application names: EDGE Consortium and Judiverse Work Suite. - `tok:decision:01m6acbx9rv6x4yxh5gxkz1x30` EDGE Finance is called EDGE. - `tok:decision:01m6aeafjpfwmamsqm7apf2wfj` Application names: EDGE Public (was Public EDGE) and Judiverse Suite (was Judiverse Work Suite). - `tok:decision:01m6aennyqz91exa4a2sda0019` The finance application is called EDGE Bank. - `tok:decision:01m6bgf72jtm0n7xyba043bqh4` Screens come from the reference flows; the estate adds only its controls and journeys. - `tok:decision:01m6bgf72knqzc3da9n20c282a` EDGE Public Consortium's mark is consortium.svg from the approved brand register. ### spec/vocabulary.json `spec/vocabulary.json` · sha256 `a6a489907549bba5a87976a27a67f62b5793c2a94292c6b0e4bcfc7d93b78940` ```json { "schema": "tok.vocabulary.v1", "statement_class": "AGREEMENT", "what_this_is": "Plain-language meanings for every word the kernel uses: kinds of object, relations between them, statuses and states. Compiled into definition objects so every word is defined.", "terms": [ {"term": "Intention", "meaning": "What the governing organisation wants the estate to achieve, in its own words; every capability serves at least one"}, {"term": "Outcome", "meaning": "A measurable result the estate exists for: time, money, lives, planet, resources or value, saved or made more"}, {"term": "Principle", "meaning": "A rule of the estate, with how it is enforced (a gate, a measurement, conduct or a decision)"}, {"term": "Decision", "meaning": "A recorded choice by a competent authority, with the options considered and what it supersedes"}, {"term": "Application", "meaning": "One of the four products: EDGE Finance, EDGE Clinic, EDGE Public Consortium or Judiverse"}, {"term": "Foundation", "meaning": "A shared part every application uses, such as EDGE ID or the Evidence portal"}, {"term": "Capability", "meaning": "Something the estate lets a person, organisation or agent do, specified completely before it is built"}, {"term": "Journey", "meaning": "The ordered steps one kind of actor takes through screens and actions to reach a capability's outcome"}, {"term": "Screen", "meaning": "One view a person uses, specified completely: fields, controls, actions, transitions, conditions, states, copy and components"}, {"term": "State", "meaning": "A condition a screen or action can be in, such as loading, error or refused"}, {"term": "Control", "meaning": "Something that prevents an unsafe or unauthorised effect before it happens"}, {"term": "Operation", "meaning": "One action that changes or reads something, described by the behaviour model, reachable by a person and by an agent"}, {"term": "Endpoint", "meaning": "The address through which an operation is called: an HTTP endpoint for software or an MCP tool for agents"}, {"term": "Organisation", "meaning": "A legal body such as a company, charity, public body or university"}, {"term": "Person", "meaning": "A human, recorded by role only; contact details never enter the estate"}, {"term": "Source", "meaning": "Where a piece of knowledge came from, with its date and licence"}, {"term": "Code module", "meaning": "A piece of software in the estate, with its declared dependencies and tests"}, {"term": "Test", "meaning": "An automatic check with a known correct answer and a command anyone can run to reproduce it"}, {"term": "Evidence", "meaning": "A record produced with an effect, confirmed by the receiving system, bound by digest"}, {"term": "Measurement", "meaning": "A number with its unit, method, time and how complete it is"}, {"term": "Benchmark", "meaning": "An agreed, version-pinned test of performance with its scoring rules"}, {"term": "Competitor", "meaning": "Another product or organisation offering something similar"}, {"term": "Competitor feature", "meaning": "One observed screen, flow or capability of a competitor, with its source and date; the floor the estate must match and exceed"}, {"term": "External artifact", "meaning": "A file kept in storage rather than the repository, recorded by digest, location, size, licence and access"}, {"term": "Incident", "meaning": "Something that went wrong, with its cause, its correction and the test that stops it recurring"}, {"term": "Lifecycle event", "meaning": "A dated event in an agent's life: birth, mandate change, upgrade, suspension, incident, handover or death"}, {"term": "Embodiment", "meaning": "A device, robot or avatar an agent acts through, with its own safety controls and stop control"}, {"term": "Serves", "meaning": "This exists to achieve that intention or outcome, or this endpoint serves that operation"}, {"term": "Belongs to", "meaning": "This is part of that application, foundation or competitor"}, {"term": "Realises", "meaning": "This journey delivers that capability or application"}, {"term": "Realised by", "meaning": "This capability is delivered by that journey"}, {"term": "Has journey", "meaning": "This application has that whole-of-life journey"}, {"term": "Shows", "meaning": "This screen appears in that journey"}, {"term": "Shown by", "meaning": "This journey appears on that screen"}, {"term": "Passes through", "meaning": "This journey includes that screen or operation as a step"}, {"term": "Transitions to", "meaning": "This state can move to that state"}, {"term": "Guarded by", "meaning": "This state or operation is protected by that control"}, {"term": "Enforces", "meaning": "This control enforces that principle"}, {"term": "Governed by", "meaning": "This is subject to that principle"}, {"term": "Authorised by", "meaning": "That role's authority is what permits this"}, {"term": "Answerable", "meaning": "That role answers if this goes wrong; never an agent"}, {"term": "Exposed at", "meaning": "This operation can be called at that endpoint"}, {"term": "Implements", "meaning": "This code delivers that operation or capability"}, {"term": "Tested by", "meaning": "That test checks this"}, {"term": "Verified by", "meaning": "That evidence shows this is true"}, {"term": "Measured by", "meaning": "That outcome or measurement shows whether this is achieved"}, {"term": "Derived from", "meaning": "This came from that source"}, {"term": "Designed by", "meaning": "This uses that approved design or logo"}, {"term": "Floors", "meaning": "This must match and exceed that competitor or competitor feature"}, {"term": "Mirrors", "meaning": "This screen is measured against that competitor feature"}, {"term": "Depends on", "meaning": "This needs that to work"}, {"term": "Supersedes", "meaning": "This replaces that rule or decision, which is retired"}, {"term": "Narrows", "meaning": "This limits the scope of that rule or decision, which stays in force for the rest"}, {"term": "Agrees with", "meaning": "This claim or agreement supports that one"}, {"term": "Disagrees with", "meaning": "This claim or agreement contradicts that one; both are kept and the disagreement is shown"}, {"term": "Rehearsed by", "meaning": "That rehearsal test simulates this before it runs for real"}, {"term": "Recovers by", "meaning": "That operation undoes or compensates for this one"}, {"term": "Owned by", "meaning": "This agent is owned by that organisation or role"}, {"term": "Used by", "meaning": "This foundation is used by that application"}, {"term": "Compiled from", "meaning": "This procedure was produced from those sources, code and tests"}, {"term": "Retired by", "meaning": "That decision retired this; its identity is kept and never reused"}, {"term": "Granted by", "meaning": "That role or organisation granted this mandate"}, {"term": "Granted to", "meaning": "This mandate was granted to that agent"}, {"term": "Mandated by", "meaning": "This agent acts under that mandate"}, {"term": "Certified by", "meaning": "This agent's birth certificate is that lifecycle event"}, {"term": "Concerns", "meaning": "This lifecycle event is about that agent or embodiment"}, {"term": "Embodies", "meaning": "This embodiment carries that agent"}, {"term": "Defines", "meaning": "This definition gives the meaning of that"}, {"term": "Verified", "meaning": "Truth status: reproduced from evidence by a party that did not produce it, with the independence level stated"}, {"term": "Observed", "meaning": "Truth status: seen directly, not yet reproduced"}, {"term": "Hypothesis", "meaning": "Truth status: a testable proposition not yet tested"}, {"term": "Aspiration", "meaning": "Truth status: a hope about the future, allowed only on a claim explicitly about the future"}, {"term": "Prediction", "meaning": "Truth status: a forecast with its uncertainty"}, {"term": "Simulation", "meaning": "Truth status: shown in a rehearsal, not in the real world"}, {"term": "Refuted", "meaning": "Truth status: contradicted by sufficient evidence"}, {"term": "Open obligation", "meaning": "Something not yet known or done, named, owned by a role and dated; counted, never hidden"} ] } ``` ## 3. Specification ### spec/SPECIFICATION.md `spec/SPECIFICATION.md` · sha256 `25e3970bb9adba341f7e54637a72a7757b529b3ca0187d1723d682039beef094` #### Specification This file says how everything in the estate is specified, what each application would be, and what "complete" means. It is an AGREEMENT. It claims no readiness and verifies nothing. It contains no count: every number about the estate is produced live by `tools/gate.py` and `tools/build.py` from the objects. ##### 1. Reading order 1. `README.md`, the one instruction file. 2. `tok/KERNEL.md` for the ontology. 3. This file for the method and what each application would be. 4. `contract/estate-contract.json` for the contract itself. ##### 2. Every object is a complete executable contract Every capability, journey, screen, state, control, operation, endpoint and agent resolves to a complete contract before it is built: | Part | Content | |---|---| | Intention | the intention and outcome it serves, the beneficiary, and why it is worth its cost in time, money, compute, energy and attention | | Requirement | what it must do, in the principal's words where they exist, with source links | | Steps | ordered steps, inputs, exact executor, authority (issuer, principal, purpose, resource, expiry, revocation) | | Design | the approved design asset and the competitor features it floors | | Acceptance | accepted and rejected outcomes, the authoritative receiver, the proof checker, the test with its oracle | | Evidence | what is produced with the effect, and the receiving-system readback that establishes the outcome | | Resources | the resource envelope and the ledger units per verified outcome | | Recovery | what happens on failure, the recovery operation, whether the consequence is reversible | | Answerability | the role that answers if it goes wrong; never an agent | No guessed, missing or placeholder binding is admitted; a missing part is an `OPEN` obligation with an owner, counted live and visible. ##### 3. Statement classes in specifications Requirements, journeys, screens, controls and decisions are AGREEMENTS. Measurements, evidence, test results, competitor observations and sources are OBSERVATIONS. Vendor and competitor statements are CLAIMS until observed. A requirement's realisation carries a truth status, UNKNOWN until observed; it is never labelled an aspiration. ##### 4. The required journey set Every capability has all of these, or `OPEN` obligations with owners (gate G11): | Journey | Actor | What it must show | |---|---|---| | HUMAN | the person the capability serves | essential result first; depth on demand; accessible (WCAG 2.2); every state and control; completes for a person with no sight | | AGENT | a Jude agent under its own EDGE ID | the same outcome through MCP with the same authority, preview, audit and undo; portal, mailbox, racetrack, rehearsal before commit | | REGULATOR | an assessor through the Regulator portal | the evidence, authority, proof and recovery for the capability, inspectable without the builder's help | | ENGINEER | the builder or maintainer | the contract, tests, dependencies, monitors and the change-impact set from TOK | | OPERATOR | the answerer and operations role | monitoring, incidents, recovery, resource ledger, handover | Each of the three applications (EDGE Finance, EDGE Clinic, EDGE Public Consortium) plus Judiverse, which comes with all three additionally has a WHOLE_OF_LIFE journey: first contact, onboarding, use, change, incident, recovery, offboarding and retention, for a person and for an organisation, with agents native throughout. People and agents share each workflow: how they hand off, approve, see each other's actions, pause, take over and undo is stated, and conflicts are resolved by lease and fence. ##### 5. The outcome ledger Six dimensions, two directions: time, money, lives, planet, resources, value; SAVED (preservation) or MORE (abundance). Every capability serves at least one (G10). Each outcome passes through OPPORTUNITY_IDENTIFIED, WORK_ATTEMPTED, RESULT_ACHIEVED and OUTCOME_VERIFIED and is counted only when verified, with its measurement. Resource to outcome (tokens, GBP, kWh, gCO2e, litres, seconds per verified outcome) is published for every agent and capability. ##### 6. What each application would be Each is a requirement set whose realisation is UNKNOWN until observed. Named competitors are floors: matched, then exceeded, made easier and faster. **EDGE Finance, JudeFI, Jude Financial Intelligence.** Banking and financial intelligence where people, organisations and agents are all real customers: accounts, payments, cards through a processor, lending and borrowing (agents as lenders and borrowers), KYC, KYB, KYA and know-your-domain in one journey, sanctions and PEP screening with a guaranteed miss rate, transaction monitoring on measured rates, a fraud-sharing network, credit reference on consented Open Banking data, agent payments and mandates, and a sponsor bank route before own permissions. Floors: Starling, Monzo, Revolut and the leading crypto platforms. **EDGE Clinic, Jude Clinical Intelligence.** Social prescribing with early detection: the front door before the GP, where anyone can come about anything and find out what the problem is. It brings together detection (including retinal screening and other early signals), validated questionnaires, triage including online A&E, an open clinical-grade drug-interaction source, prescribing safety on real terminology and referral to the right service, aimed at the costliest burden on the health service: the time and capacity it takes to be seen. Every clinical capability carries its safety and device obligations as OPEN until cleared. **EDGE Public Consortium.** The public and institutional home of the estate, founded on the declaration that education is where all capabilities come from and that all research and development is worth investing in. It houses cyber governance (attack made unprofitable), procurement intelligence, vetting and verification, logistics for defence and government, regulator and evidence portals, research consortia with universities, the fraud-sharing and data-sharing networks, and public knowledge from the Research Commons. Capabilities expand here without breaking the form. **Judiverse.** Where people and organisations work with their agents: Jude as assistant, researcher and operator; documents, mail, calendar, meetings, appointment and interview scheduling, tasks, files, forms and workflows generated from each organisation's versioned workflows; the intenJude lab where intention becomes computational instruction, agents build agents, models are trained, scheduled and compiled into certified procedures; and JudeBox for artists' own music with direct royalties. Floors: every Microsoft product and service, the leading AI assistants, GitHub and GitLab. **Shared foundations.** EDGE ID; Research Commons; Tree of Knowledge; Evidence portal; Regulator portal; Versioned organisation workflows; Agent APIs and MCPs; Governed release and adoption. Sovereign intelligence runs through all of them: data, embeddings, prompts, memory, logs and agents stay under each organisation's control and never go to model providers. **Named scheduling and intelligence capabilities.** Compute scheduling across edge, GPU and cloud by real hardware, carbon and sovereignty; model scheduling and switching; appointment and interview scheduling; procurement intelligence and verification; the evidence portal; the regulator portal; KYC, KYB and KYA. Each is placed in one of the three applications (EDGE Finance, EDGE Clinic, EDGE Public Consortium) plus Judiverse, which comes with all three, or in a foundation and specified under §2. ##### 7. Agents as native users Every workflow has a human path and an agent path to the same outcome with equal capability (G12). Every agent has an EDGE ID, a mailbox, a portal, a racetrack and a purpose-limited authority from a legitimate issuer, with an accountable human or organisational owner. Consequential actions are rehearsed with the other agents involved before they commit; the rehearsal (SIMULATION) and the committed result are both recorded. Agents cannot crash or collide: each runs on its own racetrack with lease and fence on shared resources. Agents never verify their own authority, invent success or suppress contradictory evidence. Agents are always on, inside the estate, so data never leaves it. ##### 8. Screens: clear for people, deep for agents Every screen declares its essential-first content, its help depth as markdown, and its accessibility (G14). The essential result is shown; everything else is behind a help control, the same markdown served to people and to agents on demand. Completeness lives in TOK and the help depth, never in clutter on the screen. Every screen links an approved design asset from UX Pilot, Figma or Untitled UI, or an Owner original; nothing is hand-drawn and no logo is recreated. ##### 9. Research inputs and competitor floors Research defines scope. Competitor screens and flows (Mobbin), design references (Refero), approved designs (Figma, UX Pilot, Untitled UI), diagrams (Mermaid, generated from TOK relations), benchmarks and baselines, regulations and standards are admitted as sources and competitor observations with dates and links. Every application, capability and screen links the competitor features it floors, or records none_found with its search evidence (G15). Third-party pages and screens are held in storage under their terms; the facts extracted from them are OBSERVATIONS with a date. Whether a screen mirrors a competitor's exact layout, or only its behaviour and features, is an open decision. ##### 10. Runtime: compiled governed intelligence AI is the factory, not the product. In the lab, neural models read sources, propose procedures, write tests and verify against known answers. What ships is a `procedure`: deterministic, compute-light, with a certified scope, an error bound, a calibration set and a byte-identical rerun (G17). Answers are rendered from proofs with their sources and bounds; outside a certified scope the answer is UNKNOWN or a clarifying question. Uncovered requests are logged and compiled offline; every release raises the AI-free rate. Small certified models may perceive images and speech; no large general model is required at runtime. Published measures: AI-free rate; cost and latency per request against a large model; certified versus measured error rate; reproducibility. ##### 11. What complete means Complete means: every stated intention is represented and placed, including capabilities implied and never spelled out; every capability has its journey set, screens, states, controls, operations, endpoints, tests, evidence, answerer and competitor floors or `OPEN` obligations with owners; every file is accounted for by an object; every number is produced live with its method; and every gate passes. "Nothing missing" is enforced by G03, G10, G11 and G15; "nothing broken" by G05, G07 and G17. The count of objects, open obligations and gate results is never written; `tools/gate.py` produces it at the moment it is asked. ##### 12. Research Research is pinned input held in storage, referenced by digest, from which the factory produces new objects. It never enters the repository and is never authority. Kinds: statements of intention from authorised principals (kept in their words, secrets replaced by keyed-hash markers, human-written text separated from agent-written text); regulations and standards; competitor documentation and screens; baselines and benchmarks; source code; diagrams; research documents; datasets and models with their licences. Every build starts from the current sources and the current agreement; nothing is inherited. ##### 13. First proofs Two domains, each built by the factory and run on the compiled runtime, benchmarked head-to-head against the best available large models on expert-graded questions, with four published numbers each: certified versus measured error rate, cost per answer, latency, reproducibility. 1. Sanctions and PEP name screening (EDGE Bank): a guaranteed miss rate on official lists with transliteration. 2. The open drug-interaction source (EDGE Clinic): a guaranteed false-assertion rate on a pharmacist-graded held-out set, with its abstention rate. One verified domain with a published, re-checkable error rate establishes the method; each further domain added without a kernel change (G18) is evidence that the same rules apply everywhere. ##### 14. What this specification is not It is not evidence that anything exists, works or is ready. Every status in the estate is set from observation and reproduction, never from this document. ### spec/UNIT-FORMAT.md `spec/UNIT-FORMAT.md` · sha256 `0f69ddc7d72cfb0dfbad4a23890e5d1a9ecff669e3f2571f570c99bd02a078f6` #### Specification units: how to specify a capability completely A **unit** is one JSON file that specifies one capability completely: every screen's fields, controls, actions, transitions, conditions, components, states and copy; every operation's full behaviour, HTTP endpoint and agent tool; and every journey for every kind of actor. `tools/compile.py` turns units into TOK objects and refuses any unit that is incomplete, listing exactly what is missing. The worked example is [`units/finance/send-money.json`](units/finance/send-money.json). Copy its shape. ##### Where units live `spec/units//.json`, where application is one of `finance`, `clinic`, `public-consortium`, `judiverse`. One capability per file. The file name is the capability's short name in lower case with hyphens. ##### How to check and compile ```bash python3 tools/compile.py --check ``` ```bash python3 tools/compile.py && python3 tools/build.py && python3 tools/gate.py ``` `--check` writes nothing and lists every problem. The second command writes the objects, regenerates `generated/` (including `JOURNEYS.md` with Mermaid diagrams and `ENDPOINTS.md`) and runs every gate. ##### The parts of a unit | Key | What it holds | |---|---| | `unit` | `/`, the same as the file path | | `application` | `finance`, `clinic`, `public-consortium` or `judiverse` | | `capability` | `label`; `requirement` (what it must do, in plain language); `serves_intentions` (ids from `contract/estate-contract.json`); `outcomes` (each with `dimension` time/money/lives/planet/resources/value, `direction` SAVED/MORE, and `measure`); `answerable_role` (a human or organisational role, never an agent); `competitor_floor`; `research` (the research files it rests on; each must be an existing file, or the compiler refuses the unit) | | `actors` | Every actor by id: `kind` (person, agent, organisation, system), `role`, and for agents their `mandate` | | `data` | Every entity and its fields: `type`, `required`, `validation`, `sensitivity` (none, personal, financial, health, special-category). Screens may only bind to fields declared here | | `operations` | Every action by id: `label`, `authorised_role`, `endpoint` (method, path, auth, request, response, errors), `mcp_tool` (name, input, output), and `behaviour` with every part of the behaviour model (below) | | `screens` | Every screen by id, with every blueprint part (below) | | `journeys` | At least one each of `HUMAN`, `AGENT`, `REGULATOR`, `ENGINEER`, `OPERATOR`: `type`, `actor` (an actor id), `summary`, `steps` as `screen:` or `operation:` | | `definitions` | Every domain term the unit uses that is not already defined: `term`, `meaning` in plain language | ###### Behaviour model (every operation) `actor`, `mandate`, `paths` (`human` and `agent`), `permission`, `inputs`, `guards` (each with `name`, `source`, `refusal`), `rehearsal`, `approval`, `operation`, `receiving_confirmation`, `success`, `failure` (each with `code`, `person` message, `agent` structured message), `unknown`, `recovery` (name another operation in `recovery.operation` when one undoes or compensates), `evidence`, `batch`, `output_verification`, `monitoring`. A part that genuinely does not apply is written `{"not_applicable": ""}`. Nothing is left empty. ###### Screen blueprint (every screen) `purpose`, `actors`, `journey_steps`, `entry`, `exits`, `essential_first`, `fields`, `controls`, `actions`, `transitions`, `conditions`, `components`, `states`, `copy`, `accessibility`, `help_markdown`, `agent_equivalent`, `competitor_floor`, `design_source`. - **fields**: `id`, `label`, `type`, `required`, `validation`, `binds` (a declared `entity.field`), `help`. - **controls**: `id`, `component` (from `design/components.json`), `label`, `action` (an action id on this screen), `enabled_when` (a condition id or `always`). - **actions**: `id`, `operation` (an operation id in this unit, or `navigate` for movement only), `on_success`, `on_failure` (a screen id, a state name or `journey:end`). - **transitions**: `from`, `on`, `to`, `condition` (a condition id or `always`). - **conditions**: `id`, `expression` in plain words precise enough to implement. - **components**: `id`, `component` (from the approved list), `binds`. - **states**: copy for each of `empty`, `loading`, `success`, `partial`, `error`, `refused`, `unauthorised`, `offline`, `unknown`. - **copy**: at least `title`; every other piece of visible text by key. - **agent_equivalent**: `tool` (one of this unit's MCP tool names) and `output`. - **competitor_floor**: a list of `{product, source, what, observed_at}` from research, or `{"none_found": {"search_evidence": "..."}}` saying where you looked. - **design_source**: the approved design, or `{"OPEN": "..."}` until the Figma frame exists. ##### Rules for authors 1. **Ground every unit in research.** Competitor floors cite real flows (Refero links in `research/*/REFERO-FLOWS*.md`) or research files. Regulatory duties cite `research/*/REGULATORS*.md` or `COMPETITORS-REGULATORS-REUSE.md`. Never invent a competitor feature, a rule, a number or a link. 2. **Agents are customers too.** Every operation has an agent path under the agent's own identity and mandate, with the same guards, evidence and recovery as a person. 3. **Plain language.** Copy is short, kind and exact. A brand-new engineer and a first-time customer both understand it. Every special term is in `definitions`. 4. **No secrets, no personal data, no machine paths.** The compiler refuses them. 5. **Unknowns are honest.** If a design, a test or a regulator decision does not exist yet, say `OPEN` with what is needed; never pretend it exists. 6. **One capability per unit.** Shared screens are repeated by reference to the same competitor floor, not shared across files. ##### Format 2: every click linked, no dead controls, parity proven A unit that declares `"format": 2` at the top level is held to these rules by `tools/compile.py --check`. Every new unit is format 2; existing units are being converted. Until a unit is converted, its gaps against these rules are listed in `generated/LINK-GAPS.md`. 1. **No dead controls.** Every control fires an action on its screen, and every action is fired by at least one control. A screen has at least one control. 2. **Every outcome goes somewhere.** Every action's `on_success` and `on_failure` is a screen in this unit, one of the nine state names, `journey:end`, or `unit:/#` naming a screen in another unit that exists. 3. **Transitions are real.** Every transition's `from` is its own screen, its `on` is `.` for an action on that screen (events: `success`, `failure`, `refused`, `unknown`, `cancel`, `timeout`), and its `to` follows rule 2. 4. **Conditions are used and defined.** Every `enabled_when` and transition `condition` is `always` or a condition defined on that screen, and every defined condition is used at least once. 5. **Every screen is reachable.** A screen is the first screen of some journey, or the target of an action or transition on another screen in this unit. 6. **Journeys are click-by-click.** In every journey whose actor is a person (HUMAN, REGULATOR, OPERATOR and any other journey run by a person), the first step is a screen and each step follows from the one before: screen to screen through an action or transition on the first; screen to operation through an action on that screen whose `operation` is it; operation to screen through the `on_success` or `on_failure` of the action that called it. AGENT and ENGINEER journeys may be operations only, because agents and engineers call the MCP tool or HTTP endpoint directly. 7. **No unused operations.** Every operation is called by an action on some screen or appears in some journey. 8. **Parity is proven, item by item.** Every item in a screen's `competitor_floor` has `matched_by`: a non-empty list of ids of the fields, controls, actions or components on that screen that match what the competitor offers. Add `exceeds` (plain words, what this screen does better, and how that would be measured) wherever it does. `none_found` still needs its search evidence. `tools/build.py` turns every unit into diagrams and a click-by-click breakdown in `generated/diagrams//.md`: a map of every screen and how they connect, a state diagram per screen, a behaviour diagram per operation, and for every journey a numbered table of each step (who, where, which control, which action, which operation and endpoint, what happens on success, what happens on failure) with a sequence diagram. Authors do not draw diagrams by hand; the diagrams come from the unit, so the screen, the endpoint, the agent tool and the diagram can never disagree. ##### Shared foundation units A capability that belongs to no single application, such as signing in once across every application, lives in `spec/units/foundation/.json` with `"application": "foundation"` and `"foundation": ""`, where the id is one of the shared foundations in `contract/estate-contract.json` (`edge-id`, `research-commons`, `tok`, `evidence-portal`, `regulator-portal`, `organisation-workflows`, `agent-apis-mcps`, `governed-release`). ### spec/BRIEF-FORMAT.md `spec/BRIEF-FORMAT.md` · sha256 `d74bc4fa6c0410af598d29ff15dfbd8bd89c24e242586ba6ba4e15193853f2d7` #### Domain briefs: write only what is specific to the domain A **brief** is a short JSON file, `spec/briefs//.json`, holding only the domain knowledge of one capability. `tools/expand.py` turns it into the complete format 2 unit in `spec/units/` (every state, copy, endpoint, agent tool, behaviour part, recovery route, evidence screen, journey and link), and `tools/build.py` draws every diagram and click-by-click breakdown. Never write those parts by hand, and never edit a generated unit; edit the brief. ##### Shape ```json { "unit": "finance/joint-accounts", "application": "finance", "label": "Joint accounts", "requirement": "What it must do, in plain words, including the decisions taken for the best outcome", "serves_intentions": ["agents-as-real-customers"], "outcomes": [{"dimension": "time", "direction": "SAVED", "measure": "minutes to add a second holder"}], "answerable_role": "Accounts operations lead", "research": ["research/finance/MOBBIN-FLOWS.md"], "actors": { "holder": {"kind": "person", "role": "Account holder"}, "holder-agent": {"kind": "agent", "role": "Account holder's agent", "mandate": "what it may do, limits, expiry"} }, "data": {"invite": {"email": {"label": "Their email", "type": "email", "required": true, "validation": "a valid email address", "sensitivity": "personal"}}}, "operations": { "invite-holder": { "label": "Invite a second holder", "role": "Account holder", "method": "POST", "path": "/v1/joint-accounts/invites", "inputs": ["invite.email"], "outputs": ["invite id", "status"], "guards": [{"name": "the account allows a second holder", "source": "account type", "refusal": "NOT_ELIGIBLE"}], "effect": "What changes in the world", "confirms": "What the receiving system confirms", "success": "What the person sees when it worked", "failures": [{"code": "INVITE_FAILED", "person": "Plain words, what happened and what to do"}], "recovery": "cancel-invite" } }, "screens": { "invite": { "title": "Add someone to your account", "purpose": "Why this screen exists", "fields": ["invite.email"], "shows": [{"component": "Summary.Review", "binds": "invite"}], "controls": [ {"label": "Send invite", "op": "invite-holder", "go": "invite-sent"}, {"label": "Not now", "go": "end"} ], "floor": [{"product": "Monzo", "source": "https://mobbin.com/flows/...", "what": "What they offer here", "matched_by": ["Send invite", "invite.email"], "exceeds": "What we do better and how it is measured"}] } }, "journeys": { "human": {"actor": "holder", "summary": "A person adds a second holder", "screens": ["invite", "invite-sent"]}, "agent": {"actor": "holder-agent", "summary": "An agent invites a holder within its mandate", "operations": ["invite-holder"]} }, "definitions": [{"term": "Joint account", "meaning": "An account held by two or more people, each with full access"}] } ``` ##### Rules - **Controls**: `label`; `op` (an operation in this brief) and/or `go` (a screen id in this brief, `end`, or `unit:/#` for another unit's screen); optional `fail` (screen id or state name; default `error`), `when` (a plain-words condition that enables it) and `component` (from `design/components.json`; otherwise chosen automatically). A control with `op` and no `go` stays on the screen. Give every screen at least the controls the best competitor offers, then ours that do better. - **Fields** name declared data as `entity.field`. Data field `type` picks the input: text, textarea, money, number, date, email, phone, select, choice, multiselect, boolean, toggle, file, image, chip, qr, search, password, code, address. - **shows**: read-only components from `design/components.json` (Summary.Review, Receipt, Table, List, Timeline, Chart, Card.Metric, Evidence.Panel, Confidence.Meter, Rehearsal.Result, Agent.Card, Mandate.Editor, Map, Calendar, Chat.Thread, Chat.Composer, Alert, Badge ...). - **floor**: real competitor features only, each with a real `source` (a Refero or Mobbin link from `research/`, a `research/parity/` clause source, or a research file), and `matched_by`: the control labels, `entity.field` names or `shows` binds on this screen that match it. If none exist, write `{"none_found": "where you searched"}`. - **journeys**: `human` (a person, as screen ids in click order; consecutive screens must be joined by a control's `go`; a control with `op` and `go` puts the operation between them) and `agent` (operations in order) are required. Add more person journeys (for staff, carers, officers) and agent journeys as the domain needs, each with `"type"`: HUMAN, AGENT, REGULATOR, OPERATOR or ENGINEER. The regulator, operator and engineer journeys, the evidence screen and the inspect operation are added automatically. - **Recovery**: name the operation that undoes or compensates (`recovery`), or say in words how it is put right. - **Every operation** is reachable by a person through a control and by an agent through its tool. Agents hold accounts in their own name with a named accountable owner (decision `tok:decision:01m696xzzpcbnsexpcdtz7ak9n`). - **Intentions** are ids from `contract/estate-contract.json`. Outcomes use dimension time, money, lives, planet, resources or value, and direction SAVED or MORE. - Unknown legal or clinical facts: write them in the requirement as "OPEN: what is needed; decided by ". Otherwise decide for the best outcome for the user and the work. - Plain English. No secrets, personal data or machine paths. ##### Build ```bash python3 tools/expand.py spec/briefs//.json && python3 tools/compile.py --check ``` ### ledger/SCOPE-CHANGES.md `ledger/SCOPE-CHANGES.md` · sha256 `1d58e4ad331483d58fd9e851891a7d582fda9a6a0228132747ceb5090c59998b` #### Changes since the scope freeze Generated by `tools/freeze.py check` against `spec/SCOPE-FREEZE.json` (frozen 2026-10-31, 5734 briefs). Changes: **1796**; without a change request: **0**. Research sources changed since the freeze: 4739. | Capability | Change | Admitted by | |---|---|---| | `clinic/accept-a-proxy-mandate` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/acquire-and-send-images` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/act-on-a-drug-safety-alert` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/act-on-a-high-risk-answer` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/add-an-appointment-to-my-calendar` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/adjust-a-premium-at-renewal` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/administer-a-medicine` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/allocate-a-breach-between-trusts` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/annotate-variants` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/answer-a-patient-call` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/apply-a-national-data-opt-out` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/apply-a-treatment-rule` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/apply-for-hra-and-rec-approval` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/approve-a-data-release` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/approve-a-leave-or-swap-request` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/approve-a-questionnaire-or-test-threshold` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/approve-a-staff-request` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/arrange-a-cross-trust-mdt` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/ask-for-advice-and-guidance` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/ask-for-help-now` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/assess-a-safeguarding-concern` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/attend-an-appointment-with-the-person` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/audit-access-to-records` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/audit-antimicrobial-use` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/audit-consent-records` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/audit-data-use-against-mandates` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/audit-mdt-attendance` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/audit-reporting-quality` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/audit-response-times` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/audit-waiting-time-data` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/authorise-a-result` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/book-a-pathway-test` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/book-an-appointment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/book-an-interpreter` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/book-for-the-person-i-care-for` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/book-from-the-referral-queue` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/book-transport` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/build-and-publish-the-rota` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/build-the-clinic-template` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/build-the-mdt-list` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/build-the-variant-to-trial-graph` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/cancel-a-clinic-session` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/change-my-pharmacy` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/chase-a-pathway-clock` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/chase-missing-results` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/check-a-chemotherapy-prescription` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/check-access-controls` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/check-fragment-size-before-sequencing` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/check-identity-and-consent-before-scanning` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/check-in-an-arriving-patient` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/check-in-on-arrival` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/check-laboratory-submissions-for-a-variant` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/check-record-accuracy` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/check-research-integrity` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/check-staff-registration-and-training` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/claim-my-edge-id-and-record` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/classify-a-variant` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/classify-software-as-a-medical-device` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/clinically-check-a-prescription` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/close-a-referral` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/close-a-study` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/collect-case-report-data` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/compare-two-pipeline-runs` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/complete-a-mental-health-check-in` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/complete-a-pre-clinic-questionnaire` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/complete-a-risk-formulation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/complete-mandatory-training` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/complete-the-data-security-and-protection-toolkit` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/confirm-attendance-and-prepare` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/connect-a-lims` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/connect-a-pacs-or-imaging-system` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/counsel-a-patient-on-a-new-medicine` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/decide-a-re-run` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/deliver-cancer-nurse-specialist-support` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/deploy-a-release` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/design-a-smart-questionnaire` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/design-a-trial` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/detect-claim-anomalies` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/discharge-from-the-pathway` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/discharge-from-the-stream` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/dispense-a-medicine` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/escalate-to-emergency-services` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/explain-a-result-in-plain-language` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/explain-a-result-to-the-patient` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/find-a-recruiting-trial-near-me` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/flag-a-critical-finding` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/flag-never-block` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/follow-up-after-discharge-from-crisis-care` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/follow-up-an-action-plan` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/forecast-compliance` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/forecast-days-to-treatment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/front-door-check-in` | changed | 2026-11-01-research-is-not-scope | | `clinic/fulfil-a-subject-access-request` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/give-a-vaccination` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/give-consent-on-behalf` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/give-or-withdraw-consent` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/grade-a-retinal-image` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/hand-off-to-codex-or-jude` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/handle-a-complaint` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/handle-a-controlled-drug` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/handle-a-did-not-attend` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/handle-a-disclosure-of-abuse` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/handle-a-freedom-of-information-request` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/handle-a-policyholder-complaint` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/handle-a-stock-shortage` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/handle-the-record-after-death` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/hold-a-remote-session` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/hold-a-slot-for-a-tertiary-handover` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/hold-the-duty-rota` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/import-my-practice-register` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/inspect-the-hazard-log-and-safety-case` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/investigate-a-clinical-safety-incident` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/issue-a-clinical-safety-case-report` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/issue-a-field-safety-notice` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/issue-a-genomic-report` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/issue-a-policy` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/maintain-laboratory-accreditation-evidence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/maintain-the-hazard-log` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/maintain-the-suicide-prevention-plan` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/make-a-follow-up-call` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/make-an-urgent-same-day-referral` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/manage-an-incident` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/manage-freezer-alarms` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/manage-patient-group-directions` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/manage-patient-transport-and-interpreter-contracts` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/manage-retention-and-destruction` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/manage-room-allocation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/manage-the-information-asset-register` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/manage-the-outbound-allow-list` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/manage-the-reporting-backlog` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/manage-the-room-and-equipment-register` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/manage-the-waiting-list` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/measure-a-nodule-and-its-growth` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/measure-compute-energy-and-carbon` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/measure-outcomes-over-time` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/medication-safety-check` | changed | 2026-11-01-research-is-not-scope | | `clinic/mental-health-care-and-crisis` | changed | 2026-11-01-research-is-not-scope | | `clinic/message-my-care-team` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/monitor-alerts-and-escalate` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/monitor-data-feeds` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/monitor-post-deployment-safety` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/nominate-a-carer-or-proxy` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/obtain-confirmation-of-capacity-and-capability` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/offer-a-cancellation-slot` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/offer-self-help-resources` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/onboard-a-staff-member` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/onboard-a-wearable-device-type` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/online-urgent-triage` | changed | 2026-11-01-research-is-not-scope | | `clinic/order-a-diagnostic-test` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/order-a-direct-access-test` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/order-a-genomic-test` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/order-a-repeat-medicine` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/order-repeat-medicines-on-behalf` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/pair-my-wearable-device` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/person-held-health-record` | changed | 2026-11-01-reference-records-pass5 | | `clinic/phone-a-critical-result` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/plan-weekly-capacity` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/post-a-receipt-to-the-agent-channel` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/predict-fields-in-a-form` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/prepare-a-clinic-letter-for-a-clinician` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/prepare-a-patient-for-imaging` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/prepare-for-a-cqc-inspection` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/prescribe` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/prescribe-under-the-pathway-protocol` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/price-a-policy-from-clinic-data` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/problem-annual-offshore-statistics-and-regulatory-activity` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/problem-medicare-will-pay-50k-for-your-heart-attack-it-sho` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/problem-students-tackle-public-policy-challenges-with-supp` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/problem-working-for-the-national-cyber-force-ncf-case-stud` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/process-a-claim` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/pseudonymise-a-dataset` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/publish-results` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/raise-a-complaint-or-feedback` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/raise-a-safeguarding-concern-from-a-response` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/randomise-a-participant` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/read-assistive-output-on-an-image` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/rebook-or-cancel-an-appointment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/receive-a-crisis-alert` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/receive-a-crisis-alert-for-the-person` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/receive-a-critical-result-for-my-patient` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/receive-a-did-not-attend-notice` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/receive-a-pre-authorisation-request` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/receive-a-statutory-notification` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/receive-and-book-in-a-sample` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/receive-the-outcome-letter` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/reclassify-a-variant` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/reconcile-medicines` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/reconcile-medicines-at-admission` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/record-a-clinic-consultation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/record-a-dose-incident` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/record-a-patient-deferred` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/record-a-patient-s-own-view` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/record-a-point-of-care-urine-result` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/record-a-safeguarding-concern` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/record-a-self-check-in-exception` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/record-a-wound-or-skin-check` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/record-an-adverse-drug-reaction` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/record-attendance-outcome` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/record-capacity-and-best-interests` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/record-decision-to-treat` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/record-sample-consent-version` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/record-the-faster-diagnosis-outcome` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/record-the-lawful-basis` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/record-the-mdt-recommendation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/record-trial-eligibility-at-the-mdt` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/refer-on-a-suspected-cancer-pathway` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/refer-to-a-crisis-team` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/refer-to-tertiary-care` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/refer-to-the-mental-health-stream` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/refer-to-the-multidisciplinary-team` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/refresh-public-data-caches` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/register-a-child-as-a-patient` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/register-a-device-fault` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/register-a-walk-in-patient` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/register-an-imaging-device` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/register-my-practice-with-the-clinic` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/register-the-trial` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/reject-an-unlabelled-sample` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/report-a-change-in-the-person-s-condition` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/report-a-data-breach` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/report-a-dispensing-error` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/report-a-safety-incident` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/report-a-side-effect` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/report-a-study` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/report-an-adverse-event` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/report-an-incident` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/report-histology` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/report-monthly-performance` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/report-to-the-regulator` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/report-usage-and-efficiency` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/request-a-copy-of-my-record` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/request-a-rota-swap-or-leave` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/request-a-second-opinion` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/request-a-second-read` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/request-a-second-read-or-discrepancy-review` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/request-a-young-person-s-own-access` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/request-clinic-data-for-a-quote` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/request-evidence-against-fundamental-standards` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/retinal-screening` | changed | 2026-11-01-research-is-not-scope | | `clinic/review-a-breach-notification` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/review-a-death-by-suicide` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/review-a-real-result` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/review-a-release-for-clinical-safety` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/review-a-repeat-request` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/review-a-trial-for-compliance` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/review-a-wearable-alert` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/review-assistive-output-labelling` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/review-payslips-and-expenses-approvals` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/review-the-caseload` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/review-the-dashboard` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/review-the-waiting-list-position` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/review-the-worklist` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/revoke-access-when-a-mandate-ends` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/revoke-my-own-proxy-role` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/reward-early-detection-participation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/roll-back-a-release` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/rotate-keys-and-certificates` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/run-a-business-continuity-drill` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/run-a-catalogue-workload` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/run-a-clinical-safety-training-check` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/run-a-data-protection-impact-assessment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/run-a-pipeline` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/run-a-pre-assessment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/run-a-safeguarding-review` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/run-a-synthetic-pathway-drill` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/run-a-urine-test-likelihood-ledger` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/run-an-interim-analysis` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/run-an-online-early-symptom-test` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/run-quality-control` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/run-scanner-quality-assurance` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/run-sequencing-qc` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/run-the-day-s-capacity-check` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/run-the-mdt-meeting` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/run-the-training-controller` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/run-vulnerability-scans` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/schedule-the-scan` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/score-a-model-on-a-held-split` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/score-and-interpret-a-response` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/screen-a-patient-for-eligibility` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/search-recruiting-trials` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/see-my-day` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/send-a-check-in-to-a-caseload` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/send-a-results-ready-notice` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/send-appointment-letters-and-reminders` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/serve-a-vision-model` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/set-a-surveillance-plan` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/set-my-communication-preferences` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/set-my-data-sharing-choices` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/set-role-permissions-and-expiry` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/settle-with-the-clinic` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/share-a-medicines-list-with-the-clinic` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/share-a-wearable-signal` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/share-clinic-data-with-my-insurer` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/share-images-with-another-trust` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/share-risk-information-with-partners` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/sign-a-clinic-letter` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/sign-the-clinical-risk-management-plan` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/simulate-trial-monitoring` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/social-prescribing-referral` | changed | 2026-11-01-research-is-not-scope | | `clinic/stop-on-revoked-authority` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/store-and-retrieve-genomic-data` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/submit-a-cancer-registration` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/submit-cancer-waiting-times-data` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/submit-cosd-and-other-datasets` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/submit-mdt-data` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/submit-the-mental-health-dataset` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/supervise-and-sign-off-a-trainee` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/support-deploying-organisations` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/symptom-check-and-triage` | changed | 2026-11-01-research-is-not-scope | | `clinic/take-a-blood-or-urine-sample` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/take-a-mandate-ticket` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/take-an-online-early-symptom-test` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/take-informed-consent-for-research` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/take-observations` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/take-part-in-external-quality-assessment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/tell-the-patient-a-result` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/test-a-hypothesis-on-an-open-cohort` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/test-a-pre-registered-hypothesis` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/trace-ingress-and-egress` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/track-a-referral` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/track-actions-from-the-mdt` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/track-histology-turnaround` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/track-my-cancer-pathway` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/track-sample-custody` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/track-the-62-day-clock-across-the-mdt` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/train-staff-in-information-governance` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/train-staff-in-suicide-awareness` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/triage-a-mental-health-referral` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/triage-a-patient-message` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/triage-a-referral` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/triage-inbox-items-for-a-clinician` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/tune-alert-thresholds` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/update-patient-demographics` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/validate-a-new-assay` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/validate-the-test-against-outcomes` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/verify-a-device-classification` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/verify-an-applicant-s-identity` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/verify-an-evidence-statement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/verify-backups-and-restore` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/vet-an-imaging-request` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/view-my-medicines-list` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/view-my-results` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/view-results-in-scope` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `clinic/wearable-signals` | changed | 2026-11-01-research-is-not-scope | | `clinic/wellbeing-questionnaires` | changed | 2026-11-01-research-is-not-scope | | `clinic/write-a-care-plan` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/accept-a-review-assignment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/accept-supervision-of-a-learner` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/acknowledge-the-principles-version` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/add-jude-to-a-channel-under-mandate` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/add-members-and-guests-with-expiry` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/administer-a-database` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/admit-or-refuse-a-guest` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/agree-a-data-sharing-agreement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/allocate-indirect-costs` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/answer-a-reference-request` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/answer-reviewer-questions-on-a-contribution` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/answer-with-sources-in-a-channel` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/apply-a-non-disclosure-agreement-to-a-room` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/apply-an-open-licence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/apply-for-ethics-approval` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/apply-retention-and-deletion-to-records` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/apply-to-become-a-verified-reviewer` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/approve-a-compute-request` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/approve-a-connector-or-integration` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/approve-a-contribution-before-submission` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/approve-a-director-passkey-action` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/approve-a-federated-query-request` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/approve-a-guest-account` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/approve-a-reviewer-payment-batch` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/approve-a-team-member-data-access-request` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/approve-a-visiting-researcher-affiliation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/approve-an-agent-consequential-action` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/approve-an-open-source-release` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/approve-and-mark-a-contribution-verified` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/approve-compute-spend-against-a-grant` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/approve-expenditure-against-a-grant` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/archive-a-channel` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/archive-a-module-version` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/ask-a-clarification-question` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/ask-jude-for-a-sourced-answer` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/assess-ownership-of-an-output` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/assign-a-persistent-identifier` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/attribute-carbon-to-a-project` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/attribute-cost-to-a-grant-and-project` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/audit-reviewer-quality-by-re-review` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/audit-the-signing-log` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/award-and-publish-the-contract-award-notice` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/bind-a-device` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/book-a-room-or-resource` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/build-a-project-team-across-universities` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/build-an-interactive-prototype` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/build-the-research-profile-and-skills-record` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/bulk-enrol-students-and-staff` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/catalogue-a-library-item` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/charge-compute-and-storage-to-grant-codes` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/chase-an-overdue-review` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/check-exclusion-grounds` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/choose-a-specialism` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/claim-payment-for-a-completed-review` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/classify-data-sensitivity` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/close-a-grant` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/close-a-project-and-archive-its-records` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/co-edit-a-markdown-in-real-time` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/complete-a-right-to-work-check` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/complete-research-integrity-training` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/compose-an-image-in-the-imaging-studio` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/compose-audio-in-judio` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/conduct-a-probation-review` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/confirm-a-delivery` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/confirm-a-person-s-enrolment-status` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/confirm-a-time-and-send-the-invitation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/contribute-a-library-entry-as-an-outsider` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/correct-a-student-record` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/cost-a-grant-application` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/count-agents-and-subagents-per-task` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/create-a-contract-from-a-template` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/create-a-job-role-and-vacancy` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/create-a-markdown-document` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/create-a-module` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/create-a-team-and-its-channels` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/create-an-availability-poll` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/curate-a-collection` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/declare-a-conflict-of-interest` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/declare-exclusion-grounds-and-conflicts` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/decline-an-assignment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/define-a-purpose-view` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/define-or-change-a-schema` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/delete-data-at-retention-end` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/deploy-a-release` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/deposit-a-dataset-with-a-persistent-identifier` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/deprecate-a-feature` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/detect-unlicensed-reuse` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/draw-on-judecanvas` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/edit-a-video` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/end-employment-and-revoke-access` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/enforce-the-code-of-conduct` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/enrol-on-a-doctoral-training-module` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/escalate-a-dispute-to-the-editor` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/evaluate-bids-with-a-scoring-panel` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/expire-or-revoke-a-licence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/explore-the-knowledge-graph-visually` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/export-a-creation-with-licence-and-attribution` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/export-a-meeting-record-and-audit-chain` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/export-and-verify-an-audit-chain` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/find-a-time-across-attendees` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/find-and-follow-tender-opportunities` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/find-funding-calls` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/flag-a-concern-without-blocking` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/flag-plagiarism-or-an-integrity-concern` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/governed-intelligence-degree` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/grant-a-clearance-with-expiry` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/grant-a-licence-to-a-third-party` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/grant-a-maintainer-role-with-expiry` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/hand-over-to-another-agent` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/hand-over-work-with-the-baton` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/handle-a-contract-dispute` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/handle-a-data-subject-access-request` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/handle-an-academic-appeal-on-marks` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/handle-an-appeal` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/hold-an-audit-interview-by-secure-call` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/import-external-records` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/inspect-carbon-and-energy-records` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/inspect-release-gates-and-evidence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/inspect-retention-schedule-compliance` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/inspect-reviewer-payment-records` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/international-technical-students` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/issue-a-certificate-of-membership` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/issue-a-takedown-or-attribution-request` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/issue-a-transcript-of-completed-modules` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/issue-an-employment-contract` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/issue-an-offer-letter` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/issue-findings-and-track-remediation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/issue-the-decision-letter` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/issue-the-written-statement-of-particulars` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/join-a-secure-call-through-the-lobby` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/join-consortium-through-a-university` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/judiverse-included-in-tuition` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/link-orcid-to-the-profile` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/log-a-receipt-for-every-action` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/make-a-below-threshold-direct-award` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/manage-a-controlled-vocabulary` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/manage-a-grant-extension-or-variation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/manage-an-embargo-and-release-date` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/manage-consent-records` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/manage-institutional-roles-and-delegates` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/map-a-dataset-to-the-knowledge-graph` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/map-a-module-to-a-challenge-specification` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/map-a-module-to-programme-and-standard` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/mark-and-moderate-submissions` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/match-reviewers-to-a-submission` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/measure-live-energy-per-job-and-service` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/merge-duplicate-entities` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/merge-with-required-checks` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/mirror-a-repository-to-the-source-repository` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/moderate-a-post` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/monitor-data-quality` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/monitor-job-progress-and-output` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/monitor-tokens-prompts-and-time-to-evidence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/negotiate-clauses-with-the-counterparty` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/nominate-a-learner-for-reviewer-training` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/offer-the-degree-with-a-partner-institution` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/onboard-a-new-starter` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/open-an-issue` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/paid-student-work-with-tuition-offset` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/pass-reviewer-credential-verification` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/pause-or-pre-empt-a-job` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/pay-reviewer-fees` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/pay-salary` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/petition-for-student-finance` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/pin-edit-or-delete-a-post` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/plan-a-maintenance-window` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/plan-a-procurement-and-publish-pipeline-notice` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/post-an-announcement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/pre-register-a-study` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/prepare-a-review-summary-for-a-reviewer` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/prepare-an-employment-contract-template` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/preserve-and-migrate-file-formats` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/problem-brown-student-maps-the-hidden-global-labor-powerin` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/problem-research-agentic-ai-and-consumers` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/process-a-supplier-invoice` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/produce-a-carbon-reduction-plan` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/produce-a-graphic-in-visual-studio` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/produce-a-statutory-data-return` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/propose-a-change-from-a-fork` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/propose-a-real-world-problem-project` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/provide-a-funder-audit-trail` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/provide-a-reference` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/publish-a-data-dictionary` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/publish-a-markdown-as-deck-or-book` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/publish-a-new-dataset-version` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/publish-a-tender-notice` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/publish-a-terms-version-and-require-re-acceptance` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/publish-an-open-access-version` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/publish-documentation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/publish-module-materials-to-the-library` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/publish-the-catalogue-to-open-indexes` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/publish-the-contract-details-notice` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/publish-the-monthly-carbon-and-cost-statement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/raise-a-challenge-or-complaint` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/raise-a-grievance` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/raise-a-progress-or-wellbeing-concern` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/rank-agents-and-models-by-efficiency` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/read-a-verified-entry-with-its-provenance` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/real-project-assignments` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/receive-a-mandate` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/receive-bids-in-a-sealed-queue` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/receive-the-award-outcome` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/receive-time-boxed-auditor-access` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/reconcile-monthly-spend` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/reconcile-provider-invoices-with-metered-use` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/reconcile-with-the-student-record-system` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/record-a-call-with-consent` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/record-a-conflict-of-interest-assessment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/record-a-data-processing-agreement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/record-a-priority-date` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/record-a-trademark-and-brand-use-rule` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/record-an-invention-disclosure` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/record-applications-and-outcomes-in-a-portfolio` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/record-attendance-and-engagement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/record-authorship-and-contributor-roles` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/record-background-and-foreground-ip` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/record-copyright-and-attribution` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/record-energy-per-job` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/record-payment-compliance` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/record-provenance-and-lineage` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/record-working-hours-and-leave` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/refuse-an-action-outside-the-mandate` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/register-a-compute-resource` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/register-a-dataset-in-the-catalogue` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/register-a-delegated-authority` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/register-a-research-project-with-an-id` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/register-a-university-as-a-member-institution` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/register-as-a-supplier` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/register-with-the-office-for-students` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/reject-a-contribution-with-reasons` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/release-a-version` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/release-marks-and-feedback` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/release-results-with-disclosure-control` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/renew-annual-institutional-membership` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/renew-reviewer-verification` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/report-a-personal-data-breach` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/report-a-research-misconduct-concern` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/report-carbon-intensity-per-accepted-outcome` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/report-expenditure-to-the-funder` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/report-monitoring-measures` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/report-performance-indicator-evidence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/report-review-throughput-and-times` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/report-tokens-prompts-and-time-to-evidence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/report-utilisation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/request-a-compute-allowance-for-a-project` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/request-a-purpose-view-on-federated-data` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/request-a-sandbox-with-gpu-and-data` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/request-approval-for-a-consequential-action` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/request-changes-from-the-contributor` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/request-supervisor-allocation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/research-and-development-partnership` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/research-consortium` | changed | 2026-11-01-research-is-not-scope | | `edge-consortium/reserve-capacity-for-a-deadline` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/resolve-conflicting-reviews` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/respond-to-a-security-incident` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/respond-to-an-invitation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/retract-or-correct-a-verified-entry` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/retry-a-failed-job` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/review-a-contribution-against-criteria` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/review-a-funder-expenditure-statement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/review-a-pull-request` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/review-access-logs-and-anomalies` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/review-agent-receipts` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/review-module-evaluation-feedback` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/review-quality-and-standards-evidence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/review-the-dpia-and-record-of-processing` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/revoke-a-purpose-view` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/revoke-access-for-a-leaver` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/rotate-keys` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/run-a-clarification-round` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/run-a-compute-job-under-mandate` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/run-a-cyber-recovery-drill` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/run-a-live-assessment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/run-a-live-seminar` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/run-a-security-classified-procurement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/run-a-training-job-on-scheduled-compute` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/run-a-vulnerability-scan-on-a-release` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/run-an-approved-query-where-data-lives` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/run-code-in-the-sandbox` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/run-data-quality-checks-on-a-database` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/sample-verified-entries-for-quality` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/save-sources-to-a-reading-list` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/scale-with-an-external-provider` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/schedule-a-job` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/schedule-a-meeting-on-request` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/schedule-a-secure-government-resource` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/schedule-a-send-and-set-a-signature` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/schedule-and-record-supervision-meetings` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/schedule-deliveries-and-logistics` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/search-across-mail-channels-and-meetings` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/search-libraries-and-the-knowledge-graph` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/send-a-contract-for-signature` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/send-mail-from-an-estate-mailbox` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/set-a-carbon-budget-and-threshold` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/set-a-project-access-control-policy` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/set-agent-authority-scope` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/set-and-publish-key-performance-indicators` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/set-assessment-and-marking-scheme` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/set-contract-parties-permissions-and-expiry` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/set-evaluation-criteria-and-weightings` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/set-mail-rules-flags-and-snooze` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/set-mention-and-tag-rules` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/set-out-of-office-and-working-hours` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/set-permissions-on-project-records` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/set-presence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/set-project-governance` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/set-quotas-per-project` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/set-renewal-and-expiry-alerts` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/set-retention-and-export-a-channel` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/set-review-criteria-for-a-field` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/set-review-fields-and-availability` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/set-scheduling-policy-and-queues` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/set-the-access-tier-on-an-item` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/set-the-review-deadline-and-reminders` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/set-up-an-awarded-grant` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/share-a-calendar-with-a-delegate` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/share-a-campus-with-a-partner-university` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/share-a-creation-in-the-judebox-exchange` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/share-a-document-in-a-call` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/share-a-workspace-folder-across-universities` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/shift-jobs-to-low-carbon-windows` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/show-undisclosed-figures-as-gaps` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/sign-a-collaboration-agreement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/sign-a-contract-with-edge-id` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/sign-off-a-progression-milestone` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/sign-the-contributor-agreement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/sign-the-employment-contract` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/sign-the-public-contract` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/stop-or-kill-a-job` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/store-the-signed-contract-in-the-vault` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/student-to-contractor-pathway` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/submit-a-bid` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/submit-a-contribution-for-review` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/submit-a-grant-application` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/submit-a-paper-to-consortium-review` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/submit-an-annual-report-to-the-funder` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/submit-an-invoice-and-track-payment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/submit-staff-for-security-vetting` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/suspend-or-end-a-membership` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/terminate-a-contract` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/terminate-a-contract-and-publish-the-notice` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/trace-network-ingress-and-egress` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/track-research-outputs-for-a-ref-return` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/track-royalties-and-revenue-share` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/triage-a-security-vulnerability-report` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/triage-the-submissions-queue` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/university-enrolment` | changed | 2026-11-01-research-is-not-scope | | `edge-consortium/update-supplier-information` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/use-the-terminal` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/vary-a-contract` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/verify-a-fact-before-stating-it` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/verify-a-fact-in-the-knowledge-graph` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/verify-a-person-s-identity` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/verify-a-signature` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/verify-institutional-accreditation-status` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/verify-procurement-notice-publication` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/verify-the-accessibility-statement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/verify-the-audit-chain-offline` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/version-and-roll-back-a-creation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/vet-a-reviewer-or-supplier-staff-member` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/view-the-diary-by-day-week-month` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/view-the-public-reviewer-track-record` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/vote-on-a-poll` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/write-a-data-management-plan` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `edge-consortium/write-a-paper-in-markdowns-with-co-authors` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/accept-a-cheque-deposit` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/access-a-read-only-evidence-room` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/account-for-agent-resource-use` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/accrue-deposit-interest-expense` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/acknowledge-a-complaint` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/act-under-a-power-of-attorney` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/action-an-erasure-or-rectification-request` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/activate-a-monitoring-rule` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/activate-a-proposed-rate` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/add-a-joint-account-holder` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/add-a-payee-with-name-check` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/add-or-remove-a-signatory` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/agent-mandate` | changed | 2026-11-01-research-is-not-scope | | `finance/agent-onboarding` | changed | 2026-11-01-research-is-not-scope | | `finance/agent-retirement` | changed | 2026-11-01-research-is-not-scope | | `finance/agree-a-forbearance-plan` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/amend-a-policy-mid-term` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/amend-a-repayment-schedule` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/amend-or-cancel-a-standing-order` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/analyse-complaint-root-causes` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/answer-a-customer-question-from-records` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/answer-a-subject-access-request` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/apply-a-database-migration` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/apply-a-payment-holiday` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/apply-enhanced-due-diligence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/apply-for-a-business-account` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/apply-for-a-business-loan-or-facility` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/apply-for-a-personal-loan` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/apply-for-loan` | changed | 2026-11-01-research-is-not-scope | | `finance/apply-reasonable-adjustments` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/apply-retention-and-deletion` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/approve-a-financial-promotion` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/approve-a-new-integration-or-api-consumer` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/approve-or-decline-an-application` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/approve-or-reject-a-payment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/assess-a-claim` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/assess-an-outsourcing-or-third-party` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/assess-income-and-expenditure` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/assess-source-of-funds-and-wealth` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/audit-partner-usage` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/authorise-card-transactions-in-real-time` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/back-up-the-database` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/bill-a-partner` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/block-a-payment-in-flight` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/book-an-appointment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/bulk-payments` | changed | 2026-11-01-research-is-not-scope | | `finance/business-roles-and-approvals` | changed | 2026-11-01-reference-records-pass5 | | `finance/calculate-and-post-interest-accrual` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/calculate-the-liquidity-coverage-ratio` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/cancel-a-direct-debit` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/cancel-a-policy` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/capital-to-outcomes-ledger` | added | 2026-10-31-money-provenance | | `finance/cards` | changed | 2026-11-01-research-is-not-scope | | `finance/change-contact-details` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/claim-a-direct-debit-indemnity-refund` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/close-a-business-account` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/close-a-period` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/close-account` | changed | 2026-11-01-research-is-not-scope | | `finance/close-an-account` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/collect-direct-debits-as-an-originator` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/collect-premiums` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/compare-cover-options` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/complete-entity-tax-classification` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/complete-tax-residency-self-certification` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/compute-expected-credit-losses` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/confidence-report` | changed | 2026-11-01-research-is-not-scope | | `finance/configure-scheduled-jobs` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/confirm-balances-as-external-auditor` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/connect-external-accounts` | changed | 2026-11-01-research-is-not-scope | | `finance/contact-a-customer-in-arrears` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/continue-a-joint-account-after-a-death` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/create-a-monitoring-rule` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/create-a-prospect-record` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/create-a-savings-product` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/create-and-release-a-hold` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/crypto-wallet` | changed | 2026-11-01-research-is-not-scope | | `finance/decide-a-treasury-proposal` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/decide-an-app-reimbursement-claim` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/decide-and-file-a-sar` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/decline-with-reasons` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/deploy-a-release-as-jude` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/deploy-a-release-through-deploy-py` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/design-a-new-lending-product` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/detect-a-missed-repayment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/direct-debits` | changed | 2026-11-01-research-is-not-scope | | `finance/disburse-a-loan` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/disclose-a-vulnerability-or-need` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/disposition-an-alert` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/dispute-a-card-transaction` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/disputes` | changed | 2026-11-01-research-is-not-scope | | `finance/draw-down-from-a-facility` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/enforce-strong-customer-authentication` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/escalate-a-customer-concern` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/execute-a-money-market-placement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/exit-a-customer-for-financial-crime` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/explain-a-transaction` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/export-the-evidence-bundle` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/export-transactions-to-accounting-software` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/financing-graph-independent-demand` | added | 2026-10-31-money-provenance | | `finance/flag-a-mule-account` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/fraud-sharing-network` | changed | 2026-11-01-research-is-not-scope | | `finance/freeze-or-unfreeze-a-card` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/fund-settlement-accounts-for-rails` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/give-or-revoke-open-banking-consent` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/grant-an-operator-role` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/hand-a-case-to-a-specialist-team` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/hand-work-between-agents` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/handle-a-banking-protocol-concern` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/handle-a-breathing-space-notice` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/handle-a-debt-sale` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/handle-a-gambling-block-request` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/handle-a-mental-capacity-concern` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/handle-a-partner-data-request-or-deletion` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/handle-a-security-incident` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/handle-a-trigger-event-review` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/handle-a-whistleblowing-report` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/handle-an-indemnity-claim-as-paying-bank` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/hedge-interest-rate-exposure` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/import-a-bank-or-rail-statement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/inspect-a-sar-audit-trail` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/introduce-a-customer-to-a-specialist` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/investigate-a-complaint` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/investigate-a-payment-exception` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/investigate-a-production-error` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/investigate-a-suspected-fraud` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/invite-a-user-with-a-role` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/issue-a-default-notice` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/issue-a-final-response` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/issue-a-finding-to-the-bank` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/issue-a-policy` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/issue-a-replacement-card-in-branch` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/issue-an-agent-mandate` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/issue-annual-loan-statements` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/joint-accounts` | changed | 2026-11-01-research-is-not-scope | | `finance/lend` | changed | 2026-11-01-research-is-not-scope | | `finance/lift-an-account-freeze` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/log-an-operational-incident` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/log-an-outbound-contact` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/maintain-internal-and-nostro-accounts` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/maintain-the-business-wide-risk-assessment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/maintain-the-chart-of-accounts` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/maintain-the-record-of-processing` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/maintain-the-regulatory-obligations-register` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/maintain-the-risk-register` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/make-a-complaint` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/make-a-faster-payment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/make-a-regulatory-notification` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/make-an-internal-transfer` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/make-an-international-payment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/manage-a-syndicated-participation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/manage-capacity` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/manage-encryption-and-device-binding` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/manage-environments-and-secrets-free-config` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/manage-foreign-exchange-positions` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/manage-partner-user-access` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/manage-payee-name-check-responses` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/manage-payment-limits` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/manage-senior-manager-responsibilities` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/manage-the-central-bank-reserve-account` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/map-important-business-services` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/mark-a-customer-as-a-fraud-victim` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/match-statement-rows-to-ledger-legs` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/measure-carbon-tokens-and-time-to-evidence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/message-a-customer-in-a-case` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/money-provenance-certificate` | added | 2026-10-31-money-provenance | | `finance/monitor-chaps-cut-offs-and-queue` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/monitor-conduct-and-quality` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/monitor-facility-covenants` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/monitor-intraday-liquidity` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/monitor-job-runs-and-leases` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/monitor-privileged-actions` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/monitor-product-outcomes` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/monitor-service-health` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/monitor-without-blocking` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/notify-a-claim` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/notify-a-death-and-restrict-accounts` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/notify-a-personal-data-breach` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/offboard-a-business-customer` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/onboard-a-partner-organisation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/open-a-fixed-term-deposit` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/open-a-foreign-currency-account` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/open-a-personal-current-account` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/open-a-savings-account` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/open-a-service-case` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/open-a-transaction-receipt` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/open-business-account` | changed | 2026-11-01-research-is-not-scope | | `finance/open-personal-account` | changed | 2026-11-01-research-is-not-scope | | `finance/open-virtual-accounts-under-a-master` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/order-business-cards-for-staff` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/overdraft` | changed | 2026-11-01-research-is-not-scope | | `finance/pay-redress` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/perform-a-data-protection-impact-assessment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/perform-an-internal-audit` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/place-a-temporary-account-freeze` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/plan-a-portfolio-review` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/plan-the-internal-audit-year` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/post-a-manual-journal` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/prepare-a-business-proposition` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/prepare-a-compliance-return` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/prepare-a-payment-for-approval` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/prepare-a-payment-for-human-approval` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/prepare-funds-transfer-pricing` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/prepare-regulatory-returns` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/prepare-statutory-accounts` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-6-3bn-in-30-minutes-how-the-world-s-biggest-heist` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-alcohol-duty-penalties-for-late-filing-and-late-pa` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-applications-open-for-next-director-of-the-serious` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-bank-of-england-explores-easing-key-limit-on-riski` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-bereaved-families-will-not-face-tax-bill-over-pens` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-beyond-borders-regulating-the-risks-we-share` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-binance-rushes-to-recover-customer-money-after-6-3` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-binance-s-6-3bn-crypto-hack-biggest-in-history` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-binance-s-6-3bn-heist-is-the-biggest-but-not-the-f` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-brits-nearly-twice-as-likely-to-spot-pressure-tact` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-crypto-giant-binance-attempting-to-recover-6-3bn-s` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-crypto-giant-binance-hit-by-6-3bn-hack-suspected-t` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-defence-smes-need-banking-services-that-work` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-defined-benefit-pension-scheme-surplus-payments-to` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-fca-restricts-6-firms-following-concerns-over-anti` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-former-chief-executive-banned-after-misconduct-led` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-france-s-supermarket-fridges-were-vulnerable-onlin` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-french-supermarkets-suffer-fridge-shutdown-after-u` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-from-bitcoin-to-binance-crypto-terms-and-what-they` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-government-unleashes-1-2-billion-crackdown-as-part` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-guidance-framework-agreement-between-the-law-offic` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-guidance-make-a-corporate-self-report` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-hackers-activate-automatic-defrosting-in-supermark` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-how-ai-might-be-deployed-for-better-bank-supervisi` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-how-big-a-deal-is-the-6-3bn-binance-crypto-heist-w` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-how-crypto-hacking-funds-isolated-north-korea-s-nu` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-how-poundland-grew-from-a-market-stall-to-a-high-s` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-i-read-anthropic-s-92-page-risk-factor-document-so` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-israeli-settlers-attack-european-diplomatic-cars-i` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-israeli-settlers-attack-olive-framers-washington-p` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-ledger-confirms-spy-chip-implant-behind-wallet-cry` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-minutes-of-the-sonia-stakeholder-advisory-group-25` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-north-korea-is-likely-culprit-behind-6-3bn-crypto` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-ofqual-fines-innovate-awarding-185-000-over-failur` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-pip-customers-experiences-of-the-award-review-proc` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-response-to-supreme-court-judgment` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-scam-american-companies-are-using-to-manipulate-in` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-sfo-strengthens-corporate-cooperation-guidance` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-the-staggering-446m-elon-musk-has-spent-on-the-us` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-the-timms-review-co-chair-update-october-2026` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-three-arrests-in-major-fca-operation-against-crimi` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-trump-beware-us-voters-are-concerned-about-governm` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-two-plead-guilty-after-clients-lose-2m-in-fraud-ca` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-tyneside-pizza-boss-jailed-after-gambling-with-mon` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-what-we-know-about-the-6-3bn-binance-hack` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-whistleblowing` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-who-could-be-behind-the-binance-hack-and-what-we-k` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/problem-woman-charged-with-computer-fraud-due-to-scam-cash` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/process-a-bacs-input-and-output-cycle` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/process-a-cheque-clearing-exception` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/process-a-counter-cash-deposit` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/process-a-counter-cash-withdrawal` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/process-a-loan-repayment-manually` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/process-an-inbound-faster-payment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/process-card-clearing-and-settlement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/produce-a-kyc-file-for-review` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/produce-a-trial-balance` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/produce-the-annual-mlro-report` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/produce-the-fscs-single-customer-view` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/propose-a-future-rate` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/propose-a-treasury-credit` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/provenance-coverage-report` | added | 2026-10-31-money-provenance | | `finance/provide-extra-evidence-on-request` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/proximity-verification` | changed | 2026-11-01-research-is-not-scope | | `finance/publish-terms-and-conditions` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/pull-and-record-a-credit-reference` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/query-a-mule-account-indicator` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/raise-an-internal-suspicion-report` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/receive-a-fraud-intelligence-signal` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/receive-a-rate-change-notice` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/recognise-loan-interest-income` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/reconcile-rail-settlement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/reconcile-suspense-accounts` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/record-a-credit-decision-override` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/record-a-customer-meeting` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/record-a-customer-risk-assessment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/record-a-declined-application-reason` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/record-a-regulatory-breach` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/record-a-support-need` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/record-a-vulnerability-disclosure` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/record-aml-training-completion` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/record-and-test-a-control` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/recover-account-access` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/recover-funds-from-a-receiving-account` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/refer-a-complaint-to-fos` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/refer-to-a-debt-collection-agency` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/refer-to-debt-advice` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/refund-an-unauthorised-transaction` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/register-a-court-appointed-deputy` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/register-a-default-with-agencies` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/register-a-lasting-power-of-attorney` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/register-a-new-device` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/register-a-third-party-mandate` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/register-and-value-collateral` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/register-beneficial-owners` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/regulator-inspection` | changed | 2026-11-01-research-is-not-scope | | `finance/release-estate-funds` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/rely-on-another-firm-s-due-diligence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/remove-a-joint-holder` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/rename-an-account` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/renew-a-policy` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/reopen-a-period` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/repair-a-misdirected-payment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/repay-or-overpay-a-loan` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/report-a-bereavement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/report-a-change-of-company-details` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/report-a-lost-or-stolen-card` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/report-an-app-scam` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/report-an-unauthorised-payment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/report-complaints-data` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/report-fraud-losses-to-the-regulator` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/report-fraud-to-action-fraud-or-police` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/report-insurance-data-to-the-regulator` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/report-payment-statistics` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/report-to-credit-reference-agencies` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/request-a-bank-statement-letter` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/request-a-company-audit-letter` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/request-a-covenant-waiver` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/request-a-date-of-death-balance-letter` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/request-a-defence-against-money-laundering` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/request-a-loan-settlement-quote` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/request-a-trade-finance-instrument` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/request-affordability-evidence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/request-an-arranged-overdraft` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/request-an-insurance-quote-from-kyc` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/request-further-information` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/request-money` | changed | 2026-11-01-research-is-not-scope | | `finance/request-my-personal-data` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/request-owner-approval-for-a-consequential-action` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/request-records-from-the-bank` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/reset-access-for-a-customer` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/resolve-within-3-business-days` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/respond-to-a-production-order` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/respond-to-an-open-banking-payment-request` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/responsible-lending-decision` | added | 2026-10-31-money-provenance | | `finance/restore-from-a-backup` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/retry-a-failed-job-step` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/retry-a-failed-occurrence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/return-an-unapplied-payment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/reverse-a-journal-with-a-linked-entry` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/review-a-complaint-file-as-fos` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/review-a-loan-application` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/review-access-rights` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/review-an-application-queue` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/review-app-reimbursement-performance` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/review-customer-financial-statements` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/review-device-and-session-signals` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/review-operational-resilience-evidence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/review-sm-cr-responsibilities` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/review-the-vulnerability-register` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/revoke-a-role-on-leaving` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/revoke-an-agent-mandate` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/roll-back-a-release` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/rotate-keys-and-device-bindings` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/run-a-consumer-duty-outcomes-review` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/run-a-disaster-recovery-test` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/run-a-fair-value-assessment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/run-a-kyc-check-on-a-subject` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/run-a-ledger-report-for-a-period` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/run-a-liquidity-stress-test` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/run-an-annual-relationship-review` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/run-due-jobs-as-service-actor` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/run-due-standing-orders` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/run-end-of-day-payment-reconciliation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/run-payroll-payments` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/run-penetration-testing` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/run-periodic-kyc-refresh` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/run-protected-reveals` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/run-sanctions-screening` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/run-transaction-monitoring` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/sample-test-receipts-as-external-auditor` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/sanctions-and-pep-screening` | changed | 2026-11-01-research-is-not-scope | | `finance/save-an-unfinished-application` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/savings-pots` | changed | 2026-11-01-research-is-not-scope | | `finance/schedule-a-future-payment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/scheduled-payments` | changed | 2026-11-01-research-is-not-scope | | `finance/screen-a-list-against-sanctions` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/screen-existing-customers-against-list-updates` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/screen-for-politically-exposed-persons` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/search-transactions` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/security-and-devices` | changed | 2026-11-01-research-is-not-scope | | `finance/send-a-chaps-payment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/send-a-receiving-bank-app-notice` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/send-money` | changed | 2026-11-01-research-is-not-scope | | `finance/set-and-monitor-exposure-limits` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/set-balance-and-payment-alerts` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/set-card-limits-and-controls` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/set-eligibility-rules` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/set-fees-and-charges` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/set-joint-operating-rules` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/set-loan-terms-and-produce-an-offer` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/set-negotiated-pricing` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/set-up-a-direct-debit` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/set-up-a-mandate` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/set-up-a-standing-order` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/set-up-merchant-acquiring` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/settle-a-claim` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/settle-and-close-a-loan` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/solution-bank-funding` | changed | 2026-11-01-research-is-not-scope | | `finance/stop-a-savings-enrolment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/submit-transactions-for-monitoring` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/support-a-customer-in-financial-difficulty` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/support-the-external-audit` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/switch-account-to-edge-bank` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/terminate-a-partnership` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/trace-ingress-and-egress` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/track-remediation-actions` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/transaction-monitoring` | changed | 2026-11-01-research-is-not-scope | | `finance/transition-a-case-state` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/triage-a-case-queue` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/tune-a-fraud-rule` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/underwrite-a-business-facility` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/underwrite-a-policy` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/update-communication-preferences` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/upload-a-bulk-payment-file` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/validate-a-credit-risk-model` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/verify-a-caller-or-visitor` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/verify-a-company-and-its-officers` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/verify-a-regulatory-return-against-the-ledger` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/verify-a-trust-or-charity` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/verify-address-evidence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/verify-an-identity-through-edge-id` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/verify-executor-or-administrator` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/verify-identity-evidence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/view-and-download-a-statement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/view-balances-and-available-funds` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/view-multi-account-cash-position` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/view-pin-securely` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/void-an-imported-statement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/withdraw-a-product` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/withdraw-a-proposed-rate` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/withdraw-an-application` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/work-the-fraud-alert-queue` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/write-a-fraud-case-outcome-letter` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `finance/write-off-a-debt` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/automated-review-outreach-and-scheduling` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/bottleneck-and-blocker-removal` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/claim-organisation-by-magic-link` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/evaluation-and-commercial-licensing` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/licence-enforcement-evidence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/organisation-directory-and-claim` | changed | 2026-11-01-research-is-not-scope | | `foundation/pre-onboarded-organisations` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-12-000-fine-for-cheltenham-waste-operator-for-disp` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-a-28-year-old-doge-operative-gutted-usaid-hr-had-n` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-a-complete-bypass-of-emudhra-s-1fa-otp-login-syste` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-about-the-compromised-emudhra-intermediate-cas` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-air-pollution-trends-reports` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-azure-s-weakest-link-five-full-cross-tenant-compro` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-bedfordshire-soldiers-laid-to-rest-on-the-western` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-binance-hacker-1029-5-6b-stolen-from-binance` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-blockchain-intelligence-firms-link-north-korea-to` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-britain-s-housing-crisis-won-t-be-fixed-without-a` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-bulgaria-votes-as-russian-attacks-in-black-sea-rea` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-burnham-ready-to-support-poundland-staff-as-retail` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-capital-gains-tax-returns-for-disposals-of-uk-land` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-codex-cli-installs-a-persistent-self-updating-back` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-companies-house-third-party-code-of-conduct` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-company-director-disqualification-following-breach` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-compromised-litellm-1-107-0-on-pypi-installs-a-cre` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-cops-can-bypass-iphone-s-automatic-reboot-to-get-i` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-cve-2026-111943-compromised-intermediate-certifica` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-defending-odihr-s-independence-and-the-osce-human` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-dungeon-generation-in-the-gnomish-mines-of-nethack` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-emudhra-intermediate-ca-compromise` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-excise-duty-penalties-for-wrongdoing` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-eye-of-sauron-long-range-hidden-spy-camera-detecti` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-florida-s-hoas-and-condo-associations-are-making-t` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-git-3-0-s-upcoming-sha-256-default-will-be-a-costl` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-guidance-chickerell-firing-times` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-guidance-tregantle-firing-times` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-gwyndod-supply-chain-mobilised-to-deliver-uk-s-fir` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-hackers-expose-the-identities-of-more-than-40k-dhs` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-how-a-1-deal-dragged-poundland-s-owner-into-a-poli` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-how-oil-rich-alberta-s-economy-might-fare-if-it-br` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-how-the-military-will-hack-the-election` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-i-m-on-a-war-footing-terry-tao-and-the-crisis-in-m` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-icc-judge-on-what-u-s-sanctions-mean-for-her-and-g` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-imposing-sanctions-on-the-international-criminal-c` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-independent-review-of-the-prison-system-terms-of-r` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-investigation-launched-as-dispute-risks-bringing-d` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-it-took-me-nine-years-to-hack-my-xbox-360` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-jared-kushner-hacked-phone-reveals-secret-negotiat` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-london-loft-and-extension-company-which-left-unfin` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-making-registration-applications-easier-and-more-e` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-middlesex-regiment-soldier-laid-to-rest-on-the-som` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-modernising-the-correction-of-errors` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-nicotine-approval-scheme` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-no-more-icons-in-every-menu-by-default` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-northflank-breached` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-opencode-hacked-user-credentials-leaked` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-professor-akande-brings-the-independence-legal-rig` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-publishing-details-of-deliberate-defaulters-reform` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-radioactivity-in-food-and-the-environment-rife-rep` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-restrictions-on-the-performance-of-high-risk-cosme` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-show-hn-wallhop-12ft-io-is-gone-so-i-built-a-repla` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-telegram-desktop-vulnerability-allowed-any-user-s` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-tell-hn-scribd-breached` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-the-6-3b-binance-hack-was-pulled-off-in-just-20-mi` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-today-we-hacked-what-we-d-set-out-to-hack-the-ipho` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-uk-sanctions-senior-commanders-implicated-in-campa` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-uk-ukraine-and-latvia-extend-crackdown-on-russian` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-us-federal-budget-deficit-jumps-12-in-one-year-to` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-us-imposes-sanctions-on-icc-hours-after-former-jud` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-various-projects-find-hidden-sdr-capabilities-in-e` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-we-re-going-to-need-default-hard-budget-caps-on-pr` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-why-north-korea-s-elite-hackers-are-the-prime-susp` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-why-windows-11-still-ships-with-windows-95-s-memor` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/problem-zod-4-6-6-compromised-by-malware` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/regulator-auditor-reviewer-workspace` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `foundation/scope-change-intake` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/agent-builder` | changed | 2026-11-01-reference-records-pass5 | | `judiverse/calendar-and-scheduling` | changed | 2026-11-01-reference-records-pass5 | | `judiverse/contacts-directory-and-search` | changed | 2026-11-01-reference-records-pass5 | | `judiverse/deep-research` | changed | 2026-11-01-reference-records-pass5 | | `judiverse/documents-and-forms` | changed | 2026-11-01-reference-records-pass5 | | `judiverse/jude-assistant` | changed | 2026-11-01-reference-records-pass5 | | `judiverse/judebox-agents` | added | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/judebox-calls` | added | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/judebox-connections` | added | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/judebox-contacts` | added | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/judebox-email` | added | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/judebox-memory-and-vault` | added | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/judebox-messaging` | added | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/judebox-music` | changed | 2026-11-01-reference-records-pass5 | | `judiverse/judebox-numbers-and-mailboxes` | added | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/judebox-webhooks` | added | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/judiverse-included-with-every-licence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/judiverse-standalone-subscription` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/mail-and-agent-mailboxes` | changed | 2026-11-01-reference-records-pass5 | | `judiverse/messages-and-chat` | changed | 2026-11-01-reference-records-pass5 | | `judiverse/problem-123456-password-used-in-danish-cpr-data-breach` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-afd-leak-what-421-179-hacked-emails-show-about-the` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-agentic-ai-and-consumers` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-ai-and-leadership` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-amazon-uses-ai-to-monitor-union-risk-in-uk-warehou` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-apple-and-a-hacker-s-future` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-apple-music-404-not-found` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-apple-s-git-still-trusts-the-compromised-emudhra-c` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-brown-student-maps-the-hidden-global-labor-powerin` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-cisco-gives-engineer-unlimited-ai-tokens-to-rewrit` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-clef-open-weight-decision-models-and-new-rl-fine-t` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-containment-and-monitoring-failures-are-to-blame-f` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-cross-request-kv-cache-exposure-in-the-glm-5-3-wor` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-denmark-data-breach-exposes-8-8m-people-s-personal` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-draft-legislation-better-use-of-new-and-improved-t` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-electrocutions-disease-and-no-pay-how-workers-risk` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-faster-planning-decisions-as-new-rules-and-ai-tack` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-ftc-is-investigating-openai-anthropic-and-other-ai` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-hacker-leaks-281-gb-of-data-from-germany-s-far-rig` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-i-quit-openai-because-its-culture-is-broken` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-inside-the-make-or-break-ai-mega-project-behind-ns` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-investors-trim-growth-bets-over-fears-of-us-econom` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-is-a-job-guarantee-the-solution-to-mass-ai-unemplo` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-meta-hit-with-25b-penalty-for-deceiving-users-abou` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-near-to-real-time-suspected-suicide-surveillance-n` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-openai-safety-leader-quits-warning-ai-company-s-cu` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-postgres-19-delayed-after-ai-reviewers-find-more-b` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-research-agentic-ai-and-consumers` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-run-full-ios-16-5-firmware-on-android-devices` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-turn-a-wifi-7-router-into-a-320msps-software-defin` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-vote-on-which-of-hacker-news-challenges-for-ai-hav` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-wanted-to-download-an-iso-ended-up-writing-a-kerne` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-why-aren-t-ai-companies-being-charged-for-hacking` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-with-most-information-hidden-the-game-stratego-had` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/problem-yandex-bot-traffic-down-99-after-data-center-attac` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `judiverse/procedure-compilation` | changed | 2026-11-01-reference-records-pass5 | | `judiverse/tasks-and-workflows` | changed | 2026-11-01-reference-records-pass5 | | `public-edge/accept-a-package-referral` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/act-on-a-deadline-stack-item` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/activate-the-emergency-centre-and-keep-the-decision-log` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/adopt-a-risk-based-reserves-strategy-with-minimum-level-and` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/advise-on-legality-and-sign-the-legal-implications-of-a-repo` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/agree-a-care-and-support-plan-and-personal-budget` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/agree-a-deferred-payment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/agree-a-personalised-housing-plan` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/agree-an-information-sharing-agreement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/agree-and-seal-a-planning-obligation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/agree-the-council-plan-and-its-outcomes` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/agree-the-officer-and-member-protocol` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/agree-the-political-group-position-before-a-meeting` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/agree-the-post-election-timetable-and-transition-protocol` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/allocate-committee-seats-by-political-balance` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/analog-human-switch-for-irreversible-actions` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/answer-a-best-value-notice-or-mhclg-engagement-with-one-evid` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/answer-a-code-of-conduct-complaint` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/answer-a-corporate-complaint` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/answer-a-freedom-of-information-request` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/answer-a-local-land-charges-search` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/answer-a-media-enquiry-and-issue-a-press-release` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/answer-a-standards-complaint` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/answer-a-subject-access-request` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/answer-a-written-parliamentary-question` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/answer-an-appeal-to-the-valuation-tribunal` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/answer-an-environmental-information-request` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/answer-ministerial-correspondence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/answer-questions-at-a-scrutiny-meeting` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/appeal-a-planning-decision` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/apply-an-empty-home-or-second-home-premium` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/apply-an-exemption-and-issue-a-refusal-notice` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/apply-for-a-blue-badge` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/apply-for-a-deprivation-of-liberty-authorisation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/apply-for-a-discount-or-council-tax-reduction` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/apply-for-a-licence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/apply-for-a-postal-or-proxy-vote` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/apply-for-a-school-place-and-appeal-a-refusal` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/apply-for-a-voter-authority-certificate` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/apply-for-an-export-licence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/apply-for-free-school-meals` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/apply-for-planning-permission-and-track-it` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/apply-the-armed-forces-covenant-duty-in-a-council-decision` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/apply-the-monthly-savings-test` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/apply-the-national-pay-award` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/apply-the-one-page-decision-standard-and-report-standard-to` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/apply-the-pre-election-period` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/apply-the-retention-schedule-and-dispose-of-records` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/appoint-an-election-agent` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/appoint-an-independent-member-to-the-committee` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/appoint-and-train-polling-staff` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/appoint-or-dismiss-a-statutory-officer` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/appoint-the-cabinet-and-allocate-portfolios` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/appoint-the-external-auditor-through-psaa` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/appoint-the-internal-auditor-and-receive-the-report` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/approve-a-business-case-through-the-assurance-framework` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/approve-a-change-through-the-change-board` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/approve-a-direct-payment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/approve-a-disabled-facilities-grant` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/approve-a-new-system-against-the-architecture-standard` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/approve-a-severance-package-over-100-000` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/approve-a-virement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/approve-the-internal-audit-charter` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/approve-the-internal-audit-plan` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/approve-the-medium-term-financial-strategy` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/approve-the-pension-fund-accounts-and-valuation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/approve-the-senior-structure-and-statutory-officer-designati` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/approve-the-treasury-capital-and-minimum-revenue-provision-s` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/ask-a-question-of-the-leader-or-a-cabinet-member` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/ask-a-question-or-present-a-petition-at-a-meeting` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/ask-for-a-care-needs-assessment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/ask-for-help-with-homelessness` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/ask-for-own-personal-data` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/assess-a-carer-s-needs` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/assess-a-code-of-conduct-complaint` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/assess-a-housing-benefit-claim` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/assess-and-decide-bus-franchising` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/assess-council-tax-reduction` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/assess-risk-and-publish-the-community-risk-register` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/assess-the-committee-s-effectiveness-and-report-to-council` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/assess-the-needs-of-an-adult` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/attend-film-or-report-a-public-meeting` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/attend-the-count-as-an-agent` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/authenticate-and-seal-a-document` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/authorise-and-make-a-payment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/authorise-covert-surveillance-under-ripa` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/award-a-discount-or-exemption` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/award-a-discretionary-housing-payment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/award-and-monitor-a-voluntary-sector-grant` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/award-business-rates-relief` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/award-under-a-framework-or-dynamic-market` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/back-up-and-test-a-restore` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/bid-for-and-report-on-a-transport-grant` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/bill-a-new-business-rates-property` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/build-a-resident-service-through-discovery-alpha-and-beta` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/build-and-sign-the-savings-ledger-line-by-line` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/calculate-the-council-tax-base-and-requirement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/call-in-an-executive-decision` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/canvass-voters-lawfully` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/carry-out-a-data-protection-impact-assessment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/carry-out-a-financial-assessment-and-charge` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/carry-out-an-education-health-and-care-needs-assessment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/carry-out-an-internal-review` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/cast-a-recorded-vote-on-the-budget` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/certify-a-grant-claim` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/chair-full-council-and-rule-on-order` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/challenge-a-council-tax-band` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/challenge-a-provider-rate-and-decide-an-uplift` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/check-supply-chain-labour-and-modern-slavery-risk` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/claim-emergency-financial-assistance` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/claim-members-allowances-and-expenses` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/claim-staffing-and-business-costs` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/clerk-a-school-admission-appeal-panel` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/comment-on-a-planning-application` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/comment-on-a-planning-application-2` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/commission-a-placement-for-a-looked-after-child` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/commission-a-public-health-service-under-the-provider-select` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/commission-supported-living-from-a-charity-provider` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/commission-temporary-accommodation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/commission-the-external-quality-assessment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/complain-to-the-council-and-the-ombudsman` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/complete-induction-and-training-on-reserves-debt-efs-and-ris` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/complete-mandatory-member-training` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/complete-the-annual-governance-and-accountability-return` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/complete-the-data-security-and-protection-toolkit` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/comply-with-a-secretary-of-state-direction` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/conduct-a-marriage-or-civil-partnership` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/confirm-mrp-and-efs-accounting-with-the-external-auditor` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/consider-an-objection-from-a-local-elector` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/consider-social-value-before-procurement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/consolidate-published-recommendations-into-owned-operating-s` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/consult-and-publicise-an-application` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/consult-on-and-make-redundancies` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/contract-management` | changed | 2026-11-01-research-is-not-scope | | `public-edge/convene-the-right-people-by-mail-and-call` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/correct-a-register-entry` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/council-decisions` | changed | 2026-11-01-research-is-not-scope | | `public-edge/cyber-posture` | changed | 2026-11-01-research-is-not-scope | | `public-edge/deal-with-a-dangerous-structure` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/decide-a-parking-penalty-challenge` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/decide-an-application-at-planning-committee` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/decide-on-sharing-confidential-care-information-as-caldicott` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/declare-an-impairment-to-independence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/declare-an-interest-at-a-meeting-and-withdraw` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/declare-an-interest-in-proceedings` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/declare-the-collection-fund-surplus-or-deficit` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/decommission-a-legacy-system` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/defend-a-planning-appeal` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/deliver-the-risk-based-audit-plan-from-substantive-capacity` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/design-and-run-a-grant-scheme` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/determine-a-building-regulations-application` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/determine-a-postal-or-proxy-vote-application` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/determine-an-application-to-register` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/discharge-planning-conditions` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/dispose-of-or-acquire-land-at-best-consideration` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/employ-staff-in-the-mp-s-office` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/employment` | changed | 2026-11-01-research-is-not-scope | | `public-edge/enforce-joiner-mover-and-leaver-with-segregation-of-duties` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/enforce-publication-deadlines-with-an-exception-register` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/enrol-staff-in-the-pension-scheme-and-send-monthly-data` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/ensure-continuity-of-care-on-provider-failure` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/evaluate-and-grade-a-job` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/fill-a-casual-vacancy-by-election-or-co-option` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/find-a-council-borough-or-government-department-in-the-insti` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/find-the-right-council-and-service-for-an-address` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/flow-obligations-into-a-contract-or-commission` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/give-notice-to-hand-back-a-contract` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/give-the-opinion-on-the-financial-statements` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/grant-a-dispensation-for-a-pecuniary-interest` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/grant-a-premises-licence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/grant-a-street-works-permit` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/handle-a-benefit-reconsideration-or-appeal` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/handle-a-highway-claim` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/handle-a-housing-complaint` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/handle-a-request-for-erasure-or-restriction` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/hear-a-call-in-and-decide` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/hear-a-councillor-call-for-action` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/hold-a-citizenship-ceremony` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/hold-a-constituency-surgery` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/hold-a-licensing-sub-committee-hearing` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/hold-an-initial-child-protection-conference` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/hold-and-record-the-monthly-golden-triangle-meeting` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/hold-identity-and-operational-evidence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/hold-pre-decision-scrutiny-of-a-key-decision` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/hold-the-annual-parish-meeting` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/host-a-social-housing-consumer-standards-inspection` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/host-an-lga-corporate-peer-challenge` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/independently-test-an-implemented-control-under-the-self-rev` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/inspect-a-food-business-and-award-a-hygiene-rating` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/inspect-a-highway-and-repair-a-defect` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/inspect-a-private-rented-home-and-serve-a-notice` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/inspect-a-workplace-for-health-and-safety` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/inspect-the-council-s-accounts-and-object` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/inspect-work-and-issue-a-completion-certificate` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/instruct-enforcement-agents-with-a-vulnerability-check` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/investigate-a-breach-and-serve-an-enforcement-notice` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/investigate-a-consumer-complaint-and-take-enforcement-action` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/investigate-a-flood-as-lead-local-flood-authority` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/investigate-a-statutory-nuisance-and-serve-an-abatement-noti` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/investigate-a-suspected-fraud` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/investigate-tenancy-fraud` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/issue-a-best-value-notice-and-run-engagement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/issue-a-certified-copy-of-an-entry` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/issue-a-cil-liability-notice-and-demand` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/issue-a-mayoral-decision` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/issue-a-refund-or-transfer-a-credit` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/issue-a-section-114-report` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/issue-a-section-5-report-on-unlawfulness-or-maladministratio` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/issue-a-security-aspects-letter-to-a-contractor` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/issue-a-statutory-recommendation-or-public-interest-report` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/issue-a-voter-authority-certificate` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/issue-annual-council-tax-bills` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/issue-the-annual-internal-audit-opinion` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/issue-the-audit-plan` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/issue-the-decision-notice` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/issue-the-role-statement-train-and-attribute-challenge` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/issue-the-summons-and-agenda` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/issue-the-weekly-one-page-exception-report` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/join-a-council-s-approved-provider-list` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/join-a-strategic-coordinating-group` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/join-the-housing-register-and-assess-priority` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/keep-business-continuity-plans-and-advise-businesses` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/keep-spending-within-the-limit-and-file-the-return` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/keep-statutory-notification-readiness-and-named-incident-rol` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/keep-the-device-and-software-asset-register` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/keep-the-outside-bodies-appointments-register` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/keep-the-record-of-processing-activities` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/lay-a-statutory-instrument` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/lay-the-annual-report-and-accounts` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/lead-recovery` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/lead-the-response-to-a-section-114-notice` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/let-a-social-home` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/licence-a-taxi-or-private-hire-driver` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/licence-an-animal-activity` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/license-a-house-in-multiple-occupation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/logistics-tracking` | changed | 2026-11-01-research-is-not-scope | | `public-edge/maintain-and-publish-the-register-of-members-interests` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/maintain-the-brownfield-and-self-build-registers` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/maintain-the-constitution-and-scheme-of-delegation-with-live` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/maintain-the-contract-procedure-rules` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/maintain-the-publication-scheme-and-disclosure-log` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/maintain-the-single-procurement-and-commissioning-pipeline` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/make-a-freedom-of-information-request-or-ask-for-review` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/make-a-public-spaces-protection-order` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/make-a-safeguarding-adults-enquiry` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/make-a-traffic-regulation-order` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/make-a-tree-preservation-order` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/manage-a-contract-and-hold-review-meetings` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/manage-a-persistent-or-unreasonable-complainant` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/manage-a-public-question-or-petition-at-a-meeting` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/manage-allotments` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/manage-cash-flow-and-borrowing` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/manage-sickness-absence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/mark-and-check-politically-restricted-posts` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/mark-exempt-information-and-run-a-private-item` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/mark-store-and-send-classified-information` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/meet-the-auditors-without-officers-present` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/modify-a-contract-and-publish-the-change-notice` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/monitor-a-grant-and-recover-misused-funds` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/monitor-a-provider-s-quality` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/name-and-number-a-street` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/nominate-an-asset-of-community-value` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/nominate-an-asset-of-community-value-2` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/notify-the-regulator-of-an-incident` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/obtain-facility-security-clearance-for-a-supplier-site` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/open-a-constituent-case-with-consent` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/open-a-council-tax-account-on-a-move-in` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/open-a-rest-centre-and-register-evacuees` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/open-and-conclude-an-inquest` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/open-postal-votes-and-verify-personal-identifiers` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/open-the-government-station-and-choose-a-journey` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/organise-evidence-and-records-for-a-public-body` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/pass-a-proposal-through-the-business-case-gateway` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/pay-council-tax-and-set-up-a-direct-debit` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/pay-or-challenge-a-parking-penalty` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/pay-the-data-protection-fee` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/place-a-household-in-interim-accommodation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/pool-a-budget-with-the-nhs` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/prepare-a-care-proceedings-application` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/prepare-a-green-book-business-case` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/prepare-a-neighbourhood-plan` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/prepare-a-procurement-and-agreement-requirements-brief-for-a` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/prepare-an-impact-assessment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/prepare-an-off-site-plan-for-a-major-hazard-site` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/prepare-and-examine-the-local-plan` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/prepare-for-a-cqc-assessment-of-adult-social-care` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/prepare-for-a-public-accounts-committee-hearing` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/prepare-for-an-inspection-of-children-s-services` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/prepare-main-and-supplementary-estimates` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/prepare-the-local-transport-plan` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/prepare-the-market-sustainability-plan` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/price-a-single-source-defence-contract` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-cyberattack-halts-construction-of-tunnel-between-d` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-details-of-thousands-of-us-immigration-agents-leak` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-dutch-government-had-hacked-into-jared-kushner-s-p` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-dutch-national-cyber-security-center-is-offline-be` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-experts-warn-ai-copyright-changes-could-breach-aus` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-help-the-ncsc-understand-how-organisations-approac` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-how-north-korea-became-a-hacking-superpower` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-how-sin-bin-for-red-light-jumpers-saves-cyclists-f` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-how-the-military-will-hack-the-election` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-incident-affecting-asos-customers` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-introducing-a-new-ai-chapter-in-the-ncsc-research` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-is-this-the-new-cyber-normal` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-joint-ministerial-statement-on-the-risk-of-famine` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-labor-and-the-coalition-are-poised-to-take-a-reckl` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-legacy-system-local-plan-extension-letter-from-the` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-meta-ordered-to-pay-25b-penalty-in-new-mexico-camb` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-minister-mcneill-speech-from-hunger-hotspots-to-po` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-more-people-protected-as-government-acts-to-make-m` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-procurement-compliance-oversight` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-public-electric-vehicle-charge-points-cyber-securi` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-russia-s-decision-to-veto-this-draft-resolution-ri` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-supporting-young-people-s-independence-through-tra` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-the-united-kingdom-remains-committed-to-working-wi` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-what-we-know-about-binance-s-6-3bn-hack` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/problem-william-gibson-and-the-trademarking-of-cyberspace` | removed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/process-a-right-to-buy-application` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/process-a-temporary-event-notice` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/procure-a-separately-procurable-engagement-with-an-aggregati` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/procurement-intelligence` | changed | 2026-11-01-research-is-not-scope | | `public-edge/propose-the-budget-and-council-tax-to-full-council` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/publish-an-accessibility-statement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/publish-an-algorithmic-transparency-record` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/publish-an-executive-decision-record` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/publish-and-maintain-privacy-notices` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/publish-government-procurement-card-spend` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/publish-members-allowances-for-the-year` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/publish-notice-of-a-key-decision-28-clear-days-ahead` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/publish-open-data` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/publish-spending-over-500` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/publish-the-agenda-and-reports-five-clear-days-ahead` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/publish-the-annual-pay-policy-statement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/publish-the-contracts-register` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/publish-the-draft-and-audited-statement-of-accounts` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/publish-the-forward-plan-twelve-weeks-ahead` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/publish-the-gambling-statement-of-principles` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/publish-the-gender-pay-gap` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/publish-the-market-position-statement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/publish-the-notice-of-election` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/publish-the-regulatory-enforcement-policy` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/publish-the-revised-register` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/publish-the-updated-constitution` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/publish-transparency-data` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/put-a-digital-imprint-on-campaign-material` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/raise-a-purchase-order-before-package-spend` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/raise-a-safeguarding-concern` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/raise-a-service-complaint` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/read-a-committee-pack-within-the-page-limit-and-lodge-questi` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/rebase-the-adult-social-care-budget-from-the-care-ledger` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/receive-and-decide-on-nominations` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/receive-and-handle-a-whistleblowing-disclosure` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/receive-and-publish-candidates-spending-returns` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/receive-quarterly-monitoring-with-a-demand-bridge` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/receive-the-auditor-s-annual-report-and-value-for-money-comm` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/receive-the-counter-fraud-annual-report` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/receive-the-monthly-evidence-pack-rated-on-evidence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/reconcile-access-and-contain-unexplained-accounts` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/reconcile-the-balance-sheet-exposure-statement-quarterly` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/record-a-gift-or-hospitality` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/record-a-price-exception-with-the-finance-position` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/record-an-exemption-from-contract-procedure-rules` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/recover-a-benefit-overpayment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/recover-unpaid-council-tax-through-a-liability-order` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/recruit-safely-with-dbs-checks` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/refer-a-band-challenge-to-the-valuation-office-agency` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/refer-higher-risk-building-work-to-the-building-safety-regul` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/register-a-birth` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/register-a-birth-death-or-marriage-notice` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/register-a-death-and-offer-tell-us-once` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/register-a-financial-interest` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/register-a-foreign-activity-arrangement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/register-a-political-party` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/register-as-a-consultant-lobbyist-and-file-returns` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/register-as-a-non-party-campaigner` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/register-disclosable-pecuniary-interests` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/register-to-vote` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/reimburse-operators-for-concessionary-travel` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/renew-public-services-network-compliance` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/reperform-a-closed-control-from-source` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/report-a-defence-security-incident` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/report-a-street-problem` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/report-complaints-and-compliments-annually` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/report-contract-performance-indicators` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/report-digital-service-performance` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/report-donations-and-loans` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/report-limited-or-no-assurance-to-the-audit-committee` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/report-on-a-council-owned-company-to-the-shareholder-committ` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/report-on-staffing-and-organisation-as-head-of-paid-service` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/report-on-the-robustness-of-estimates-and-adequacy-of-reserv` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/report-on-value-for-money-arrangements` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/represent-the-council-on-the-combined-authority-or-a-regiona` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/request-a-rate-uplift` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/request-a-review-of-a-housing-decision` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/request-and-track-audit-evidence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/request-exceptional-financial-support` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/require-a-cabinet-member-or-officer-to-attend` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/resolve-a-service-desk-incident` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/resolve-eligibility-for-the-general-power-of-competence` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/respond-to-a-child-protection-referral` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/respond-to-a-damp-mould-or-repair-report` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/respond-to-a-judicial-review-pre-action-letter` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/respond-to-a-parliamentary-petition` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/respond-to-a-statutory-recommendation-from-the-auditor` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/respond-to-an-area-send-inspection` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/respond-to-an-ico-decision-notice` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/respond-to-an-ombudsman-enquiry-and-carry-out-the-remedy` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/respond-to-an-ombudsman-finding` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/return-post-election-data-to-the-electoral-commission` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/review-a-homelessness-decision` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/review-a-looked-after-child-s-care-plan` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/review-local-air-quality` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/review-polling-districts-and-places` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/review-the-annual-governance-statement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/review-the-corporate-risk-register-quarterly` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/review-the-draft-statement-of-accounts` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/review-the-information-risk-register-as-senior-information-r` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/review-the-risk-management-framework` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-a-combined-authority-board-meeting` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-a-deep-dive-with-recorded-influence-on-direction` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-a-demand-control-cell-weekly` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-a-disaster-recovery-test` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-a-disciplinary-process` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-a-neighbourhood-planning-referendum` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-a-public-consultation-and-publish-the-response` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-a-public-consultation-and-report-the-results` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-a-recall-petition-as-petition-officer` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-a-standards-hearing-and-publish-the-decision` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-a-task-and-finish-review-and-secure-the-executive-respon` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-an-assurance-review-on-a-major-project` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-an-audit-engagement-from-terms-of-reference-to-report` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-monthly-budget-monitoring-with-budget-holders` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-monthly-payroll-and-report-to-hmrc` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-national-fraud-initiative-matches` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-security-awareness-and-phishing-tests` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-the-annual-canvass` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-the-annual-council-meeting` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-the-cross-committee-forum-with-other-chairs` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-the-decision-clock-on-stalled-reforms` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-the-make-or-buy-test-before-a-material-re-let` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-the-out-of-hours-duty-officer-rota` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-the-period-for-exercise-of-public-rights` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-the-public-inspection-period-for-the-accounts` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-the-reorganisation-programme-to-vesting-day` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/run-the-verification-and-count-and-declare-the-result` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/score-the-thirty-cyber-criteria-with-evidence-references` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/scrutinise-a-substantial-nhs-service-change` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/scrutinise-the-budget-proposals-before-council` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/scrutinise-the-combined-authority` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/scrutinise-the-community-safety-partnership` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/scrutinise-treasury-management` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/security-incident-response` | changed | 2026-11-01-research-is-not-scope | | `public-edge/seek-a-cabinet-office-spend-control-approval` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/seek-hm-treasury-approval-above-delegated-limits` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/select-a-candidate-and-authorise-the-party-description` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/send-apologies-and-keep-attendance-within-the-six-month-rule` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/separate-cashable-saving-cost-avoidance-and-deferred-demand` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/set-and-collect-council-housing-rents` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/set-the-annual-calendar-of-meetings` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/set-the-annual-work-programme` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/set-the-local-government-finance-settlement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/set-the-mayoral-budget-and-precept` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/set-the-precept-and-notify-the-billing-authority` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/set-the-scheme-of-executive-delegation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/set-up-an-instalment-plan-or-direct-debit` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/share-information-with-other-responders` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/sign-the-annual-governance-statement` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/sign-the-declaration-of-acceptance-of-office` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/speak-in-a-debate-and-check-the-official-report` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/spend-under-the-section-137-power-within-the-limit` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/submit-a-motion-to-full-council` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/submit-an-innovation-proposal-to-the-defence-and-security-ac` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/submit-planning-statistics-returns` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/submit-revenue-and-capital-returns-to-mhclg` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/submit-single-source-contract-reports` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/submit-the-business-rates-forecast-and-outturn-returns` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/submit-the-homelessness-statistics-return` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/submit-the-party-s-annual-accounts` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/submit-the-provider-cost-return` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/submit-the-tenant-satisfaction-measures` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/supplier-vetting` | changed | 2026-11-01-research-is-not-scope | | `public-edge/supply-the-register-to-an-entitled-person` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/support-a-care-leaver-with-a-pathway-plan` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/table-a-motion-or-seek-a-debate` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/table-a-written-question` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/take-a-bill-through-parliament` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/take-a-financial-monitoring-item-with-decision-grade-evidenc` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/take-a-homelessness-application-and-decide-the-duty` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/take-a-key-decision-on-the-one-page-standard` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/take-a-ward-casework-enquiry-through-to-reply` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/take-an-individual-portfolio-decision` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/take-an-urgent-decision-under-special-urgency` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/take-and-publish-the-minutes` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/take-and-publish-the-minutes-2` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/take-evidence-and-agree-a-select-committee-report` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/take-notice-of-marriage-or-civil-partnership` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/take-the-six-month-progress-report-on-statutory-recommendati` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/tender-and-award` | changed | 2026-11-01-research-is-not-scope | | `public-edge/track-audit-recommendations-to-owner-and-date-and-escalate-o` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/track-the-executive-s-responses-to-recommendations` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/transfer-staff-under-tupe` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/transfer-the-commissioning-toolkit-into-a-service` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/validate-a-planning-application` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/verify-weighing-and-measuring-equipment` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/vet-and-train-care-staff` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/view-the-public-outcome-scorecard` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/vote-in-a-division` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/warn-and-inform-the-public` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/webcast-and-allow-reporting-of-a-public-meeting` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/write-a-submission-to-a-minister` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/write-and-exercise-the-emergency-plan` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/write-off-an-irrecoverable-debt` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/write-the-officer-report-and-decide-under-delegation` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | | `public-edge/write-to-a-public-body-on-a-constituent-s-behalf` | changed | 2026-11-01-typed-steps-honest-outcomes-reference-flows | ### spec/changes/2026-10-31-money-provenance.json `spec/changes/2026-10-31-money-provenance.json` · sha256 `c7680815f0a592ac3249fd690da890d3134359700ea4c68ef252e5360e472970` ```json { "id": "2026-10-31-money-provenance", "request": "Owner, 31 October 2026 22:59Z: add money provenance to the scope and contract for the financial intelligence, confidence reporting and credit reference capabilities: trust in money comes from where it was obtained and what activity supports it, never from discounting face value; four typed assessments (provenance coverage, evidence reliability, financial-crime findings, investment risk) kept separate; signed provenance certificates with stated verification scope; exact pro-rata attribution so moving money cannot manufacture credibility; the financing graph with independent demand share to reveal circular funding; capital followed to delivered work and resources; a responsible lending decision that uses these with reasons, challenge and a named person. Decision tok:decision:01m6a4t0fxra5avfxf0pr2p692.", "admits": ["finance/money-provenance-certificate", "finance/provenance-coverage-report", "finance/financing-graph-independent-demand", "finance/capital-to-outcomes-ledger", "finance/responsible-lending-decision"], "decided_by": "X-X.TECH", "date": "2026-10-31" } ``` ### spec/changes/2026-11-01-reference-records-pass5.json `spec/changes/2026-11-01-reference-records-pass5.json` · sha256 `0e258a2e131fba670c489005ab890c3e9d28d582a7ebd105b162f9f82523d06d` ```json { "id": "2026-11-01-reference-records-pass5", "request": "Owner, 1 November 2026 04:10Z to 04:14Z: every layout and screen is mapped to a reference (Mobbin flows: Notion Home, ChatGPT, Google Drive, Spotify, Revolut Business, Base44 V7BRO), the only difference being the estate's added controls or journeys. The reference records written in research/ during finishing-loop passes 1 to 5 (research/judiverse/REFERENCE-*.md, research/finance/REFERENCE-CARDS.md, research/edge-consortium and public-edge REFERENCE-STUDIO.md) are carried into these twelve frozen briefs by tools/match.py as screen pattern references and the typed fields the references show; nothing in scope is added or removed. Found by the clean-clone make (tools/reproduce.sh) on 1 Nov 2026 09:4xZ. Clinic's person-held health record is carried the same way (its reference flows were recorded in pass 4; found by the make of 1 Nov 10:1xZ).", "admits": [ "clinic/person-held-health-record", "finance/business-roles-and-approvals", "judiverse/agent-builder", "judiverse/calendar-and-scheduling", "judiverse/contacts-directory-and-search", "judiverse/deep-research", "judiverse/documents-and-forms", "judiverse/jude-assistant", "judiverse/judebox-music", "judiverse/mail-and-agent-mailboxes", "judiverse/messages-and-chat", "judiverse/procedure-compilation", "judiverse/tasks-and-workflows" ], "decided_by": "X-X.TECH (Owner order; recorded by Claude-Uriel 21:3531102026-1)", "date": "2026-11-01" } ``` ### spec/changes/2026-11-01-research-is-not-scope.json `spec/changes/2026-11-01-research-is-not-scope.json` · sha256 `6d9db088c530db59d90dbcb9c0bec9212d99044b3818508d751ecbd02380016f` ```json { "id": "2026-11-01-research-is-not-scope", "request": "Owner, 1 November 2026 (11:29Z to 11:43Z): one estate built from one source, nothing hand-drawn or blindly imported; the specification drives working applications. Research is not scope: the 4,132 briefs that tools/match.py (parity clusters), tools/problems.py (problem records) and tools/carry.py (findings) had generated from research leave spec/briefs and spec/units and are kept as research/records//.json; on the 1,616 capability briefs the 8,975 operations, controls, entities and fields that tools/match.py had placed by word score become source-linked research_refs. tools/expand.py no longer invents 'Continue to' or 'Open' controls or an operator journey the brief does not declare: a journey step's own operation leads to the next step, and anything still unreachable is recorded as a gap on the unit. tools/research_not_scope.py applies the rule on every make.", "admits": [ "clinic/findings-competitors-regulators-reuse-0-the-ten-things-to-know-first", "clinic/findings-competitors-regulators-reuse-1-1-digital-front-doors-and-prim", "clinic/findings-competitors-regulators-reuse-1-2-ai-clinical-intelligence-ven", "clinic/findings-competitors-regulators-reuse-1-3-clinical-decision-support-an", "clinic/findings-competitors-regulators-reuse-1-4-retinal-ai-vendors", "clinic/findings-competitors-regulators-reuse-2-1-united-kingdom", "clinic/findings-competitors-regulators-reuse-2-3-united-states", "clinic/findings-competitors-regulators-reuse-3-agent-assurance-gaps-in-health", "clinic/findings-competitors-regulators-reuse-4-1-standards-and-specifications", "clinic/findings-competitors-regulators-reuse-4-3-nhs-national-services-reuse", "clinic/findings-competitors-regulators-reuse-4-reusable-systems-open-standard", "clinic/findings-competitors-regulators-reuse-5-open-questions-unknown-to-reso", "clinic/findings-competitors-regulators-reuse-edge-clinic-competitors-regulato", "clinic/findings-competitors-regulators-reuse-how-to-read-the-labels", "clinic/findings-mobbin-flows-coverage-what-mobbin-held-and-what-it-did-not", "clinic/findings-mobbin-flows-edge-clinic-research-real-product-flows-from-mob", "clinic/findings-mobbin-flows-symptom-checking", "clinic/findings-mobbin-flows-test-results", "clinic/findings-mobbin-flows-wearables-scores-and-data-sharing", "clinic/findings-refero-flows-edge-clinic-research-real-product-flows-from-ref", "clinic/findings-refero-flows-task-access-review-permissions-audit-log", "clinic/findings-refero-flows-task-access-review-permissions-audit-log-2", "clinic/findings-refero-flows-task-account-closure", "clinic/findings-refero-flows-task-ai-chat-assistant", "clinic/findings-refero-flows-task-apply-for-loan-credit", "clinic/findings-refero-flows-task-book-gp-appointment", "clinic/findings-refero-flows-task-book-gp-appointment-2", "clinic/findings-refero-flows-task-book-gp-appointment-3", "clinic/findings-refero-flows-task-book-gp-appointment-4", "clinic/findings-refero-flows-task-budget-spending-insights", "clinic/findings-refero-flows-task-budget-spending-insights-2", "clinic/findings-refero-flows-task-business-account-onboarding-company-verific", "clinic/findings-refero-flows-task-calendar-scheduling-meeting", "clinic/findings-refero-flows-task-care-plan", "clinic/findings-refero-flows-task-care-plan-2", "clinic/findings-refero-flows-task-care-plan-3", "clinic/findings-refero-flows-task-care-plan-4", "clinic/findings-refero-flows-task-care-plan-5", "clinic/findings-refero-flows-task-case-management", "clinic/findings-refero-flows-task-code-repository-pull-request", "clinic/findings-refero-flows-task-compliance-dashboard-audit", "clinic/findings-refero-flows-task-contract-management", "clinic/findings-refero-flows-task-doctor-video-consultation", "clinic/findings-refero-flows-task-doctor-video-consultation-2", "clinic/findings-refero-flows-task-eye-test-vision", "clinic/findings-refero-flows-task-health-onboarding-profile", "clinic/findings-refero-flows-task-health-onboarding-profile-2", "clinic/findings-refero-flows-task-health-onboarding-profile-3", "clinic/findings-refero-flows-task-health-onboarding-profile-4", "clinic/findings-refero-flows-task-health-record-medical-history", "clinic/findings-refero-flows-task-health-record-medical-history-2", "clinic/findings-refero-flows-task-insurance-claim-health", "clinic/findings-refero-flows-task-learning-management-course-progress", "clinic/findings-refero-flows-task-medication-reminder", "clinic/findings-refero-flows-task-medication-reminder-2", "clinic/findings-refero-flows-task-medication-reminder-3", "clinic/findings-refero-flows-task-medication-reminder-4", "clinic/findings-refero-flows-task-medication-reminder-5", "clinic/findings-refero-flows-task-mental-health-questionnaire-mood-check", "clinic/findings-refero-flows-task-mental-health-questionnaire-mood-check-2", "clinic/findings-refero-flows-task-prescription-refill-repeat", "clinic/findings-refero-flows-task-referral-to-specialist", "clinic/findings-refero-flows-task-savings-pot-goal", "clinic/findings-refero-flows-task-symptom-checker", "clinic/findings-refero-flows-task-test-results", "clinic/findings-refero-flows-task-test-results-2", "clinic/findings-refero-flows-task-therapy-booking", "clinic/findings-refero-flows-task-triage-questionnaire", "clinic/findings-refero-flows-task-wellbeing-activities-community", "clinic/findings-reversibility-how-the-register-grows-deterministically", "clinic/findings-reversibility-impact-reversal-for-conditions-that-are-part-of", "clinic/findings-reversibility-research-targets-conditions-still-treated-as-ir", "clinic/findings-reversibility-reversals-with-measured-evidence", "clinic/findings-reversibility-reversibility-in-health-the-register", "clinic/findings-social-prescribing-and-detection-0-the-twelve-things-to-know", "clinic/findings-social-prescribing-and-detection-1-1-scale", "clinic/findings-social-prescribing-and-detection-1-2-what-link-workers-actual", "clinic/findings-social-prescribing-and-detection-1-3-nhs-england-and-governme", "clinic/findings-social-prescribing-and-detection-1-4-outcome-and-cost-evidenc", "clinic/findings-social-prescribing-and-detection-1-5-digital-platforms-for-so", "clinic/findings-social-prescribing-and-detection-1-uk-social-prescribing", "clinic/findings-social-prescribing-and-detection-2-1-the-services", "clinic/findings-social-prescribing-and-detection-2-2-babylon-health-why-it-fa", "clinic/findings-social-prescribing-and-detection-2-3-accuracy-studies-with-nu", "clinic/findings-social-prescribing-and-detection-2-4-what-makes-triage-safe", "clinic/findings-social-prescribing-and-detection-2-symptom-checkers-and-digit", "clinic/findings-social-prescribing-and-detection-3-1-the-nhs-diabetic-eye-scr", "clinic/findings-social-prescribing-and-detection-3-2-atrial-fibrillation-af", "clinic/findings-social-prescribing-and-detection-3-3-hypertension-high-blood", "clinic/findings-social-prescribing-and-detection-3-4-diabetes", "clinic/findings-social-prescribing-and-detection-3-5-cancer-symptoms", "clinic/findings-social-prescribing-and-detection-3-6-mental-health-and-valida", "clinic/findings-social-prescribing-and-detection-3-7-cost-of-late-diagnosis", "clinic/findings-social-prescribing-and-detection-4-1-what-is-clinically-valid", "clinic/findings-social-prescribing-and-detection-4-2-nhs-plans-and-pilots", "clinic/findings-social-prescribing-and-detection-4-3-portable-and-smartphone", "clinic/findings-social-prescribing-and-detection-4-4-the-false-positive-probl", "clinic/findings-social-prescribing-and-detection-5-physical-presence-places-a", "clinic/findings-social-prescribing-and-detection-6-1-medical-device-regulatio", "clinic/findings-social-prescribing-and-detection-6-2-nhs-assurance", "clinic/findings-social-prescribing-and-detection-6-3-health-data-protection", "clinic/findings-social-prescribing-and-detection-6-4-emergency-signposting-co", "clinic/findings-social-prescribing-and-detection-7-competitors-and-the-gap", "clinic/findings-social-prescribing-and-detection-8-the-costliest-burdens-edge", "clinic/findings-social-prescribing-and-detection-8-the-costliest-burdens-edge-2", "clinic/findings-social-prescribing-and-detection-9-gaps-and-unknowns-what-i-c", "clinic/findings-social-prescribing-and-detection-bare-minimum-vs-ambitious-pr", "clinic/findings-social-prescribing-and-detection-edge-clinic-social-prescribi", "clinic/findings-social-prescribing-and-detection-how-to-read-the-labels", "clinic/parity-16-june-2025-clinic", "clinic/parity-2-august-2026-clinic", "clinic/parity-222-assign-close-friends-clinic", "clinic/parity-26-march-2025-clinic", "clinic/parity-27-july-2026-clinic", "clinic/parity-29-july-2026-clinic", "clinic/parity-6-april-2026-clinic", "clinic/parity-6-january-2026-clinic", "clinic/parity-9-december-2026-clinic", "clinic/parity-ada-dizziness-results-and-condition-detail-clinic", "clinic/parity-ada-manage-allergies-clinic", "clinic/parity-ada-neck-mobility-question-and-clarification-clinic", "clinic/parity-ada-select-primary-symptom-clinic", "clinic/parity-ada-sign-in-to-start-assessment-clinic", "clinic/parity-ada-symptom-search-and-selection-clinic", "clinic/parity-adidas-running-clinic", "clinic/parity-adobe-photoshop-color-balance-adjustment-clinic", "clinic/parity-adobe-photoshop-layer-blend-and-opacity-edit-clinic", "clinic/parity-alan-clinic", "clinic/parity-alo-class-detail-and-playback-clinic", "clinic/parity-alo-class-finder-browsing-clinic", "clinic/parity-alo-prenatal-program-detail-clinic", "clinic/parity-alo-refer-a-friend-sharing-clinic", "clinic/parity-alo-wellness-club-registration-and-personalization-clinic", "clinic/parity-anam-live-conversation-session-clinic", "clinic/parity-apple-image-playground-report-generated-image-clinic", "clinic/parity-apple-maps-place-details-and-street-view-clinic", "clinic/parity-bahmni-bahmni", "clinic/parity-binge-viewing-statistics-breakdown-clinic", "clinic/parity-blackbird-referral-invite-clinic", "clinic/parity-blackbird-referral-share-fly-reward-clinic", "clinic/parity-brink-enable-health-integration-clinic", "clinic/parity-butter-share-referral-modal-clinic", "clinic/parity-byborre-full-repeat-sample-checkout-clinic", "clinic/parity-care-caregiver-review-submission-clinic", "clinic/parity-care-pet-care-caregiver-search-clinic", "clinic/parity-care-pet-caregiver-search-clinic", "clinic/parity-chance-ai-visual-product-lookup-clinic", "clinic/parity-chance-ai-visual-snap-interpreter-clinic", "clinic/parity-chatgpt-delete-saved-memory-clinic", "clinic/parity-cushion-direct-message-conversation-clinic", "clinic/parity-dailyart-magazine-image-based-quiz-with-results-clinic", "clinic/parity-deezer-report-lyrics-problem-clinic", "clinic/parity-delphi-add-user-memory-clinic", "clinic/parity-delphi-digital-persona-voice-call-clinic", "clinic/parity-doctronic-ai-healthcare-subscription-signup-clinic", "clinic/parity-doctronic-ai-sleep-concern-assessment-clinic", "clinic/parity-duolingo-practice-word-question-clinic", "clinic/parity-elevenmusic-invite-friends-referral-clinic", "clinic/parity-elevenmusic-track-published-confirmation-clinic", "clinic/parity-emis-ownership-reported-clinic", "clinic/parity-emis-web-emis-web", "clinic/parity-end-2026-mhra-clinic", "clinic/parity-epic-ehr-epic-ehr", "clinic/parity-epic-ehr-epic-ehr-part-2", "clinic/parity-eyeart-clinic", "clinic/parity-fal-search-recent-endpoints-clinic", "clinic/parity-farfetch-refer-a-friend-invite-clinic", "clinic/parity-foodllama-nutrition-and-body-stats-review-clinic", "clinic/parity-foodllama-nutrition-insight-completion-clinic", "clinic/parity-fuse-adjust-slippage-tolerance-clinic", "clinic/parity-go-club-plan-review-daily-reflection-clinic", "clinic/parity-gofundme-fundraiser-details-and-organization-follow-clinic", "clinic/parity-google-gemini-voice-query-to-suggestions-clinic", "clinic/parity-google-translate-lens-text-translation-clinic", "clinic/parity-ground-news-invite-friends-referral-sharing-clinic", "clinic/parity-hapi-fhir-clinic", "clinic/parity-has-the-government-clinic", "clinic/parity-ikea-chatbot-post-conversation-survey-clinic", "clinic/parity-instacart-configure-out-of-stock-substitution-clinic", "clinic/parity-instacart-referral-invite-clinic", "clinic/parity-introspect-friend-decision-response-clinic", "clinic/parity-july-2027-hipaa-clinic", "clinic/parity-june-2027-uk-clinic", "clinic/parity-lex-clinic", "clinic/parity-longevity-deck-comment-interactions-clinic", "clinic/parity-longevity-deck-first-launch-introduction-clinic", "clinic/parity-lovable-generated-landing-page-review-clinic", "clinic/parity-lovable-plan-review-and-approval-clinic", "clinic/parity-lovable-referral-invite-modal-clinic", "clinic/parity-manus-send-referral-invite-clinic", "clinic/parity-medplum-clinic", "clinic/parity-melius-send-referral-invitation-clinic", "clinic/parity-melius-skincare-hero-image-generation-clinic", "clinic/parity-meta-ai-daily-trivia-quiz-clinic", "clinic/parity-meta-ai-save-recipe-instructions-clinic", "clinic/parity-meta-ai-weekend-activity-suggestions-clinic", "clinic/parity-mymind-visual-reference-search-clinic", "clinic/parity-n8n-ai-summary-automation-clinic", "clinic/parity-n8n-evaluation-setup-and-debugging-clinic", "clinic/parity-napper-baby-schedule-invite-clinic", "clinic/parity-napper-daily-question-poll-clinic", "clinic/parity-new-balance-coming-soon-alert-signup-clinic", "clinic/parity-new-balance-post-purchase-survey-clinic", "clinic/parity-nextdoor-clinic", "clinic/parity-nextdoor-local-hobby-recommendations-clinic", "clinic/parity-nextdoor-rate-ask-answer-clinic", "clinic/parity-nhs-app-nhs-app", "clinic/parity-nlm-rxnav-drug-interaction-api-clinic", "clinic/parity-ohdsi-webapi-atlas-clinic", "clinic/parity-on-order-details-review-clinic", "clinic/parity-openehr-clinic", "clinic/parity-openemr-openemr", "clinic/parity-openmrs-o3-openmrs", "clinic/parity-oracle-health-ehr-cerner-oracle-health-ehr", "clinic/parity-orbit-notifications-settings-and-test-clinic", "clinic/parity-oura-clinic", "clinic/parity-patient-access-patient-access", "clinic/parity-paypal-open-a-transaction-dispute-clinic", "clinic/parity-pinterest-image-text-editor-clinic", "clinic/parity-podimo-live-poll-participation-clinic", "clinic/parity-podimo-search-episodes-in-show-clinic", "clinic/parity-polaroid-gift-camera-recommendation-quiz-clinic", "clinic/parity-pool-place-details-to-directions-clinic", "clinic/parity-pool-refero-mcp-search-clinic", "clinic/parity-poppy-recommended-connection-setup-clinic", "clinic/parity-poppy-support-and-privacy-settings-clinic", "clinic/parity-preply-placement-test-and-speaking-assessment-clinic", "clinic/parity-preply-pre-class-device-check-clinic", "clinic/parity-preply-pre-lesson-device-check-clinic", "clinic/parity-preply-vocab-self-check-and-results-clinic", "clinic/parity-product-liability-directive-clinic", "clinic/parity-queue-subscription-cards-list-clinic", "clinic/parity-rad-weather-ambient-weather-connection-clinic", "clinic/parity-raycast-question-search-and-select-clinic", "clinic/parity-resend-test-email-from-broadcast-composer-clinic", "clinic/parity-riverside-test-recording-clinic", "clinic/parity-rork-model-catalog-search-clinic", "clinic/parity-seed-personalized-supplement-recommendation-clinic", "clinic/parity-seed-supplement-purchase-and-cart-review-clinic", "clinic/parity-shazam-location-based-event-filtering-clinic", "clinic/parity-sky-guide-tonight-s-best-observation-mode-selector-clinic", "clinic/parity-stardust-archetype-quiz-reveal-clinic", "clinic/parity-stardust-custom-symptom-creation-clinic", "clinic/parity-stardust-cycle-day-symptom-logging-clinic", "clinic/parity-stardust-topic-discovery-feelings-search-clinic", "clinic/parity-subo-add-subscription-clinic", "clinic/parity-teal-work-style-assessment-clinic", "clinic/parity-the-outsiders-first-launch-setup-clinic", "clinic/parity-tidal-artist-biography-clinic", "clinic/parity-tidal-track-report-reason-selection-clinic", "clinic/parity-tinder-safety-center-quiz-clinic", "clinic/parity-train-fitness-feed-sentiment-prompt-clinic", "clinic/parity-train-fitness-workout-detail-clinic", "clinic/parity-typeform-ai-lead-qualification-form-builder-clinic", "clinic/parity-typeform-ai-question-generation-clinic", "clinic/parity-vinyls-turntable-calibration-clinic", "clinic/parity-vocabulary-custom-mix-builder-clinic", "clinic/parity-vocabulary-vocabulary-practice-drill-clinic", "clinic/parity-voicenotes-voice-query-to-notes-clinic", "clinic/parity-wabi-popular-category-browse-clinic", "clinic/parity-wabi-submit-personal-analysis-clinic", "clinic/parity-weather-mini-trip-planning-and-save-clinic", "clinic/parity-whatsapp-active-call-controls-clinic", "clinic/parity-whatsapp-create-and-review-poll-clinic", "clinic/parity-whatsapp-group-voice-conversation-clinic", "clinic/parity-whatsapp-schedule-call-and-invite-participants-clinic", "clinic/parity-whatsapp-whatsapp-call-transition-clinic", "clinic/parity-willow-friend-referral-invite-clinic", "clinic/parity-willow-refer-a-friend-clinic", "clinic/parity-willow-referral-card-share-clinic", "clinic/parity-willow-referral-card-sharing-clinic", "clinic/parity-wispr-flow-action-button-dictation-setup-clinic", "clinic/parity-wispr-flow-friend-referral-invitation-clinic", "clinic/problem-annual-offshore-statistics-and-regulatory-activity", "clinic/problem-medicare-will-pay-50k-for-your-heart-attack-it-sho", "clinic/problem-students-tackle-public-policy-challenges-with-supp", "clinic/problem-working-for-the-national-cyber-force-ncf-case-stud", "edge-consortium/findings-competitors-regulators-reuse-2-5-education-and-research-platf", "edge-consortium/findings-competitors-regulators-reuse-2-competitors", "edge-consortium/findings-competitors-regulators-reuse-3-5-education-and-research-engla", "edge-consortium/findings-competitors-regulators-reuse-edge-public-consortium-competito", "edge-consortium/findings-governed-intelligence-degree-assessment-real-work-never-recal", "edge-consortium/findings-governed-intelligence-degree-awarding-it", "edge-consortium/findings-governed-intelligence-degree-governed-intelligence-the-degree", "edge-consortium/findings-governed-intelligence-degree-level-4-bsc-year-1", "edge-consortium/findings-governed-intelligence-degree-level-5-bsc-year-2", "edge-consortium/findings-governed-intelligence-degree-level-6-bsc-year-3", "edge-consortium/findings-governed-intelligence-degree-level-7-msc", "edge-consortium/findings-governed-intelligence-degree-partner-institutions", "edge-consortium/findings-governed-intelligence-degree-scope-discipline-and-bottleneck", "edge-consortium/findings-governed-intelligence-degree-set-against-leading-programmes-i", "edge-consortium/findings-governed-intelligence-degree-set-against-leading-programmes-i-2", "edge-consortium/findings-governed-intelligence-degree-specialising-only-after-the-whol", "edge-consortium/findings-governed-intelligence-degree-specialising-only-after-the-whol-2", "edge-consortium/findings-governed-intelligence-degree-studying-it-judiverse-included-r", "edge-consortium/findings-governed-intelligence-degree-the-research-it-is-built-on", "edge-consortium/findings-mobbin-flows-adding-and-onboarding-suppliers", "edge-consortium/findings-mobbin-flows-edge-public-consortium-research-real-product-flo", "edge-consortium/findings-mobbin-flows-identity-checks-and-sign-in", "edge-consortium/findings-mobbin-flows-incident-response", "edge-consortium/findings-mobbin-flows-learning-and-course-enrolment", "edge-consortium/findings-mobbin-flows-roles-and-permissions", "edge-consortium/findings-refero-flows-edge-public-consortium-research-real-product-flo", "edge-consortium/findings-refero-flows-task-access-review-permissions-audit-log-1-3", "edge-consortium/findings-refero-flows-task-access-review-permissions-audit-log-4", "edge-consortium/findings-refero-flows-task-access-review-permissions-audit-log-6", "edge-consortium/findings-refero-flows-task-ai-agent-setup-permissions-1-2", "edge-consortium/findings-refero-flows-task-ai-agent-setup-permissions-2-2", "edge-consortium/findings-refero-flows-task-ai-chat-assistant-1-3", "edge-consortium/findings-refero-flows-task-api-keys-developer-settings-1-2", "edge-consortium/findings-refero-flows-task-apply-for-loan-credit-1-3", "edge-consortium/findings-refero-flows-task-apply-for-loan-credit-2", "edge-consortium/findings-refero-flows-task-approval-workflow-2-2", "edge-consortium/findings-refero-flows-task-approval-workflow-3-2", "edge-consortium/findings-refero-flows-task-book-gp-appointment-1-3", "edge-consortium/findings-refero-flows-task-budget-spending-insights-1-2", "edge-consortium/findings-refero-flows-task-budget-spending-insights-2-2", "edge-consortium/findings-refero-flows-task-business-account-onboarding-company-verific-1-3", "edge-consortium/findings-refero-flows-task-business-account-onboarding-company-verific-2", "edge-consortium/findings-refero-flows-task-business-account-onboarding-company-verific-3", "edge-consortium/findings-refero-flows-task-business-account-onboarding-company-verific-4", "edge-consortium/findings-refero-flows-task-buy-crypto", "edge-consortium/findings-refero-flows-task-calendar-scheduling-meeting-1-3", "edge-consortium/findings-refero-flows-task-care-plan-1-2", "edge-consortium/findings-refero-flows-task-care-plan-2-2", "edge-consortium/findings-refero-flows-task-case-management-1-2", "edge-consortium/findings-refero-flows-task-case-management-2", "edge-consortium/findings-refero-flows-task-case-management-3", "edge-consortium/findings-refero-flows-task-case-management-4", "edge-consortium/findings-refero-flows-task-case-management-5", "edge-consortium/findings-refero-flows-task-case-management-7", "edge-consortium/findings-refero-flows-task-code-repository-pull-request-1-3", "edge-consortium/findings-refero-flows-task-compliance-dashboard-audit-2", "edge-consortium/findings-refero-flows-task-contract-management-1-3", "edge-consortium/findings-refero-flows-task-contract-management-2", "edge-consortium/findings-refero-flows-task-contract-management-3", "edge-consortium/findings-refero-flows-task-course-enrolment-university", "edge-consortium/findings-refero-flows-task-course-enrolment-university-2", "edge-consortium/findings-refero-flows-task-course-enrolment-university-3", "edge-consortium/findings-refero-flows-task-course-enrolment-university-4", "edge-consortium/findings-refero-flows-task-course-enrolment-university-5", "edge-consortium/findings-refero-flows-task-course-enrolment-university-6", "edge-consortium/findings-refero-flows-task-course-enrolment-university-7", "edge-consortium/findings-refero-flows-task-credit-score-report", "edge-consortium/findings-refero-flows-task-crm-contacts-pipeline-1-2", "edge-consortium/findings-refero-flows-task-crm-contacts-pipeline-2-2", "edge-consortium/findings-refero-flows-task-dispute-transaction-chargeback", "edge-consortium/findings-refero-flows-task-doctor-video-consultation-1-3", "edge-consortium/findings-refero-flows-task-document-editor-collaboration-1-2", "edge-consortium/findings-refero-flows-task-email-inbox-1-2", "edge-consortium/findings-refero-flows-task-expense-management-receipts", "edge-consortium/findings-refero-flows-task-eye-test-vision-1-2", "edge-consortium/findings-refero-flows-task-fleet-management-2", "edge-consortium/findings-refero-flows-task-fleet-management-3", "edge-consortium/findings-refero-flows-task-fleet-management-5", "edge-consortium/findings-refero-flows-task-fraud-alert-suspicious-transaction-1-2", "edge-consortium/findings-refero-flows-task-government-service-identity-verification", "edge-consortium/findings-refero-flows-task-government-service-identity-verification-4", "edge-consortium/findings-refero-flows-task-grant-application-1-2", "edge-consortium/findings-refero-flows-task-grant-application-2", "edge-consortium/findings-refero-flows-task-grant-application-3", "edge-consortium/findings-refero-flows-task-grant-application-4", "edge-consortium/findings-refero-flows-task-grant-application-5", "edge-consortium/findings-refero-flows-task-grant-application-6", "edge-consortium/findings-refero-flows-task-grant-application-7", "edge-consortium/findings-refero-flows-task-health-onboarding-profile-1-3", "edge-consortium/findings-refero-flows-task-health-record-medical-history-1-3", "edge-consortium/findings-refero-flows-task-insurance-claim-health-1-3", "edge-consortium/findings-refero-flows-task-inventory-warehouse-2", "edge-consortium/findings-refero-flows-task-inventory-warehouse-3", "edge-consortium/findings-refero-flows-task-inventory-warehouse-4", "edge-consortium/findings-refero-flows-task-inventory-warehouse-5", "edge-consortium/findings-refero-flows-task-inventory-warehouse-6", "edge-consortium/findings-refero-flows-task-investing-buy-stock", "edge-consortium/findings-refero-flows-task-learning-management-course-progress-1-2", "edge-consortium/findings-refero-flows-task-learning-management-course-progress-2", "edge-consortium/findings-refero-flows-task-learning-management-course-progress-3", "edge-consortium/findings-refero-flows-task-learning-management-course-progress-4", "edge-consortium/findings-refero-flows-task-learning-management-course-progress-5", "edge-consortium/findings-refero-flows-task-medication-reminder-1-2", "edge-consortium/findings-refero-flows-task-open-bank-account-identity-verification", "edge-consortium/findings-refero-flows-task-pay-invoice-bill", "edge-consortium/findings-refero-flows-task-payroll", "edge-consortium/findings-refero-flows-task-policy-management-approval", "edge-consortium/findings-refero-flows-task-policy-management-approval-2", "edge-consortium/findings-refero-flows-task-policy-management-approval-3", "edge-consortium/findings-refero-flows-task-prescription-refill-repeat-1-2", "edge-consortium/findings-refero-flows-task-procurement-tender-bid-2", "edge-consortium/findings-refero-flows-task-research-collaboration", "edge-consortium/findings-refero-flows-task-research-collaboration-2", "edge-consortium/findings-refero-flows-task-risk-register-2", "edge-consortium/findings-refero-flows-task-security-incident-alert-1-2", "edge-consortium/findings-refero-flows-task-security-incident-alert-3", "edge-consortium/findings-refero-flows-task-shipment-tracking-logistics-3", "edge-consortium/findings-refero-flows-task-shipment-tracking-logistics-4", "edge-consortium/findings-refero-flows-task-supplier-onboarding-vetting", "edge-consortium/findings-refero-flows-task-supplier-onboarding-vetting-2", "edge-consortium/findings-refero-flows-task-two-factor-security-device", "edge-consortium/findings-refero-flows-task-virtual-card-freeze-card", "edge-consortium/findings-refero-flows-task-vulnerability-management", "edge-consortium/findings-sources-2026-10-31-sources-given-by-the-owner-31-october-2026", "edge-consortium/parity-agreement-policy-public-consortium", "edge-consortium/parity-brightspace-brightspace", "edge-consortium/parity-brilliant-brilliant", "edge-consortium/parity-canvas-lms-canvas-lms", "edge-consortium/parity-canvas-lms-canvas-lms-part-2", "edge-consortium/parity-canvas-lms-canvas-lms-part-3", "edge-consortium/parity-canvas-lms-canvas-lms-part-4", "edge-consortium/parity-canvas-lms-canvas-lms-part-5", "edge-consortium/parity-canvas-lms-public-consortium", "edge-consortium/parity-cloudflare-zero-trust-cloudflare-zero-trust", "edge-consortium/parity-companies-house-companies-house", "edge-consortium/parity-coursera-for-business-coursera-for-business", "edge-consortium/parity-credo-ai-credo-ai", "edge-consortium/parity-crowdstrike-crowdstrike", "edge-consortium/parity-darktrace-darktrace", "edge-consortium/parity-datadog-datadog", "edge-consortium/parity-dependabot-dependabot", "edge-consortium/parity-drata-drata", "edge-consortium/parity-elastic-elastic", "edge-consortium/parity-fixmystreet-fixmystreet", "edge-consortium/parity-futurelearn-futurelearn", "edge-consortium/parity-github-advanced-security-github-advanced-security", "edge-consortium/parity-gov-uk-forms-gov-uk-forms", "edge-consortium/parity-gov-uk-notify-gov-uk-notify", "edge-consortium/parity-gov-uk-pay-gov-uk-pay", "edge-consortium/parity-grafana-grafana", "edge-consortium/parity-honeycomb-honeycomb", "edge-consortium/parity-human-control-is-a-principle-public-consortium", "edge-consortium/parity-hyperproof-hyperproof", "edge-consortium/parity-ibm-watsonx-governance-ibm-watsonx-governance", "edge-consortium/parity-innovate-uk-funding-service-innovate-uk-funding-service", "edge-consortium/parity-innovate-uk-ukri-funding-service-innovate-uk-ukri-funding-servi", "edge-consortium/parity-logicgate-logicgate", "edge-consortium/parity-metricstream-metricstream", "edge-consortium/parity-microsoft-defender-microsoft-defender", "edge-consortium/parity-mitre-d3fend-public-consortium", "edge-consortium/parity-moodle-lms-moodle-lms", "edge-consortium/parity-onetrust-onetrust", "edge-consortium/parity-opentelemetry-opentelemetry", "edge-consortium/parity-palantir-apollo-apollo", "edge-consortium/parity-palantir-apollo-apollo-part-2", "edge-consortium/parity-palantir-foundry-foundry", "edge-consortium/parity-palantir-foundry-foundry-part-2", "edge-consortium/parity-palantir-gotham-gotham", "edge-consortium/parity-palantir-technologies-apollo-apollo", "edge-consortium/parity-palantir-technologies-apollo-apollo-part-2", "edge-consortium/parity-palantir-technologies-apollo-apollo-part-3", "edge-consortium/parity-palantir-technologies-apollo-apollo-part-4", "edge-consortium/parity-palantir-technologies-apollo-apollo-part-5", "edge-consortium/parity-palantir-technologies-apollo-apollo-part-6", "edge-consortium/parity-palantir-technologies-foundry-foundry", "edge-consortium/parity-palantir-technologies-foundry-foundry-part-2", "edge-consortium/parity-palantir-technologies-foundry-foundry-part-3", "edge-consortium/parity-palantir-technologies-foundry-foundry-part-4", "edge-consortium/parity-palantir-technologies-foundry-foundry-part-5", "edge-consortium/parity-palantir-technologies-gotham-gotham", "edge-consortium/parity-prisma-cloud-prisma-cloud", "edge-consortium/parity-prometheus-prometheus", "edge-consortium/parity-proposed-pilot-public-consortium", "edge-consortium/parity-qualys-qualys", "edge-consortium/parity-secureframe-secureframe", "edge-consortium/parity-sentinelone-sentinelone", "edge-consortium/parity-servicenow-grc-servicenow-grc", "edge-consortium/parity-snyk-snyk", "edge-consortium/parity-sonarqube-sonarqube", "edge-consortium/parity-splunk-splunk", "edge-consortium/parity-sprinto-sprinto", "edge-consortium/parity-tenable-tenable", "edge-consortium/parity-twingate-twingate", "edge-consortium/parity-udemy-business-udemy-business", "edge-consortium/parity-vanta-public-consortium", "edge-consortium/parity-vanta-vanta", "edge-consortium/parity-wiz-public-consortium", "edge-consortium/parity-wiz-wiz", "edge-consortium/parity-workiva-workiva", "edge-consortium/parity-zscaler-zscaler", "edge-consortium/problem-brown-student-maps-the-hidden-global-labor-powerin", "edge-consortium/problem-research-agentic-ai-and-consumers", "finance/findings-assurance-and-reuse-0-1-evidence-labels", "finance/findings-assurance-and-reuse-0-2-how-the-facts-were-checked", "finance/findings-assurance-and-reuse-0-3-acronyms-first-use", "finance/findings-assurance-and-reuse-0-3-acronyms-first-use-2", "finance/findings-assurance-and-reuse-1-1-1-summary-table", "finance/findings-assurance-and-reuse-1-1-1-summary-table-2", "finance/findings-assurance-and-reuse-1-1-2-what-this-means-for-edge-finance", "finance/findings-assurance-and-reuse-1-2-agent-identity-and-authority", "finance/findings-assurance-and-reuse-1-2-agent-identity-and-authority-2", "finance/findings-assurance-and-reuse-1-3-agent-protocols", "finance/findings-assurance-and-reuse-1-4-open-banking-and-api-security-standar", "finance/findings-assurance-and-reuse-1-5-1-implication-for-edge-finance-assess", "finance/findings-assurance-and-reuse-1-5-what-is-missing-for-agents-to-be-safe", "finance/findings-assurance-and-reuse-1-5-what-is-missing-for-agents-to-be-safe-2", "finance/findings-assurance-and-reuse-2-1-ledgers", "finance/findings-assurance-and-reuse-2-10-agent-assurance-tooling-to-reuse", "finance/findings-assurance-and-reuse-2-11-licence-flags-what-blocks-or-constra", "finance/findings-assurance-and-reuse-2-11-licence-flags-what-blocks-or-constra-2", "finance/findings-assurance-and-reuse-2-2-core-banking", "finance/findings-assurance-and-reuse-2-3-payments", "finance/findings-assurance-and-reuse-2-4-open-banking-and-api-security", "finance/findings-assurance-and-reuse-2-5-kyc-kyb-and-identity-verification-ide", "finance/findings-assurance-and-reuse-2-6-1-official-sanctions-sources-free-pri", "finance/findings-assurance-and-reuse-2-6-fraud-aml-and-transaction-monitoring", "finance/findings-assurance-and-reuse-2-7-policy-and-authorisation-engines", "finance/findings-assurance-and-reuse-2-8-workflow-and-durability", "finance/findings-assurance-and-reuse-2-9-journey-and-diagram-tooling-mermaid", "finance/findings-assurance-and-reuse-3-suggested-reuse-stack-assessment", "finance/findings-assurance-and-reuse-4-things-we-could-not-confirm-unknown-and", "finance/findings-assurance-and-reuse-edge-finance-agent-assurance-landscape-an", "finance/findings-assurance-and-reuse-executive-summary", "finance/findings-assurance-and-reuse-part-2-existing-subsystems-to-reuse-inste", "finance/findings-competitors-0-1-evidence-labels", "finance/findings-competitors-0-2-plain-english-glossary-acronyms-defined-once", "finance/findings-competitors-0-2-plain-english-glossary-acronyms-defined-once-2", "finance/findings-competitors-1-executive-summary-the-ten-things-that-matter-mo", "finance/findings-competitors-10-top-10-features-competitors-have-that-edge-fin", "finance/findings-competitors-2-1-summary-table", "finance/findings-competitors-2-10-other-uk-eu-players-notably-agent-ready", "finance/findings-competitors-2-2-starling-bank-and-engine-by-starling", "finance/findings-competitors-2-4-revolut-banking", "finance/findings-competitors-2-5-wise", "finance/findings-competitors-2-9-allica-bank-sme-i-e-small-and-medium-sized-en", "finance/findings-competitors-3-2-coinbase-agentic-wallets-x402-agentkit", "finance/findings-competitors-3-4-revolut-crypto", "finance/findings-competitors-4-2-openai", "finance/findings-competitors-4-4-google", "finance/findings-competitors-4-5-microsoft", "finance/findings-competitors-4-6-oracle-financial-services", "finance/findings-competitors-4-7-palantir", "finance/findings-competitors-4-8-finance-specific-intelligence-and-financial-c", "finance/findings-competitors-5-2-stripe", "finance/findings-competitors-5-5-paypal", "finance/findings-competitors-5-7-x402-coinbase-origin-and-mpp", "finance/findings-competitors-7-regulatory-context-that-shapes-the-lane", "finance/findings-competitors-8-1-what-nobody-does-today-based-on-what-i-could", "finance/findings-competitors-8-2-concrete-positioning-statement-draft", "finance/findings-competitors-8-3-honest-constraints-and-risks", "finance/findings-competitors-9-bare-minimum-table-stakes-vs-ambitious-differen", "finance/findings-competitors-9-bare-minimum-table-stakes-vs-ambitious-differen-2", "finance/findings-competitors-edge-finance-competitor-and-positioning-report", "finance/findings-confidence-reporting-0-1-evidence-labels", "finance/findings-confidence-reporting-0-2-method-and-limits", "finance/findings-confidence-reporting-0-3-plain-english-glossary", "finance/findings-confidence-reporting-0-3-plain-english-glossary-2", "finance/findings-confidence-reporting-1-summary", "finance/findings-confidence-reporting-2-1-the-three-cras-experian-equifax-tran", "finance/findings-confidence-reporting-2-2-fico-and-vantagescore-us-scoring-mod", "finance/findings-confidence-reporting-2-3-open-banking-and-cash-flow-challenge", "finance/findings-confidence-reporting-2-4-business-credit", "finance/findings-confidence-reporting-3-1-credit-invisible-and-thin-file-peopl", "finance/findings-confidence-reporting-3-2-harm-from-single-events", "finance/findings-confidence-reporting-3-3-fca-financial-lives-vulnerability-an", "finance/findings-confidence-reporting-3-4-what-lenders-learn-from-communicatio", "finance/findings-confidence-reporting-4-1-fca-credit-information-market-study", "finance/findings-confidence-reporting-4-2-consumer-credit-act-rights-and-fca-p", "finance/findings-confidence-reporting-4-3-uk-gdpr-article-22-and-the-data-use", "finance/findings-confidence-reporting-4-4-equality-act-indirect-discrimination", "finance/findings-confidence-reporting-4-5-united-states-fcra-and-ecoa-adverse", "finance/findings-confidence-reporting-4-6-eu-ai-act-creditworthiness-is-high-r", "finance/findings-confidence-reporting-4-7-risks-of-behavioural-and-communicati", "finance/findings-confidence-reporting-5-evidence-that-richer-data-improves-out", "finance/findings-confidence-reporting-6-1-contents-of-a-confidence-report-pers", "finance/findings-confidence-reporting-6-2-how-a-one-off-event-is-handled", "finance/findings-confidence-reporting-6-3-explanation-contest-and-audit", "finance/findings-confidence-reporting-6-4-how-it-serves-kyc", "finance/findings-confidence-reporting-6-5-organisations", "finance/findings-confidence-reporting-6-6-ai-agents-as-customers", "finance/findings-confidence-reporting-6-the-design-what-a-confidence-report-co", "finance/findings-confidence-reporting-7-1-sourced-figures", "finance/findings-confidence-reporting-7-2-own-estimates-prediction", "finance/findings-confidence-reporting-7-3-how-to-prove-it-measured-outcomes", "finance/findings-confidence-reporting-8-how-to-become-the-standard-setter", "finance/findings-confidence-reporting-9-open-questions-unknown", "finance/findings-confidence-reporting-edge-finance-confidence-reporting-resear", "finance/findings-fca-agentic-framework-sources", "finance/findings-fca-agentic-framework-the-fca-s-agentic-finance-framework-aga", "finance/findings-fca-agentic-framework-the-four-headings-of-the-trusted-framew", "finance/findings-fca-agentic-framework-what-the-fca-and-the-mills-review-add", "finance/findings-fca-agentic-framework-what-the-framework-does-not-have-and-th", "finance/findings-mobbin-flows-3-sending-money-and-payee-checks", "finance/findings-mobbin-flows-coverage", "finance/findings-mobbin-flows-coverage-2", "finance/findings-mobbin-flows-what-this-is", "finance/findings-presence-and-solution-banks-0-1-the-two-intentions-this-note", "finance/findings-presence-and-solution-banks-0-2-evidence-labels", "finance/findings-presence-and-solution-banks-0-3-method-and-limits", "finance/findings-presence-and-solution-banks-0-4-acronyms-first-use", "finance/findings-presence-and-solution-banks-0-4-acronyms-first-use-2", "finance/findings-presence-and-solution-banks-1-summary", "finance/findings-presence-and-solution-banks-2-1-passport-chips-and-icao-doc-9", "finance/findings-presence-and-solution-banks-2-2-uk-digital-identity-and-attri", "finance/findings-presence-and-solution-banks-2-3-gov-uk-one-login-gov-uk-walle", "finance/findings-presence-and-solution-banks-2-4-post-office-identity-services", "finance/findings-presence-and-solution-banks-2-5-eidas-2-wallets-and-mobile-dr", "finance/findings-presence-and-solution-banks-2-6-liveness-and-deepfake-defence", "finance/findings-presence-and-solution-banks-3-1-branch-closures", "finance/findings-presence-and-solution-banks-3-2-banking-hubs-cash-access-uk-a", "finance/findings-presence-and-solution-banks-3-3-mobile-vans-and-community-ban", "finance/findings-presence-and-solution-banks-3-4-shared-counters-through-the-p", "finance/findings-presence-and-solution-banks-3-5-access-to-cash-rules-fsma-202", "finance/findings-presence-and-solution-banks-4-intention-a-3-edge-hardware-and", "finance/findings-presence-and-solution-banks-4-intention-a-3-edge-hardware-and-2", "finance/findings-presence-and-solution-banks-5-1-community-development-finance", "finance/findings-presence-and-solution-banks-5-2-community-wealth-building", "finance/findings-presence-and-solution-banks-5-3-social-impact-investment-and", "finance/findings-presence-and-solution-banks-5-4-challenge-prizes", "finance/findings-presence-and-solution-banks-5-5-venture-studios", "finance/findings-presence-and-solution-banks-5-6-public-development-banks", "finance/findings-presence-and-solution-banks-6-1-late-payments", "finance/findings-presence-and-solution-banks-6-2-sme-payment-cost-and-speed", "finance/findings-presence-and-solution-banks-6-3-sme-lending-gaps", "finance/findings-presence-and-solution-banks-6-4-open-banking-payments-and-var", "finance/findings-presence-and-solution-banks-6-5-stablecoins-and-tokenised-dep", "finance/findings-presence-and-solution-banks-7-1-inputs-where-solutions-come-f", "finance/findings-presence-and-solution-banks-7-2-selection-deciding-what-to-fu", "finance/findings-presence-and-solution-banks-7-3-funding-instruments", "finance/findings-presence-and-solution-banks-7-4-people-empowered", "finance/findings-presence-and-solution-banks-7-5-outcome-measurement-jobs-time", "finance/findings-presence-and-solution-banks-7-6-governance", "finance/findings-presence-and-solution-banks-7-7-biggest-risks-prediction", "finance/findings-presence-and-solution-banks-7-intention-b-6-how-a-solution-ba", "finance/findings-presence-and-solution-banks-8-open-questions-unknown", "finance/findings-presence-and-solution-banks-edge-finance-physical-presence-an", "finance/findings-refero-flows-all-edge-finance-research-real-product-flows-fro", "finance/findings-refero-flows-all-task-access-review-permissions-audit-log", "finance/findings-refero-flows-all-task-access-review-permissions-audit-log-2", "finance/findings-refero-flows-all-task-account-closure", "finance/findings-refero-flows-all-task-account-closure-2", "finance/findings-refero-flows-all-task-account-closure-3", "finance/findings-refero-flows-all-task-account-closure-4", "finance/findings-refero-flows-all-task-account-closure-5", "finance/findings-refero-flows-all-task-ai-agent-setup-permissions", "finance/findings-refero-flows-all-task-ai-agent-setup-permissions-2", "finance/findings-refero-flows-all-task-apply-for-loan-credit", "finance/findings-refero-flows-all-task-apply-for-loan-credit-2", "finance/findings-refero-flows-all-task-apply-for-loan-credit-3", "finance/findings-refero-flows-all-task-book-gp-appointment", "finance/findings-refero-flows-all-task-budget-spending-insights", "finance/findings-refero-flows-all-task-budget-spending-insights-2", "finance/findings-refero-flows-all-task-budget-spending-insights-3", "finance/findings-refero-flows-all-task-budget-spending-insights-4", "finance/findings-refero-flows-all-task-business-account-onboarding-company-ver", "finance/findings-refero-flows-all-task-business-account-onboarding-company-ver-2", "finance/findings-refero-flows-all-task-business-account-onboarding-company-ver-3", "finance/findings-refero-flows-all-task-business-account-onboarding-company-ver-4", "finance/findings-refero-flows-all-task-business-account-onboarding-company-ver-5", "finance/findings-refero-flows-all-task-business-account-onboarding-company-ver-6", "finance/findings-refero-flows-all-task-buy-crypto", "finance/findings-refero-flows-all-task-buy-crypto-2", "finance/findings-refero-flows-all-task-buy-crypto-3", "finance/findings-refero-flows-all-task-buy-crypto-4", "finance/findings-refero-flows-all-task-calendar-scheduling-meeting", "finance/findings-refero-flows-all-task-care-plan", "finance/findings-refero-flows-all-task-compliance-dashboard-audit", "finance/findings-refero-flows-all-task-connect-bank-account-open-banking", "finance/findings-refero-flows-all-task-connect-bank-account-open-banking-2", "finance/findings-refero-flows-all-task-contract-management", "finance/findings-refero-flows-all-task-contract-management-2", "finance/findings-refero-flows-all-task-credit-score-report", "finance/findings-refero-flows-all-task-crm-contacts-pipeline", "finance/findings-refero-flows-all-task-crypto-wallet-send-receive", "finance/findings-refero-flows-all-task-crypto-wallet-send-receive-2", "finance/findings-refero-flows-all-task-direct-debit", "finance/findings-refero-flows-all-task-direct-debit-2", "finance/findings-refero-flows-all-task-dispute-transaction-chargeback", "finance/findings-refero-flows-all-task-doctor-video-consultation", "finance/findings-refero-flows-all-task-expense-management-receipts", "finance/findings-refero-flows-all-task-expense-management-receipts-2", "finance/findings-refero-flows-all-task-eye-test-vision", "finance/findings-refero-flows-all-task-fraud-alert-suspicious-transaction", "finance/findings-refero-flows-all-task-fraud-alert-suspicious-transaction-2", "finance/findings-refero-flows-all-task-government-service-identity-verificatio", "finance/findings-refero-flows-all-task-government-service-identity-verificatio-2", "finance/findings-refero-flows-all-task-government-service-identity-verificatio-3", "finance/findings-refero-flows-all-task-grant-application", "finance/findings-refero-flows-all-task-health-onboarding-profile", "finance/findings-refero-flows-all-task-international-transfer-currency-exchang", "finance/findings-refero-flows-all-task-international-transfer-currency-exchang-2", "finance/findings-refero-flows-all-task-international-transfer-currency-exchang-3", "finance/findings-refero-flows-all-task-inventory-warehouse", "finance/findings-refero-flows-all-task-investing-buy-stock", "finance/findings-refero-flows-all-task-investing-buy-stock-2", "finance/findings-refero-flows-all-task-investing-buy-stock-3", "finance/findings-refero-flows-all-task-investing-buy-stock-4", "finance/findings-refero-flows-all-task-invoice-create-send", "finance/findings-refero-flows-all-task-invoice-create-send-2", "finance/findings-refero-flows-all-task-invoice-create-send-3", "finance/findings-refero-flows-all-task-invoice-create-send-4", "finance/findings-refero-flows-all-task-open-bank-account-identity-verification", "finance/findings-refero-flows-all-task-pay-invoice-bill", "finance/findings-refero-flows-all-task-request-money-split-bill", "finance/findings-refero-flows-all-task-savings-pot-goal", "finance/findings-refero-flows-all-task-savings-pot-goal-2", "finance/findings-refero-flows-all-task-scheduled-payment-standing-order", "finance/findings-refero-flows-all-task-statements-export", "finance/findings-refero-flows-all-task-statements-export-2", "finance/findings-refero-flows-all-task-task-project-management", "finance/findings-refero-flows-all-task-team-permissions-roles-access", "finance/findings-refero-flows-all-task-team-permissions-roles-access-2", "finance/findings-refero-flows-all-task-two-factor-security-device", "finance/findings-refero-flows-all-task-two-factor-security-device-2", "finance/findings-refero-flows-all-task-virtual-card-freeze-card", "finance/findings-refero-flows-all-task-virtual-card-freeze-card-2", "finance/findings-regulators-0-the-headline-for-engineers", "finance/findings-regulators-2-1-consumer-duty", "finance/findings-regulators-2-10-operational-resilience-and-critical-third-par", "finance/findings-regulators-2-11-cryptoasset-regime-and-financial-promotion-ru", "finance/findings-regulators-2-12-consumer-credit", "finance/findings-regulators-2-13-credit-reference-agency-permission", "finance/findings-regulators-2-14-sanctions-screening-and-reporting-to-ofsi", "finance/findings-regulators-2-15-suspicious-activity-reports-and-tipping-off", "finance/findings-regulators-2-2-senior-managers-and-certification-regime-sm-cr", "finance/findings-regulators-2-3-money-laundering-regulations-2017-mlrs", "finance/findings-regulators-2-3-money-laundering-regulations-2017-mlrs-2", "finance/findings-regulators-2-4-payment-services-regulations-2017-psrs-and-str", "finance/findings-regulators-2-5-electronic-money-regulations-2011-emrs", "finance/findings-regulators-2-6-faster-payments-and-mandatory-app-fraud-reimbu", "finance/findings-regulators-2-7-confirmation-of-payee-cop", "finance/findings-regulators-2-8-direct-debit-scheme-and-the-direct-debit-guara", "finance/findings-regulators-2-9-open-banking-consent-oversight-smart-data-and", "finance/findings-regulators-3-routes-to-operate-from-bare-minimum-to-full", "finance/findings-regulators-4-1-european-union", "finance/findings-regulators-4-1-european-union-2", "finance/findings-regulators-4-2-united-states", "finance/findings-regulators-4-3-international-standard-setters", "finance/findings-regulators-4-3-international-standard-setters-2", "finance/findings-regulators-6-obligations-table", "finance/findings-regulators-6-obligations-table-2", "finance/findings-regulators-7-what-to-check-next-unknown-items", "finance/findings-regulators-7-what-to-check-next-unknown-items-2", "finance/findings-regulators-bank-of-england-boe", "finance/findings-regulators-competition-and-markets-authority-cma", "finance/findings-regulators-financial-conduct-authority-fca", "finance/findings-regulators-financial-ombudsman-service-fos", "finance/findings-regulators-hm-revenue-customs-hmrc", "finance/findings-regulators-hm-treasury-hmt-and-the-office-of-financial-sancti", "finance/findings-regulators-information-commissioner-s-office-ico", "finance/findings-regulators-national-crime-agency-nca-suspicious-activity-repo", "finance/findings-regulators-payment-systems-regulator-psr-and-its-merger-into", "finance/findings-regulators-prudential-regulation-authority-pra", "finance/findings-regulators-regulators-and-rules-for-edge-finance", "finance/findings-regulators-what-is-unresolved-for-ai-agents-as-customers", "finance/findings-regulators-what-is-unresolved-for-ai-agents-as-customers-2", "finance/findings-regulators-what-they-have-said", "finance/findings-screens-and-flows-adding-people-to-an-account-the-closest-thi", "finance/findings-screens-and-flows-edge-finance-research-competitors-real-scre", "finance/findings-screens-and-flows-linking-an-outside-bank-account-and-funding", "finance/findings-screens-and-flows-linking-an-outside-bank-account-and-funding-2", "finance/findings-screens-and-flows-opening-a-business-account-and-proving-the", "finance/findings-screens-and-flows-opening-an-account-and-proving-who-you-are", "finance/findings-screens-and-flows-opening-an-account-and-proving-who-you-are-2", "finance/findings-screens-and-flows-other-finance-flows-found", "finance/findings-screens-and-flows-summary-of-what-the-floor-is", "finance/findings-screens-and-flows-what-none-of-them-do-the-gap-edge-finance-f", "finance/parity-2-customer-data-finance", "finance/parity-222-account-how-222-works-finance", "finance/parity-222-event-invite-checkout-finance", "finance/parity-222-subscription-purchase-finance", "finance/parity-aboard-add-and-complete-1-on-1-topic-finance", "finance/parity-aboard-add-employee-salary-finance", "finance/parity-aboard-employee-sign-up-and-verification-finance", "finance/parity-account-and-transaction-api-specification-4-0-1-account-and-tra", "finance/parity-acctual-add-and-configure-invoice-payment-method-finance", "finance/parity-acctual-add-flexible-finance", "finance/parity-acctual-add-flexible-payment-option-finance", "finance/parity-acctual-country-selector-modal-finance", "finance/parity-acctual-invoice-creation-and-preview-finance", "finance/parity-acctual-record-invoice-payment-finance", "finance/parity-acctual-select-or-add-invoice-recipient-finance", "finance/parity-agent-bot-detection-at-finance", "finance/parity-apple-books-gift-card-redemption-finance", "finance/parity-apple-stocks-add-symbol-from-watchlist-search-finance", "finance/parity-apple-stocks-add-symbol-to-watchlist-finance", "finance/parity-apple-stocks-create-and-populate-new-watchlist-finance", "finance/parity-apple-stocks-submit-stocks-feedback-finance", "finance/parity-apple-stocks-watchlist-item-quick-actions-finance", "finance/parity-apple-stocks-watchlist-search-add-finance", "finance/parity-artsy-contact-gallery-inquiry-finance", "finance/parity-artsy-create-artist-alert-with-detailed-filters-finance", "finance/parity-artsy-email-sign-in-with-2fa-finance", "finance/parity-artsy-select-billing-country-finance", "finance/parity-avnier-checkout-and-payment-finance", "finance/parity-avnier-product-detail-and-purchase-decision-finance", "finance/parity-barclays-business-barclays-business", "finance/parity-binge-apple-sign-in-in-settings-finance", "finance/parity-binge-viewing-statistics-breakdown-finance", "finance/parity-blackbird-buy-fly-purchase-flow-finance", "finance/parity-blackbird-default-payment-preferences-finance", "finance/parity-blackbird-pass-purchase-gift-checkout-finance", "finance/parity-blackbird-rewards-wallet-overview-finance", "finance/parity-brilliant-chart-based-question-selection-finance", "finance/parity-brilliant-streak-charges-confirmation-finance", "finance/parity-british-business-bank-british-business-bank", "finance/parity-cake-equity-add-transaction-comment-finance", "finance/parity-cake-equity-create-seed-raise-finance", "finance/parity-cake-equity-issuing-shares-to-a-shareholder-finance", "finance/parity-cashcaded-share-purchase-receipt-finance", "finance/parity-cashcaded-transaction-period-review-finance", "finance/parity-chance-ai-image-insight-recommendations-finance", "finance/parity-chance-ai-palmistry-camera-to-report-finance", "finance/parity-claim-allica-reported-finance", "finance/parity-claim-banks-build-finance", "finance/parity-claim-claude-for-finance", "finance/parity-claim-critique-because-finance", "finance/parity-claim-hmrc-the-finance", "finance/parity-claim-it-says-finance", "finance/parity-claim-okx-launched-finance", "finance/parity-claim-oracle-positions-finance", "finance/parity-claim-starling-calls-finance", "finance/parity-claim-stripe-and-finance", "finance/parity-claim-the-new-finance", "finance/parity-claim-the-pra-finance", "finance/parity-claim-the-templates-finance", "finance/parity-claim-there-is-finance", "finance/parity-claim-trulioo-s-kya-finance", "finance/parity-claim-wise-published-finance", "finance/parity-claim-worldpay-partnered-finance", "finance/parity-claim-zopa-reported-finance", "finance/parity-claim-zopa-says-finance", "finance/parity-column-ach-transfer-creation-finance", "finance/parity-column-initiate-transfer-return-finance", "finance/parity-column-simulate-a-transfer-in-sandbox-finance", "finance/parity-comet-agent-card-gallery-finance", "finance/parity-confirmation-of-funds-api-specification-4-0-1-confirmation-of-f", "finance/parity-copperx-crypto-card-unlock-finance", "finance/parity-copperx-email-payment-to-new-contact-finance", "finance/parity-copperx-multi-recipient-email-payout-finance", "finance/parity-curater-report-ai-generated-content-finance", "finance/parity-cushion-weekly-goals-check-in-finance", "finance/parity-dailyart-account-deletion-finance", "finance/parity-dailyart-checkout-payment-details-finance", "finance/parity-dailyart-checkout-payment-entry-finance", "finance/parity-dailyart-tip-jar-donation-finance", "finance/parity-digg-account-deletion-finance", "finance/parity-district-account-deletion-review-finance", "finance/parity-district-add-card-finance", "finance/parity-district-add-card-for-booking-payment-finance", "finance/parity-district-booking-payment-selection-finance", "finance/parity-district-booking-review-voucher-offer-finance", "finance/parity-district-closed-support-rating-finance", "finance/parity-district-dining-save-and-reserve-finance", "finance/parity-district-district-money-top-up-finance", "finance/parity-district-event-booking-checkout-finance", "finance/parity-district-event-booking-confirmation-finance", "finance/parity-district-gift-card-redemption-finance", "finance/parity-district-invite-guests-after-booking-finance", "finance/parity-district-movie-ticket-booking-finance", "finance/parity-district-offers-search-and-selection-finance", "finance/parity-district-payment-method-selection-finance", "finance/parity-district-payment-method-settings-finance", "finance/parity-district-rename-saved-card-finance", "finance/parity-district-restaurant-q-a-assistant-finance", "finance/parity-district-restaurant-question-assistant-finance", "finance/parity-district-restaurant-savings-calculator-finance", "finance/parity-district-salon-service-booking-finance", "finance/parity-district-save-to-hotlist-finance", "finance/parity-district-save-to-list-finance", "finance/parity-district-state-selection-finance", "finance/parity-district-store-offer-notification-subscription-finance", "finance/parity-district-wallet-balance-and-history-finance", "finance/parity-doji-account-creation-likeness-submission-finance", "finance/parity-doordash-schedule-pickup-time-finance", "finance/parity-dribbble-account-deletion-and-redirect-finance", "finance/parity-drinkit-apply-welcome-drink-offer-finance", "finance/parity-drinkit-invite-a-friend-reward-finance", "finance/parity-dropbox-create-activity-report-finance", "finance/parity-dropbox-dropbox-password-recovery-finance", "finance/parity-dropbox-share-file-request-finance", "finance/parity-dropbox-share-file-request-link-finance", "finance/parity-elevenlabs-developer-usage-analytics-finance", "finance/parity-elevenlabs-generation-details-finance", "finance/parity-elevenlabs-request-license-for-iconic-voice-finance", "finance/parity-elevenlabs-subscription-billing-cancellation-finance", "finance/parity-elevenlabs-upgrade-to-creator-subscription-finance", "finance/parity-elevenmusic-artist-issue-report-finance", "finance/parity-event-notification-api-specification-tpp-endpoints-4-0-1-event", "finance/parity-events-api-specification-aspsp-endpoints-4-0-1-events-api-speci", "finance/parity-family-create-account-and-wallet-finance", "finance/parity-family-create-and-finance", "finance/parity-family-enable-authenticator-method-finance", "finance/parity-family-secret-recovery-phrase-education-finance", "finance/parity-family-swap-eth-to-token-finance", "finance/parity-family-token-overview-finance", "finance/parity-farfetch-address-to-payment-checkout-finance", "finance/parity-fuse-add-external-recovery-key-finance", "finance/parity-fuse-token-selector-finance", "finance/parity-fuse-token-selector-finance-2", "finance/parity-glossgenius-analytics-with-ai-growth-analyst-finance", "finance/parity-glossgenius-send-sale-receipt-to-client-finance", "finance/parity-gocardless-payments-api-bank-payment-collection-apis", "finance/parity-gofundme-create-and-launch-a-fundraiser-finance", "finance/parity-gofundme-donate-and-share-confirmation-finance", "finance/parity-gofundme-fundraiser-details-and-organization-follow-finance", "finance/parity-gofundme-fundraiser-story-review-finance", "finance/parity-gofundme-fundraiser-support-chat-finance", "finance/parity-gofundme-report-content-submission-finance", "finance/parity-gofundme-sharehub-co-organizer-invitation-finance", "finance/parity-google-gemini-report-legal-content-from-gemini-chat-finance", "finance/parity-google-gemini-sign-in-and-launch-gemini-finance", "finance/parity-google-translate-dictionary-translation-entry-finance", "finance/parity-google-translate-download-language-pack-finance", "finance/parity-google-translate-lens-text-translation-finance", "finance/parity-google-translate-source-language-picker-finance", "finance/parity-google-translate-word-translation-details-finance", "finance/parity-grammarly-configure-writing-goals-finance", "finance/parity-grammarly-writing-performance-report-finance", "finance/parity-granola-copy-generated-response-finance", "finance/parity-granola-share-note-with-editor-access-finance", "finance/parity-grok-chat-to-voice-activation-finance", "finance/parity-ground-news-alerts-inbox-cleanup-finance", "finance/parity-ground-news-delete-single-alert-finance", "finance/parity-gumroad-membership-charge-refund-finance", "finance/parity-gumroad-process-sale-refund-finance", "finance/parity-human-in-the-loop-confirmation-for-key-actions-finance", "finance/parity-hype-account-deletion-finance", "finance/parity-hype-upcoming-product-details-finance", "finance/parity-ikea-contact-support-chatbot-interaction-finance", "finance/parity-introspect-friend-decision-response-finance", "finance/parity-kann-in-game-coin-life-purchase-finance", "finance/parity-kickstarter-manage-pledge-cancel-finance", "finance/parity-lego-account-editing-and-deletion-finance", "finance/parity-literal-club-add-books-to-yearly-reading-target-finance", "finance/parity-loowner-add-new-item-with-photo-and-receipt-finance", "finance/parity-melius-request-a-template-finance", "finance/parity-mercury-create-and-select-recipient-finance", "finance/parity-mercury-create-company-card-finance", "finance/parity-mercury-delete-payment-finance", "finance/parity-mercury-delete-payment-from-inbox-finance", "finance/parity-mercury-invoice-customer-edit-finance", "finance/parity-mercury-payment-approval-review-finance", "finance/parity-mercury-remove-transaction-attachment-finance", "finance/parity-mercury-set-up-recurring-ach-payment-finance", "finance/parity-meta-ai-meta-ai-email-sign-in-finance", "finance/parity-meta-ai-weekend-activity-suggestions-finance", "finance/parity-monzo-current-accounts-current-accounts", "finance/parity-monzo-monzo", "finance/parity-moonly-compatibility-check-finance", "finance/parity-moonly-tarot-chat-with-recommendations-finance", "finance/parity-mymind-account-deletion-finance", "finance/parity-mymind-delete-saved-card-finance", "finance/parity-navan-manage-travel-payment-methods-finance", "finance/parity-nudea-stockist-locator-finance", "finance/parity-obl-vrp-profile-4-0-1-obl-vrp-profile", "finance/parity-observed-agentmail-calls-finance", "finance/parity-observed-bunq-and-finance", "finance/parity-observed-excerpt-chase-finance", "finance/parity-observed-excerpt-kroo-s-finance", "finance/parity-observed-excerpt-revolut-finance", "finance/parity-observed-excerpt-revolut-s-finance", "finance/parity-observed-excerpt-stripe-finance", "finance/parity-observed-excerpt-visa-finance", "finance/parity-observed-it-is-finance", "finance/parity-observed-resend-s-post-finance", "finance/parity-observed-that-post-finance", "finance/parity-observed-the-2026-finance", "finance/parity-observed-the-catalogue-finance", "finance/parity-observed-the-readme-finance", "finance/parity-omio-discount-card-selection-finance", "finance/parity-omio-recommended-train-selection-finance", "finance/parity-omio-select-discount-card-finance", "finance/parity-on-account-deletion-finance", "finance/parity-on-apply-gift-card-at-checkout-finance", "finance/parity-opensea-add-weth-to-make-offer-finance", "finance/parity-oracle-fusion-netsuite-oracle-fusion", "finance/parity-orbit-change-subscription-payment-method-finance", "finance/parity-orbit-subscription-import-and-review-finance", "finance/parity-payment-initiation-api-4-0-1-payment-initiation-api", "finance/parity-paypal-activity-transaction-search-finance", "finance/parity-paypal-add-customer-to-invoice-finance", "finance/parity-paypal-add-payment-card-finance", "finance/parity-paypal-invoice-search-and-results-finance", "finance/parity-paypal-loyalty-card-finance", "finance/parity-paypal-manage-loyalty-cards-finance", "finance/parity-paypal-manage-loyalty-finance", "finance/parity-paypal-paypal-assistant-conversation-finance", "finance/parity-paypal-peer-payment-request-finance", "finance/parity-paypal-request-payment-finance", "finance/parity-paypal-set-preferred-online-payment-finance", "finance/parity-paypal-set-preferred-payment-method-finance", "finance/parity-paypal-sign-in-and-account-overview-finance", "finance/parity-paypal-sign-in-and-biometric-enablement-finance", "finance/parity-paypal-transaction-conversation-thread-finance", "finance/parity-phantom-bookmark-page-finance", "finance/parity-phantom-bookmark-page-from-swap-screen-finance", "finance/parity-phantom-create-new-wallet-account-finance", "finance/parity-phantom-token-search-and-results-finance", "finance/parity-phantom-token-selector-finance", "finance/parity-plane-finder-account-deletion-finance", "finance/parity-podimo-account-deletion-finance", "finance/parity-podimo-premium-subscription-purchase-finance", "finance/parity-pool-place-details-to-directions-finance", "finance/parity-preply-manage-payment-methods-finance", "finance/parity-queue-subscription-cards-list-finance", "finance/parity-rad-weather-support-credit-purchase-finance", "finance/parity-rad-weather-weather-intelligence-details-finance", "finance/parity-revolut-add-money-card-top-up-finance", "finance/parity-revolut-business-business-account-and-business-api", "finance/parity-revolut-business-revolut-business", "finance/parity-revolut-gbp-statement-generator-finance", "finance/parity-revolut-help-centre-article-discovery-finance", "finance/parity-revolut-merchant-api-merchant-api", "finance/parity-revolut-vault-detail-and-edit-finance", "finance/parity-rovr-account-deactivation-finance", "finance/parity-rox-news-insight-with-suggested-actions-finance", "finance/parity-runey-invoice-payment-finance", "finance/parity-shop-product-reporting-finance", "finance/parity-shop-shop-account-deletion-finance", "finance/parity-spoil-me-invite-and-earn-finance", "finance/parity-spyglass-audience-scaling-recommendations-finance", "finance/parity-square-requesting-payment-from-conversation-finance", "finance/parity-starling-starling", "finance/parity-stocktwits-add-symbol-to-watchlist-finance", "finance/parity-stocktwits-add-symbols-to-watchlist-finance", "finance/parity-stocktwits-earnings-call-chat-interaction-finance", "finance/parity-stocktwits-poll-reply-with-price-prediction-and-media-finance", "finance/parity-stocktwits-poll-reply-with-price-prediction-finance", "finance/parity-stripe-payments-payments-apis", "finance/parity-subo-add-payment-method-finance", "finance/parity-subo-add-subscription-finance", "finance/parity-subo-find-subscription-finance", "finance/parity-teal-final-resume-check-finance", "finance/parity-teal-set-weekly-application-target-finance", "finance/parity-tidal-apple-sign-in-to-home-finance", "finance/parity-tidal-artist-biography-finance", "finance/parity-tidal-artist-credits-browser-finance", "finance/parity-tidal-email-account-sign-in-finance", "finance/parity-tidal-playlist-privacy-change-finance", "finance/parity-tidal-report-incorrect-track-credits-finance", "finance/parity-tidal-track-report-reason-selection-finance", "finance/parity-tide-save-card-to-photos-finance", "finance/parity-tide-start-a-finance", "finance/parity-tide-tide", "finance/parity-tinder-account-pause-and-deletion-options-finance", "finance/parity-tinder-help-center-account-appeals-finance", "finance/parity-tinder-safety-tips-carousel-finance", "finance/parity-travelperk-new-account-signup-finance", "finance/parity-trip-multi-city-flight-booking-finance", "finance/parity-turf-send-friend-request-finance", "finance/parity-uglycash-deposit-usdc-to-portfolio-finance", "finance/parity-uglycash-p2p-usd-send-finance", "finance/parity-unknown-pricing-finance", "finance/parity-unknown-timing-of-finance", "finance/parity-urban-outfitters-gift-card-browse-and-search-finance", "finance/parity-vivid-finance", "finance/parity-vocabulary-mark-vocabulary-card-favorite-finance", "finance/parity-voicenotes-email-verification-finance", "finance/parity-wabi-home-card-actions-finance", "finance/parity-wabi-submit-personal-analysis-finance", "finance/parity-wealthsimple-accounts-overview-and-view-toggle-finance", "finance/parity-wealthsimple-accounts-overview-finance", "finance/parity-wealthsimple-add-money-to-investment-account-finance", "finance/parity-wealthsimple-contact-transfer-request-finance", "finance/parity-wealthsimple-enable-options-trading-finance", "finance/parity-wealthsimple-manage-linked-account-nickname-finance", "finance/parity-wealthsimple-password-reset-with-two-step-verification-finance", "finance/parity-willow-referral-card-sharing-finance", "finance/parity-wise-card-ordering-and-payment-finance", "finance/parity-wise-create-a-jar-finance", "finance/parity-wise-help-centre-search-and-results-finance", "finance/parity-wise-platform-global-payments-platform", "finance/parity-wise-schedule-a-transfer-finance", "finance/parity-wittl-create-job-listing-finance", "finance/parity-xero-xero", "finance/parity-youtube-music-report-a-comment-finance", "finance/problem-6-3bn-in-30-minutes-how-the-world-s-biggest-heist", "finance/problem-alcohol-duty-penalties-for-late-filing-and-late-pa", "finance/problem-applications-open-for-next-director-of-the-serious", "finance/problem-bank-of-england-explores-easing-key-limit-on-riski", "finance/problem-bereaved-families-will-not-face-tax-bill-over-pens", "finance/problem-beyond-borders-regulating-the-risks-we-share", "finance/problem-binance-rushes-to-recover-customer-money-after-6-3", "finance/problem-binance-s-6-3bn-crypto-hack-biggest-in-history", "finance/problem-binance-s-6-3bn-heist-is-the-biggest-but-not-the-f", "finance/problem-brits-nearly-twice-as-likely-to-spot-pressure-tact", "finance/problem-crypto-giant-binance-attempting-to-recover-6-3bn-s", "finance/problem-crypto-giant-binance-hit-by-6-3bn-hack-suspected-t", "finance/problem-defence-smes-need-banking-services-that-work", "finance/problem-defined-benefit-pension-scheme-surplus-payments-to", "finance/problem-fca-restricts-6-firms-following-concerns-over-anti", "finance/problem-former-chief-executive-banned-after-misconduct-led", "finance/problem-france-s-supermarket-fridges-were-vulnerable-onlin", "finance/problem-french-supermarkets-suffer-fridge-shutdown-after-u", "finance/problem-from-bitcoin-to-binance-crypto-terms-and-what-they", "finance/problem-government-unleashes-1-2-billion-crackdown-as-part", "finance/problem-guidance-framework-agreement-between-the-law-offic", "finance/problem-guidance-make-a-corporate-self-report", "finance/problem-hackers-activate-automatic-defrosting-in-supermark", "finance/problem-how-ai-might-be-deployed-for-better-bank-supervisi", "finance/problem-how-big-a-deal-is-the-6-3bn-binance-crypto-heist-w", "finance/problem-how-crypto-hacking-funds-isolated-north-korea-s-nu", "finance/problem-how-poundland-grew-from-a-market-stall-to-a-high-s", "finance/problem-i-read-anthropic-s-92-page-risk-factor-document-so", "finance/problem-israeli-settlers-attack-european-diplomatic-cars-i", "finance/problem-israeli-settlers-attack-olive-framers-washington-p", "finance/problem-ledger-confirms-spy-chip-implant-behind-wallet-cry", "finance/problem-minutes-of-the-sonia-stakeholder-advisory-group-25", "finance/problem-north-korea-is-likely-culprit-behind-6-3bn-crypto", "finance/problem-ofqual-fines-innovate-awarding-185-000-over-failur", "finance/problem-pip-customers-experiences-of-the-award-review-proc", "finance/problem-response-to-supreme-court-judgment", "finance/problem-scam-american-companies-are-using-to-manipulate-in", "finance/problem-sfo-strengthens-corporate-cooperation-guidance", "finance/problem-the-staggering-446m-elon-musk-has-spent-on-the-us", "finance/problem-the-timms-review-co-chair-update-october-2026", "finance/problem-three-arrests-in-major-fca-operation-against-crimi", "finance/problem-trump-beware-us-voters-are-concerned-about-governm", "finance/problem-two-plead-guilty-after-clients-lose-2m-in-fraud-ca", "finance/problem-tyneside-pizza-boss-jailed-after-gambling-with-mon", "finance/problem-what-we-know-about-the-6-3bn-binance-hack", "finance/problem-whistleblowing", "finance/problem-who-could-be-behind-the-binance-hack-and-what-we-k", "finance/problem-woman-charged-with-computer-fraud-due-to-scam-cash", "foundation/findings-blueprints-features-journeys-0-how-i-read-the-corpus-and-how", "foundation/findings-blueprints-features-journeys-1-inventory-blueprints-specifica", "foundation/findings-blueprints-features-journeys-2-1-the-short-answer-on-near-100", "foundation/findings-blueprints-features-journeys-2-2-recorded-counts-against-my-r", "foundation/findings-blueprints-features-journeys-2-2-recorded-counts-against-my-r-2", "foundation/findings-blueprints-features-journeys-2-3-per-application-mapped-to-th", "foundation/findings-blueprints-features-journeys-2-3-per-application-mapped-to-th-2", "foundation/findings-blueprints-features-journeys-3-1-fields-used", "foundation/findings-blueprints-features-journeys-3-1-fields-used-2", "foundation/findings-blueprints-features-journeys-3-2-consistently-missing-or-only", "foundation/findings-blueprints-features-journeys-3-2-consistently-missing-or-only-2", "foundation/findings-blueprints-features-journeys-4-behaviour-and-generator-rules", "foundation/findings-blueprints-features-journeys-4-behaviour-and-generator-rules-2", "foundation/findings-blueprints-features-journeys-4-behaviour-and-generator-rules-3", "foundation/findings-blueprints-features-journeys-4-behaviour-and-generator-rules-4", "foundation/findings-blueprints-features-journeys-5-1-edge-finance", "foundation/findings-blueprints-features-journeys-5-1-edge-finance-2", "foundation/findings-blueprints-features-journeys-5-1-edge-finance-3", "foundation/findings-blueprints-features-journeys-5-2-edge-clinic", "foundation/findings-blueprints-features-journeys-5-2-edge-clinic-2", "foundation/findings-blueprints-features-journeys-5-3-edge-public-consortium", "foundation/findings-blueprints-features-journeys-5-3-edge-public-consortium-2", "foundation/findings-blueprints-features-journeys-5-4-judiverse", "foundation/findings-blueprints-features-journeys-5-4-judiverse-2", "foundation/findings-blueprints-features-journeys-5-5-cross-cutting-edge-id-and-sh", "foundation/findings-blueprints-features-journeys-5-old-estate-capabilities-and-fe", "foundation/findings-blueprints-features-journeys-6-what-is-superseded", "foundation/findings-blueprints-features-journeys-6-what-is-superseded-2", "foundation/findings-blueprints-features-journeys-7-notes-for-whoever-specifies-ne", "foundation/findings-blueprints-features-journeys-old-estate-blueprints-features-a", "foundation/findings-competitors-benchmarks-readiness-0-method-and-coverage-stated", "foundation/findings-competitors-benchmarks-readiness-1-inventory-by-topic", "foundation/findings-competitors-benchmarks-readiness-2-3-edge-clinic-social-presc", "foundation/findings-competitors-benchmarks-readiness-2-competitors-already-resear", "foundation/findings-competitors-benchmarks-readiness-3-1-how-the-old-estate-defin", "foundation/findings-competitors-benchmarks-readiness-3-3-published-best-scores-ob", "foundation/findings-competitors-benchmarks-readiness-3-3-published-best-scores-ob-2", "foundation/findings-competitors-benchmarks-readiness-3-6-which-benchmarks-apply-t", "foundation/findings-competitors-benchmarks-readiness-4-2-evidence-required-per-le", "foundation/findings-competitors-benchmarks-readiness-4-2-evidence-required-per-le-2", "foundation/findings-competitors-benchmarks-readiness-5-1-the-most-valuable-findin", "foundation/findings-competitors-benchmarks-readiness-5-2-superseded-or-unsafe-to", "foundation/findings-competitors-benchmarks-readiness-5-3-gaps-the-new-estate-must", "foundation/findings-competitors-benchmarks-readiness-acronyms-used", "foundation/findings-competitors-benchmarks-readiness-acronyms-used-2", "foundation/findings-competitors-benchmarks-readiness-legacy-competitors-benchmark", "foundation/findings-costliest-problems-0-the-twelve-things-to-know-first", "foundation/findings-costliest-problems-1-1-payment-fraud-in-the-uk-uk-finance-ann", "foundation/findings-costliest-problems-1-1-payment-fraud-in-the-uk-uk-finance-ann-2", "foundation/findings-costliest-problems-1-2-reimbursement-of-app-fraud", "foundation/findings-costliest-problems-1-3-the-wider-cost-of-fraud", "foundation/findings-costliest-problems-1-4-identity-fraud-synthetic-identities-an", "foundation/findings-costliest-problems-1-4-identity-fraud-synthetic-identities-an-2", "foundation/findings-costliest-problems-1-5-deepfakes-and-identity-spoofing", "foundation/findings-costliest-problems-1-6-crypto-hacks-and-theft-2025-2026", "foundation/findings-costliest-problems-1-7-sanctions-evasion-and-illicit-crypto-f", "foundation/findings-costliest-problems-1-8-poor-or-incomplete-customer-data-and-t", "foundation/findings-costliest-problems-1-9-finance-what-is-going-wrong-and-what-i", "foundation/findings-costliest-problems-2-1-what-the-ncsc-says-is-getting-worse", "foundation/findings-costliest-problems-2-2-the-largest-uk-attacks-of-2025-2026-an", "foundation/findings-costliest-problems-2-3-global-attacks-and-ransomware", "foundation/findings-costliest-problems-2-4-supply-chain", "foundation/findings-costliest-problems-2-5-the-cost-of-incomplete-asset-informati", "foundation/findings-costliest-problems-2-6-cyber-what-is-going-wrong-and-what-is", "foundation/findings-costliest-problems-3-1-waiting-lists-england", "foundation/findings-costliest-problems-3-2-gp-access-and-the-cost-of-being-seen", "foundation/findings-costliest-problems-3-3-missed-appointments", "foundation/findings-costliest-problems-3-4-late-diagnosis", "foundation/findings-costliest-problems-3-5-preventable-illness-and-ill-health-out", "foundation/findings-costliest-problems-3-6-sight-loss", "foundation/findings-costliest-problems-3-7-loneliness", "foundation/findings-costliest-problems-3-8-social-prescribing-what-the-evidence-s", "foundation/findings-costliest-problems-3-9-health-what-is-going-wrong-and-what-is", "foundation/findings-costliest-problems-4-1-public-sector-fraud-and-error", "foundation/findings-costliest-problems-4-2-major-projects-and-procurement", "foundation/findings-costliest-problems-4-3-defence-and-defence-logistics", "foundation/findings-costliest-problems-4-4-security-vetting-backlogs", "foundation/findings-costliest-problems-4-5-government-what-is-going-wrong-and-wha", "foundation/findings-costliest-problems-5-education-and-research", "foundation/findings-costliest-problems-6-1-documented-failures-incidents-and-loss", "foundation/findings-costliest-problems-6-2-missing-assurance", "foundation/findings-costliest-problems-6-3-ai-what-is-going-wrong-and-what-is-mis", "foundation/findings-costliest-problems-7-synthesis", "foundation/findings-costliest-problems-7-synthesis-2", "foundation/findings-costliest-problems-7-synthesis-3", "foundation/findings-costliest-problems-8-what-i-could-not-find-unknown-for-the-ne", "foundation/findings-costliest-problems-how-to-read-this-file", "foundation/findings-costliest-problems-the-costliest-real-problems-what-is-going", "foundation/findings-funding-and-procurement-0-the-ten-things-to-know-first", "foundation/findings-funding-and-procurement-1-1-innovate-uk-competitions-open-or", "foundation/findings-funding-and-procurement-1-1-innovate-uk-competitions-open-or-2", "foundation/findings-funding-and-procurement-1-2-what-ukri-has-already-funded-gate", "foundation/findings-funding-and-procurement-1-2-what-ukri-has-already-funded-gate-2", "foundation/findings-funding-and-procurement-1-3-aria-advanced-research-and-invent", "foundation/findings-funding-and-procurement-1-4-ai-security-institute-grants", "foundation/findings-funding-and-procurement-1-5-defence-innovation-dasa-now-uk-de", "foundation/findings-funding-and-procurement-1-6-sbri-healthcare-and-nhs-ai-progra", "foundation/findings-funding-and-procurement-1-7-sovereign-ai", "foundation/findings-funding-and-procurement-2-1-uk-open-opportunities-found-deadl", "foundation/findings-funding-and-procurement-2-2-uk-who-is-already-paying-and-for", "foundation/findings-funding-and-procurement-2-2-uk-who-is-already-paying-and-for-2", "foundation/findings-funding-and-procurement-2-2-uk-who-is-already-paying-and-for-3", "foundation/findings-funding-and-procurement-2-2-uk-who-is-already-paying-and-for-4", "foundation/findings-funding-and-procurement-2-3-uk-frameworks-how-public-bodies-b", "foundation/findings-funding-and-procurement-2-4-european-union-ted", "foundation/findings-funding-and-procurement-2-5-united-states-sam-gov-and-sbir-st", "foundation/findings-funding-and-procurement-3-1-trend-reports", "foundation/findings-funding-and-procurement-3-2-notable-rounds-by-theme", "foundation/findings-funding-and-procurement-3-2-notable-rounds-by-theme-2", "foundation/findings-funding-and-procurement-3-2-notable-rounds-by-theme-3", "foundation/findings-funding-and-procurement-4-synthesis-who-is-paying-for-which-p", "foundation/findings-funding-and-procurement-4-synthesis-who-is-paying-for-which-p-2", "foundation/findings-funding-and-procurement-5-dates-to-act-on-from-the-evidence-a", "foundation/findings-funding-and-procurement-6-what-i-could-not-access-or-confirm", "foundation/findings-funding-and-procurement-how-to-read-this-file", "foundation/findings-funding-and-procurement-method-and-limits", "foundation/findings-funding-and-procurement-what-is-actually-being-funded-and-bou", "foundation/findings-headlines-2026-10-ai-agents-and-regulation", "foundation/findings-headlines-2026-10-crypto-the-largest-theft-on-record", "foundation/findings-headlines-2026-10-energy", "foundation/findings-headlines-2026-10-fraud-scams-and-cyber-uk", "foundation/findings-headlines-2026-10-headlines-october-2026", "foundation/findings-headlines-2026-10-nhs", "foundation/findings-readme-coverage", "foundation/findings-readme-how-to-read-it", "foundation/findings-readme-real-product-flows", "foundation/findings-readme-research", "foundation/findings-refero-flows-1-customer-experience-all-applications-research", "foundation/findings-refero-flows-1-task-access-review-permissions-audit-log", "foundation/findings-refero-flows-1-task-access-review-permissions-audit-log-2", "foundation/findings-refero-flows-1-task-access-review-permissions-audit-log-3", "foundation/findings-refero-flows-1-task-accessibility-settings-text-size", "foundation/findings-refero-flows-1-task-accessibility-settings-text-size-2", "foundation/findings-refero-flows-1-task-ai-agent-setup-permissions", "foundation/findings-refero-flows-1-task-ai-chat-assistant", "foundation/findings-refero-flows-1-task-api-keys-developer-settings", "foundation/findings-refero-flows-1-task-apply-for-loan-credit", "foundation/findings-refero-flows-1-task-approval-workflow", "foundation/findings-refero-flows-1-task-approval-workflow-2", "foundation/findings-refero-flows-1-task-budget-spending-insights", "foundation/findings-refero-flows-1-task-business-account-onboarding-company-verif", "foundation/findings-refero-flows-1-task-business-account-onboarding-company-verif-2", "foundation/findings-refero-flows-1-task-calendar-scheduling-meeting", "foundation/findings-refero-flows-1-task-care-plan", "foundation/findings-refero-flows-1-task-case-management", "foundation/findings-refero-flows-1-task-case-management-2", "foundation/findings-refero-flows-1-task-code-repository-pull-request", "foundation/findings-refero-flows-1-task-compliance-dashboard-audit", "foundation/findings-refero-flows-1-task-compliance-dashboard-audit-2", "foundation/findings-refero-flows-1-task-connect-bank-account-open-banking", "foundation/findings-refero-flows-1-task-contract-management", "foundation/findings-refero-flows-1-task-course-enrolment-university", "foundation/findings-refero-flows-1-task-crm-contacts-pipeline", "foundation/findings-refero-flows-1-task-doctor-video-consultation", "foundation/findings-refero-flows-1-task-document-editor-collaboration", "foundation/findings-refero-flows-1-task-email-inbox", "foundation/findings-refero-flows-1-task-email-inbox-2", "foundation/findings-refero-flows-1-task-expense-management-receipts", "foundation/findings-refero-flows-1-task-export-download-data", "foundation/findings-refero-flows-1-task-file-sharing-storage", "foundation/findings-refero-flows-1-task-fleet-management", "foundation/findings-refero-flows-1-task-fraud-alert-suspicious-transaction", "foundation/findings-refero-flows-1-task-government-service-identity-verification", "foundation/findings-refero-flows-1-task-government-service-identity-verification-2", "foundation/findings-refero-flows-1-task-government-service-identity-verification-3", "foundation/findings-refero-flows-1-task-grant-application", "foundation/findings-refero-flows-1-task-grant-application-2", "foundation/findings-refero-flows-1-task-health-onboarding-profile", "foundation/findings-refero-flows-1-task-integrations-connect-apps", "foundation/findings-refero-flows-1-task-international-transfer-currency-exchange", "foundation/findings-refero-flows-1-task-interview-scheduling", "foundation/findings-refero-flows-1-task-inventory-warehouse", "foundation/findings-refero-flows-1-task-inventory-warehouse-2", "foundation/findings-refero-flows-1-task-inventory-warehouse-3", "foundation/findings-refero-flows-2-customer-experience-all-applications-research", "foundation/findings-refero-flows-2-task-open-bank-account-identity-verification", "foundation/findings-refero-flows-2-task-policy-management-approval", "foundation/findings-refero-flows-2-task-policy-management-approval-2", "foundation/findings-refero-flows-2-task-prescription-refill-repeat", "foundation/findings-refero-flows-2-task-pricing-plans-compare", "foundation/findings-refero-flows-2-task-risk-register", "foundation/findings-refero-flows-2-task-savings-pot-goal", "foundation/findings-refero-flows-2-task-security-incident-alert", "foundation/findings-refero-flows-2-task-shipment-tracking-logistics", "foundation/findings-refero-flows-2-task-shipment-tracking-logistics-2", "foundation/findings-refero-flows-2-task-supplier-onboarding-vetting", "foundation/findings-refero-flows-2-task-task-project-management", "foundation/findings-refero-flows-2-task-vulnerability-management", "foundation/findings-standards-papers-guides-1-1-papers-white-papers-research-pape", "foundation/findings-standards-papers-guides-1-2-standards-regulatory-registers-re", "foundation/findings-standards-papers-guides-1-3-manuals-handbooks-guides-playbook", "foundation/findings-standards-papers-guides-1-4-record-families-in-scope-by-categ", "foundation/findings-standards-papers-guides-1-inventory", "foundation/findings-standards-papers-guides-2-1-the-five-public-papers-of-13-octo", "foundation/findings-standards-papers-guides-2-1-the-five-public-papers-of-13-octo-2", "foundation/findings-standards-papers-guides-2-1-the-five-public-papers-of-13-octo-3", "foundation/findings-standards-papers-guides-2-2-research-white-papers-of-13-octob", "foundation/findings-standards-papers-guides-2-5-what-the-papers-show-as-a-body-of", "foundation/findings-standards-papers-guides-2-papers-and-white-papers-the-organis", "foundation/findings-standards-papers-guides-3-0-where-the-mapping-lives", "foundation/findings-standards-papers-guides-3-1-finance-edge-finance", "foundation/findings-standards-papers-guides-3-1-finance-edge-finance-2", "foundation/findings-standards-papers-guides-3-2-health-edge-clinic", "foundation/findings-standards-papers-guides-3-3-public-sector-defence-procurement", "foundation/findings-standards-papers-guides-3-4-cyber-security", "foundation/findings-standards-papers-guides-3-5-ai-assurance", "foundation/findings-standards-papers-guides-3-6-cross-cutting-data-protection-acc", "foundation/findings-standards-papers-guides-3-7-how-this-can-fill-the-new-estate", "foundation/findings-standards-papers-guides-3-8-status-of-the-mapping-as-research", "foundation/findings-standards-papers-guides-4-1-product-specification-and-operati", "foundation/findings-standards-papers-guides-4-2-engineering-whole-life-manuals-17", "foundation/findings-standards-papers-guides-4-3-sector-operator-and-user-guides", "foundation/findings-standards-papers-guides-4-4-risk-resilience-and-operations", "foundation/findings-standards-papers-guides-4-5-tools-agents-and-engineering", "foundation/findings-standards-papers-guides-4-manuals-handbooks-guides-playbooks", "foundation/findings-standards-papers-guides-5-1-worth-admitting-as-research-with", "foundation/findings-standards-papers-guides-5-2-superseded-or-not-to-be-reused", "foundation/findings-standards-papers-guides-5-3-gaps-the-old-estate-did-not-close", "foundation/findings-standards-papers-guides-acronyms-used", "foundation/findings-standards-papers-guides-acronyms-used-2", "foundation/findings-standards-papers-guides-acronyms-used-3", "foundation/findings-standards-papers-guides-legacy-research-standards-regulations", "foundation/findings-standards-papers-guides-scope-method-and-coverage", "foundation/findings-strategy-funding-other-0-method-and-coverage-stated-honestly", "foundation/findings-strategy-funding-other-1-inventory-of-strategy-vision-researc", "foundation/findings-strategy-funding-other-2-2-problems-recorded-with-the-evidenc", "foundation/findings-strategy-funding-other-2-4-pricing-and-competitor-price-resea", "foundation/findings-strategy-funding-other-2-market-and-customer-research-recorde", "foundation/findings-strategy-funding-other-3-3-buyers-partners-and-relationships", "foundation/findings-strategy-funding-other-3-4-outcomes-what-the-corpus-does-and", "foundation/findings-strategy-funding-other-4-valuable-material-in-other-that-the", "foundation/findings-strategy-funding-other-5-1-reuse", "foundation/findings-strategy-funding-other-5-2-superseded-or-not-to-be-carried-ov", "foundation/findings-strategy-funding-other-5-3-open-questions-the-corpus-cannot-a", "foundation/findings-strategy-funding-other-5-reuse-in-the-new-estate-and-what-is", "foundation/findings-strategy-funding-other-legacy-strategy-funding-and-other-docu", "foundation/parity-222-decline-invite-with-feedback-experience", "foundation/parity-222-notification-delivery-selection-experience", "foundation/parity-222-resume-invites-in-account-settings-experience", "foundation/parity-222-set-preferred-city-experience", "foundation/parity-aboard-create-and-view-a-goal-experience", "foundation/parity-aboard-create-and-view-time-off-report-experience", "foundation/parity-aboard-create-device-report-experience", "foundation/parity-aboard-document-category-creation-experience", "foundation/parity-aboard-document-review-and-approval-experience", "foundation/parity-aboard-filter-documents-by-type-and-status-experience", "foundation/parity-aboard-form-creation-to-draft-editor-experience", "foundation/parity-abode-group-dashboard-home-experience", "foundation/parity-abode-group-theme-selection-experience", "foundation/parity-abode-theme-selection-experience", "foundation/parity-aby-journal-first-time-onboarding-and-conversion-experience", "foundation/parity-aby-journal-journal-scan-capture-experience", "foundation/parity-aby-journal-premium-upgrade-modal-experience", "foundation/parity-aby-journal-scan-handwritten-journal-entry-experience", "foundation/parity-aby-journal-settings-to-timeline-scrub-experience", "foundation/parity-acctual-company-document-verification-experience", "foundation/parity-acctual-country-selector-modal-experience", "foundation/parity-acctual-create-bill-from-uploaded-invoice-experience", "foundation/parity-acctual-manage-approval-rules-delete-experience", "foundation/parity-acctual-manage-contact-record-experience", "foundation/parity-acne-studios-account-sign-in-registration-experience", "foundation/parity-acne-studios-checkout-with-gift-wrap-and-klarna-payment-experie", "foundation/parity-acne-studios-product-delivery-check-experience", "foundation/parity-acuity-scheduling-account-deletion-experience", "foundation/parity-acuity-scheduling-new-account-signup-and-first-run-setup-experi", "foundation/parity-ada-change-language-preference-experience", "foundation/parity-ada-medical-quality-article-walkthrough-experience", "foundation/parity-ada-onboarding-and-account-creation-experience", "foundation/parity-ada-profile-setup-and-symptom-logging-experience", "foundation/parity-ada-profile-setup-wizard-experience", "foundation/parity-ada-sign-in-to-start-assessment-experience", "foundation/parity-adidas-back-in-stock-notification-experience", "foundation/parity-adidas-leave-a-product-review-experience", "foundation/parity-adidas-store-locator-with-map-and-details-experience", "foundation/parity-adobe-photoshop-add-adobe-stock-search-and-empty-result-experie", "foundation/parity-adobe-photoshop-general-settings-theme-and-interface-experience", "foundation/parity-adobe-photoshop-image-skew-correction-experience", "foundation/parity-adobe-photoshop-immediate-cloud-save-experience", "foundation/parity-adobe-photoshop-subscription-purchase-wizard-experience", "foundation/parity-adobe-photoshop-version-history-comment-review-experience", "foundation/parity-aesop-aesop-account-signup-experience", "foundation/parity-aesop-gift-finder-guided-quiz-experience", "foundation/parity-aesop-store-locator-and-store-detail-experience", "foundation/parity-agora-2-onboarding-welcome-card-experience", "foundation/parity-agora-2-theme-display-settings-experience", "foundation/parity-airbnb-account-deletion-experience", "foundation/parity-airbnb-airbnb", "foundation/parity-airbnb-artist-experience-request-experience", "foundation/parity-airbnb-booking-payment-verification-experience", "foundation/parity-airbnb-browse-and-save-listing-experience", "foundation/parity-airbnb-calendar-view-switcher-experience", "foundation/parity-airbnb-conversation-message-compose-experience", "foundation/parity-airbnb-date-range-pricing-and-rules-editor-experience", "foundation/parity-airbnb-edit-personal-profile-experience", "foundation/parity-airbnb-event-booking-and-verification-experience", "foundation/parity-airbnb-filter-help-search-by-article-type-experience", "foundation/parity-airbnb-id-capture-and-name-confirmation-experience", "foundation/parity-airbnb-inbox-feedback-submission-experience", "foundation/parity-airbnb-inbox-message-management-experience", "foundation/parity-airbnb-listing-save-share-experience", "foundation/parity-airbnb-map-area-search-and-filter-experience", "foundation/parity-airbnb-map-based-filtered-search-experience", "foundation/parity-airbnb-map-side-listing-discovery-experience", "foundation/parity-airbnb-message-interaction-react-actions-experience", "foundation/parity-airbnb-messages-inbox-conversation-list-experience", "foundation/parity-airbnb-phone-sign-in-experience", "foundation/parity-airbnb-photographer-booking-experience", "foundation/parity-airbnb-private-chef-booking-experience", "foundation/parity-airbnb-profile-introduction-and-interests-edit-experience", "foundation/parity-airbnb-profile-prompts-experience", "foundation/parity-airbnb-refer-a-host-referral-creation-experience", "foundation/parity-airbnb-reservation-checkout-and-payment-experience", "foundation/parity-airbnb-review-search-results-experience", "foundation/parity-airbnb-reviews-sorting-and-viewing-experience", "foundation/parity-airbnb-save-listing-to-new-wishlist-experience", "foundation/parity-airbnb-save-listing-to-wishlist-experience", "foundation/parity-airbnb-selfie-verification-capture-experience", "foundation/parity-airbnb-share-listing-to-messages-experience", "foundation/parity-airbnb-support-chat-message-experience", "foundation/parity-airbnb-switch-calendar-view-experience", "foundation/parity-airtable-add-extension-from-marketplace-experience", "foundation/parity-airtable-adjust-grid-row-height-experience", "foundation/parity-airtable-delete-asset-from-kanban-columns-experience", "foundation/parity-airtable-invite-link-creation-experience", "foundation/parity-allbirds-rate-assistant-response-experience", "foundation/parity-alo-personalized-fitness-dashboard-experience", "foundation/parity-alo-prenatal-program-detail-experience", "foundation/parity-alo-refer-a-friend-sharing-experience", "foundation/parity-alo-wellness-club-registration-and-personalization-experience", "foundation/parity-anam-avatar-selection-setup-experience", "foundation/parity-anam-call-feedback-submission-experience", "foundation/parity-anam-ended-conversation-actions-experience", "foundation/parity-anam-feedback-submission-experience", "foundation/parity-anam-persona-conversation-test-experience", "foundation/parity-anam-persona-llm-selection-experience", "foundation/parity-anam-session-history-export-experience", "foundation/parity-anam-virtual-host-testing-and-handoff-experience", "foundation/parity-apple-books-book-catalog-search-experience", "foundation/parity-apple-books-book-reviews-and-rating-experience", "foundation/parity-apple-books-customer-reviews-experience", "foundation/parity-apple-books-daily-goal-achieved-experience", "foundation/parity-apple-books-daily-reading-goal-adjustment-experience", "foundation/parity-apple-books-reader-theme-customization-experience", "foundation/parity-apple-books-reading-appearance-settings-experience", "foundation/parity-apple-games-best-distance-leaderboard-experience", "foundation/parity-apple-invites-add-invite-to-calendar-experience", "foundation/parity-apple-invites-event-guest-management-experience", "foundation/parity-apple-invites-event-invitation-upgrade-experience", "foundation/parity-apple-maps-gas-station-search-refinement-experience", "foundation/parity-apple-maps-map-search-and-place-details-experience", "foundation/parity-apple-maps-place-details-and-street-view-experience", "foundation/parity-apple-maps-recent-places-management-experience", "foundation/parity-apple-maps-search-and-directions-in-maps-experience", "foundation/parity-apple-podcasts-podcast-rating-review-submission-experience", "foundation/parity-apple-stocks-add-earnings-event-to-calendar-experience", "foundation/parity-apple-tv-content-search-and-filtering-experience", "foundation/parity-apple-tv-mls-club-directory-interaction-experience", "foundation/parity-apple-tv-movie-download-and-completion-experience", "foundation/parity-apple-tv-team-detail-kansas-city-experience", "foundation/parity-appwrite-account-deletion-experience", "foundation/parity-appwrite-create-team-and-view-details-experience", "foundation/parity-appwrite-project-overview-integrations-experience", "foundation/parity-arket-register-new-account-experience", "foundation/parity-artsy-account-deletion-confirmation-experience", "foundation/parity-artsy-account-signup-and-onboarding-experience", "foundation/parity-artsy-add-artwork-to-collection-experience", "foundation/parity-artsy-artist-auction-lookup-experience", "foundation/parity-artsy-artist-auction-results-search-experience", "foundation/parity-artsy-contact-gallery-inquiry-experience", "foundation/parity-artsy-dark-mode-settings-in-account-experience", "foundation/parity-artsy-email-sign-in-with-2fa-experience", "foundation/parity-artsy-gallery-conversation-reply-experience", "foundation/parity-artsy-onboarding-preference-to-curated-picks-experience", "foundation/parity-asos-access-returns-refunds-policy-experience", "foundation/parity-asos-add-credit-debit-card-experience", "foundation/parity-asos-brand-search-to-product-listing-experience", "foundation/parity-asos-contact-to-virtual-assistant-order-query-experience", "foundation/parity-asos-guest-to-account-sign-up-from-wishlist-experience", "foundation/parity-asos-initial-onboarding-regional-settings-experience", "foundation/parity-asos-mobile-checkout-bag-to-payment-experience", "foundation/parity-athleats-recipe-browsing-and-search-experience", "foundation/parity-auth0-auth0", "foundation/parity-avnier-cart-to-checkout-experience", "foundation/parity-avnier-customer-account-access-experience", "foundation/parity-avnier-product-detail-and-purchase-decision-experience", "foundation/parity-b150-theme-preference-experience", "foundation/parity-ballpark-in-product-support-chat-account-help-experience", "foundation/parity-ballpark-share-protect-test-link-experience", "foundation/parity-bereal-account-creation-and-onboarding-experience", "foundation/parity-bereal-contact-support-with-attachment-experience", "foundation/parity-bereal-in-conversation-emoji-reactions-experience", "foundation/parity-bereal-manage-friend-requests-experience", "foundation/parity-bereal-save-bereal-to-photos-experience", "foundation/parity-bezi-account-deletion-experience", "foundation/parity-bezi-create-new-scene-experience", "foundation/parity-bezi-share-modal-controls-experience", "foundation/parity-bezi-signup-first-time-onboarding-experience", "foundation/parity-bezi-switch-to-team-workspace-experience", "foundation/parity-binge-apple-sign-in-in-settings-experience", "foundation/parity-binge-binge-onboarding-and-pro-subscription-experience", "foundation/parity-binge-binge-pro-upgrade-experience", "foundation/parity-binge-cast-people-display-experience", "foundation/parity-binge-entertainment-onboarding-experience", "foundation/parity-binge-viewing-statistics-breakdown-experience", "foundation/parity-binge-where-to-watch-providers-experience", "foundation/parity-blackbird-blackbird-tap-to-pay-tutorial-experience", "foundation/parity-blackbird-change-appearance-mode-experience", "foundation/parity-blackbird-new-user-onboarding-and-wallet-setup-experience", "foundation/parity-blackbird-pass-purchase-gift-checkout-experience", "foundation/parity-blackbird-promo-code-checkout-experience", "foundation/parity-blackbird-referral-invite-experience", "foundation/parity-blackbird-referral-share-fly-reward-experience", "foundation/parity-bond-bond-account-onboarding-experience", "foundation/parity-bond-create-photo-story-experience", "foundation/parity-bond-find-friends-from-contacts-experience", "foundation/parity-bond-like-memory-response-experience", "foundation/parity-bond-restaurant-recommendation-approval-experience", "foundation/parity-booking-com-booking-com", "foundation/parity-box-box-club-default-widget-setup-experience", "foundation/parity-box-box-club-glossary-search-and-filter-experience", "foundation/parity-box-box-club-view-preference-setup-experience", "foundation/parity-breathwrk-onboard-sign-in-experience", "foundation/parity-breathwrk-password-recovery-via-email-experience", "foundation/parity-brilliant-analyzing-bar-charts-lesson-experience", "foundation/parity-brilliant-calculus-lesson-rate-of-change-experience", "foundation/parity-brilliant-lesson-connecting-gears-experience", "foundation/parity-brilliant-open-help-article-from-premium-experience", "foundation/parity-brilliant-premium-subscription-checkout-experience", "foundation/parity-brilliant-sign-in-and-entry-to-home-experience", "foundation/parity-brilliant-start-7-day-premium-trial-experience", "foundation/parity-brink-download-and-save-event-recording-experience", "foundation/parity-brink-start-walking-session-experience", "foundation/parity-butter-share-referral-modal-experience", "foundation/parity-cake-equity-add-personal-equity-goal-experience", "foundation/parity-cake-equity-export-shareholders-list-experience", "foundation/parity-cake-equity-inbox-message-react-and-confirm-experience", "foundation/parity-cal-automated-reminder-setup-experience", "foundation/parity-cal-bookings-filter-and-narrow-list-experience", "foundation/parity-calendly-contact-meeting-booking-experience", "foundation/parity-calendly-gmail-connection-and-email-composer-experience", "foundation/parity-calendly-update-conflict-calendars-experience", "foundation/parity-canopi-create-multi-day-event-template-experience", "foundation/parity-canopi-permanent-account-deletion-experience", "foundation/parity-canopi-sign-in-and-first-time-onboarding-experience", "foundation/parity-canva-brand-typography-editor-experience", "foundation/parity-canva-clear-notifications-experience", "foundation/parity-canva-revise-cover-title-via-voice-or-text-experience", "foundation/parity-canva-share-design-with-team-experience", "foundation/parity-canva-share-link-and-set-permission-experience", "foundation/parity-canva-text-font-selection-for-vertical-video-experience", "foundation/parity-capwords-daily-review-scheduler-experience", "foundation/parity-carbon-export-lossless-track-experience", "foundation/parity-carbon-subscription-purchase-experience", "foundation/parity-carbon-two-step-feedback-contact-capture-experience", "foundation/parity-care-activity-search-and-filtering-experience", "foundation/parity-care-caregiver-review-submission-experience", "foundation/parity-care-pet-care-caregiver-search-experience", "foundation/parity-care-profile-photo-update-experience", "foundation/parity-cashcaded-offline-setup-and-apple-pay-introduction-experience", "foundation/parity-chance-ai-account-onboarding-access-experience", "foundation/parity-chance-ai-assistant-answer-follow-up-question-experience", "foundation/parity-chance-ai-image-based-assistant-answer-and-follow-ups-experienc", "foundation/parity-chance-ai-image-to-insight-camera-article-experience", "foundation/parity-chance-ai-signup-and-first-run-onboarding-experience", "foundation/parity-chance-ai-visual-product-lookup-experience", "foundation/parity-chargetrip-create-and-select-workspace-experience", "foundation/parity-chargetrip-schedule-account-deletion-24-hour-hold-experience", "foundation/parity-chargetrip-workspace-analytics-overview-experience", "foundation/parity-chatgpt-attach-photo-and-summarize-document-experience", "foundation/parity-chatgpt-conversation-search-open-experience", "foundation/parity-chatgpt-create-image-from-photos-experience", "foundation/parity-chatgpt-sign-in-to-chat-experience", "foundation/parity-chatgpt-sign-up-and-email-verification-experience", "foundation/parity-chatgpt-voice-prompt-to-generated-email-experience", "foundation/parity-chatgpt-voice-to-text-assistant-interaction-experience", "foundation/parity-checker-boards-view-and-list-switch-experience", "foundation/parity-citymapper-citymapper", "foundation/parity-claude-account-deletion-confirmation-experience", "foundation/parity-claude-interactive-budget-dashboard-creation-experience", "foundation/parity-claude-profile-settings-prompt-suggestions-experience", "foundation/parity-clime-first-run-onboarding-and-permissions-experience", "foundation/parity-clipchamp-project-sharing-link-invite-experience", "foundation/parity-clipchamp-submit-feedback-slide-over-experience", "foundation/parity-cocoonweaver-first-time-recording-and-transcription-experience", "foundation/parity-cofounder-daily-meeting-summary-test-experience", "foundation/parity-cohere-account-registration-and-onboarding-experience", "foundation/parity-cohere-document-attachment-in-conversation-experience", "foundation/parity-cohere-share-conversation-link-experience", "foundation/parity-cohere-tool-assisted-charting-with-python-interpreter-experienc", "foundation/parity-column-developer-onboarding-sandbox-setup-experience", "foundation/parity-comet-agent-card-gallery-experience", "foundation/parity-comet-browser-sync-settings-experience", "foundation/parity-comet-browser-sync-setup-experience", "foundation/parity-comet-comet-first-run-onboarding-and-sign-in-experience", "foundation/parity-comet-safari-data-import-experience", "foundation/parity-comet-sign-in-from-settings-experience", "foundation/parity-comet-theme-appearance-settings-experience", "foundation/parity-copperx-invite-earn-experience", "foundation/parity-craft-account-deletion-confirmation-experience", "foundation/parity-cursor-agent-prompt-composer-experience", "foundation/parity-cursor-change-privacy-to-privacy-mode-experience", "foundation/parity-cursor-configure-repository-routing-rule-experience", "foundation/parity-cursor-enterprise-sales-contact-form-experience", "foundation/parity-cushion-add-message-reaction-experience", "foundation/parity-cushion-cancel-invitation-experience", "foundation/parity-cushion-email-login-to-inbox-experience", "foundation/parity-cushion-inbox-management-experience", "foundation/parity-cushion-resolve-channel-post-experience", "foundation/parity-cushion-weekly-goals-check-in-experience", "foundation/parity-cushion-workspace-logo-upload-experience", "foundation/parity-cushion-workspace-members-management-experience", "foundation/parity-cushion-workspace-search-management-experience", "foundation/parity-cyberark-cyberark", "foundation/parity-dailyart-account-deletion-experience", "foundation/parity-dailyart-billing-country-selection-experience", "foundation/parity-dailyart-checkout-payment-details-experience", "foundation/parity-dailyart-checkout-payment-entry-experience", "foundation/parity-dailyart-create-account-and-patron-onboarding-experience", "foundation/parity-dailyart-magazine-account-authentication-and-registration-exper", "foundation/parity-dailyart-magazine-account-order-management-experience", "foundation/parity-dailyart-magazine-cart-review-and-empty-cart-recovery-experienc", "foundation/parity-dailyart-magazine-checkout-billing-payment-experience", "foundation/parity-dailyart-magazine-course-rating-and-review-experience", "foundation/parity-dailyart-magazine-guest-billing-payment-checkout-experience", "foundation/parity-dailyart-museum-selector-experience", "foundation/parity-dailyart-save-image-to-photos-experience", "foundation/parity-dailyart-settings-experience", "foundation/parity-daylish-day-timeline-filtering-experience", "foundation/parity-daylish-edit-scheduled-item-calendar-experience", "foundation/parity-daylish-first-time-onboarding-subscription-setup-experience", "foundation/parity-daylish-inline-event-edit-peeking-ahead-experience", "foundation/parity-daylish-subscription-paywall-purchase-experience", "foundation/parity-days-create-event-title-and-optional-calendar-import-experience", "foundation/parity-days-support-chat-conversation-experience", "foundation/parity-daze-account-creation-friend-onboarding-experience", "foundation/parity-daze-message-interaction-and-sticker-insertion-experience", "foundation/parity-daze-profile-sharing-confirmation-experience", "foundation/parity-delinea-delinea", "foundation/parity-deliveroo-deliveroo", "foundation/parity-delphi-discover-relevant-perspectives-experience", "foundation/parity-delphi-pdf-question-submission-experience", "foundation/parity-delphi-profile-personalization-experience", "foundation/parity-delphi-voice-reply-review-experience", "foundation/parity-denim-gallery-permissions-onboarding-experience", "foundation/parity-descript-account-deletion-experience", "foundation/parity-descript-remove-participant-from-meeting-experience", "foundation/parity-digg-account-deletion-experience", "foundation/parity-digg-newsletter-edition-selection-experience", "foundation/parity-digg-trending-story-voting-experience", "foundation/parity-district-add-card-experience", "foundation/parity-district-add-card-for-booking-payment-experience", "foundation/parity-district-basketball-booking-payment-experience", "foundation/parity-district-booking-payment-selection-experience", "foundation/parity-district-cinema-booking-payment-experience", "foundation/parity-district-closed-support-rating-experience", "foundation/parity-district-dining-save-and-reserve-experience", "foundation/parity-district-edit-profile-and-select-state-experience", "foundation/parity-district-edit-profile-details-experience", "foundation/parity-district-edit-profile-settings-experience", "foundation/parity-district-event-feedback-submission-experience", "foundation/parity-district-event-ticket-booking-experience", "foundation/parity-district-favourite-stores-management-experience", "foundation/parity-district-find-a-restaurant-experience", "foundation/parity-district-invite-guests-after-booking-experience", "foundation/parity-district-movie-booking-add-on-support-experience", "foundation/parity-district-movie-seat-selection-experience", "foundation/parity-district-movie-ticket-booking-experience", "foundation/parity-district-restaurant-q-a-assistant-experience", "foundation/parity-district-restaurant-question-assistant-experience", "foundation/parity-district-restaurant-review-submission-experience", "foundation/parity-district-restaurant-reviews-and-ratings-experience", "foundation/parity-district-restaurant-savings-calculator-experience", "foundation/parity-district-restaurant-table-booking-and-guest-invite-experience", "foundation/parity-district-search-discovery-experience", "foundation/parity-district-share-feedback-from-profile-experience", "foundation/parity-district-swimming-slot-reservation-experience", "foundation/parity-district-wallet-balance-and-history-experience", "foundation/parity-doctronic-ai-passwordless-ai-doctor-sign-in-experience", "foundation/parity-doctronic-ai-response-style-preference-experience", "foundation/parity-doctronic-ai-start-health-consultation-experience", "foundation/parity-doctronic-ai-timezone-selection-experience", "foundation/parity-doji-phone-verification-and-onboarding-experience", "foundation/parity-dona-task-dashboard-theme-switch-experience", "foundation/parity-doordash-cancel-scheduled-order-with-refund-and-feedback-experi", "foundation/parity-doordash-pickup-discovery-and-merchant-inspection-experience", "foundation/parity-doordash-pickup-results-filtering-experience", "foundation/parity-doordash-store-review-submission-experience", "foundation/parity-dopper-business-account-sign-in-and-dashboard-experience", "foundation/parity-dopper-store-locator-map-experience", "foundation/parity-douze-account-registration-and-onboarding-experience", "foundation/parity-dribbble-account-deletion-and-redirect-experience", "foundation/parity-dribbble-messages-read-experience", "foundation/parity-dribbble-set-work-availability-and-profile-details-experience", "foundation/parity-drinkit-cart-review-and-edit-experience", "foundation/parity-drinkit-post-order-rating-and-feedback-experience", "foundation/parity-dropbox-activity-log-filtering-experience", "foundation/parity-dropbox-add-file-to-folder-experience", "foundation/parity-dropbox-create-folder-view-link-experience", "foundation/parity-dropbox-create-new-folder-experience", "foundation/parity-dropbox-delete-dropbox-account-experience", "foundation/parity-dropbox-delete-selected-photos-experience", "foundation/parity-dropbox-delete-shared-file-experience", "foundation/parity-dropbox-document-scan-and-save-experience", "foundation/parity-dropbox-document-scan-crop-filter-experience", "foundation/parity-dropbox-dropbox-business-onboarding-experience", "foundation/parity-dropbox-file-upload-to-folder-experience", "foundation/parity-dropbox-folder-access-management-experience", "foundation/parity-dropbox-make-file-available-offline-experience", "foundation/parity-dropbox-new-account-first-run-home-experience", "foundation/parity-dropbox-password-recovery-via-email-experience", "foundation/parity-dropbox-pdf-editing-feedback-submission-experience", "foundation/parity-dropbox-recover-deleted-file-experience", "foundation/parity-dropbox-recover-deleted-files-search-experience", "foundation/parity-dropbox-resend-member-invite-experience", "foundation/parity-dropbox-search-with-type-filter-experience", "foundation/parity-dropbox-share-file-request-experience", "foundation/parity-dropbox-subscription-selection-and-trial-start-experience", "foundation/parity-dropbox-team-folder-upload-experience", "foundation/parity-dropbox-team-folder-viewing-link-settings-experience", "foundation/parity-dropbox-vault-setup-and-biometric-enablement-experience", "foundation/parity-dropset-apple-sign-in-and-profile-setup-experience", "foundation/parity-dropset-edit-workout-calendar-appearance-experience", "foundation/parity-dropset-exercise-selection-list-experience", "foundation/parity-dropset-post-message-in-public-chat-experience", "foundation/parity-dropset-workout-exercise-editor-experience", "foundation/parity-dub-contact-support-modal-experience", "foundation/parity-dub-in-product-help-search-experience", "foundation/parity-duo-duo", "foundation/parity-duolingo-account-login-and-learning-dashboard-experience", "foundation/parity-duolingo-friend-search-and-invite-experience", "foundation/parity-duolingo-friend-update-reaction-experience", "foundation/parity-duolingo-leaderboard-status-setup-experience", "foundation/parity-duolingo-practice-word-question-experience", "foundation/parity-duolingo-section-unlock-assessment-experience", "foundation/parity-elevenlabs-account-deletion-process-experience", "foundation/parity-elevenlabs-account-deletion-sign-out-experience", "foundation/parity-elevenlabs-caption-styling-and-font-selection-experience", "foundation/parity-elevenlabs-chat-conversation-rating-and-feedback-experience", "foundation/parity-elevenlabs-developer-usage-analytics-experience", "foundation/parity-elevenlabs-email-sign-up-first-run-onboarding-experience", "foundation/parity-elevenlabs-narration-editing-and-export-setup-experience", "foundation/parity-elevenlabs-record-or-upload-for-voice-isolation-experience", "foundation/parity-elevenlabs-request-license-for-iconic-voice-experience", "foundation/parity-elevenlabs-share-project-link-and-permissions-experience", "foundation/parity-elevenlabs-signup-experience", "foundation/parity-elevenlabs-text-to-speech-voice-selection-experience", "foundation/parity-elevenlabs-voice-isolation-upload-record-experience", "foundation/parity-elevenlabs-workspace-home-voice-actions-experience", "foundation/parity-elevenmusic-account-settings-edit-experience", "foundation/parity-elevenmusic-account-settings-feedback-experience", "foundation/parity-elevenmusic-artist-profile-editing-experience", "foundation/parity-elevenmusic-artist-profile-setup-experience", "foundation/parity-elevenmusic-audio-reference-recording-experience", "foundation/parity-elevenmusic-generated-tracks-management-experience", "foundation/parity-elevenmusic-invite-friends-referral-experience", "foundation/parity-elevenmusic-music-creation-dashboard-experience", "foundation/parity-elevenmusic-music-creation-workspace-experience", "foundation/parity-elevenmusic-pro-subscription-upgrade-experience", "foundation/parity-elevenmusic-subscription-upgrade-and-account-settings-experienc", "foundation/parity-elevenmusic-track-published-confirmation-experience", "foundation/parity-elevenmusic-track-selection-for-creation-experience", "foundation/parity-elevenmusic-track-studio-search-and-publish-experience", "foundation/parity-elevenreader-account-deletion-and-confirmation-experience", "foundation/parity-elevenreader-account-signup-and-personalization-experience", "foundation/parity-elevenreader-add-document-for-listening-experience", "foundation/parity-elevenreader-create-and-listen-to-narrated-text-experience", "foundation/parity-elevenreader-create-genfm-episode-experience", "foundation/parity-elevenreader-genre-based-book-filtering-experience", "foundation/parity-elevenreader-new-account-onboarding-experience", "foundation/parity-elevenreader-publish-manuscript-to-elevenreader-experience", "foundation/parity-elevenreader-scan-import-document-experience", "foundation/parity-elevenreader-set-daily-listening-goal-experience", "foundation/parity-elevenreader-share-bookcard-experience", "foundation/parity-elevenreader-share-referral-link-flow-experience", "foundation/parity-endless-onboarding-to-first-canvas-experience", "foundation/parity-endless-tools-3d-text-material-background-customization-experie", "foundation/parity-endless-tools-subscription-and-account-management-experience", "foundation/parity-enode-api-credential-generation-experience", "foundation/parity-enode-support-ticket-submission-experience", "foundation/parity-epidemic-sound-plan-selection-to-trial-activation-experience", "foundation/parity-epidemic-sound-select-and-play-stem-mix-experience", "foundation/parity-exa-developer-workspace-home-experience", "foundation/parity-exa-search-response-feedback-experience", "foundation/parity-exa-usage-date-range-selection-experience", "foundation/parity-exactly-ai-account-signup-to-prompt-composer-experience", "foundation/parity-excalidraw-text-properties-editing-experience", "foundation/parity-exoplan-edit-calendar-event-experience", "foundation/parity-exoplan-health-calendar-onboarding-experience", "foundation/parity-exoplan-in-app-feedback-rating-experience", "foundation/parity-exoplan-natural-language-event-creation-experience", "foundation/parity-exoplan-pro-subscription-purchase-experience", "foundation/parity-expedia-account-deletion-experience", "foundation/parity-expedia-report-a-guest-review-experience", "foundation/parity-factory-documentation-discovery-and-article-view-experience", "foundation/parity-factory-in-document-assistant-experience", "foundation/parity-factory-knowledge-droid-session-start-experience", "foundation/parity-factory-new-account-setup-and-onboarding-experience", "foundation/parity-faire-image-search-product-browse-experience", "foundation/parity-faire-kitchen-tabletop-product-browsing-experience", "foundation/parity-faire-product-reviews-sorting-experience", "foundation/parity-faire-product-sharing-experience", "foundation/parity-faire-request-unlisted-product-experience", "foundation/parity-faire-retailer-account-deactivation-experience", "foundation/parity-faire-seller-storefront-browsing-experience", "foundation/parity-faire-share-product-with-teammates-experience", "foundation/parity-faire-ship-date-selection-experience", "foundation/parity-faire-trending-products-browsing-experience", "foundation/parity-fal-delete-model-set-experience", "foundation/parity-fal-generation-sharing-settings-experience", "foundation/parity-fal-recent-history-display-experience", "foundation/parity-fal-recent-history-preview-visibility-experience", "foundation/parity-family-connect-wallet-theme-builder-experience", "foundation/parity-family-enable-authenticator-method-experience", "foundation/parity-family-explore-suggested-sites-experience", "foundation/parity-family-help-support-search-experience", "foundation/parity-family-report-bug-via-support-experience", "foundation/parity-family-sign-in-to-wallets-dashboard-experience", "foundation/parity-family-token-send-initiation-experience", "foundation/parity-farfetch-refer-a-friend-invite-experience", "foundation/parity-fernand-account-details-settings-experience", "foundation/parity-fernand-authenticator-mfa-setup-experience", "foundation/parity-fernand-conversation-search-experience", "foundation/parity-fernand-help-center-search-and-article-reading-experience", "foundation/parity-fernand-knowledge-base-search-experience", "foundation/parity-fernand-message-options-and-deletion-experience", "foundation/parity-fernand-send-new-support-email-experience", "foundation/parity-fernand-support-inbox-review-experience", "foundation/parity-fernand-support-tool-sign-in-experience", "foundation/parity-figma-about-section-text-edit-experience", "foundation/parity-figma-delete-file-from-recents-experience", "foundation/parity-figma-rename-team-header-experience", "foundation/parity-figma-slide-sharing-permissions-experience", "foundation/parity-figma-summarize-team-sticky-notes-experience", "foundation/parity-fingerprint-create-environment-with-usage-notification-experien", "foundation/parity-fingerprint-manage-email-notifications-experience", "foundation/parity-fingerprint-rename-workspace-experience", "foundation/parity-fingerprint-sign-in-and-getting-started-onboarding-experience", "foundation/parity-fingerprint-usage-insights-overview-experience", "foundation/parity-focusflight-measurement-unit-settings-experience", "foundation/parity-focusflight-pause-and-give-up-confirmation-experience", "foundation/parity-focuspomo-first-time-onboarding-and-timer-start-experience", "foundation/parity-focuspomo-icloud-sync-confirmation-help-experience", "foundation/parity-foodllama-daily-nutrition-review-experience", "foundation/parity-foodllama-foodllama-free-trial-subscription-experience", "foundation/parity-foodllama-meal-logging-and-scan-review-experience", "foundation/parity-foodllama-nutrition-and-body-stats-review-experience", "foundation/parity-foodllama-nutrition-insight-completion-experience", "foundation/parity-foodllama-weight-logging-and-progress-experience", "foundation/parity-foodllama-weight-logs-management-experience", "foundation/parity-foodllama-weight-progress-overview-experience", "foundation/parity-fourmula-ai-account-billing-and-credits-experience", "foundation/parity-fourmula-ai-account-sign-up-and-dashboard-entry-experience", "foundation/parity-fourmula-ai-background-generation-support-experience", "foundation/parity-fourmula-ai-project-collection-management-experience", "foundation/parity-fourthwall-creator-account-signup-and-onboarding-experience", "foundation/parity-fourthwall-marketing-to-creator-dashboard-experience", "foundation/parity-fresha-account-deletion-confirmation-experience", "foundation/parity-fresha-add-appointment-to-calendar-experience", "foundation/parity-fresha-buy-and-send-gift-card-experience", "foundation/parity-fresha-country-region-selector-experience", "foundation/parity-fresha-country-region-selector-experience-2", "foundation/parity-fresha-customer-signup-and-return-to-homepage-experience", "foundation/parity-fresha-edit-personal-profile-details-experience", "foundation/parity-fresha-email-social-sign-up-experience", "foundation/parity-fresha-help-center-search-and-results-experience", "foundation/parity-fresha-membership-purchase-and-checkout-experience", "foundation/parity-fresha-reviews-filtering-experience", "foundation/parity-fresha-sales-help-navigation-experience", "foundation/parity-fresha-salon-detail-listing-experience", "foundation/parity-fresha-search-and-filter-results-experience", "foundation/parity-fresha-search-availability-experience", "foundation/parity-fresha-search-book-venue-listings-experience", "foundation/parity-fresha-search-results-to-venue-details-experience", "foundation/parity-fresha-sign-in-to-personal-profile-experience", "foundation/parity-fuse-enable-notifications-and-offer-experience", "foundation/parity-fuse-new-wallet-signup-and-secure-key-setup-experience", "foundation/parity-fuse-recipient-selection-experience", "foundation/parity-genie-ask-genie-with-photos-experience", "foundation/parity-genie-conversation-photo-reimagining-experience", "foundation/parity-genie-dinner-recommendation-chat-experience", "foundation/parity-genie-photo-annotation-editor-experience", "foundation/parity-genie-voice-message-in-chat-experience", "foundation/parity-givingli-add-card-and-confirm-payment-experience", "foundation/parity-givingli-brand-kit-upgrade-experience", "foundation/parity-givingli-premium-subscription-checkout-experience", "foundation/parity-givingli-schedule-pay-for-gift-experience", "foundation/parity-givingli-send-and-view-a-digital-card-experience", "foundation/parity-gizmo-manage-conversation-from-inbox-experience", "foundation/parity-glorify-dashboard-onboarding-tour-experience", "foundation/parity-glorify-new-user-signup-and-workspace-setup-experience", "foundation/parity-glorify-start-subscription-trial-experience", "foundation/parity-glorify-subscription-cancellation-and-downgrade-experience", "foundation/parity-glossgenius-analytics-with-ai-growth-analyst-experience", "foundation/parity-glossgenius-client-quick-sale-cash-experience", "foundation/parity-glossgenius-sales-refund-handling-experience", "foundation/parity-glossgenius-send-sale-receipt-to-client-experience", "foundation/parity-go-club-theme-selection-and-preview-experience", "foundation/parity-gofundme-add-cover-media-experience", "foundation/parity-gofundme-donate-and-share-confirmation-experience", "foundation/parity-gofundme-fundraiser-details-and-organization-follow-experience", "foundation/parity-gofundme-fundraiser-settings-and-delete-confirmation-experience", "foundation/parity-gofundme-gofundme-account-creation-experience", "foundation/parity-gofundme-personal-profile-customization-experience", "foundation/parity-gofundme-profile-visibility-settings-experience", "foundation/parity-gofundme-report-content-submission-experience", "foundation/parity-gofundme-support-chat-feedback-experience", "foundation/parity-google-gemini-access-gemini-help-center-experience", "foundation/parity-google-gemini-create-public-share-link-experience", "foundation/parity-google-gemini-gemini-welcome-consent-experience", "foundation/parity-google-gemini-report-legal-content-from-gemini-chat-experience", "foundation/parity-google-gemini-request-java-ping-pong-example-experience", "foundation/parity-google-gemini-save-full-size-image-to-library-experience", "foundation/parity-google-gemini-voice-query-to-suggestions-experience", "foundation/parity-google-translate-camera-text-translation-experience", "foundation/parity-google-translate-download-language-pack-experience", "foundation/parity-google-translate-downloaded-languages-management-experience", "foundation/parity-google-translate-translation-home-settings-experience", "foundation/parity-google-translate-word-translation-details-experience", "foundation/parity-gov-uk-one-login-gov-uk-one-login", "foundation/parity-grammarly-account-deletion-confirmation-experience", "foundation/parity-grammarly-password-recovery-email-experience", "foundation/parity-grammarly-premium-plan-upgrade-annual-purchase-experience", "foundation/parity-grammarly-proofreading-suggestion-review-experience", "foundation/parity-grammarly-writing-performance-report-experience", "foundation/parity-granola-copy-generated-response-experience", "foundation/parity-granola-delete-account-experience", "foundation/parity-granola-granola-account-setup-experience", "foundation/parity-granola-search-chat-history-experience", "foundation/parity-grok-account-creation-and-verification-experience", "foundation/parity-grok-appearance-settings-experience", "foundation/parity-grok-attach-and-summarize-document-experience", "foundation/parity-grok-latest-news-assistant-query-experience", "foundation/parity-grok-private-chat-compose-and-receive-response-experience", "foundation/parity-ground-news-alerts-inbox-cleanup-experience", "foundation/parity-ground-news-edit-source-bias-rating-experience", "foundation/parity-ground-news-enable-local-news-experience", "foundation/parity-ground-news-help-center-search-experience", "foundation/parity-ground-news-night-mode-selection-experience", "foundation/parity-gumroad-help-center-find-articles-experience", "foundation/parity-gumroad-membership-charge-refund-experience", "foundation/parity-gumroad-seller-sign-in-and-dashboard-experience", "foundation/parity-handshake-career-interests-setup-experience", "foundation/parity-handshake-document-upload-preview-experience", "foundation/parity-handshake-event-registration-and-confirmation-experience", "foundation/parity-handshake-profile-edit-and-add-link-experience", "foundation/parity-handshake-report-and-archive-conversation-experience", "foundation/parity-handshake-student-signup-onboarding-experience", "foundation/parity-haptic-activity-timeline-experience", "foundation/parity-haptic-enable-health-access-experience", "foundation/parity-hashnode-account-deletion-process-experience", "foundation/parity-hashnode-email-magic-link-signup-and-first-run-setup-experience", "foundation/parity-hashnode-search-with-empty-state-transition-experience", "foundation/parity-hbo-max-hbo-max-subscription-and-personalization-experience", "foundation/parity-hbo-max-help-center-search-results-experience", "foundation/parity-hbo-max-replay-key-moments-navigation-experience", "foundation/parity-hbo-max-subscription-checkout-experience", "foundation/parity-hbo-max-subscription-signup-and-profile-setup-experience", "foundation/parity-hbo-max-subtitle-style-customization-experience", "foundation/parity-headspace-headspace", "foundation/parity-hello-ivy-asset-approval-experience", "foundation/parity-hello-ivy-delete-page-from-workspace-experience", "foundation/parity-hello-ivy-pause-notifications-temporarily-experience", "foundation/parity-hello-ivy-share-document-link-experience", "foundation/parity-hello-ivy-sign-in-to-project-overview-experience", "foundation/parity-hello-ivy-workspace-overview-dashboard-experience", "foundation/parity-hello-klean-product-review-submission-experience", "foundation/parity-homerun-candidate-interview-scheduler-experience", "foundation/parity-hype-account-deletion-experience", "foundation/parity-hype-article-comments-experience", "foundation/parity-hype-comment-discussion-panel-experience", "foundation/parity-hype-hype-app-sign-in-entry-experience", "foundation/parity-hype-password-recovery-and-reset-experience", "foundation/parity-hype-profile-completion-and-newsletter-selection-experience", "foundation/parity-hype-share-article-by-message-experience", "foundation/parity-hype-upcoming-product-details-experience", "foundation/parity-ikea-account-deletion-self-serve-experience", "foundation/parity-ikea-add-co-registrant-to-gift-registry-experience", "foundation/parity-ikea-back-in-stock-notification-signup-experience", "foundation/parity-ikea-chatbot-post-conversation-survey-experience", "foundation/parity-ikea-contact-support-chatbot-interaction-experience", "foundation/parity-ikea-select-collection-store-experience", "foundation/parity-ikea-select-delivery-or-collection-method-experience", "foundation/parity-ikea-share-design-via-qr-experience", "foundation/parity-ikea-write-a-product-review-experience", "foundation/parity-imprint-account-deletion-experience", "foundation/parity-imprint-commit-schedule-reminder-experience", "foundation/parity-imprint-new-user-onboarding-and-course-starter-experience", "foundation/parity-imprint-start-7-day-trial-subscribe-experience", "foundation/parity-incogni-incogni", "foundation/parity-incogni-incogni-part-2", "foundation/parity-incogni-incogni-part-3", "foundation/parity-incogni-incogni-part-4", "foundation/parity-incogni-ironwall-ironwall", "foundation/parity-instacart-add-product-to-cart-and-review-experience", "foundation/parity-instacart-flyer-notifications-settings-experience", "foundation/parity-instacart-in-store-pickup-checkout-and-confirmation-experience", "foundation/parity-instacart-instacart", "foundation/parity-instacart-recipe-store-selection-experience", "foundation/parity-instacart-referral-invite-experience", "foundation/parity-instacart-smart-shop-preferences-editor-experience", "foundation/parity-instagram-account-recovery-find-account-experience", "foundation/parity-instagram-account-switch-and-messages-experience", "foundation/parity-instagram-call-summary-and-rating-experience", "foundation/parity-instagram-delete-a-direct-message-experience", "foundation/parity-instagram-direct-message-audio-calling-experience", "foundation/parity-instagram-enable-face-id-with-pin-setup-experience", "foundation/parity-instagram-help-center-article-navigation-experience", "foundation/parity-instagram-in-call-camera-effects-experience", "foundation/parity-instagram-in-shop-product-search-and-empty-state-experience", "foundation/parity-instagram-in-shop-product-search-experience", "foundation/parity-instagram-instagram-account-signup-and-onboarding-experience", "foundation/parity-instagram-manage-favorites-experience", "foundation/parity-instagram-map-based-place-discovery-experience", "foundation/parity-instagram-post-ephemeral-note-experience", "foundation/parity-instagram-post-share-qr-generator-experience", "foundation/parity-instagram-profile-nametag-card-experience", "foundation/parity-instagram-reel-qr-share-color-selector-experience", "foundation/parity-instagram-report-a-reel-and-take-support-actions-experience", "foundation/parity-instagram-send-voice-note-experience", "foundation/parity-instagram-sign-in-to-instagram-experience", "foundation/parity-intercom-help-center-messenger-branding-experience", "foundation/parity-intercom-install-clearbit-reveal-integration-experience", "foundation/parity-intercom-installed-integrations-browse-experience", "foundation/parity-intercom-new-account-signup-and-verification-experience", "foundation/parity-intercom-profile-edit-and-photo-update-experience", "foundation/parity-intercom-profile-editing-photo-and-fields-experience", "foundation/parity-intercom-search-and-filter-articles-experience", "foundation/parity-intercom-subscription-purchase-advanced-plan-experience", "foundation/parity-intercom-subscription-purchase-experience", "foundation/parity-introspect-friend-decision-response-experience", "foundation/parity-introspect-impulse-purchase-cooldown-and-pro-subscription-exper", "foundation/parity-introspect-item-verification-share-experience", "foundation/parity-introspect-pro-subscription-purchase-experience", "foundation/parity-jace-ai-archive-message-mark-done-experience", "foundation/parity-jace-ai-find-messages-by-contact-experience", "foundation/parity-jace-ai-jace-trial-onboarding-experience", "foundation/parity-jace-ai-onboarding-assistant-conversation-experience", "foundation/parity-jasper-ai-brand-voice-preview-and-sample-generation-experience", "foundation/parity-jasper-ai-new-user-signup-to-dashboard-experience", "foundation/parity-jasper-ai-restore-deleted-content-experience", "foundation/parity-jasper-ai-slogan-creation-with-conversational-assistant-experie", "foundation/parity-jasper-ai-subscription-feedback-experience", "foundation/parity-joi-calendar-selection-experience", "foundation/parity-joi-first-launch-setup-experience", "foundation/parity-joi-subscribe-to-joi-pro-experience", "foundation/parity-julienne-account-signup-experience", "foundation/parity-julienne-manage-saved-recipe-experience", "foundation/parity-julienne-saved-recipe-search-experience", "foundation/parity-kann-in-game-coin-life-purchase-experience", "foundation/parity-kann-onboarding-and-kana-entry-experience", "foundation/parity-keycloak-keycloak", "foundation/parity-kickstarter-inbox-message-search-experience", "foundation/parity-kickstarter-project-search-and-empty-state-experience", "foundation/parity-kickstarter-sign-up-and-enable-project-reminders-experience", "foundation/parity-kin-appearance-settings-experience", "foundation/parity-kin-calendar-integration-permissions-experience", "foundation/parity-kin-onboarding-to-chat-reminder-setup-experience", "foundation/parity-kit-account-deletion-via-in-product-support-experience", "foundation/parity-kit-manage-recommendations-experience", "foundation/parity-kit-template-sharing-and-link-copy-experience", "foundation/parity-komoot-account-deletion-experience", "foundation/parity-komoot-delete-saved-route-experience", "foundation/parity-komoot-invite-participants-and-manage-qr-invite-experience", "foundation/parity-komoot-invite-participants-experience", "foundation/parity-komoot-manage-offline-data-experience", "foundation/parity-komoot-manage-offline-regions-experience", "foundation/parity-komoot-password-recovery-via-reset-email-experience", "foundation/parity-komoot-places-selector-experience", "foundation/parity-komoot-report-tip-comment-experience", "foundation/parity-komoot-revoke-pending-invitation-experience", "foundation/parity-komoot-save-and-download-route-experience", "foundation/parity-komoot-share-profile-qr-copy-link-experience", "foundation/parity-komoot-take-and-save-photo-during-navigation-experience", "foundation/parity-lego-builder-collaborative-build-party-experience", "foundation/parity-lego-builder-create-lego-account-and-consent-experience", "foundation/parity-lego-builder-qr-scanner-onboarding-and-scan-experience", "foundation/parity-lego-builder-scan-qr-to-open-set-experience", "foundation/parity-lego-builder-set-search-filter-experience", "foundation/parity-lego-find-lego-store-locator-experience", "foundation/parity-lemni-conversation-status-update-experience", "foundation/parity-lemni-inbox-conversation-triage-experience", "foundation/parity-leonardo-ai-ai-creation-intro-guide-experience", "foundation/parity-leonardo-ai-ai-music-generation-experience", "foundation/parity-leonardo-ai-essential-plan-upgrade-experience", "foundation/parity-leonardo-ai-image-iteration-and-prompt-edit-experience", "foundation/parity-leonardo-ai-password-reset-recovery-experience", "foundation/parity-leonardo-ai-profile-settings-management-experience", "foundation/parity-leonardo-ai-subscription-cancellation-and-feedback-experience", "foundation/parity-linear-account-preferences-and-theme-settings-experience", "foundation/parity-linear-assign-project-experience", "foundation/parity-linear-assign-teammate-to-task-experience", "foundation/parity-linear-attach-document-to-draft-experience", "foundation/parity-linear-customize-theme-colors-experience", "foundation/parity-linear-edit-issue-priority-experience", "foundation/parity-linear-issue-status-update-experience", "foundation/parity-linear-members-management-experience", "foundation/parity-linear-passwordless-email-sign-in-to-home-experience", "foundation/parity-linear-project-selector-experience", "foundation/parity-linear-project-statuses-management-experience", "foundation/parity-linear-quick-find-search-filter-experience", "foundation/parity-linear-sign-in-to-project-workspace-experience", "foundation/parity-linear-user-preferences-center-experience", "foundation/parity-linkedin-account-help-discovery-experience", "foundation/parity-linkedin-choose-career-premium-and-start-trial-experience", "foundation/parity-linkedin-easy-apply-to-a-job-experience", "foundation/parity-linkedin-new-member-onboarding-and-job-preferences-experience", "foundation/parity-linktree-digital-business-card-setup-experience", "foundation/parity-linktree-insights-dashboard-and-content-assistant-experience", "foundation/parity-linktree-linktree-qr-code-customization-experience", "foundation/parity-linktree-password-recovery-and-reset-experience", "foundation/parity-literal-club-account-creation-onboarding-experience", "foundation/parity-literal-club-add-books-to-yearly-reading-target-experience", "foundation/parity-literal-club-reading-goal-setup-experience", "foundation/parity-longevity-deck-air-purifier-community-commenting-experience", "foundation/parity-longevity-deck-comment-interactions-experience", "foundation/parity-longevity-deck-discover-search-filter-experience", "foundation/parity-longevity-deck-first-launch-introduction-experience", "foundation/parity-longevity-deck-product-comments-interactions-experience", "foundation/parity-longevity-deck-product-comments-management-experience", "foundation/parity-loom-new-account-setup-and-workspace-onboarding-experience", "foundation/parity-loowner-add-new-item-with-photo-and-receipt-experience", "foundation/parity-loowner-first-time-guided-tour-experience", "foundation/parity-loowner-inventory-search-and-filter-experience", "foundation/parity-loowner-new-user-onboarding-and-setup-experience", "foundation/parity-loowner-set-personal-goal-experience", "foundation/parity-lovable-create-landing-page-experience", "foundation/parity-lovable-design-preview-comments-experience", "foundation/parity-lovable-edit-profile-and-banner-experience", "foundation/parity-lovable-generated-landing-page-review-experience", "foundation/parity-lovable-help-and-resources-center-experience", "foundation/parity-lovable-lovable-sign-in-and-builder-entry-experience", "foundation/parity-lovable-missing-connector-request-experience", "foundation/parity-lovable-password-recovery-and-reset-experience", "foundation/parity-lovable-personalize-public-profile-experience", "foundation/parity-lovable-plan-review-and-approval-experience", "foundation/parity-lovable-plans-and-credits-management-experience", "foundation/parity-lovable-qr-asset-label-implementation-experience", "foundation/parity-lovable-referral-invite-modal-experience", "foundation/parity-lovable-start-project-prompt-experience", "foundation/parity-lovable-workspace-people-management-experience", "foundation/parity-lovable-workspace-project-search-experience", "foundation/parity-luminar-account-to-support-help-experience", "foundation/parity-luminar-plugin-compatibility-license-lookup-experience", "foundation/parity-luminar-support-search-activation-experience", "foundation/parity-luminary-account-creation-and-subscription-experience", "foundation/parity-luminary-onboarding-to-authenticated-home-experience", "foundation/parity-lummi-create-account-from-photo-browse-experience", "foundation/parity-macaw-keyword-lookup-by-url-experience", "foundation/parity-make-signup-experience", "foundation/parity-manus-account-profile-update-experience", "foundation/parity-manus-documentation-search-and-ai-query-experience", "foundation/parity-manus-help-center-search-and-browse-experience", "foundation/parity-manus-meeting-summary-share-document-experience", "foundation/parity-manus-send-referral-invite-experience", "foundation/parity-manus-upgrade-to-manus-pro-7-day-trial-experience", "foundation/parity-manus-voice-to-task-composer-experience", "foundation/parity-manychat-account-deletion-confirmation-experience", "foundation/parity-manychat-get-started-to-accounts-experience", "foundation/parity-manychat-inbox-message-reply-experience", "foundation/parity-mapbox-get-started-preferences-experience", "foundation/parity-mapbox-map-typography-font-selector-experience", "foundation/parity-mapbox-mapbox-account-creation-and-onboarding-experience", "foundation/parity-mapbox-publish-confirmation-modal-experience", "foundation/parity-mapbox-sign-in-to-account-overview-experience", "foundation/parity-marathon-invite-teammates-experience", "foundation/parity-matter-appearance-settings-experience", "foundation/parity-matter-bulk-tag-selected-queue-items-experience", "foundation/parity-matter-sign-in-to-queue-experience", "foundation/parity-maze-feedback-submission-experience", "foundation/parity-maze-in-meeting-chat-message-reaction-experience", "foundation/parity-maze-interview-session-supervision-experience", "foundation/parity-maze-new-account-onboarding-experience", "foundation/parity-maze-react-to-meeting-chat-message-experience", "foundation/parity-maze-report-export-to-pdf-experience", "foundation/parity-maze-share-response-reel-experience", "foundation/parity-md-vinyl-music-player-onboarding-experience", "foundation/parity-medium-account-deletion-experience", "foundation/parity-medium-author-mute-from-profile-experience", "foundation/parity-medium-edit-profile-header-image-experience", "foundation/parity-medium-image-alt-text-editor-experience", "foundation/parity-medium-passwordless-email-sign-in-to-personalized-feed-experien", "foundation/parity-medium-writer-follow-modal-experience", "foundation/parity-mela-change-appearance-theme-experience", "foundation/parity-mela-create-and-save-recipe-experience", "foundation/parity-mela-document-photo-capture-and-edit-experience", "foundation/parity-mela-scan-import-printed-recipe-experience", "foundation/parity-mela-share-recipe-to-messages-experience", "foundation/parity-melius-copy-asset-link-experience", "foundation/parity-melius-create-project-workspace-experience", "foundation/parity-melius-image-upload-assistance-experience", "foundation/parity-melius-melius-account-and-workspace-setup-experience", "foundation/parity-melius-send-referral-invitation-experience", "foundation/parity-melius-share-canvas-link-experience", "foundation/parity-melius-subscription-upgrade-experience", "foundation/parity-memotron-appearance-customization-experience", "foundation/parity-memotron-create-interview-with-audio-and-transcription-experien", "foundation/parity-mercury-application-completion-checklist-experience", "foundation/parity-mercury-bulk-payment-approval-experience", "foundation/parity-mercury-check-image-upload-experience", "foundation/parity-mercury-company-information-with-inline-help-experience", "foundation/parity-mercury-formation-document-upload-and-validation-experience", "foundation/parity-mercury-invite-and-onboard-team-member-experience", "foundation/parity-mercury-locate-help-article-and-view-detail-experience", "foundation/parity-mercury-preferred-name-edit-experience", "foundation/parity-mercury-recipient-banking-and-upload-experience", "foundation/parity-mercury-refer-a-contact-via-email-experience", "foundation/parity-mercury-schedule-payment-amount-recurrence-experience", "foundation/parity-mercury-sign-in-to-business-dashboard-experience", "foundation/parity-mercury-upload-and-verify-government-id-experience", "foundation/parity-meta-ai-ai-image-creation-experience", "foundation/parity-meta-ai-chat-response-feedback-experience", "foundation/parity-meta-ai-created-media-management-experience", "foundation/parity-meta-ai-gmail-connection-authorization-experience", "foundation/parity-meta-ai-gmail-email-audit-experience", "foundation/parity-meta-ai-incognito-ai-conversation-experience", "foundation/parity-meta-ai-meta-ai-account-sign-in-experience", "foundation/parity-meta-ai-meta-ai-problem-report-experience", "foundation/parity-meta-ai-meta-ai-sign-in-experience", "foundation/parity-meta-ai-mute-user-from-comment-menu-experience", "foundation/parity-meta-ai-mythical-creature-field-guide-experience", "foundation/parity-meta-ai-permanent-meta-account-deletion-experience", "foundation/parity-meta-ai-recipe-response-rating-experience", "foundation/parity-meta-ai-save-recipe-instructions-experience", "foundation/parity-meta-ai-trending-recipe-search-experience", "foundation/parity-microsoft-copilot-image-review-with-copilot-experience", "foundation/parity-microsoft-copilot-referral-rewards-modal-experience", "foundation/parity-microsoft-copilot-start-new-copilot-conversation-experience", "foundation/parity-microsoft-copilot-voice-conversation-with-message-copilot-exper", "foundation/parity-microsoft-entra-id-microsoft-entra-id", "foundation/parity-microsoft-marketplace-product-discovery-experience", "foundation/parity-microsoft-microsoft-365-plan-recommender-experience", "foundation/parity-microsoft-search-from-order-tracking-experience", "foundation/parity-microsoft-teams-change-presence-status-experience", "foundation/parity-microsoft-teams-chat-request-handling-experience", "foundation/parity-microsoft-teams-in-call-reactions-experience", "foundation/parity-microsoft-teams-post-call-feedback-survey-experience", "foundation/parity-miles-first-time-onboarding-and-health-permission-setup-experie", "foundation/parity-miles-preferences-running-appearance-settings-experience", "foundation/parity-miles-run-session-start-to-finish-experience", "foundation/parity-mindllama-activity-record-deletion-experience", "foundation/parity-mindllama-delete-activity-from-day-experience", "foundation/parity-mindllama-focus-session-setup-experience", "foundation/parity-mindllama-theme-selection-dark-mode-experience", "foundation/parity-miro-board-deletion-experience", "foundation/parity-missive-appearance-sidebar-settings-experience", "foundation/parity-missive-configure-swipe-actions-experience", "foundation/parity-missive-conversation-labeling-experience", "foundation/parity-missive-help-center-search-and-article-viewing-experience", "foundation/parity-missive-help-center-search-and-empty-result-state-experience", "foundation/parity-missive-schedule-snooze-for-email-experience", "foundation/parity-missive-search-and-refine-results-experience", "foundation/parity-missive-shared-account-setup-experience", "foundation/parity-missive-upgrade-to-starter-plan-experience", "foundation/parity-mocha-apps-dashboard-empty-state-and-account-menu-experience", "foundation/parity-mocha-gallery-review-and-feedback-experience", "foundation/parity-mocha-new-project-creation-experience", "foundation/parity-modyfi-in-context-commenting-attachment-thread-experience", "foundation/parity-modyfi-new-account-registration-and-first-run-onboarding-experi", "foundation/parity-moises-account-deletion-flow-experience", "foundation/parity-moises-subscription-upgrade-for-audio-separation-experience", "foundation/parity-monodesk-conversational-task-planner-experience", "foundation/parity-monodesk-daily-planner-mark-task-complete-experience", "foundation/parity-monodesk-kanban-task-completion-experience", "foundation/parity-moonlitt-first-launch-onboarding-and-subscription-flow-experien", "foundation/parity-moonly-subscription-purchase-ios-hero-offer-experience", "foundation/parity-mozi-invite-connect-people-experience", "foundation/parity-mozi-invite-via-qr-and-share-link-experience", "foundation/parity-mozi-mozi-onboarding-setup-experience", "foundation/parity-mozi-set-preferred-messaging-experience", "foundation/parity-mwm-account-deletion-experience", "foundation/parity-mwm-mwm-ai-workspace-onboarding-experience", "foundation/parity-mwm-product-request-refinement-experience", "foundation/parity-mymind-create-new-space-experience", "foundation/parity-n8n-account-deletion-permanent-flow-experience", "foundation/parity-n8n-message-composer-with-collaborator-experience", "foundation/parity-napper-account-management-delete-account-experience", "foundation/parity-napper-baby-schedule-invite-experience", "foundation/parity-napper-confirm-personal-details-experience", "foundation/parity-napper-contact-support-with-attachments-experience", "foundation/parity-napper-create-first-personalized-schedule-experience", "foundation/parity-napper-first-schedule-onboarding-experience", "foundation/parity-napper-invite-caregiver-to-baby-profile-experience", "foundation/parity-napper-journal-calendar-entry-experience", "foundation/parity-napper-sleep-session-event-review-experience", "foundation/parity-navan-add-company-credit-card-experience", "foundation/parity-navan-create-travel-policy-experience", "foundation/parity-navan-sign-in-to-admin-dashboard-experience", "foundation/parity-navia-30-day-marketing-plan-experience", "foundation/parity-navia-account-management-and-sign-out-experience", "foundation/parity-netflix-downloads-settings-experience", "foundation/parity-netflix-game-details-experience", "foundation/parity-netflix-new-account-signup-and-onboarding-experience", "foundation/parity-netflix-signup-and-first-time-onboarding-experience", "foundation/parity-neuecast-appearance-selection-experience", "foundation/parity-new-balance-checkout-and-order-confirmation-experience", "foundation/parity-nextdoor-change-appearance-theme-experience", "foundation/parity-nextdoor-contact-listing-seller-experience", "foundation/parity-nextdoor-copy-listing-for-sharing-experience", "foundation/parity-nextdoor-create-and-publish-listing-experience", "foundation/parity-nextdoor-delete-marketplace-listing-experience", "foundation/parity-nextdoor-mark-listing-sold-experience", "foundation/parity-nextdoor-marketplace-category-filtering-experience", "foundation/parity-nextdoor-publish-a-neighborhood-event-experience", "foundation/parity-nextdoor-publish-neighborhood-poll-experience", "foundation/parity-nextdoor-rate-ask-answer-experience", "foundation/parity-nextdoor-remove-chat-experience", "foundation/parity-nextdoor-reply-to-existing-chat-experience", "foundation/parity-nextdoor-share-listing-in-chat-experience", "foundation/parity-nike-checkout-chat-support-experience", "foundation/parity-nike-checkout-purchase-experience", "foundation/parity-nike-locate-view-nike-store-experience", "foundation/parity-nike-select-pickup-location-experience", "foundation/parity-nooka-account-creation-experience", "foundation/parity-nooka-nooka-trial-subscription-onboarding-experience", "foundation/parity-nooka-sign-in-to-join-live-session-experience", "foundation/parity-not-boring-camera-subscription-purchase-paywall-experience", "foundation/parity-not-boring-habits-remove-habit-check-experience", "foundation/parity-not-boring-weather-subscription-paywall-and-purchase-experience", "foundation/parity-nudea-report-review-experience", "foundation/parity-nudea-stockist-locator-experience", "foundation/parity-okta-okta", "foundation/parity-okuri-calendar-appearance-change-experience", "foundation/parity-okuri-collectors-edition-subscription-experience", "foundation/parity-okuri-collectors-edition-upgrade-experience", "foundation/parity-okuri-okuri-first-time-setup-experience", "foundation/parity-okuri-schedule-team-standup-experience", "foundation/parity-omio-booking-retrieval-by-reference-experience", "foundation/parity-omio-booking-retrieval-via-help-center-experience", "foundation/parity-omio-bus-ticket-selection-and-passenger-entry-experience", "foundation/parity-omio-checkout-ticket-retrieval-experience", "foundation/parity-omio-customer-help-ticket-support-experience", "foundation/parity-omio-fare-selection-and-add-on-experience", "foundation/parity-omio-first-time-onboarding-experience", "foundation/parity-omio-flight-extras-seat-selection-experience", "foundation/parity-omio-help-center-cancel-ticket-experience", "foundation/parity-omio-passenger-details-review-experience", "foundation/parity-omio-payment-selection-and-card-entry-experience", "foundation/parity-omio-seat-selection-experience", "foundation/parity-omio-ticket-booking-and-checkout-experience", "foundation/parity-on-apply-gift-card-at-checkout-experience", "foundation/parity-on-checkout-and-payment-completion-experience", "foundation/parity-on-checkout-payment-experience", "foundation/parity-on-country-region-selector-experience", "foundation/parity-on-favorite-toggle-on-category-listing-experience", "foundation/parity-on-friend-referral-sharing-experience", "foundation/parity-on-new-user-sign-up-and-onboarding-experience", "foundation/parity-on-order-cancellation-experience", "foundation/parity-on-order-details-review-experience", "foundation/parity-on-product-purchase-checkout-experience", "foundation/parity-on-refine-product-listing-by-filters-experience", "foundation/parity-on-restock-notification-signup-experience", "foundation/parity-on-sign-in-and-enter-home-experience", "foundation/parity-one-year-subscription-purchase-experience", "foundation/parity-onefinestay-contact-widget-message-submission-experience", "foundation/parity-onefinestay-sign-up-and-confirmation-experience", "foundation/parity-ontra-create-a-service-listing-experience", "foundation/parity-ontra-email-preference-settings-experience", "foundation/parity-ontra-import-existing-project-experience", "foundation/parity-ontra-passwordless-email-sign-in-to-dashboard-experience", "foundation/parity-ontra-project-follow-toggle-experience", "foundation/parity-ontra-request-recommendation-from-client-experience", "foundation/parity-open-sign-in-and-account-access-experience", "foundation/parity-openai-audio-chat-translator-experience", "foundation/parity-openai-cookie-preferences-management-experience", "foundation/parity-openai-image-insertion-via-url-experience", "foundation/parity-openai-playground-chat-voice-model-configuration-experience", "foundation/parity-openai-record-and-preview-voice-message-experience", "foundation/parity-openai-speech-to-multilingual-translation-with-tts-experience", "foundation/parity-openai-submit-feedback-on-assistant-message-experience", "foundation/parity-openai-tts-playground-generate-play-speech-experience", "foundation/parity-oppie-pro-subscription-purchase-experience", "foundation/parity-orbit-edit-subscription-details-experience", "foundation/parity-ozone-import-and-preview-media-experience", "foundation/parity-ozone-invite-share-project-experience", "foundation/parity-pampam-account-menu-to-empty-state-transition-experience", "foundation/parity-pampam-share-map-link-experience", "foundation/parity-parallel-add-profile-video-experience", "foundation/parity-parallel-candidate-profile-edit-modal-experience", "foundation/parity-parallel-executive-profile-enrichment-experience", "foundation/parity-parallel-profile-enrichment-setup-experience", "foundation/parity-patreon-email-code-sign-in-experience", "foundation/parity-patreon-privacy-erasure-request-experience", "foundation/parity-pattern-brands-account-creation-and-profile-landing-experience", "foundation/parity-pattern-brands-guest-checkout-with-shipping-and-payment-experie", "foundation/parity-paypal-account-deletion-and-data-removal-experience", "foundation/parity-paypal-activity-transaction-search-experience", "foundation/parity-paypal-add-tracking-information-experience", "foundation/parity-paypal-country-region-selector-experience", "foundation/parity-paypal-edit-about-you-experience", "foundation/parity-paypal-help-article-feedback-experience", "foundation/parity-paypal-help-center-search-to-article-experience", "foundation/parity-paypal-help-centre-search-and-article-experience", "foundation/parity-paypal-international-bank-transfer-setup-experience", "foundation/parity-paypal-password-recovery-and-verification-experience", "foundation/parity-paypal-paypal-assistant-conversation-experience", "foundation/parity-paypal-profile-editing-header-experience", "foundation/parity-paypal-save-personal-qr-code-experience", "foundation/parity-paypal-search-documentation-from-docs-page-experience", "foundation/parity-paypal-seller-qr-checkout-settings-experience", "foundation/parity-paypal-set-preferred-online-payment-experience", "foundation/parity-paypal-set-preferred-payment-method-experience", "foundation/parity-paypal-sign-in-and-biometric-enablement-experience", "foundation/parity-paypal-tip-jar-qr-setup-experience", "foundation/parity-paypal-transaction-activity-experience", "foundation/parity-paypal-transaction-filter-and-date-range-selection-experience", "foundation/parity-peak-performance-account-creation-right-slide-panel-experience", "foundation/parity-peak-performance-desktop-checkout-3-step-purchase-experience", "foundation/parity-peak-performance-sign-in-to-account-experience", "foundation/parity-peerlist-job-preferences-setup-experience", "foundation/parity-peerlist-start-and-send-direct-message-experience", "foundation/parity-phantom-buy-bitcoin-with-third-party-provider-experience", "foundation/parity-phantom-change-display-language-experience", "foundation/parity-phantom-create-new-wallet-account-experience", "foundation/parity-phantom-export-diagnostic-logs-experience", "foundation/parity-phantom-icon-selection-experience", "foundation/parity-phantom-new-wallet-setup-experience", "foundation/parity-phantom-send-via-qr-experience", "foundation/parity-pi-account-sign-in-and-status-experience", "foundation/parity-pi-export-personal-chat-history-experience", "foundation/parity-pi-new-user-onboarding-and-setup-experience", "foundation/parity-pi-pi-onboarding-and-chat-entry-experience", "foundation/parity-picnic-picnic-photo-organization-onboarding-experience", "foundation/parity-picnic-troubleshooting-hub-experience", "foundation/parity-pico-cam-premium-subscription-purchase-experience", "foundation/parity-pillowtalk-enable-night-mode-experience", "foundation/parity-pillowtalk-pillowtalk-plus-subscription-experience", "foundation/parity-pillowtalk-streak-sharing-experience", "foundation/parity-ping-identity-ping-identity", "foundation/parity-pinkllama-apple-health-sync-setup-experience", "foundation/parity-pinkllama-consume-and-complete-a-recommended-insight-experience", "foundation/parity-pinkllama-initial-onboarding-and-subscription-experience", "foundation/parity-pinkllama-subscription-purchase-trial-setup-experience", "foundation/parity-pinterest-account-deletion-experience", "foundation/parity-pinterest-add-and-mix-recorded-audio-experience", "foundation/parity-pinterest-business-manager-onboarding-tour-experience", "foundation/parity-pinterest-chat-feedback-on-help-article-experience", "foundation/parity-pinterest-hide-or-report-pin-from-feed-experience", "foundation/parity-pinterest-hide-pin-and-provide-feedback-experience", "foundation/parity-pinterest-pin-creation-onboarding-experience", "foundation/parity-pinterest-search-results-filter-by-content-type-experience", "foundation/parity-pinterest-search-results-profiles-empty-state-experience", "foundation/parity-pinterest-video-split-editor-experience", "foundation/parity-pitch-manage-account-settings-experience", "foundation/parity-plane-finder-appearance-map-color-settings-experience", "foundation/parity-plane-finder-flight-tracking-setup-experience", "foundation/parity-play-create-text-component-with-state-and-style-experience", "foundation/parity-playground-annotation-color-edit-experience", "foundation/parity-playground-report-template-issue-experience", "foundation/parity-playstation-playstation-account-signup-and-verification-experie", "foundation/parity-pocketshelf-book-progress-session-manager-experience", "foundation/parity-pocketshelf-data-backup-import-experience", "foundation/parity-pocketshelf-subscription-purchase-checkout-experience", "foundation/parity-podimo-audiobook-detail-and-download-management-experience", "foundation/parity-podimo-delete-comment-from-thread-experience", "foundation/parity-podimo-live-poll-participation-experience", "foundation/parity-podimo-podcast-comment-moderation-experience", "foundation/parity-podimo-podcast-comments-thread-experience", "foundation/parity-podimo-search-episodes-in-show-experience", "foundation/parity-podimo-settings-and-downloads-management-experience", "foundation/parity-podimo-show-sharing-to-messages-experience", "foundation/parity-polar-account-preferences-delete-confirmation-experience", "foundation/parity-polar-organization-review-and-submission-experience", "foundation/parity-polar-passwordless-sign-in-to-admin-overview-experience", "foundation/parity-polaroid-account-photo-profile-edit-experience", "foundation/parity-polaroid-email-magic-link-sign-in-experience", "foundation/parity-pool-place-details-to-directions-experience", "foundation/parity-pool-refero-mcp-item-review-experience", "foundation/parity-pop-site-invite-to-earn-pro-experience", "foundation/parity-pop-site-manage-my-personal-information-experience", "foundation/parity-pop-site-upgrade-to-pro-experience", "foundation/parity-poppy-create-location-reminder-experience", "foundation/parity-poppy-poppy-daily-preferences-setup-experience", "foundation/parity-poppy-poppy-subscription-upgrade-experience", "foundation/parity-poppy-recommended-connection-setup-experience", "foundation/parity-poppy-voice-event-creation-experience", "foundation/parity-portrait-portrait-editor-setup-experience", "foundation/parity-preply-account-deletion-confirmation-experience", "foundation/parity-preply-lesson-pre-join-and-entry-experience", "foundation/parity-preply-pre-class-device-check-experience", "foundation/parity-prime-video-browse-and-hide-title-experience", "foundation/parity-product-hunt-follow-product-profile-experience", "foundation/parity-product-hunt-permanent-account-deletion-experience", "foundation/parity-product-hunt-submit-a-product-review-experience", "foundation/parity-programa-meeting-summary-feedback-experience", "foundation/parity-programa-update-project-status-experience", "foundation/parity-project-indigo-project-indigo-first-launch-setup-experience", "foundation/parity-prose-help-center-search-and-faq-navigation-experience", "foundation/parity-prose-referral-program-share-confirm-experience", "foundation/parity-pryzm-pryzm-pro-subscription-experience", "foundation/parity-pushr-feedback-rating-experience", "foundation/parity-pushr-positive-rating-response-experience", "foundation/parity-pushr-pushr-first-workout-setup-experience", "foundation/parity-queue-account-deletion-confirmation-experience", "foundation/parity-rad-weather-forecast-preferences-experience", "foundation/parity-rad-weather-preference-personalization-experience", "foundation/parity-rad-weather-preference-setup-experience", "foundation/parity-rad-weather-premium-subscription-onboarding-experience", "foundation/parity-rad-weather-weather-theme-selection-experience", "foundation/parity-rarible-report-nft-listing-experience", "foundation/parity-raum-home-location-setup-experience", "foundation/parity-raum-travel-poster-palette-selection-experience", "foundation/parity-raycast-account-sign-in-onboarding-experience", "foundation/parity-raycast-new-user-signup-and-onboarding-experience", "foundation/parity-raycast-save-ai-message-to-note-experience", "foundation/parity-raycast-settings-theme-selector-experience", "foundation/parity-raycast-sign-in-to-main-experience", "foundation/parity-raycast-subscription-upgrade-and-purchase-experience", "foundation/parity-raycast-two-factor-setup-backup-experience", "foundation/parity-reclaim-ai-email-referral-invite-experience", "foundation/parity-reclaim-ai-personal-time-stats-review-experience", "foundation/parity-reclaim-ai-reclaim-account-setup-experience", "foundation/parity-reeder-article-search-and-filter-experience", "foundation/parity-reeder-display-layout-settings-experience", "foundation/parity-reeder-first-time-onboarding-setup-experience", "foundation/parity-reeder-in-feed-search-and-filter-experience", "foundation/parity-reeder-reader-typography-settings-experience", "foundation/parity-remy-calendar-event-enablement-experience", "foundation/parity-remy-remy-first-launch-onboarding-experience", "foundation/parity-remy-share-from-rating-prompt-experience", "foundation/parity-reown-cloud-onboarding-carousel-experience", "foundation/parity-reown-metamask-qr-connect-experience", "foundation/parity-reown-project-authentication-to-workspace-experience", "foundation/parity-reown-theme-and-notifications-settings-experience", "foundation/parity-replo-move-project-between-workspaces-experience", "foundation/parity-replo-password-recovery-to-projects-experience", "foundation/parity-replo-responsive-preview-control-experience", "foundation/parity-replo-sign-in-to-projects-experience", "foundation/parity-resend-mfa-enrollment-experience", "foundation/parity-revolut-download-revolut-app-via-sms-or-qr-experience", "foundation/parity-revolut-help-center-search-and-results-experience", "foundation/parity-revolut-help-centre-article-discovery-experience", "foundation/parity-revolut-joint-account-withdrawal-help-and-chat-support-experien", "foundation/parity-revolut-phone-number-sign-in-experience", "foundation/parity-revolut-revolut-login-and-welcome-onboarding-experience", "foundation/parity-revolut-vault-goal-deadline-setup-experience", "foundation/parity-rive-delete-account-confirmation-experience", "foundation/parity-rive-user-account-profile-editing-experience", "foundation/parity-riverside-export-recording-video-audio-experience", "foundation/parity-riverside-in-call-teleprompter-setup-use-experience", "foundation/parity-riverside-marketing-to-studio-sign-in-experience", "foundation/parity-riverside-pre-join-studio-invite-experience", "foundation/parity-riverside-recording-asset-export-experience", "foundation/parity-riverside-remove-project-experience", "foundation/parity-riverside-sign-up-first-run-onboarding-experience", "foundation/parity-riverside-snapshot-sharing-experience", "foundation/parity-riverside-studio-branding-editor-experience", "foundation/parity-riverside-subscribe-to-plus-annual-experience", "foundation/parity-riverside-subscription-management-cancellation-experience", "foundation/parity-roots-developer-tools-logs-experience", "foundation/parity-roots-new-user-onboarding-and-subscription-signup-experience", "foundation/parity-roots-resume-global-blocking-experience", "foundation/parity-roots-streak-introduction-and-detail-experience", "foundation/parity-roots-subscription-trial-onboarding-experience", "foundation/parity-rork-ai-mobile-product-creation-experience", "foundation/parity-rork-analytics-review-and-date-range-change-experience", "foundation/parity-rork-app-idea-creation-experience", "foundation/parity-rork-app-project-creation-dashboard-experience", "foundation/parity-rork-collaborator-invitation-experience", "foundation/parity-rork-spanish-learning-dashboard-experience", "foundation/parity-rork-vocabu-home-dashboard-experience", "foundation/parity-rork-vocabulary-dashboard-experience", "foundation/parity-rovr-archive-date-filtering-experience", "foundation/parity-rovr-archive-search-refinement-experience", "foundation/parity-rox-email-template-management-experience", "foundation/parity-rox-google-calendar-gmail-connect-experience", "foundation/parity-rox-manage-account-next-steps-experience", "foundation/parity-rox-manage-billing-and-upgrade-experience", "foundation/parity-rox-share-conversation-link-experience", "foundation/parity-runey-account-deletion-experience", "foundation/parity-runey-account-login-experience", "foundation/parity-runey-business-dashboard-overview-experience", "foundation/parity-runey-project-sharing-settings-experience", "foundation/parity-runey-project-status-update-experience", "foundation/parity-runey-project-tasks-overview-experience", "foundation/parity-runey-receipt-creation-experience", "foundation/parity-runey-task-status-filtering-experience", "foundation/parity-runey-task-time-tracking-on-board-experience", "foundation/parity-runey-update-profile-picture-experience", "foundation/parity-runway-create-account-with-verification-and-first-run-experienc", "foundation/parity-runway-empty-projects-image-library-experience", "foundation/parity-runway-text-to-image-generator-experience", "foundation/parity-sailpoint-sailpoint", "foundation/parity-saint-laurent-account-registration-and-confirmation-experience", "foundation/parity-saint-laurent-checkout-live-chat-assistance-experience", "foundation/parity-saint-laurent-store-appointment-booking-experience", "foundation/parity-savee-email-social-login-onboarding-experience", "foundation/parity-saviynt-saviynt", "foundation/parity-scape-brochure-request-and-confirmation-experience", "foundation/parity-sharewillow-sign-in-and-access-plans-experience", "foundation/parity-shazam-library-search-and-results-experience", "foundation/parity-shazam-location-based-event-filtering-experience", "foundation/parity-shop-delete-order-history-item-experience", "foundation/parity-shop-lamp-product-search-experience", "foundation/parity-shop-mark-order-delivered-experience", "foundation/parity-shop-product-reporting-experience", "foundation/parity-shop-product-reviews-experience", "foundation/parity-shopify-analytics-dashboard-tour-experience", "foundation/parity-shopify-change-theme-typography-experience", "foundation/parity-shopify-export-products-csv-experience", "foundation/parity-shopify-install-klaviyo-from-marketplace-experience", "foundation/parity-shopify-purchase-order-mark-as-ordered-experience", "foundation/parity-shopify-request-weekly-store-metrics-experience", "foundation/parity-shopify-shopify", "foundation/parity-showcase-discover-search-content-experience", "foundation/parity-showcase-showcase-plus-subscription-sign-up-experience", "foundation/parity-shuttle-create-and-open-new-space-experience", "foundation/parity-skiff-appearance-settings-experience", "foundation/parity-skiff-create-team-workspace-experience", "foundation/parity-skiff-delete-document-to-trash-experience", "foundation/parity-skiff-select-and-move-messages-to-trash-experience", "foundation/parity-sky-guide-date-time-selector-experience", "foundation/parity-sky-guide-pro-subscription-purchase-experience", "foundation/parity-skyscanner-account-deletion-experience", "foundation/parity-slite-product-roadmap-sharing-experience", "foundation/parity-sora-account-data-export-experience", "foundation/parity-sora-account-profile-deletion-experience", "foundation/parity-sora-cameo-preferences-editor-experience", "foundation/parity-sora-comment-reply-to-specific-commenter-experience", "foundation/parity-sora-onboarding-to-first-feed-experience", "foundation/parity-sora-profile-avatar-update-experience", "foundation/parity-sora-profile-header-editing-experience", "foundation/parity-sora-send-and-play-messages-experience", "foundation/parity-soundcloud-profile-editing-experience", "foundation/parity-soundcloud-search-results-track-discovery-experience", "foundation/parity-soundtrap-guest-call-recording-experience", "foundation/parity-spline-account-deletion-experience", "foundation/parity-spline-create-first-team-project-experience", "foundation/parity-spline-invite-collaborator-experience", "foundation/parity-spoil-me-add-listing-image-experience", "foundation/parity-spoil-me-invite-and-earn-experience", "foundation/parity-spotify-search-and-discover-artist-content-experience", "foundation/parity-spotify-spotify-new-user-onboarding-experience", "foundation/parity-spotify-spotify-premium-plan-cancellation-experience", "foundation/parity-spotify-submit-episode-reply-experience", "foundation/parity-spyglass-audience-scaling-recommendations-experience", "foundation/parity-spyglass-creative-strategy-follow-up-experience", "foundation/parity-square-merchant-onboarding-and-account-setup-experience", "foundation/parity-square-report-a-review-experience", "foundation/parity-square-virtual-terminal-quick-charge-experience", "foundation/parity-ssense-account-registration-and-login-entry-experience", "foundation/parity-ssense-add-credit-card-during-checkout-experience", "foundation/parity-ssense-cart-checkout-address-payment-experience", "foundation/parity-ssense-checkout-add-address-card-experience", "foundation/parity-ssense-new-customer-sign-up-experience", "foundation/parity-ssense-onboarding-department-selection-notifications-experience", "foundation/parity-ssense-password-recovery-and-reset-experience", "foundation/parity-stardust-apple-sign-in-to-cycle-dashboard-experience", "foundation/parity-stardust-cycle-day-logging-experience", "foundation/parity-stardust-stardust-first-run-sign-up-onboarding-experience", "foundation/parity-steel-account-deletion-experience", "foundation/parity-steel-connect-github-account-experience", "foundation/parity-steel-create-and-monitor-browser-session-experience", "foundation/parity-steel-create-organization-experience", "foundation/parity-steel-create-usage-alert-experience", "foundation/parity-stellate-account-deletion-confirmation-experience", "foundation/parity-stellate-billing-plan-upgrade-to-business-experience", "foundation/parity-stellate-sign-in-to-organization-services-experience", "foundation/parity-stellate-user-profile-edit-experience", "foundation/parity-stemz-record-audio-import-experience", "foundation/parity-steps-activity-summary-experience", "foundation/parity-steps-appearance-mode-selection-experience", "foundation/parity-steps-download-steps-qr-experience", "foundation/parity-steps-filter-monthly-activities-by-type-experience", "foundation/parity-steps-first-run-setup-pro-purchase-experience", "foundation/parity-steps-history-list-calendar-toggle-experience", "foundation/parity-stocktwits-profile-edit-personal-info-preferences-experience", "foundation/parity-strava-strava", "foundation/parity-stridegate-onboarding-experience", "foundation/parity-stripe-mailchimp-installation-and-permissions-experience", "foundation/parity-stripe-test-data-review-experience", "foundation/parity-structured-checklist-modal-interaction-experience", "foundation/parity-structured-customization-settings-experience", "foundation/parity-structured-daily-timeline-checklist-experience", "foundation/parity-structured-first-use-onboarding-setup-experience", "foundation/parity-structured-handwritten-planner-extraction-experience", "foundation/parity-structured-help-center-getting-started-guide-experience", "foundation/parity-structured-scan-parse-schedule-experience", "foundation/parity-structured-structured-ai-onboarding-experience", "foundation/parity-structured-structured-ai-suggestions-review-experience", "foundation/parity-structured-theme-layout-customization-experience", "foundation/parity-subo-icloud-sync-status-check-experience", "foundation/parity-substack-appearance-settings-experience", "foundation/parity-substack-appearance-visual-preferences-experience", "foundation/parity-substack-article-typography-customization-experience", "foundation/parity-substack-comment-thread-interactions-experience", "foundation/parity-substack-create-and-send-first-chat-thread-experience", "foundation/parity-substack-global-search-with-empty-note-results-experience", "foundation/parity-substack-hide-note-from-feed-experience", "foundation/parity-substack-post-moderation-controls-experience", "foundation/parity-substack-restack-a-post-experience", "foundation/parity-substack-send-photo-in-direct-message-experience", "foundation/parity-sunsama-ai-assisted-task-capture-experience", "foundation/parity-sunsama-referral-invitation-experience", "foundation/parity-synthesia-io-ai-presentation-review-and-summary-experience", "foundation/parity-synthesia-io-avatar-consent-recording-experience", "foundation/parity-synthesia-io-copy-project-share-link-experience", "foundation/parity-synthesia-io-discovery-call-practice-experience", "foundation/parity-synthesia-io-password-recovery-experience", "foundation/parity-synthesia-io-profile-picture-update-experience", "foundation/parity-tango-remove-draft-to-trash-experience", "foundation/parity-tango-submit-button-step-review-experience", "foundation/parity-tango-subscription-plan-purchase-experience", "foundation/parity-tango-team-library-display-toggle-experience", "foundation/parity-tango-upgrade-to-tango-pro-experience", "foundation/parity-tango-viewer-activity-modal-experience", "foundation/parity-teal-job-tracker-onboarding-experience", "foundation/parity-teal-set-weekly-application-target-experience", "foundation/parity-teal-signup-to-resume-builder-onboarding-experience", "foundation/parity-teal-target-job-preferences-experience", "foundation/parity-teal-work-style-assessment-experience", "foundation/parity-telegram-in-chat-voice-call-experience", "foundation/parity-telegram-manage-conversation-experience", "foundation/parity-telegram-telegram-web-sign-in-experience", "foundation/parity-telescope-copy-referral-link-experience", "foundation/parity-telescope-join-and-customize-membership-experience", "foundation/parity-tella-manage-and-delete-videos-experience", "foundation/parity-tella-new-account-signup-and-welcome-experience", "foundation/parity-ten-ten-invite-first-friend-experience", "foundation/parity-the-leap-creator-store-signup-and-setup-experience", "foundation/parity-the-leap-referral-sharing-center-experience", "foundation/parity-the-outsiders-first-launch-setup-experience", "foundation/parity-the-outsiders-perceived-exertion-rating-experience", "foundation/parity-the-outsiders-training-goal-creation-experience", "foundation/parity-the-outsiders-workouts-overview-experience", "foundation/parity-the-washington-post-birthday-front-page-lookup-experience", "foundation/parity-the-washington-post-change-display-theme-experience", "foundation/parity-the-washington-post-save-recommended-story-experience", "foundation/parity-threads-change-appearance-theme-settings-experience", "foundation/parity-threads-edit-profile-change-avatar-experience", "foundation/parity-threads-enable-private-profile-experience", "foundation/parity-threads-manage-profile-notifications-experience", "foundation/parity-threads-profile-bio-edit-experience", "foundation/parity-threads-profile-editing-photo-settings-experience", "foundation/parity-threads-remove-post-from-feed-experience", "foundation/parity-tidal-account-deletion-experience", "foundation/parity-tidal-apple-sign-in-to-home-experience", "foundation/parity-tidal-artist-profile-and-follow-experience", "foundation/parity-tidal-audio-playback-preferences-experience", "foundation/parity-tidal-music-subscription-purchase-experience", "foundation/parity-tidal-track-report-reason-selection-experience", "foundation/parity-tide-ambient-timer-tagging-experience", "foundation/parity-tide-email-sign-up-to-subscription-onboarding-experience", "foundation/parity-tide-save-download-meditation-experience", "foundation/parity-tide-sleep-recordings-manager-experience", "foundation/parity-tidycal-create-booking-type-experience", "foundation/parity-tidycal-pro-subscription-upgrade-experience", "foundation/parity-tidycal-provider-account-setup-experience", "foundation/parity-tiktok-account-deletion-and-verification-experience", "foundation/parity-tiktok-activity-status-visibility-experience", "foundation/parity-tiktok-add-and-select-card-at-checkout-experience", "foundation/parity-tiktok-checkout-add-address-payment-experience", "foundation/parity-tiktok-direct-message-settings-experience", "foundation/parity-tiktok-direct-message-stickers-reactions-experience", "foundation/parity-tiktok-edit-profile-modal-experience", "foundation/parity-tiktok-find-and-open-place-profile-experience", "foundation/parity-tiktok-image-based-product-search-experience", "foundation/parity-tiktok-manage-offline-videos-experience", "foundation/parity-tiktok-offline-videos-manager-experience", "foundation/parity-tiktok-password-reset-via-email-experience", "foundation/parity-tiktok-remove-visual-effect-from-clip-experience", "foundation/parity-tiktok-sign-up-and-verify-via-mobile-experience", "foundation/parity-tiktok-visual-product-search-camera-experience", "foundation/parity-tilt-countdown-reveal-experience", "foundation/parity-time2book-class-reservation-at-studio-experience", "foundation/parity-time2book-edit-account-profile-experience", "foundation/parity-tinder-account-creation-and-first-time-onboarding-experience", "foundation/parity-tinder-account-pause-and-deletion-options-experience", "foundation/parity-tinder-help-center-account-appeals-experience", "foundation/parity-tinder-photo-verification-consent-experience", "foundation/parity-tinder-photo-verification-video-selfie-experience", "foundation/parity-tinder-profile-quiz-setup-ask-me-about-experience", "foundation/parity-tinder-report-a-safety-concern-experience", "foundation/parity-tinder-safety-basics-help-experience", "foundation/parity-tinder-swipe-gesture-onboarding-experience", "foundation/parity-tinder-unmatch-conversation-experience", "foundation/parity-tone-save-edited-image-experience", "foundation/parity-too-good-to-go-too-good-to-go", "foundation/parity-touch-time-support-tip-purchase-experience", "foundation/parity-train-fitness-comment-list-and-composer-experience", "foundation/parity-train-fitness-exercise-search-with-filter-chips-experience", "foundation/parity-train-fitness-export-workout-data-experience", "foundation/parity-train-fitness-feed-sentiment-prompt-experience", "foundation/parity-train-fitness-user-profile-actions-experience", "foundation/parity-train-fitness-workout-detail-experience", "foundation/parity-travelperk-edit-personal-profile-experience", "foundation/parity-travelperk-export-people-list-experience", "foundation/parity-travelperk-new-account-signup-experience", "foundation/parity-trip-chatbot-feedback-submission-experience", "foundation/parity-trip-flight-status-lookup-experience", "foundation/parity-trip-group-tour-booking-experience", "foundation/parity-trip-hotel-property-review-and-room-selection-experience", "foundation/parity-trip-hotel-room-selection-experience", "foundation/parity-trip-hotel-search-results-filtering-experience", "foundation/parity-trip-tour-search-results-experience", "foundation/parity-trip-traveler-profile-and-trip-moments-experience", "foundation/parity-trip-way-add-and-import-documents-experience", "foundation/parity-trip-way-location-item-creation-experience", "foundation/parity-trip-way-manage-documents-experience", "foundation/parity-trip-way-onboarding-pro-upgrade-experience", "foundation/parity-trip-way-pro-subscription-upgrade-experience", "foundation/parity-tripadvisor-account-registration-and-sign-in-experience", "foundation/parity-tripadvisor-city-hotel-search-and-filtering-experience", "foundation/parity-tripadvisor-restaurant-discovery-and-reservation-experience", "foundation/parity-tripsy-account-deletion-experience", "foundation/parity-tripsy-filtered-places-list-by-category-experience", "foundation/parity-tripsy-onboarding-and-first-launch-experience", "foundation/parity-tripsy-open-help-article-from-settings-experience", "foundation/parity-tripsy-password-recovery-from-failed-login-experience", "foundation/parity-tripsy-places-quest-voyage-experience", "foundation/parity-tripsy-sign-up-to-my-trips-experience", "foundation/parity-tripsy-trip-date-selection-experience", "foundation/parity-tripsy-tripsy-pro-purchase-experience", "foundation/parity-turf-appearance-theme-selection-experience", "foundation/parity-turf-favorite-team-picker-experience", "foundation/parity-turf-phone-sign-in-onboarding-experience", "foundation/parity-twelvelabs-video-analysis-composer-experience", "foundation/parity-twitch-account-deletion-experience", "foundation/parity-twitch-buy-twitch-bits-experience", "foundation/parity-twitch-follow-alert-customization-experience", "foundation/parity-twitch-report-chat-messages-experience", "foundation/parity-twitch-send-customized-cheer-experience", "foundation/parity-twitch-set-stream-reminder-experience", "foundation/parity-twitch-stream-notification-preference-experience", "foundation/parity-typeform-account-deletion-with-email-mfa-experience", "foundation/parity-typeform-account-signup-and-onboarding-to-builder-experience", "foundation/parity-typeform-global-organization-search-experience", "foundation/parity-uber-eats-delivery-address-configuration-experience", "foundation/parity-uber-sign-in-to-access-rentals-experience", "foundation/parity-uglycash-change-language-setting-experience", "foundation/parity-uglycash-new-user-onboarding-and-verification-experience", "foundation/parity-uglycash-switch-to-dark-theme-experience", "foundation/parity-uglycash-trading-clubs-leaderboard-experience", "foundation/parity-understory-booking-check-in-experience", "foundation/parity-understory-email-permissions-download-experience", "foundation/parity-understory-storefront-branding-settings-experience", "foundation/parity-unhinged-export-data-and-share-experience", "foundation/parity-unhinged-monthly-calendar-filtering-experience", "foundation/parity-unhinged-person-notes-search-and-open-experience", "foundation/parity-untitled-account-deletion-via-typed-confirmation-experience", "foundation/parity-untitled-edit-profile-change-username-experience", "foundation/parity-unwind-start-pro-trial-experience", "foundation/parity-urban-outfitters-product-review-submission-experience", "foundation/parity-urban-outfitters-store-locator-search-map-interaction-experienc", "foundation/parity-variant-multi-artifact-mobile-concept-generation-experience", "foundation/parity-verse-explore-search-and-create-verse-experience", "foundation/parity-verse-feedback-issue-submission-experience", "foundation/parity-verse-text-overlay-editor-experience", "foundation/parity-verse-user-search-and-no-results-create-prompt-experience", "foundation/parity-verse-wallpaper-editor-arrange-overlays-experience", "foundation/parity-vimeo-account-deletion-privacy-settings-experience", "foundation/parity-vimeo-add-media-to-project-experience", "foundation/parity-vimeo-getting-started-tour-experience", "foundation/parity-vimeo-live-event-studio-setup-experience", "foundation/parity-vimeo-new-user-sign-up-and-first-time-onboarding-experience", "foundation/parity-vimeo-video-editing-and-export-experience", "foundation/parity-vinyls-appearance-settings-experience", "foundation/parity-vinyls-edit-purchase-place-experience", "foundation/parity-vinyls-monthly-listening-history-experience", "foundation/parity-vinyls-onboarding-and-sign-in-experience", "foundation/parity-vinyls-qr-sign-in-and-onboarding-experience", "foundation/parity-vinyls-subscription-purchase-and-confirmation-experience", "foundation/parity-visitors-account-deletion-confirmation-experience", "foundation/parity-visitors-passwordless-sign-in-to-analytics-experience", "foundation/parity-visitors-people-list-period-filter-experience", "foundation/parity-vocabulary-personalization-trial-onboarding-experience", "foundation/parity-voicenotes-account-deletion-request-experience", "foundation/parity-voicenotes-ask-ai-from-filtered-notes-experience", "foundation/parity-voicenotes-calendar-streak-review-experience", "foundation/parity-voicenotes-configure-ask-ai-search-experience", "foundation/parity-voicenotes-email-sign-in-experience", "foundation/parity-voicenotes-premium-trial-purchase-experience", "foundation/parity-voicenotes-suggest-a-fix-for-transcript-experience", "foundation/parity-voicenotes-teams-upgrade-and-billing-experience", "foundation/parity-voila-buy-indonesia-1gb-data-experience", "foundation/parity-voila-map-place-picker-experience", "foundation/parity-volkshotel-hourly-work-cabin-reservation-experience", "foundation/parity-volv-email-password-recovery-experience", "foundation/parity-volv-search-from-profile-experience", "foundation/parity-vsco-import-and-open-images-in-studio-experience", "foundation/parity-wabi-hidden-gems-search-experience", "foundation/parity-wabi-post-comments-experience", "foundation/parity-wabi-recipe-comments-experience", "foundation/parity-wabi-recipe-rating-and-note-experience", "foundation/parity-wabi-voice-input-transcription-experience", "foundation/parity-wabi-wabi-search-and-discovery-experience", "foundation/parity-waterllama-apple-health-sync-setup-experience", "foundation/parity-waterllama-hydration-goal-tracking-experience", "foundation/parity-waterllama-hydration-onboarding-first-sip-flow-experience", "foundation/parity-waterllama-hydration-summary-experience", "foundation/parity-wealthsimple-account-creation-and-phone-verification-experience", "foundation/parity-wealthsimple-help-centre-hub-experience", "foundation/parity-wealthsimple-help-centre-search-and-browse-experience", "foundation/parity-wealthsimple-language-preference-change-experience", "foundation/parity-wealthsimple-purchase-tax-plan-experience", "foundation/parity-wealthsimple-self-serve-2024-tax-setup-experience", "foundation/parity-weather-mini-weather-mini-plus-purchase-experience", "foundation/parity-weekrise-embedded-support-chat-conversation-experience", "foundation/parity-weekrise-new-account-onboarding-setup-experience", "foundation/parity-west-elm-check-local-availability-experience", "foundation/parity-west-elm-checkout-shipping-pickup-selection-experience", "foundation/parity-west-elm-holiday-search-results-experience", "foundation/parity-whatsapp-active-call-controls-experience", "foundation/parity-whatsapp-add-whatsapp-contact-experience", "foundation/parity-whatsapp-announcement-replies-experience", "foundation/parity-whatsapp-announcement-replies-thread-experience", "foundation/parity-whatsapp-archive-chat-experience", "foundation/parity-whatsapp-calls-onboarding-experience", "foundation/parity-whatsapp-chat-theme-preview-experience", "foundation/parity-whatsapp-chat-theme-selection-experience", "foundation/parity-whatsapp-delete-chat-message-experience", "foundation/parity-whatsapp-document-search-in-whatsapp-experience", "foundation/parity-whatsapp-edit-profile-details-experience", "foundation/parity-whatsapp-encrypted-call-controls-experience", "foundation/parity-whatsapp-group-voice-conversation-experience", "foundation/parity-whatsapp-list-theme-customization-experience", "foundation/parity-whatsapp-meta-ai-message-feedback-experience", "foundation/parity-whatsapp-meta-ai-voice-chat-setup-experience", "foundation/parity-whatsapp-meta-ai-voice-selection-experience", "foundation/parity-whatsapp-multi-file-document-sharing-experience", "foundation/parity-whatsapp-region-search-selector-experience", "foundation/parity-whatsapp-send-documents-in-chat-experience", "foundation/parity-whatsapp-share-call-link-experience", "foundation/parity-whatsapp-switch-to-video-call-experience", "foundation/parity-whatsapp-take-photo-in-chat-experience", "foundation/parity-whatsapp-update-about-status-experience", "foundation/parity-whatsapp-voice-message-recording-experience", "foundation/parity-whatsapp-weather-assistant-response-experience", "foundation/parity-whatsapp-whatsapp-call-link-sharing-and-join-experience", "foundation/parity-whatsapp-whatsapp-call-transition-experience", "foundation/parity-whatsapp-whatsapp-qr-scanner-experience", "foundation/parity-whatsapp-whatsapp-web-sign-in-experience", "foundation/parity-whereby-account-deletion-from-profile-experience", "foundation/parity-whereby-meeting-prejoin-and-entry-experience", "foundation/parity-whereby-new-account-signup-and-room-creation-experience", "foundation/parity-whereby-pre-call-settings-setup-experience", "foundation/parity-willow-action-button-dictation-setup-experience", "foundation/parity-willow-offline-mode-settings-experience", "foundation/parity-willow-offline-mode-setup-experience", "foundation/parity-willow-referral-card-sharing-experience", "foundation/parity-willow-voice-assisted-message-creation-experience", "foundation/parity-willow-voice-dictation-onboarding-experience", "foundation/parity-willow-willow-home-dashboard-experience", "foundation/parity-willow-willow-onboarding-intro-experience", "foundation/parity-willow-willow-onboarding-introduction-experience", "foundation/parity-willow-willow-pro-subscription-purchase-experience", "foundation/parity-wise-help-centre-search-and-results-experience", "foundation/parity-wise-invite-friends-experience", "foundation/parity-wishwish-account-deletion-experience", "foundation/parity-wispr-flow-account-deletion-experience", "foundation/parity-wispr-flow-default-dictation-style-experience", "foundation/parity-wispr-flow-email-dictation-tutorial-experience", "foundation/parity-wispr-flow-flow-pro-subscription-upgrade-experience", "foundation/parity-wittl-job-description-image-upload-experience", "foundation/parity-womp-new-3d-project-creation-experience", "foundation/parity-wrike-in-chat-meeting-scheduling-and-gif-send-experience", "foundation/parity-wrike-resources-export-configuration-experience", "foundation/parity-wrike-subscription-plan-selection-checkout-experience", "foundation/parity-wrike-trial-signup-and-workspace-onboarding-experience", "foundation/parity-wurrd-new-user-signup-and-premium-conversion-experience", "foundation/parity-x-account-creation-and-onboarding-experience", "foundation/parity-x-account-signup-and-initial-profile-setup-experience", "foundation/parity-x-block-participant-in-direct-messages-experience", "foundation/parity-x-community-scoped-search-experience", "foundation/parity-x-edit-profile-header-and-details-experience", "foundation/parity-x-login-and-recovery-experience", "foundation/parity-x-members-directory-search-experience", "foundation/parity-x-message-to-call-and-back-experience", "foundation/parity-x-password-recovery-via-confirmation-code-experience", "foundation/parity-x-profile-edit-update-header-and-details-experience", "foundation/parity-x-unsent-dm-with-retry-delete-experience", "foundation/parity-x-upgrade-to-verified-premium-experience", "foundation/parity-xbox-report-a-review-experience", "foundation/parity-xchat-appearance-settings-experience", "foundation/parity-xchat-conversation-filtering-and-mark-read-experience", "foundation/parity-xchat-message-request-acceptance-experience", "foundation/parity-xchat-password-recovery-and-reset-experience", "foundation/parity-xchat-pin-conversation-experience", "foundation/parity-xchat-sign-up-and-chat-onboarding-experience", "foundation/parity-xchat-start-new-conversation-experience", "foundation/parity-xchat-username-sign-in-to-chat-experience", "foundation/parity-yelp-report-a-business-post-experience", "foundation/parity-youtube-analytics-assistant-help-experience", "foundation/parity-youtube-community-post-deletion-experience", "foundation/parity-youtube-music-artist-search-results-experience", "foundation/parity-youtube-music-comment-reply-sharing-experience", "foundation/parity-youtube-music-configure-smart-downloads-experience", "foundation/parity-youtube-music-download-single-track-from-album-experience", "foundation/parity-youtube-music-manage-offline-downloads-experience", "foundation/parity-youtube-music-podcast-experience", "foundation/parity-youtube-music-report-a-comment-experience", "foundation/parity-youtube-music-single-track-download-and-feedback-experience", "foundation/parity-youtube-music-voice-search-to-artist-results-experience", "foundation/parity-youtube-search-filters-selection-experience", "foundation/parity-youtube-search-refinement-and-filter-application-experience", "foundation/parity-youtube-shorts-comment-moderation-experience", "foundation/parity-youtube-smart-downloads-onboarding-and-management-experience", "foundation/parity-youtube-voice-search-interaction-experience", "foundation/parity-zara-check-order-status-via-chat-experience", "foundation/parity-zara-checkout-with-gift-pickup-selection-experience", "foundation/parity-zara-home-gift-wrapped-item-checkout-experience", "foundation/parity-zara-home-shipping-support-conversation-experience", "foundation/parity-zara-store-locator-search-and-select-store-experience", "foundation/parity-zendesk-guided-demo-walkthrough-experience", "foundation/parity-zendesk-handle-incoming-support-ticket-experience", "foundation/parity-zendesk-help-center-article-read-and-search-experience", "foundation/parity-zendesk-in-context-help-search-and-open-article-experience", "foundation/parity-zendesk-personal-settings-editor-experience", "foundation/parity-zendesk-staff-security-settings-change-password-experience", "foundation/parity-zendesk-subscription-purchase-and-customization-experience", "foundation/parity-zendesk-talk-configuration-and-reporting-experience", "foundation/problem-12-000-fine-for-cheltenham-waste-operator-for-disp", "foundation/problem-a-28-year-old-doge-operative-gutted-usaid-hr-had-n", "foundation/problem-a-complete-bypass-of-emudhra-s-1fa-otp-login-syste", "foundation/problem-about-the-compromised-emudhra-intermediate-cas", "foundation/problem-air-pollution-trends-reports", "foundation/problem-azure-s-weakest-link-five-full-cross-tenant-compro", "foundation/problem-bedfordshire-soldiers-laid-to-rest-on-the-western", "foundation/problem-binance-hacker-1029-5-6b-stolen-from-binance", "foundation/problem-blockchain-intelligence-firms-link-north-korea-to", "foundation/problem-britain-s-housing-crisis-won-t-be-fixed-without-a", "foundation/problem-bulgaria-votes-as-russian-attacks-in-black-sea-rea", "foundation/problem-burnham-ready-to-support-poundland-staff-as-retail", "foundation/problem-capital-gains-tax-returns-for-disposals-of-uk-land", "foundation/problem-codex-cli-installs-a-persistent-self-updating-back", "foundation/problem-companies-house-third-party-code-of-conduct", "foundation/problem-company-director-disqualification-following-breach", "foundation/problem-compromised-litellm-1-107-0-on-pypi-installs-a-cre", "foundation/problem-cops-can-bypass-iphone-s-automatic-reboot-to-get-i", "foundation/problem-cve-2026-111943-compromised-intermediate-certifica", "foundation/problem-defending-odihr-s-independence-and-the-osce-human", "foundation/problem-dungeon-generation-in-the-gnomish-mines-of-nethack", "foundation/problem-emudhra-intermediate-ca-compromise", "foundation/problem-excise-duty-penalties-for-wrongdoing", "foundation/problem-eye-of-sauron-long-range-hidden-spy-camera-detecti", "foundation/problem-florida-s-hoas-and-condo-associations-are-making-t", "foundation/problem-git-3-0-s-upcoming-sha-256-default-will-be-a-costl", "foundation/problem-guidance-chickerell-firing-times", "foundation/problem-guidance-tregantle-firing-times", "foundation/problem-gwyndod-supply-chain-mobilised-to-deliver-uk-s-fir", "foundation/problem-hackers-expose-the-identities-of-more-than-40k-dhs", "foundation/problem-how-a-1-deal-dragged-poundland-s-owner-into-a-poli", "foundation/problem-how-oil-rich-alberta-s-economy-might-fare-if-it-br", "foundation/problem-how-the-military-will-hack-the-election", "foundation/problem-i-m-on-a-war-footing-terry-tao-and-the-crisis-in-m", "foundation/problem-icc-judge-on-what-u-s-sanctions-mean-for-her-and-g", "foundation/problem-imposing-sanctions-on-the-international-criminal-c", "foundation/problem-independent-review-of-the-prison-system-terms-of-r", "foundation/problem-investigation-launched-as-dispute-risks-bringing-d", "foundation/problem-it-took-me-nine-years-to-hack-my-xbox-360", "foundation/problem-jared-kushner-hacked-phone-reveals-secret-negotiat", "foundation/problem-london-loft-and-extension-company-which-left-unfin", "foundation/problem-making-registration-applications-easier-and-more-e", "foundation/problem-middlesex-regiment-soldier-laid-to-rest-on-the-som", "foundation/problem-modernising-the-correction-of-errors", "foundation/problem-nicotine-approval-scheme", "foundation/problem-no-more-icons-in-every-menu-by-default", "foundation/problem-northflank-breached", "foundation/problem-opencode-hacked-user-credentials-leaked", "foundation/problem-professor-akande-brings-the-independence-legal-rig", "foundation/problem-publishing-details-of-deliberate-defaulters-reform", "foundation/problem-radioactivity-in-food-and-the-environment-rife-rep", "foundation/problem-restrictions-on-the-performance-of-high-risk-cosme", "foundation/problem-show-hn-wallhop-12ft-io-is-gone-so-i-built-a-repla", "foundation/problem-telegram-desktop-vulnerability-allowed-any-user-s", "foundation/problem-tell-hn-scribd-breached", "foundation/problem-the-6-3b-binance-hack-was-pulled-off-in-just-20-mi", "foundation/problem-today-we-hacked-what-we-d-set-out-to-hack-the-ipho", "foundation/problem-uk-sanctions-senior-commanders-implicated-in-campa", "foundation/problem-uk-ukraine-and-latvia-extend-crackdown-on-russian", "foundation/problem-us-federal-budget-deficit-jumps-12-in-one-year-to", "foundation/problem-us-imposes-sanctions-on-icc-hours-after-former-jud", "foundation/problem-various-projects-find-hidden-sdr-capabilities-in-e", "foundation/problem-we-re-going-to-need-default-hard-budget-caps-on-pr", "foundation/problem-why-north-korea-s-elite-hackers-are-the-prime-susp", "foundation/problem-why-windows-11-still-ships-with-windows-95-s-memor", "foundation/problem-zod-4-6-6-compromised-by-malware", "judiverse/findings-competitors-regulators-reuse-0-1-labels", "judiverse/findings-competitors-regulators-reuse-0-2-limits-of-this-research-read", "judiverse/findings-competitors-regulators-reuse-0-3-acronyms-and-terms-defined-o", "judiverse/findings-competitors-regulators-reuse-0-3-acronyms-and-terms-defined-o-2", "judiverse/findings-competitors-regulators-reuse-2-2-ai-assistants-and-agent-plat", "judiverse/findings-competitors-regulators-reuse-2-4-meetings-and-scheduling", "judiverse/findings-competitors-regulators-reuse-4-3-lifecycle-creation-versionin", "judiverse/findings-competitors-regulators-reuse-4-5-prompt-injection", "judiverse/findings-competitors-regulators-reuse-5-2-candidate-systems-by-judiver", "judiverse/findings-competitors-regulators-reuse-judiverse-competitors-regulators", "judiverse/findings-mobbin-flows-code-review", "judiverse/findings-mobbin-flows-coverage-what-mobbin-held-and-what-it-did-not-1-2", "judiverse/findings-mobbin-flows-judiverse-research-real-product-flows-from-mobbi", "judiverse/findings-refero-flows-judiverse-research-real-product-flows-from-refer", "judiverse/findings-refero-flows-task-access-review-permissions-audit-log-1-2", "judiverse/findings-refero-flows-task-access-review-permissions-audit-log-2-2", "judiverse/findings-refero-flows-task-ai-agent-setup-permissions", "judiverse/findings-refero-flows-task-ai-agent-setup-permissions-2", "judiverse/findings-refero-flows-task-ai-agent-setup-permissions-3", "judiverse/findings-refero-flows-task-ai-chat-assistant-1-2", "judiverse/findings-refero-flows-task-ai-chat-assistant-2", "judiverse/findings-refero-flows-task-api-keys-developer-settings", "judiverse/findings-refero-flows-task-apply-for-loan-credit-1-2", "judiverse/findings-refero-flows-task-approval-workflow", "judiverse/findings-refero-flows-task-approval-workflow-2", "judiverse/findings-refero-flows-task-approval-workflow-3", "judiverse/findings-refero-flows-task-book-gp-appointment-1-2", "judiverse/findings-refero-flows-task-business-account-onboarding-company-verific-1-2", "judiverse/findings-refero-flows-task-calendar-scheduling-meeting-1-2", "judiverse/findings-refero-flows-task-calendar-scheduling-meeting-2", "judiverse/findings-refero-flows-task-code-repository-pull-request-1-2", "judiverse/findings-refero-flows-task-code-repository-pull-request-2", "judiverse/findings-refero-flows-task-contract-management-1-2", "judiverse/findings-refero-flows-task-crm-contacts-pipeline", "judiverse/findings-refero-flows-task-crm-contacts-pipeline-2", "judiverse/findings-refero-flows-task-doctor-video-consultation-1-2", "judiverse/findings-refero-flows-task-document-editor-collaboration", "judiverse/findings-refero-flows-task-document-editor-collaboration-2", "judiverse/findings-refero-flows-task-document-editor-collaboration-3", "judiverse/findings-refero-flows-task-document-editor-collaboration-4", "judiverse/findings-refero-flows-task-email-inbox", "judiverse/findings-refero-flows-task-email-inbox-2", "judiverse/findings-refero-flows-task-email-inbox-3", "judiverse/findings-refero-flows-task-file-sharing-storage", "judiverse/findings-refero-flows-task-fraud-alert-suspicious-transaction", "judiverse/findings-refero-flows-task-grant-application", "judiverse/findings-refero-flows-task-health-onboarding-profile-1-2", "judiverse/findings-refero-flows-task-health-record-medical-history-1-2", "judiverse/findings-refero-flows-task-insurance-claim-health-1-2", "judiverse/findings-refero-flows-task-integrations-connect-apps", "judiverse/findings-refero-flows-task-integrations-connect-apps-2", "judiverse/findings-refero-flows-task-interview-scheduling", "judiverse/findings-refero-flows-task-meeting-notes-transcription", "judiverse/findings-refero-flows-task-music-player-playlist", "judiverse/findings-refero-flows-task-notifications-inbox", "judiverse/findings-refero-flows-task-search-everything", "judiverse/findings-refero-flows-task-security-incident-alert", "judiverse/findings-refero-flows-task-shipment-tracking-logistics", "judiverse/findings-refero-flows-task-task-project-management", "judiverse/findings-refero-flows-task-task-project-management-2", "judiverse/findings-refero-flows-task-team-chat-messaging", "judiverse/findings-refero-flows-task-triage-questionnaire-1-2", "judiverse/findings-refero-flows-task-upload-track-artist", "judiverse/findings-refero-flows-task-workflow-automation-builder", "judiverse/findings-refero-flows-task-workflow-automation-builder-2", "judiverse/parity-15five-15five", "judiverse/parity-222-assign-close-friends-judiverse", "judiverse/parity-222-people-selection-for-itinerary-judiverse", "judiverse/parity-6-uk-data-judiverse", "judiverse/parity-a-paper-titled-judiverse", "judiverse/parity-a2a-judiverse", "judiverse/parity-aboard-onboardee-dashboard-chat-judiverse", "judiverse/parity-aboard-vacation-plan-approval-judiverse", "judiverse/parity-abode-voice-note-transcription-and-reaction-judiverse", "judiverse/parity-acctual-manage-approval-rules-delete-judiverse", "judiverse/parity-acctual-select-or-add-invoice-recipient-judiverse", "judiverse/parity-activepieces-activepieces", "judiverse/parity-adobe-photoshop-generative-policy-acknowledgement-judiverse", "judiverse/parity-agno-agno", "judiverse/parity-ai-omnibus-commission-judiverse", "judiverse/parity-airbnb-notification-dismissal-judiverse", "judiverse/parity-airbnb-photographer-booking-judiverse", "judiverse/parity-airbnb-support-chat-message-judiverse", "judiverse/parity-airtable-airtable", "judiverse/parity-airtable-record-created-email-notifier-judiverse", "judiverse/parity-allbirds-ai-assisted-shoe-purchase-judiverse", "judiverse/parity-amazon-sagemaker-ai-amazon-sagemaker-ai", "judiverse/parity-anam-api-key-creation-judiverse", "judiverse/parity-anam-persona-setup-and-sharing-judiverse", "judiverse/parity-anthropic-claude-code-judiverse", "judiverse/parity-anthropic-s-agentic-misalignment-judiverse", "judiverse/parity-anthropic-warns-cowork-judiverse", "judiverse/parity-apache-airflow-apache-airflow", "judiverse/parity-appflowy-judiverse", "judiverse/parity-apple-image-playground-report-generated-image-judiverse", "judiverse/parity-appwrite-api-key-removal-from-integrations-judiverse", "judiverse/parity-appwrite-create-development-key-judiverse", "judiverse/parity-appwrite-enterprise-contact-and-confirmation-judiverse", "judiverse/parity-around-live-transcript-during-gallery-meeting-judiverse", "judiverse/parity-article-50-transparency-judiverse", "judiverse/parity-asana-asana", "judiverse/parity-avoma-avoma", "judiverse/parity-basecamp-basecamp", "judiverse/parity-bentoml-bentoml", "judiverse/parity-bereal-connect-music-service-judiverse", "judiverse/parity-bezi-signup-first-time-onboarding-judiverse", "judiverse/parity-bigbluebutton-judiverse", "judiverse/parity-binsoo-settings-clear-cache-judiverse", "judiverse/parity-blackbird-referral-invite-judiverse", "judiverse/parity-bond-restaurant-recommendation-approval-judiverse", "judiverse/parity-box-box-club-account-app-settings-judiverse", "judiverse/parity-brink-enable-health-integration-judiverse", "judiverse/parity-bump-friend-discovery-and-invite-judiverse", "judiverse/parity-byborre-media-kits-catalog-and-search-judiverse", "judiverse/parity-cake-equity-grant-existing-share-offer-judiverse", "judiverse/parity-cal-com-judiverse", "judiverse/parity-cal-new-user-onboarding-and-setup-judiverse", "judiverse/parity-cal-request-reschedule-for-booking-judiverse", "judiverse/parity-calendly-copy-email-availability-suggestions-judiverse", "judiverse/parity-calendly-email-availability-sharing-judiverse", "judiverse/parity-canva-bring-google-drive-files-into-design-judiverse", "judiverse/parity-canva-manage-connected-social-accounts-judiverse", "judiverse/parity-canva-revise-cover-title-via-voice-or-text-judiverse", "judiverse/parity-canva-tokyo-travel-vlog-cover-creation-judiverse", "judiverse/parity-carbon-playlist-save-and-confirm-judiverse", "judiverse/parity-carbon-queue-reorder-judiverse", "judiverse/parity-chance-ai-whiteboard-capture-and-ai-summary-judiverse", "judiverse/parity-chargetrip-contact-sales-lead-capture-judiverse", "judiverse/parity-chatgpt-judiverse", "judiverse/parity-chatgpt-new-conversation-onboarding-response-judiverse", "judiverse/parity-chatgpt-persona-customization-judiverse", "judiverse/parity-chatgpt-voice-prompt-to-generated-email-judiverse", "judiverse/parity-claude-code-claude-code", "judiverse/parity-claude-managed-agents-judiverse", "judiverse/parity-clay-nexus-contact-recommendation-judiverse", "judiverse/parity-clickup-clickup", "judiverse/parity-clipchamp-add-music-track-to-timeline-judiverse", "judiverse/parity-clipchamp-onedrive-backup-for-project-media-judiverse", "judiverse/parity-cocoonweaver-cocoon-list-and-sorting-judiverse", "judiverse/parity-cocoonweaver-re-transcribe-audio-note-judiverse", "judiverse/parity-cocoonweaver-share-transcription-via-messages-judiverse", "judiverse/parity-cofounder-daily-meeting-summary-test-judiverse", "judiverse/parity-cofounder-sign-up-to-workspace-onboarding-judiverse", "judiverse/parity-cohere-api-request-composer-judiverse", "judiverse/parity-cohere-chat-code-snippet-language-selector-judiverse", "judiverse/parity-cohere-document-based-conversation-review-judiverse", "judiverse/parity-cohere-document-review-conversation-judiverse", "judiverse/parity-cohere-document-summarization-with-attachments-judiverse", "judiverse/parity-cohere-inspect-and-export-chat-request-judiverse", "judiverse/parity-cohere-platform-cohere-platform", "judiverse/parity-cohere-tool-assisted-charting-with-python-interpreter-judiverse", "judiverse/parity-collabora-online-judiverse", "judiverse/parity-collective-management-of-judiverse", "judiverse/parity-comet-agent-card-gallery-judiverse", "judiverse/parity-composio-judiverse", "judiverse/parity-confluence-confluence", "judiverse/parity-copperx-deposit-and-cross-chain-bridge-to-usdc-judiverse", "judiverse/parity-craft-inline-collaborative-whiteboard-judiverse", "judiverse/parity-craft-invite-collaborator-to-document-judiverse", "judiverse/parity-crewai-crewai", "judiverse/parity-crewai-judiverse", "judiverse/parity-crossref-rest-api-crossref-rest-api", "judiverse/parity-cursor-agent-prompt-composer-judiverse", "judiverse/parity-cursor-enterprise-sales-contact-form-judiverse", "judiverse/parity-cursor-generate-and-review-agents-md-judiverse", "judiverse/parity-cushion-draft-message-save-judiverse", "judiverse/parity-cushion-inbox-management-judiverse", "judiverse/parity-cushion-workspace-content-search-judiverse", "judiverse/parity-cushion-workspace-search-management-judiverse", "judiverse/parity-customer-io-customer-io", "judiverse/parity-cyrus-imap-judiverse", "judiverse/parity-daylish-create-and-configure-event-judiverse", "judiverse/parity-daylish-edit-scheduled-event-time-judiverse", "judiverse/parity-daylish-reschedule-single-event-judiverse", "judiverse/parity-days-create-visual-event-card-judiverse", "judiverse/parity-daytona-daytona", "judiverse/parity-daze-new-conversation-creation-judiverse", "judiverse/parity-deep-agents-deep-agents", "judiverse/parity-deepseek-api-deepseek-api", "judiverse/parity-deezer-home-flow-listening-and-favourite-judiverse", "judiverse/parity-deezer-personalised-flow-selection-judiverse", "judiverse/parity-deezer-personalized-track-exclusion-judiverse", "judiverse/parity-deezer-playlist-quiz-playthrough-judiverse", "judiverse/parity-descript-live-room-screen-share-with-chat-judiverse", "judiverse/parity-dify-dify", "judiverse/parity-digital-omnibus-proposal-judiverse", "judiverse/parity-disputed-terms-judiverse", "judiverse/parity-distribution-deadline-directive-judiverse", "judiverse/parity-district-closed-support-rating-judiverse", "judiverse/parity-district-event-feedback-submission-judiverse", "judiverse/parity-district-movie-share-sheet-judiverse", "judiverse/parity-district-salon-service-booking-judiverse", "judiverse/parity-district-search-discovery-judiverse", "judiverse/parity-doctronic-ai-sleep-concern-assessment-judiverse", "judiverse/parity-dropbox-add-date-text-to-scanned-form-judiverse", "judiverse/parity-dropbox-add-file-to-folder-judiverse", "judiverse/parity-dropbox-add-folder-tag-judiverse", "judiverse/parity-dropbox-add-photos-to-mobile-uploads-judiverse", "judiverse/parity-dropbox-copy-uploaded-file-link-judiverse", "judiverse/parity-dropbox-create-automated-folder-judiverse", "judiverse/parity-dropbox-create-new-folder-judiverse", "judiverse/parity-dropbox-create-team-folder-judiverse", "judiverse/parity-dropbox-delete-shared-file-judiverse", "judiverse/parity-dropbox-dropbox-upgrade-prompt-judiverse", "judiverse/parity-dropbox-folder-access-management-judiverse", "judiverse/parity-dropbox-make-file-available-offline-judiverse", "judiverse/parity-dropbox-share-file-from-folder-judiverse", "judiverse/parity-dropbox-share-file-request-judiverse", "judiverse/parity-dropbox-share-file-request-link-judiverse", "judiverse/parity-dropbox-star-folder-from-files-judiverse", "judiverse/parity-dropbox-subscription-selection-and-trial-start-judiverse", "judiverse/parity-dropbox-team-file-type-filter-judiverse", "judiverse/parity-dropset-post-message-in-public-chat-judiverse", "judiverse/parity-echoleak-cve-2025-32711-judiverse", "judiverse/parity-elevenlabs-api-key-deletion-judiverse", "judiverse/parity-elevenlabs-create-and-configure-api-key-judiverse", "judiverse/parity-elevenlabs-create-text-to-speech-project-judiverse", "judiverse/parity-elevenlabs-image-generation-settings-judiverse", "judiverse/parity-elevenlabs-project-deletion-from-home-judiverse", "judiverse/parity-elevenmusic-account-settings-feedback-judiverse", "judiverse/parity-elevenmusic-add-song-to-playlist-judiverse", "judiverse/parity-elevenmusic-artist-identity-setup-judiverse", "judiverse/parity-elevenmusic-custom-style-audio-upload-judiverse", "judiverse/parity-elevenmusic-edit-track-details-judiverse", "judiverse/parity-elevenmusic-music-creation-workspace-judiverse", "judiverse/parity-elevenmusic-music-library-browsing-judiverse", "judiverse/parity-elevenmusic-music-track-selection-judiverse", "judiverse/parity-elevenmusic-playlist-track-actions-judiverse", "judiverse/parity-elevenmusic-remix-discovery-and-playback-judiverse", "judiverse/parity-elevenmusic-track-selection-for-creation-judiverse", "judiverse/parity-elevenreader-add-document-for-listening-judiverse", "judiverse/parity-elevenreader-search-collections-judiverse", "judiverse/parity-elicit-systematic-review-elicit-systematic-review", "judiverse/parity-entry-into-force-judiverse", "judiverse/parity-epidemic-sound-add-track-to-playlist-judiverse", "judiverse/parity-epidemic-sound-find-similar-tracks-judiverse", "judiverse/parity-epidemic-sound-sign-in-and-browse-music-judiverse", "judiverse/parity-epidemic-sound-track-download-with-usage-confirmation-judiverse", "judiverse/parity-eu-dsm-directive-judiverse", "judiverse/parity-eu-european-parliament-judiverse", "judiverse/parity-eu-gema-v-judiverse", "judiverse/parity-european-accessibility-act-judiverse", "judiverse/parity-evernote-evernote", "judiverse/parity-exa-api-key-creation-judiverse", "judiverse/parity-exa-developer-workspace-home-judiverse", "judiverse/parity-exa-service-key-management-judiverse", "judiverse/parity-exa-setup-complete-judiverse", "judiverse/parity-excalidraw-judiverse", "judiverse/parity-exoplan-quick-natural-language-event-add-judiverse", "judiverse/parity-exoplan-reschedule-event-by-dragging-judiverse", "judiverse/parity-factory-attach-context-to-droid-session-judiverse", "judiverse/parity-factory-contact-sales-lead-capture-judiverse", "judiverse/parity-factory-droid-session-setup-and-repo-inspection-judiverse", "judiverse/parity-fal-create-api-key-judiverse", "judiverse/parity-family-collectible-item-management-judiverse", "judiverse/parity-fathom-fathom", "judiverse/parity-fernand-api-key-creation-judiverse", "judiverse/parity-fernand-conversation-detail-judiverse", "judiverse/parity-fernand-knowledge-base-search-judiverse", "judiverse/parity-fernand-send-new-support-email-judiverse", "judiverse/parity-fernand-support-inbox-review-judiverse", "judiverse/parity-fernand-support-tool-sign-in-judiverse", "judiverse/parity-figma-design-figma-design", "judiverse/parity-figma-summarize-team-sticky-notes-judiverse", "judiverse/parity-fingerprint-public-api-key-creation-judiverse", "judiverse/parity-fireflies-ai-fireflies-ai", "judiverse/parity-fireflies-judiverse", "judiverse/parity-forgejo-judiverse", "judiverse/parity-fourmula-ai-fashion-asset-generation-judiverse", "judiverse/parity-framer-delete-screen-with-confirmation-judiverse", "judiverse/parity-framer-manage-archived-project-judiverse", "judiverse/parity-fresha-add-appointment-to-calendar-judiverse", "judiverse/parity-fresha-help-center-search-and-no-results-judiverse", "judiverse/parity-fresha-password-reset-and-update-judiverse", "judiverse/parity-fresha-salon-appointment-booking-judiverse", "judiverse/parity-fuse-new-wallet-signup-and-secure-key-setup-judiverse", "judiverse/parity-fuse-virtual-bank-account-signup-judiverse", "judiverse/parity-genie-ask-genie-with-photos-judiverse", "judiverse/parity-genie-chat-reminder-setup-judiverse", "judiverse/parity-genie-contact-search-add-judiverse", "judiverse/parity-genie-dinner-recommendation-chat-judiverse", "judiverse/parity-genie-reminder-task-details-judiverse", "judiverse/parity-gitlab-duo-agent-platform-judiverse", "judiverse/parity-givingli-card-message-editor-judiverse", "judiverse/parity-gizmo-create-new-message-with-recipients-judiverse", "judiverse/parity-gizmo-direct-message-compose-judiverse", "judiverse/parity-gizmo-image-overlay-generator-judiverse", "judiverse/parity-gizmo-manage-conversation-from-inbox-judiverse", "judiverse/parity-gladia-api-key-creation-and-listing-judiverse", "judiverse/parity-gladia-asynchronous-transcription-job-judiverse", "judiverse/parity-glean-judiverse", "judiverse/parity-glossgenius-analytics-with-ai-growth-analyst-judiverse", "judiverse/parity-gmail-gmail", "judiverse/parity-gmail-judiverse", "judiverse/parity-go-club-go-plus-subscription-purchase-judiverse", "judiverse/parity-google-appsheet-google-appsheet", "judiverse/parity-google-calendar-google-calendar", "judiverse/parity-google-calendar-meet-google-calendar", "judiverse/parity-google-chat-google-chat", "judiverse/parity-google-docs-google-docs", "judiverse/parity-google-docs-sheets-slides-google-docs", "judiverse/parity-google-drive-google-drive", "judiverse/parity-google-forms-google-forms", "judiverse/parity-google-gemini-export-chat-response-to-google-docs-judiverse", "judiverse/parity-google-gemini-gemini-welcome-consent-judiverse", "judiverse/parity-google-gemini-request-java-ping-pong-example-judiverse", "judiverse/parity-google-gemini-voice-query-to-suggestions-judiverse", "judiverse/parity-google-meet-google-meet", "judiverse/parity-google-offers-agent-judiverse", "judiverse/parity-google-sheets-google-sheets", "judiverse/parity-google-sites-google-sites", "judiverse/parity-google-slides-google-slides", "judiverse/parity-google-workspace-google-workspace", "judiverse/parity-gpai-code-of-judiverse", "judiverse/parity-gpai-model-obligations-judiverse", "judiverse/parity-granola-copy-generated-response-judiverse", "judiverse/parity-granola-search-chat-history-judiverse", "judiverse/parity-granola-search-notes-and-meetings-judiverse", "judiverse/parity-granola-share-note-with-editor-access-judiverse", "judiverse/parity-granola-voice-note-transcription-judiverse", "judiverse/parity-grok-attach-and-summarize-document-judiverse", "judiverse/parity-grok-image-assisted-question-and-response-judiverse", "judiverse/parity-grok-personalized-voice-settings-judiverse", "judiverse/parity-grok-private-chat-compose-and-receive-response-judiverse", "judiverse/parity-grok-share-conversation-to-messages-judiverse", "judiverse/parity-ground-news-alerts-inbox-cleanup-judiverse", "judiverse/parity-ground-news-invite-friends-referral-code-judiverse", "judiverse/parity-guardrails-ai-guardrails-ai", "judiverse/parity-handshake-document-upload-preview-judiverse", "judiverse/parity-helicone-helicone", "judiverse/parity-hello-ivy-chat-gif-insertion-judiverse", "judiverse/parity-hello-ivy-inline-comment-and-highlight-judiverse", "judiverse/parity-hello-ivy-open-documentation-article-judiverse", "judiverse/parity-hello-ivy-sign-in-to-project-overview-judiverse", "judiverse/parity-high-risk-systems-embedded-judiverse", "judiverse/parity-high-risk-systems-in-judiverse", "judiverse/parity-homerun-candidate-interview-scheduler-judiverse", "judiverse/parity-homerun-candidate-management-profile-actions-judiverse", "judiverse/parity-homerun-candidate-message-template-email-send-judiverse", "judiverse/parity-homerun-candidate-review-submission-judiverse", "judiverse/parity-hugging-face-transformers-judiverse", "judiverse/parity-hype-comment-discussion-panel-judiverse", "judiverse/parity-hype-discover-search-judiverse", "judiverse/parity-hype-password-recovery-and-reset-judiverse", "judiverse/parity-ikea-room-designer-judiverse", "judiverse/parity-imprint-commit-schedule-reminder-judiverse", "judiverse/parity-instacart-payment-methods-management-judiverse", "judiverse/parity-introspect-friend-decision-response-judiverse", "judiverse/parity-introspect-item-verification-share-judiverse", "judiverse/parity-jace-ai-find-messages-by-contact-judiverse", "judiverse/parity-jace-ai-jace-trial-onboarding-judiverse", "judiverse/parity-jace-ai-select-and-mark-email-as-done-judiverse", "judiverse/parity-jace-ai-sign-in-to-inbox-judiverse", "judiverse/parity-jasper-ai-create-and-share-blog-draft-judiverse", "judiverse/parity-jasper-ai-document-sharing-and-link-copy-judiverse", "judiverse/parity-jira-jira", "judiverse/parity-joi-first-launch-setup-judiverse", "judiverse/parity-kestra-judiverse", "judiverse/parity-kestra-kestra", "judiverse/parity-keycloak-judiverse", "judiverse/parity-kickstarter-reward-pledge-and-checkout-judiverse", "judiverse/parity-kickstarter-share-project-via-messages-judiverse", "judiverse/parity-kin-calendar-integration-permissions-judiverse", "judiverse/parity-kin-journaling-draft-review-judiverse", "judiverse/parity-kin-onboarding-to-chat-reminder-setup-judiverse", "judiverse/parity-komoot-friend-discovery-invitations-judiverse", "judiverse/parity-komoot-password-recovery-via-reset-email-judiverse", "judiverse/parity-komoot-save-or-share-activity-image-judiverse", "judiverse/parity-label-led-measures-to-judiverse", "judiverse/parity-langchain-langchain", "judiverse/parity-langchain-langgraph-judiverse", "judiverse/parity-langflow-langflow", "judiverse/parity-langfuse-judiverse", "judiverse/parity-langfuse-langfuse", "judiverse/parity-langgraph-langgraph", "judiverse/parity-lemni-agent-configuration-and-enablement-judiverse", "judiverse/parity-lemni-agent-instruction-and-action-configuration-judiverse", "judiverse/parity-lemni-inbox-conversation-triage-judiverse", "judiverse/parity-lemni-webshop-support-agent-editing-judiverse", "judiverse/parity-leonardo-ai-copy-generation-api-code-judiverse", "judiverse/parity-leonardo-ai-profile-settings-management-judiverse", "judiverse/parity-linear-assign-project-judiverse", "judiverse/parity-linear-create-item-and-add-labels-judiverse", "judiverse/parity-linear-create-personal-api-key-judiverse", "judiverse/parity-linear-filter-inbox-by-issue-status-judiverse", "judiverse/parity-linear-inbox-issue-browsing-and-notification-action-judiverse", "judiverse/parity-linear-inbox-notification-delete-judiverse", "judiverse/parity-linear-issue-detail-actions-favorite-judiverse", "judiverse/parity-linear-judiverse", "judiverse/parity-linear-linear", "judiverse/parity-linear-passwordless-email-sign-in-to-home-judiverse", "judiverse/parity-linear-project-selector-judiverse", "judiverse/parity-linktree-audience-contact-filtering-judiverse", "judiverse/parity-linktree-insights-and-ai-tips-judiverse", "judiverse/parity-litellm-litellm", "judiverse/parity-livekit-judiverse", "judiverse/parity-llamaindex-llamaindex", "judiverse/parity-longevity-deck-enable-health-integration-judiverse", "judiverse/parity-loowner-inventory-search-and-filter-judiverse", "judiverse/parity-lovable-git-integrations-management-judiverse", "judiverse/parity-lovable-github-connection-management-judiverse", "judiverse/parity-lovable-missing-connector-request-judiverse", "judiverse/parity-lovable-qr-asset-label-implementation-judiverse", "judiverse/parity-lovable-start-project-prompt-judiverse", "judiverse/parity-main-application-judiverse", "judiverse/parity-main-data-protection-changes-judiverse", "judiverse/parity-make-create-and-configure-mcp-toolbox-with-ai-module-judiverse", "judiverse/parity-make-gmail-to-google-calendar-event-judiverse", "judiverse/parity-manus-api-key-creation-and-storage-judiverse", "judiverse/parity-manus-create-report-from-voice-judiverse", "judiverse/parity-manus-delete-api-key-judiverse", "judiverse/parity-manus-edit-manus-service-email-judiverse", "judiverse/parity-manus-github-connector-installation-judiverse", "judiverse/parity-manus-meeting-recording-and-auto-summary-judiverse", "judiverse/parity-manus-meeting-recording-to-summary-judiverse", "judiverse/parity-manus-project-skill-management-judiverse", "judiverse/parity-manus-project-task-visibility-judiverse", "judiverse/parity-manus-remove-task-from-sidebar-judiverse", "judiverse/parity-manus-task-assignment-to-manus-judiverse", "judiverse/parity-manychat-add-contact-via-contacts-list-judiverse", "judiverse/parity-manychat-welcome-coupon-randomizer-judiverse", "judiverse/parity-marathon-demo-booking-confirmation-judiverse", "judiverse/parity-marathon-demo-booking-judiverse", "judiverse/parity-marathon-project-overview-and-inbox-review-judiverse", "judiverse/parity-marathon-sign-in-to-my-desk-judiverse", "judiverse/parity-mastra-mastra", "judiverse/parity-matter-connect-and-manage-email-sources-judiverse", "judiverse/parity-matter-onboarding-newsletter-import-subscription-judiverse", "judiverse/parity-maze-interview-recruitment-setup-judiverse", "judiverse/parity-maze-interview-session-supervision-judiverse", "judiverse/parity-maze-manage-scheduled-interview-judiverse", "judiverse/parity-maze-recruitment-link-copy-judiverse", "judiverse/parity-md-vinyl-md-vinyl-setup-and-purchase-judiverse", "judiverse/parity-md-vinyl-music-library-browse-judiverse", "judiverse/parity-md-vinyl-music-library-browsing-judiverse", "judiverse/parity-md-vinyl-music-player-onboarding-judiverse", "judiverse/parity-md-vinyl-music-share-options-judiverse", "judiverse/parity-melius-image-comment-annotation-judiverse", "judiverse/parity-melius-reference-based-product-image-judiverse", "judiverse/parity-memotron-create-interview-with-audio-and-transcription-judivers", "judiverse/parity-memotron-global-search-modal-judiverse", "judiverse/parity-memotron-global-search-modal-judiverse-2", "judiverse/parity-meta-ai-16-bit-sprite-creation-judiverse", "judiverse/parity-meta-ai-ai-conversation-and-creative-writing-judiverse", "judiverse/parity-meta-ai-ai-image-creation-judiverse", "judiverse/parity-meta-ai-ai-presentation-creation-judiverse", "judiverse/parity-meta-ai-delete-notification-judiverse", "judiverse/parity-meta-ai-gmail-connection-authorization-judiverse", "judiverse/parity-meta-ai-gmail-email-audit-judiverse", "judiverse/parity-meta-ai-high-protein-meal-plan-artifact-judiverse", "judiverse/parity-meta-ai-incognito-ai-conversation-judiverse", "judiverse/parity-meta-ai-meta-ai-account-sign-in-judiverse", "judiverse/parity-meta-ai-mythical-biology-poster-judiverse", "judiverse/parity-meta-ai-mythical-creature-field-guide-judiverse", "judiverse/parity-meta-ai-recipe-response-rating-judiverse", "judiverse/parity-meta-ai-save-recipe-instructions-judiverse", "judiverse/parity-meta-ai-trending-recipe-search-judiverse", "judiverse/parity-microsoft-365-microsoft-365", "judiverse/parity-microsoft-365-teams-outlook-microsoft-365-teams-outlook", "judiverse/parity-microsoft-agent-framework-microsoft-agent-framework", "judiverse/parity-microsoft-and-google-judiverse", "judiverse/parity-microsoft-applies-purview-judiverse", "judiverse/parity-microsoft-copilot-document-drafting-with-copilot-judiverse", "judiverse/parity-microsoft-distinguishes-agents-judiverse", "judiverse/parity-microsoft-excel-microsoft-excel", "judiverse/parity-microsoft-forms-microsoft-forms", "judiverse/parity-microsoft-lists-microsoft-lists", "judiverse/parity-microsoft-loop-microsoft-loop", "judiverse/parity-microsoft-onedrive-microsoft-onedrive", "judiverse/parity-microsoft-outlook-judiverse", "judiverse/parity-microsoft-outlook-microsoft-outlook", "judiverse/parity-microsoft-planner-microsoft-planner", "judiverse/parity-microsoft-power-automate-microsoft-power-automate", "judiverse/parity-microsoft-powerpoint-microsoft-powerpoint", "judiverse/parity-microsoft-s-registry-sync-judiverse", "judiverse/parity-microsoft-sharepoint-microsoft-sharepoint", "judiverse/parity-microsoft-teams-add-emoji-reaction-to-message-judiverse", "judiverse/parity-microsoft-teams-chat-request-handling-judiverse", "judiverse/parity-microsoft-teams-direct-chat-message-judiverse", "judiverse/parity-microsoft-teams-file-attachment-and-send-judiverse", "judiverse/parity-microsoft-teams-global-search-query-inspection-judiverse", "judiverse/parity-microsoft-teams-microsoft-teams", "judiverse/parity-microsoft-teams-pin-contact-to-pinned-list-judiverse", "judiverse/parity-microsoft-teams-start-conversation-by-recipient-judiverse", "judiverse/parity-microsoft-word-microsoft-word", "judiverse/parity-mini-swe-agent-mini-swe-agent", "judiverse/parity-miro-ai-generated-retrospective-summary-creation-judiverse", "judiverse/parity-miro-miro", "judiverse/parity-missive-add-emoji-to-conversation-judiverse", "judiverse/parity-missive-collaborative-inbox-sign-in-to-mailbox-judiverse", "judiverse/parity-missive-create-and-configure-team-judiverse", "judiverse/parity-missive-email-conversation-with-inline-tasks-judiverse", "judiverse/parity-missive-open-and-view-attachment-in-conversation-judiverse", "judiverse/parity-missive-open-team-meeting-conversation-judiverse", "judiverse/parity-missive-organization-rules-create-outgoing-rule-judiverse", "judiverse/parity-missive-schedule-snooze-for-email-judiverse", "judiverse/parity-missive-team-meeting-draft-composition-and-commenting-judiverse", "judiverse/parity-mistral-ai-api-mistral-ai-api", "judiverse/parity-mlflow-judiverse", "judiverse/parity-mlflow-mlflow", "judiverse/parity-mobbin-mobbin", "judiverse/parity-mocha-ai-assisted-project-generation-judiverse", "judiverse/parity-moffatt-v-air-judiverse", "judiverse/parity-moises-song-library-judiverse", "judiverse/parity-moises-song-library-player-mixer-judiverse", "judiverse/parity-monday-com-monday-com", "judiverse/parity-moonly-meditation-player-full-moon-judiverse", "judiverse/parity-mozi-contact-search-and-invite-judiverse", "judiverse/parity-mozi-find-and-contact-person-judiverse", "judiverse/parity-mwm-habit-tracker-product-creation-judiverse", "judiverse/parity-mymind-add-card-to-space-judiverse", "judiverse/parity-mymind-mymind-onboarding-judiverse", "judiverse/parity-mymind-responsive-reference-library-judiverse", "judiverse/parity-mymind-visual-reference-search-judiverse", "judiverse/parity-n8n-ai-summary-agent-editor-judiverse", "judiverse/parity-n8n-chat-triggered-export-pipeline-judiverse", "judiverse/parity-n8n-create-personal-agent-judiverse", "judiverse/parity-n8n-edit-workflow-description-judiverse", "judiverse/parity-n8n-integrations-agents-browse-judiverse", "judiverse/parity-n8n-n8n", "judiverse/parity-n8n-node-based-workflow-editing-and-execution-judiverse", "judiverse/parity-n8n-personal-workflows-list-judiverse", "judiverse/parity-navia-instagram-carousel-review-judiverse", "judiverse/parity-navia-replace-carousel-image-judiverse", "judiverse/parity-navia-update-carousel-slide-judiverse", "judiverse/parity-nemo-guardrails-nemo-guardrails", "judiverse/parity-nextcloud-server-judiverse", "judiverse/parity-nextdoor-remove-chat-judiverse", "judiverse/parity-nist-s-caisi-center-judiverse", "judiverse/parity-no-in-place-migration-judiverse", "judiverse/parity-non-open-until-2030-judiverse", "judiverse/parity-nooka-password-recovery-and-reset-judiverse", "judiverse/parity-notion-judiverse", "judiverse/parity-notion-notion", "judiverse/parity-oauth-2-0-token-judiverse", "judiverse/parity-obsidian-obsidian", "judiverse/parity-okuri-daily-task-checklist-judiverse", "judiverse/parity-okuri-schedule-team-standup-judiverse", "judiverse/parity-okuri-settings-and-feature-info-judiverse", "judiverse/parity-ollama-judiverse", "judiverse/parity-ollama-ollama", "judiverse/parity-online-music-licensing-judiverse", "judiverse/parity-onlyoffice-docs-judiverse", "judiverse/parity-openai-agentkit-judiverse", "judiverse/parity-openai-agents-sdk-python-judiverse", "judiverse/parity-openai-documentation-search-modal-judiverse", "judiverse/parity-openalex-openalex", "judiverse/parity-openfga-judiverse", "judiverse/parity-openpipe-openpipe", "judiverse/parity-otter-ai-otter-ai", "judiverse/parity-owasp-top-10-judiverse", "judiverse/parity-palantir-aip-aip", "judiverse/parity-palantir-foundry-foundry", "judiverse/parity-palantir-foundry-foundry-part-10", "judiverse/parity-palantir-foundry-foundry-part-11", "judiverse/parity-palantir-foundry-foundry-part-12", "judiverse/parity-palantir-foundry-foundry-part-2", "judiverse/parity-palantir-foundry-foundry-part-3", "judiverse/parity-palantir-foundry-foundry-part-4", "judiverse/parity-palantir-foundry-foundry-part-5", "judiverse/parity-palantir-foundry-foundry-part-6", "judiverse/parity-palantir-foundry-foundry-part-7", "judiverse/parity-palantir-foundry-foundry-part-8", "judiverse/parity-palantir-foundry-foundry-part-9", "judiverse/parity-palantir-technologies-aip-aip", "judiverse/parity-palantir-technologies-aip-aip-part-2", "judiverse/parity-palantir-technologies-aip-aip-part-3", "judiverse/parity-palantir-technologies-aip-aip-part-4", "judiverse/parity-palantir-technologies-aip-aip-part-5", "judiverse/parity-palantir-technologies-aip-aip-part-6", "judiverse/parity-palantir-technologies-foundry-foundry", "judiverse/parity-palantir-technologies-foundry-foundry-part-10", "judiverse/parity-palantir-technologies-foundry-foundry-part-100", "judiverse/parity-palantir-technologies-foundry-foundry-part-101", "judiverse/parity-palantir-technologies-foundry-foundry-part-11", "judiverse/parity-palantir-technologies-foundry-foundry-part-12", "judiverse/parity-palantir-technologies-foundry-foundry-part-13", "judiverse/parity-palantir-technologies-foundry-foundry-part-14", "judiverse/parity-palantir-technologies-foundry-foundry-part-15", "judiverse/parity-palantir-technologies-foundry-foundry-part-16", "judiverse/parity-palantir-technologies-foundry-foundry-part-17", "judiverse/parity-palantir-technologies-foundry-foundry-part-18", "judiverse/parity-palantir-technologies-foundry-foundry-part-19", "judiverse/parity-palantir-technologies-foundry-foundry-part-2", "judiverse/parity-palantir-technologies-foundry-foundry-part-20", "judiverse/parity-palantir-technologies-foundry-foundry-part-21", "judiverse/parity-palantir-technologies-foundry-foundry-part-22", "judiverse/parity-palantir-technologies-foundry-foundry-part-23", "judiverse/parity-palantir-technologies-foundry-foundry-part-24", "judiverse/parity-palantir-technologies-foundry-foundry-part-25", "judiverse/parity-palantir-technologies-foundry-foundry-part-26", "judiverse/parity-palantir-technologies-foundry-foundry-part-27", "judiverse/parity-palantir-technologies-foundry-foundry-part-28", "judiverse/parity-palantir-technologies-foundry-foundry-part-29", "judiverse/parity-palantir-technologies-foundry-foundry-part-3", "judiverse/parity-palantir-technologies-foundry-foundry-part-30", "judiverse/parity-palantir-technologies-foundry-foundry-part-31", "judiverse/parity-palantir-technologies-foundry-foundry-part-32", "judiverse/parity-palantir-technologies-foundry-foundry-part-33", "judiverse/parity-palantir-technologies-foundry-foundry-part-34", "judiverse/parity-palantir-technologies-foundry-foundry-part-35", "judiverse/parity-palantir-technologies-foundry-foundry-part-36", "judiverse/parity-palantir-technologies-foundry-foundry-part-37", "judiverse/parity-palantir-technologies-foundry-foundry-part-38", "judiverse/parity-palantir-technologies-foundry-foundry-part-39", "judiverse/parity-palantir-technologies-foundry-foundry-part-4", "judiverse/parity-palantir-technologies-foundry-foundry-part-40", "judiverse/parity-palantir-technologies-foundry-foundry-part-41", "judiverse/parity-palantir-technologies-foundry-foundry-part-42", "judiverse/parity-palantir-technologies-foundry-foundry-part-43", "judiverse/parity-palantir-technologies-foundry-foundry-part-44", "judiverse/parity-palantir-technologies-foundry-foundry-part-45", "judiverse/parity-palantir-technologies-foundry-foundry-part-46", "judiverse/parity-palantir-technologies-foundry-foundry-part-47", "judiverse/parity-palantir-technologies-foundry-foundry-part-48", "judiverse/parity-palantir-technologies-foundry-foundry-part-49", "judiverse/parity-palantir-technologies-foundry-foundry-part-5", "judiverse/parity-palantir-technologies-foundry-foundry-part-50", "judiverse/parity-palantir-technologies-foundry-foundry-part-51", "judiverse/parity-palantir-technologies-foundry-foundry-part-52", "judiverse/parity-palantir-technologies-foundry-foundry-part-53", "judiverse/parity-palantir-technologies-foundry-foundry-part-54", "judiverse/parity-palantir-technologies-foundry-foundry-part-55", "judiverse/parity-palantir-technologies-foundry-foundry-part-56", "judiverse/parity-palantir-technologies-foundry-foundry-part-57", "judiverse/parity-palantir-technologies-foundry-foundry-part-58", "judiverse/parity-palantir-technologies-foundry-foundry-part-59", "judiverse/parity-palantir-technologies-foundry-foundry-part-6", "judiverse/parity-palantir-technologies-foundry-foundry-part-60", "judiverse/parity-palantir-technologies-foundry-foundry-part-61", "judiverse/parity-palantir-technologies-foundry-foundry-part-62", "judiverse/parity-palantir-technologies-foundry-foundry-part-63", "judiverse/parity-palantir-technologies-foundry-foundry-part-64", "judiverse/parity-palantir-technologies-foundry-foundry-part-65", "judiverse/parity-palantir-technologies-foundry-foundry-part-66", "judiverse/parity-palantir-technologies-foundry-foundry-part-67", "judiverse/parity-palantir-technologies-foundry-foundry-part-68", "judiverse/parity-palantir-technologies-foundry-foundry-part-69", "judiverse/parity-palantir-technologies-foundry-foundry-part-7", "judiverse/parity-palantir-technologies-foundry-foundry-part-70", "judiverse/parity-palantir-technologies-foundry-foundry-part-71", "judiverse/parity-palantir-technologies-foundry-foundry-part-72", "judiverse/parity-palantir-technologies-foundry-foundry-part-73", "judiverse/parity-palantir-technologies-foundry-foundry-part-74", "judiverse/parity-palantir-technologies-foundry-foundry-part-75", "judiverse/parity-palantir-technologies-foundry-foundry-part-76", "judiverse/parity-palantir-technologies-foundry-foundry-part-77", "judiverse/parity-palantir-technologies-foundry-foundry-part-78", "judiverse/parity-palantir-technologies-foundry-foundry-part-79", "judiverse/parity-palantir-technologies-foundry-foundry-part-8", "judiverse/parity-palantir-technologies-foundry-foundry-part-80", "judiverse/parity-palantir-technologies-foundry-foundry-part-81", "judiverse/parity-palantir-technologies-foundry-foundry-part-82", "judiverse/parity-palantir-technologies-foundry-foundry-part-83", "judiverse/parity-palantir-technologies-foundry-foundry-part-84", "judiverse/parity-palantir-technologies-foundry-foundry-part-85", "judiverse/parity-palantir-technologies-foundry-foundry-part-86", "judiverse/parity-palantir-technologies-foundry-foundry-part-87", "judiverse/parity-palantir-technologies-foundry-foundry-part-88", "judiverse/parity-palantir-technologies-foundry-foundry-part-89", "judiverse/parity-palantir-technologies-foundry-foundry-part-9", "judiverse/parity-palantir-technologies-foundry-foundry-part-90", "judiverse/parity-palantir-technologies-foundry-foundry-part-91", "judiverse/parity-palantir-technologies-foundry-foundry-part-92", "judiverse/parity-palantir-technologies-foundry-foundry-part-93", "judiverse/parity-palantir-technologies-foundry-foundry-part-94", "judiverse/parity-palantir-technologies-foundry-foundry-part-95", "judiverse/parity-palantir-technologies-foundry-foundry-part-96", "judiverse/parity-palantir-technologies-foundry-foundry-part-97", "judiverse/parity-palantir-technologies-foundry-foundry-part-98", "judiverse/parity-palantir-technologies-foundry-foundry-part-99", "judiverse/parity-parallel-create-api-key-for-default-application-judiverse", "judiverse/parity-parallel-search-api-results-judiverse", "judiverse/parity-parallel-slack-notification-integration-judiverse", "judiverse/parity-paypal-contact-search-selection-judiverse", "judiverse/parity-paypal-sales-contact-form-submission-judiverse", "judiverse/parity-pdf-js-judiverse", "judiverse/parity-penpot-judiverse", "judiverse/parity-perplexity-judiverse", "judiverse/parity-phantom-settings-search-judiverse", "judiverse/parity-phantom-token-list-manager-judiverse", "judiverse/parity-phoenix-phoenix", "judiverse/parity-pi-language-tutor-onboarding-judiverse", "judiverse/parity-pillowtalk-pillowtalk-onboarding-judiverse", "judiverse/parity-pillowtalk-voice-reflection-entry-judiverse", "judiverse/parity-pinkllama-apple-health-sync-setup-judiverse", "judiverse/parity-pinterest-board-edit-and-collaborator-management-judiverse", "judiverse/parity-pinterest-collaborator-management-judiverse", "judiverse/parity-pinterest-invite-collaborators-and-set-permissions-judiverse", "judiverse/parity-pinterest-private-message-and-media-actions-judiverse", "judiverse/parity-pinterest-story-editor-judiverse", "judiverse/parity-play-bottom-navigation-editor-judiverse", "judiverse/parity-play-create-text-component-with-state-and-style-judiverse", "judiverse/parity-play-edit-named-color-judiverse", "judiverse/parity-playground-birthday-card-edit-save-judiverse", "judiverse/parity-playground-email-sign-in-first-design-onboarding-judiverse", "judiverse/parity-pocketshelf-data-backup-import-judiverse", "judiverse/parity-podimo-podcast-comment-moderation-judiverse", "judiverse/parity-podimo-podcast-comments-thread-judiverse", "judiverse/parity-pool-refero-mcp-item-review-judiverse", "judiverse/parity-poppy-create-calendar-event-judiverse", "judiverse/parity-poppy-memories-search-judiverse", "judiverse/parity-poppy-recommended-connection-setup-judiverse", "judiverse/parity-poppy-support-and-privacy-settings-judiverse", "judiverse/parity-poppy-voice-event-creation-judiverse", "judiverse/parity-portkey-ai-gateway-portkey-ai-gateway", "judiverse/parity-preply-archive-conversation-judiverse", "judiverse/parity-production-use-needs-nvidia-ai-enterprise-judiverse", "judiverse/parity-programa-complete-to-do-item-judiverse", "judiverse/parity-programa-meeting-recording-and-summary-judiverse", "judiverse/parity-prohibited-practices-1-8-judiverse", "judiverse/parity-prohibition-9-ai-judiverse", "judiverse/parity-pushr-create-recurring-schedule-judiverse", "judiverse/parity-pydantic-ai-pydantic-ai", "judiverse/parity-qwen-qwen", "judiverse/parity-rad-weather-ambient-weather-connection-judiverse", "judiverse/parity-rad-weather-personalization-preferences-judiverse", "judiverse/parity-raycast-ai-model-preset-selection-judiverse", "judiverse/parity-raycast-create-and-use-ai-preset-judiverse", "judiverse/parity-raycast-edit-generated-prompt-judiverse", "judiverse/parity-raycast-inline-ai-command-palette-judiverse", "judiverse/parity-raycast-question-search-and-select-judiverse", "judiverse/parity-raycast-save-ai-message-to-note-judiverse", "judiverse/parity-read-ai-read-ai", "judiverse/parity-record-club-new-user-registration-and-onboarding-judiverse", "judiverse/parity-reown-contact-lead-capture-and-confirmation-judiverse", "judiverse/parity-replicate-judiverse", "judiverse/parity-replit-s-agent-deleted-judiverse", "judiverse/parity-resend-judiverse", "judiverse/parity-riverside-backstage-chat-during-recording-judiverse", "judiverse/parity-rork-ai-mobile-product-creation-judiverse", "judiverse/parity-rork-collaborator-invitation-judiverse", "judiverse/parity-rork-comments-panel-judiverse", "judiverse/parity-rork-project-bookmarking-and-actions-judiverse", "judiverse/parity-rox-add-contacts-to-campaign-judiverse", "judiverse/parity-rox-google-calendar-gmail-connect-judiverse", "judiverse/parity-rox-new-opportunity-add-associated-people-judiverse", "judiverse/parity-runey-add-customer-judiverse", "judiverse/parity-runey-project-message-thread-judiverse", "judiverse/parity-runey-project-tasks-overview-judiverse", "judiverse/parity-runway-home-global-search-judiverse", "judiverse/parity-sabre-dav-judiverse", "judiverse/parity-salesforce-agentforce-360-judiverse", "judiverse/parity-savee-email-social-login-onboarding-judiverse", "judiverse/parity-savvycal-attendee-scheduling-and-location-setup-judiverse", "judiverse/parity-semantic-kernel-semantic-kernel", "judiverse/parity-servicenow-ai-control-tower-judiverse", "judiverse/parity-sglang-judiverse", "judiverse/parity-sglang-sglang", "judiverse/parity-shuttle-access-and-workspace-onboarding-judiverse", "judiverse/parity-shuttle-create-and-open-new-space-judiverse", "judiverse/parity-shuttle-edit-filename-in-space-judiverse", "judiverse/parity-shuttle-upload-items-to-space-judiverse", "judiverse/parity-skiff-document-sharing-invite-judiverse", "judiverse/parity-skiff-inbox-search-and-results-judiverse", "judiverse/parity-skiff-inline-text-commenting-with-mention-judiverse", "judiverse/parity-skiff-manage-email-aliases-judiverse", "judiverse/parity-skiff-select-and-move-messages-to-trash-judiverse", "judiverse/parity-slack-judiverse", "judiverse/parity-slack-slack", "judiverse/parity-slite-new-user-signup-and-workspace-setup-judiverse", "judiverse/parity-soundcloud-add-track-to-playlist-judiverse", "judiverse/parity-soundcloud-email-sign-up-and-verification-judiverse", "judiverse/parity-soundcloud-judiverse", "judiverse/parity-soundcloud-save-track-to-new-playlist-judiverse", "judiverse/parity-soundcloud-search-from-playlist-to-artist-judiverse", "judiverse/parity-soundcloud-search-results-track-discovery-judiverse", "judiverse/parity-soundcloud-sign-in-to-access-uploads-and-dashboard-judiverse", "judiverse/parity-soundcloud-track-mastering-wizard-judiverse", "judiverse/parity-source-available-judiverse", "judiverse/parity-spotify-adding-a-radio-song-to-a-playlist-judiverse", "judiverse/parity-spotify-filter-library-to-my-playlists-judiverse", "judiverse/parity-spotify-judiverse", "judiverse/parity-spotify-share-track-link-judiverse", "judiverse/parity-square-password-reset-and-confirmation-judiverse", "judiverse/parity-ssense-search-results-discovery-judiverse", "judiverse/parity-steel-create-project-api-key-judiverse", "judiverse/parity-stemz-track-options-and-artwork-selection-judiverse", "judiverse/parity-structured-privacy-analytics-settings-judiverse", "judiverse/parity-structured-structured-ai-onboarding-judiverse", "judiverse/parity-structured-structured-ai-suggestions-review-judiverse", "judiverse/parity-structured-structured-planner-download-judiverse", "judiverse/parity-substack-appearance-visual-preferences-judiverse", "judiverse/parity-substack-convert-paragraph-to-heading-judiverse", "judiverse/parity-substack-global-search-with-empty-note-results-judiverse", "judiverse/parity-synthesia-io-add-music-to-library-judiverse", "judiverse/parity-synthesia-io-ai-presentation-review-and-summary-judiverse", "judiverse/parity-tana-tana", "judiverse/parity-temporal-judiverse", "judiverse/parity-the-leap-bulk-contact-import-judiverse", "judiverse/parity-the-openid-foundation-s-judiverse", "judiverse/parity-thinking-machines-10-judiverse", "judiverse/parity-tidal-add-song-to-collection-judiverse", "judiverse/parity-tidal-edit-album-tracks-judiverse", "judiverse/parity-tidal-edit-track-metadata-judiverse", "judiverse/parity-tidal-music-album-preparation-judiverse", "judiverse/parity-tidal-music-search-and-browse-judiverse", "judiverse/parity-tidal-music-upload-and-album-edit-judiverse", "judiverse/parity-tidal-profile-email-verification-judiverse", "judiverse/parity-tidal-track-metadata-editing-judiverse", "judiverse/parity-tidal-track-report-reason-selection-judiverse", "judiverse/parity-tidycal-therapy-session-booking-judiverse", "judiverse/parity-tilt-contact-tagging-selector-judiverse", "judiverse/parity-tilt-invite-and-add-friends-judiverse", "judiverse/parity-tinder-share-a-date-with-friends-judiverse", "judiverse/parity-tinder-share-spotify-track-in-conversation-judiverse", "judiverse/parity-tinder-spotify-connect-for-music-profile-judiverse", "judiverse/parity-tl-dv-tl-dv", "judiverse/parity-trello-trello", "judiverse/parity-trip-way-manage-documents-judiverse", "judiverse/parity-tripsy-onboarding-and-first-launch-judiverse", "judiverse/parity-tripsy-trip-itinerary-search-judiverse", "judiverse/parity-tripsy-venue-visit-date-duration-judiverse", "judiverse/parity-turf-discover-join-live-game-judiverse", "judiverse/parity-turf-live-game-chat-compose-judiverse", "judiverse/parity-turf-react-to-a-chat-message-judiverse", "judiverse/parity-turf-send-friend-request-judiverse", "judiverse/parity-twenty-judiverse", "judiverse/parity-uk-copyright-and-ai-training-no-change-to-the-law-judiverse", "judiverse/parity-uk-employment-rights-judiverse", "judiverse/parity-uk-getty-images-judiverse", "judiverse/parity-uk-industry-agreement-judiverse", "judiverse/parity-uk-report-on-judiverse", "judiverse/parity-unfair-terms-rules-apply-judiverse", "judiverse/parity-unhinged-share-day-entry-judiverse", "judiverse/parity-untitled-create-and-upload-audio-project-judiverse", "judiverse/parity-untitled-delete-track-from-album-judiverse", "judiverse/parity-ux-pilot-ux-pilot", "judiverse/parity-verse-connect-add-song-judiverse", "judiverse/parity-verse-connect-streaming-account-and-load-playlists-judiverse", "judiverse/parity-verse-profile-music-add-judiverse", "judiverse/parity-verse-text-overlay-editor-judiverse", "judiverse/parity-visitors-stripe-integration-setup-judiverse", "judiverse/parity-voicenotes-add-recording-by-upload-judiverse", "judiverse/parity-voicenotes-ask-ai-from-filtered-notes-judiverse", "judiverse/parity-voicenotes-configure-ask-ai-search-judiverse", "judiverse/parity-voicenotes-copy-note-summary-judiverse", "judiverse/parity-voicenotes-create-from-notes-with-custom-prompt-judiverse", "judiverse/parity-voicenotes-create-note-summary-judiverse", "judiverse/parity-voicenotes-create-summary-from-notes-judiverse", "judiverse/parity-voicenotes-email-verification-judiverse", "judiverse/parity-voicenotes-generate-ai-content-from-a-voice-note-judiverse", "judiverse/parity-voicenotes-join-and-record-meeting-judiverse", "judiverse/parity-voicenotes-record-and-transcribe-note-judiverse", "judiverse/parity-voicenotes-regenerate-voice-note-title-judiverse", "judiverse/parity-voicenotes-voice-note-capture-judiverse", "judiverse/parity-voicenotes-voice-query-to-notes-judiverse", "judiverse/parity-voicenotes-voicenotes-first-time-setup-judiverse", "judiverse/parity-volv-search-from-account-menu-judiverse", "judiverse/parity-volv-search-from-profile-judiverse", "judiverse/parity-w-b-weave-w-b-weave", "judiverse/parity-wabi-account-settings-and-deletion-judiverse", "judiverse/parity-wabi-connect-gmail-in-settings-judiverse", "judiverse/parity-wabi-home-and-activity-review-judiverse", "judiverse/parity-wabi-recipe-comments-judiverse", "judiverse/parity-wabi-voice-input-transcription-judiverse", "judiverse/parity-webex-meetings-webex-meetings", "judiverse/parity-weekrise-project-management-modal-judiverse", "judiverse/parity-west-elm-design-appointment-scheduling-judiverse", "judiverse/parity-whatsapp-add-new-contact-in-whatsapp-judiverse", "judiverse/parity-whatsapp-add-whatsapp-contact-judiverse", "judiverse/parity-whatsapp-ai-image-restyling-judiverse", "judiverse/parity-whatsapp-announcement-replies-thread-judiverse", "judiverse/parity-whatsapp-call-search-and-history-lookup-judiverse", "judiverse/parity-whatsapp-calls-onboarding-judiverse", "judiverse/parity-whatsapp-chat-message-actions-judiverse", "judiverse/parity-whatsapp-delete-chat-message-judiverse", "judiverse/parity-whatsapp-export-chat-and-share-judiverse", "judiverse/parity-whatsapp-invite-contacts-judiverse", "judiverse/parity-whatsapp-invite-contacts-to-whatsapp-judiverse", "judiverse/parity-whatsapp-message-forward-to-status-judiverse", "judiverse/parity-whatsapp-meta-ai-image-conversation-judiverse", "judiverse/parity-whatsapp-meta-ai-message-feedback-judiverse", "judiverse/parity-whatsapp-meta-ai-voice-chat-setup-judiverse", "judiverse/parity-whatsapp-meta-ai-voice-selection-judiverse", "judiverse/parity-whatsapp-photo-status-creation-judiverse", "judiverse/parity-whatsapp-pin-chat-message-judiverse", "judiverse/parity-whatsapp-schedule-call-and-invite-participants-judiverse", "judiverse/parity-whatsapp-settings-search-judiverse", "judiverse/parity-whatsapp-share-contacts-in-chat-judiverse", "judiverse/parity-wikidata-wikidata", "judiverse/parity-willow-ai-message-drafting-judiverse", "judiverse/parity-willow-offline-mode-settings-judiverse", "judiverse/parity-willow-scribe-message-drafting-judiverse", "judiverse/parity-willow-voice-prompt-drafting-judiverse", "judiverse/parity-willow-willow-home-dashboard-judiverse", "judiverse/parity-willow-willow-onboarding-intro-judiverse", "judiverse/parity-wise-newsroom-content-search-judiverse", "judiverse/parity-wise-referral-source-survey-judiverse", "judiverse/parity-wishwish-email-sign-in-and-first-box-judiverse", "judiverse/parity-wispr-flow-ai-note-summary-judiverse", "judiverse/parity-wispr-flow-copy-transcript-judiverse", "judiverse/parity-wispr-flow-password-recovery-judiverse", "judiverse/parity-wispr-flow-voice-note-capture-judiverse", "judiverse/parity-wispr-flow-voice-note-transcription-judiverse", "judiverse/parity-wittl-sign-in-to-inbox-judiverse", "judiverse/parity-womp-project-search-results-judiverse", "judiverse/parity-wrike-project-table-view-and-sidebar-toggle-judiverse", "judiverse/parity-wrike-status-triggered-assignment-judiverse", "judiverse/parity-wurrd-ai-practice-chat-judiverse", "judiverse/parity-x-conversation-leave-confirmation-judiverse", "judiverse/parity-xchat-conversation-filtering-and-mark-read-judiverse", "judiverse/parity-xchat-create-and-start-conversation-judiverse", "judiverse/parity-xchat-message-request-acceptance-judiverse", "judiverse/parity-xchat-sign-up-and-chat-onboarding-judiverse", "judiverse/parity-youtube-music-account-confirmation-and-signed-in-home-judiverse", "judiverse/parity-youtube-music-add-track-to-library-from-playlist-judiverse", "judiverse/parity-youtube-music-add-track-to-playlist-from-album-judiverse", "judiverse/parity-youtube-music-create-and-customize-mix-judiverse", "judiverse/parity-youtube-music-creator-profile-subscription-judiverse", "judiverse/parity-youtube-music-download-a-track-from-a-playlist-judiverse", "judiverse/parity-youtube-music-library-recent-activity-browsing-judiverse", "judiverse/parity-youtube-music-liked-music-collection-judiverse", "judiverse/parity-youtube-music-personalized-home-discovery-judiverse", "judiverse/parity-youtube-music-playlist-track-list-judiverse", "judiverse/parity-youtube-music-playlist-track-selection-playback-judiverse", "judiverse/parity-youtube-music-song-recognition-and-match-judiverse", "judiverse/parity-youtube-music-track-playback-like-feedback-judiverse", "judiverse/parity-zitadel-judiverse", "judiverse/parity-zoho-judiverse", "judiverse/parity-zoom-workplace-meetings-zoom-workplace-meetings", "judiverse/parity-zotero-zotero", "judiverse/problem-123456-password-used-in-danish-cpr-data-breach", "judiverse/problem-afd-leak-what-421-179-hacked-emails-show-about-the", "judiverse/problem-agentic-ai-and-consumers", "judiverse/problem-ai-and-leadership", "judiverse/problem-amazon-uses-ai-to-monitor-union-risk-in-uk-warehou", "judiverse/problem-apple-and-a-hacker-s-future", "judiverse/problem-apple-music-404-not-found", "judiverse/problem-apple-s-git-still-trusts-the-compromised-emudhra-c", "judiverse/problem-brown-student-maps-the-hidden-global-labor-powerin", "judiverse/problem-cisco-gives-engineer-unlimited-ai-tokens-to-rewrit", "judiverse/problem-clef-open-weight-decision-models-and-new-rl-fine-t", "judiverse/problem-containment-and-monitoring-failures-are-to-blame-f", "judiverse/problem-cross-request-kv-cache-exposure-in-the-glm-5-3-wor", "judiverse/problem-denmark-data-breach-exposes-8-8m-people-s-personal", "judiverse/problem-draft-legislation-better-use-of-new-and-improved-t", "judiverse/problem-electrocutions-disease-and-no-pay-how-workers-risk", "judiverse/problem-faster-planning-decisions-as-new-rules-and-ai-tack", "judiverse/problem-ftc-is-investigating-openai-anthropic-and-other-ai", "judiverse/problem-hacker-leaks-281-gb-of-data-from-germany-s-far-rig", "judiverse/problem-i-quit-openai-because-its-culture-is-broken", "judiverse/problem-inside-the-make-or-break-ai-mega-project-behind-ns", "judiverse/problem-investors-trim-growth-bets-over-fears-of-us-econom", "judiverse/problem-is-a-job-guarantee-the-solution-to-mass-ai-unemplo", "judiverse/problem-meta-hit-with-25b-penalty-for-deceiving-users-abou", "judiverse/problem-near-to-real-time-suspected-suicide-surveillance-n", "judiverse/problem-openai-safety-leader-quits-warning-ai-company-s-cu", "judiverse/problem-postgres-19-delayed-after-ai-reviewers-find-more-b", "judiverse/problem-research-agentic-ai-and-consumers", "judiverse/problem-run-full-ios-16-5-firmware-on-android-devices", "judiverse/problem-turn-a-wifi-7-router-into-a-320msps-software-defin", "judiverse/problem-vote-on-which-of-hacker-news-challenges-for-ai-hav", "judiverse/problem-wanted-to-download-an-iso-ended-up-writing-a-kerne", "judiverse/problem-why-aren-t-ai-companies-being-charged-for-hacking", "judiverse/problem-with-most-information-hidden-the-game-stratego-had", "judiverse/problem-yandex-bot-traffic-down-99-after-data-center-attac", "public-edge/findings-analog-human-switch-hardware-specification", "public-edge/findings-analog-human-switch-precedents-it-builds-on-to-verify-against", "public-edge/findings-analog-human-switch-the-analog-human-switch-no-irreversible-o", "public-edge/findings-analog-human-switch-what-the-estate-enforces-in-software-reco", "public-edge/findings-analog-human-switch-why-software-sign-off-is-not-enough", "public-edge/findings-competitors-regulators-reuse-0-1-evidence-labels", "public-edge/findings-competitors-regulators-reuse-0-2-how-the-facts-were-checked", "public-edge/findings-competitors-regulators-reuse-0-3-acronyms-first-use", "public-edge/findings-competitors-regulators-reuse-0-3-acronyms-first-use-2", "public-edge/findings-competitors-regulators-reuse-0-3-acronyms-first-use-3", "public-edge/findings-competitors-regulators-reuse-1-ten-things-a-new-engineer-shou", "public-edge/findings-competitors-regulators-reuse-2-1-intelligence-platforms-for-g", "public-edge/findings-competitors-regulators-reuse-2-2-procurement-and-supplier-vet", "public-edge/findings-competitors-regulators-reuse-2-3-cyber-governance-and-attack", "public-edge/findings-competitors-regulators-reuse-3-1-uk-public-procurement", "public-edge/findings-competitors-regulators-reuse-3-3-uk-defence-and-vetting", "public-edge/findings-competitors-regulators-reuse-4-agent-assurance-gaps-in-public", "public-edge/findings-competitors-regulators-reuse-5-reusable-systems", "public-edge/findings-mobbin-flows-audit-logs", "public-edge/findings-mobbin-flows-coverage-what-mobbin-held-and-what-it-did-not-1-3", "public-edge/findings-refero-flows-task-access-review-permissions-audit-log-2-3", "public-edge/findings-refero-flows-task-access-review-permissions-audit-log-3", "public-edge/findings-refero-flows-task-access-review-permissions-audit-log-5", "public-edge/findings-refero-flows-task-approval-workflow-1-2", "public-edge/findings-refero-flows-task-case-management-6", "public-edge/findings-refero-flows-task-compliance-dashboard-audit-1-2", "public-edge/findings-refero-flows-task-compliance-dashboard-audit-3", "public-edge/findings-refero-flows-task-connect-bank-account-open-banking", "public-edge/findings-refero-flows-task-contract-management-4", "public-edge/findings-refero-flows-task-contract-management-5", "public-edge/findings-refero-flows-task-fleet-management", "public-edge/findings-refero-flows-task-fleet-management-4", "public-edge/findings-refero-flows-task-fraud-alert-suspicious-transaction-2", "public-edge/findings-refero-flows-task-government-service-identity-verification-2", "public-edge/findings-refero-flows-task-government-service-identity-verification-3", "public-edge/findings-refero-flows-task-government-service-identity-verification-5", "public-edge/findings-refero-flows-task-inventory-warehouse", "public-edge/findings-refero-flows-task-procurement-tender-bid", "public-edge/findings-refero-flows-task-risk-register", "public-edge/findings-refero-flows-task-security-incident-alert-2", "public-edge/findings-refero-flows-task-shipment-tracking-logistics-1-2", "public-edge/findings-refero-flows-task-shipment-tracking-logistics-2", "public-edge/findings-refero-flows-task-vulnerability-management-2", "public-edge/parity-adobe-photoshop-inspect-and-filter-version-comments-public-cons", "public-edge/parity-auditboard-auditboard", "public-edge/parity-brightspace-brightspace", "public-edge/parity-customer-io-customer-io", "public-edge/parity-darktrace-darktrace", "public-edge/parity-fingerprint-inspect-event-json-payload-public-consortium", "public-edge/parity-innovate-uk-ukri-funding-service-innovate-uk-ukri-funding-servi", "public-edge/parity-metricstream-metricstream", "public-edge/parity-palantir-apollo-apollo", "public-edge/parity-palantir-foundry-foundry", "public-edge/parity-palantir-gotham-gotham", "public-edge/parity-palantir-technologies-apollo-apollo", "public-edge/parity-palantir-technologies-foundry-foundry", "public-edge/parity-palantir-technologies-gotham-gotham", "public-edge/parity-plane-finder-filter-and-inspect-flight-categories-public-consor", "public-edge/parity-twingate-twingate", "public-edge/parity-vanta-vanta", "public-edge/parity-weather-mini-trip-planning-and-save-public-consortium", "public-edge/parity-weather-mini-trip-planning-public-consortium", "public-edge/problem-cyberattack-halts-construction-of-tunnel-between-d", "public-edge/problem-details-of-thousands-of-us-immigration-agents-leak", "public-edge/problem-dutch-government-had-hacked-into-jared-kushner-s-p", "public-edge/problem-dutch-national-cyber-security-center-is-offline-be", "public-edge/problem-experts-warn-ai-copyright-changes-could-breach-aus", "public-edge/problem-help-the-ncsc-understand-how-organisations-approac", "public-edge/problem-how-north-korea-became-a-hacking-superpower", "public-edge/problem-how-sin-bin-for-red-light-jumpers-saves-cyclists-f", "public-edge/problem-how-the-military-will-hack-the-election", "public-edge/problem-incident-affecting-asos-customers", "public-edge/problem-introducing-a-new-ai-chapter-in-the-ncsc-research", "public-edge/problem-is-this-the-new-cyber-normal", "public-edge/problem-joint-ministerial-statement-on-the-risk-of-famine", "public-edge/problem-labor-and-the-coalition-are-poised-to-take-a-reckl", "public-edge/problem-legacy-system-local-plan-extension-letter-from-the", "public-edge/problem-meta-ordered-to-pay-25b-penalty-in-new-mexico-camb", "public-edge/problem-minister-mcneill-speech-from-hunger-hotspots-to-po", "public-edge/problem-more-people-protected-as-government-acts-to-make-m", "public-edge/problem-procurement-compliance-oversight", "public-edge/problem-public-electric-vehicle-charge-points-cyber-securi", "public-edge/problem-russia-s-decision-to-veto-this-draft-resolution-ri", "public-edge/problem-supporting-young-people-s-independence-through-tra", "public-edge/problem-the-united-kingdom-remains-committed-to-working-wi", "public-edge/problem-what-we-know-about-binance-s-6-3bn-hack", "public-edge/problem-william-gibson-and-the-trademarking-of-cyberspace", "clinic/accept-a-proxy-mandate", "clinic/acquire-and-send-images", "clinic/act-on-a-drug-safety-alert", "clinic/act-on-a-high-risk-answer", "clinic/add-an-appointment-to-my-calendar", "clinic/adjust-a-premium-at-renewal", "clinic/administer-a-medicine", "clinic/allocate-a-breach-between-trusts", "clinic/annotate-variants", "clinic/answer-a-patient-call", "clinic/apply-a-national-data-opt-out", "clinic/apply-a-treatment-rule", "clinic/apply-for-hra-and-rec-approval", "clinic/approve-a-data-release", "clinic/approve-a-leave-or-swap-request", "clinic/approve-a-questionnaire-or-test-threshold", "clinic/approve-a-staff-request", "clinic/arrange-a-cross-trust-mdt", "clinic/ask-for-advice-and-guidance", "clinic/ask-for-help-now", "clinic/assess-a-safeguarding-concern", "clinic/attend-an-appointment-with-the-person", "clinic/audit-access-to-records", "clinic/audit-antimicrobial-use", "clinic/audit-consent-records", "clinic/audit-data-use-against-mandates", "clinic/audit-mdt-attendance", "clinic/audit-reporting-quality", "clinic/audit-response-times", "clinic/audit-waiting-time-data", "clinic/authorise-a-result", "clinic/book-a-pathway-test", "clinic/book-an-appointment", "clinic/book-an-interpreter", "clinic/book-for-the-person-i-care-for", "clinic/book-from-the-referral-queue", "clinic/book-transport", "clinic/build-and-publish-the-rota", "clinic/build-the-clinic-template", "clinic/build-the-mdt-list", "clinic/build-the-variant-to-trial-graph", "clinic/cancel-a-clinic-session", "clinic/change-my-pharmacy", "clinic/chase-a-pathway-clock", "clinic/chase-missing-results", "clinic/check-a-chemotherapy-prescription", "clinic/check-access-controls", "clinic/check-fragment-size-before-sequencing", "clinic/check-identity-and-consent-before-scanning", "clinic/check-in-an-arriving-patient", "clinic/check-in-on-arrival", "clinic/check-laboratory-submissions-for-a-variant", "clinic/check-record-accuracy", "clinic/check-research-integrity", "clinic/check-staff-registration-and-training", "clinic/claim-my-edge-id-and-record", "clinic/classify-a-variant", "clinic/classify-software-as-a-medical-device", "clinic/clinically-check-a-prescription", "clinic/close-a-referral", "clinic/close-a-study", "clinic/collect-case-report-data", "clinic/compare-two-pipeline-runs", "clinic/complete-a-mental-health-check-in", "clinic/complete-a-pre-clinic-questionnaire", "clinic/complete-a-risk-formulation", "clinic/complete-mandatory-training", "clinic/complete-the-data-security-and-protection-toolkit", "clinic/confirm-attendance-and-prepare", "clinic/connect-a-lims", "clinic/connect-a-pacs-or-imaging-system", "clinic/counsel-a-patient-on-a-new-medicine", "clinic/decide-a-re-run", "clinic/deliver-cancer-nurse-specialist-support", "clinic/deploy-a-release", "clinic/design-a-smart-questionnaire", "clinic/design-a-trial", "clinic/detect-claim-anomalies", "clinic/discharge-from-the-pathway", "clinic/discharge-from-the-stream", "clinic/dispense-a-medicine", "clinic/escalate-to-emergency-services", "clinic/explain-a-result-in-plain-language", "clinic/explain-a-result-to-the-patient", "clinic/find-a-recruiting-trial-near-me", "clinic/flag-a-critical-finding", "clinic/flag-never-block", "clinic/follow-up-after-discharge-from-crisis-care", "clinic/follow-up-an-action-plan", "clinic/forecast-compliance", "clinic/forecast-days-to-treatment", "clinic/front-door-check-in", "clinic/fulfil-a-subject-access-request", "clinic/give-a-vaccination", "clinic/give-consent-on-behalf", "clinic/give-or-withdraw-consent", "clinic/grade-a-retinal-image", "clinic/hand-off-to-codex-or-jude", "clinic/handle-a-complaint", "clinic/handle-a-controlled-drug", "clinic/handle-a-did-not-attend", "clinic/handle-a-disclosure-of-abuse", "clinic/handle-a-freedom-of-information-request", "clinic/handle-a-policyholder-complaint", "clinic/handle-a-stock-shortage", "clinic/handle-the-record-after-death", "clinic/hold-a-remote-session", "clinic/hold-a-slot-for-a-tertiary-handover", "clinic/hold-the-duty-rota", "clinic/import-my-practice-register", "clinic/inspect-the-hazard-log-and-safety-case", "clinic/investigate-a-clinical-safety-incident", "clinic/issue-a-clinical-safety-case-report", "clinic/issue-a-field-safety-notice", "clinic/issue-a-genomic-report", "clinic/issue-a-policy", "clinic/maintain-laboratory-accreditation-evidence", "clinic/maintain-the-hazard-log", "clinic/maintain-the-suicide-prevention-plan", "clinic/make-a-follow-up-call", "clinic/make-an-urgent-same-day-referral", "clinic/manage-an-incident", "clinic/manage-freezer-alarms", "clinic/manage-patient-group-directions", "clinic/manage-patient-transport-and-interpreter-contracts", "clinic/manage-retention-and-destruction", "clinic/manage-room-allocation", "clinic/manage-the-information-asset-register", "clinic/manage-the-outbound-allow-list", "clinic/manage-the-reporting-backlog", "clinic/manage-the-room-and-equipment-register", "clinic/manage-the-waiting-list", "clinic/measure-a-nodule-and-its-growth", "clinic/measure-compute-energy-and-carbon", "clinic/measure-outcomes-over-time", "clinic/medication-safety-check", "clinic/mental-health-care-and-crisis", "clinic/message-my-care-team", "clinic/monitor-alerts-and-escalate", "clinic/monitor-data-feeds", "clinic/monitor-post-deployment-safety", "clinic/nominate-a-carer-or-proxy", "clinic/obtain-confirmation-of-capacity-and-capability", "clinic/offer-a-cancellation-slot", "clinic/offer-self-help-resources", "clinic/onboard-a-staff-member", "clinic/onboard-a-wearable-device-type", "clinic/online-urgent-triage", "clinic/order-a-diagnostic-test", "clinic/order-a-direct-access-test", "clinic/order-a-genomic-test", "clinic/order-a-repeat-medicine", "clinic/order-repeat-medicines-on-behalf", "clinic/pair-my-wearable-device", "clinic/person-held-health-record", "clinic/phone-a-critical-result", "clinic/plan-weekly-capacity", "clinic/post-a-receipt-to-the-agent-channel", "clinic/predict-fields-in-a-form", "clinic/prepare-a-clinic-letter-for-a-clinician", "clinic/prepare-a-patient-for-imaging", "clinic/prepare-for-a-cqc-inspection", "clinic/prescribe", "clinic/prescribe-under-the-pathway-protocol", "clinic/price-a-policy-from-clinic-data", "clinic/process-a-claim", "clinic/pseudonymise-a-dataset", "clinic/publish-results", "clinic/raise-a-complaint-or-feedback", "clinic/raise-a-safeguarding-concern-from-a-response", "clinic/randomise-a-participant", "clinic/read-assistive-output-on-an-image", "clinic/rebook-or-cancel-an-appointment", "clinic/receive-a-crisis-alert", "clinic/receive-a-crisis-alert-for-the-person", "clinic/receive-a-critical-result-for-my-patient", "clinic/receive-a-did-not-attend-notice", "clinic/receive-a-pre-authorisation-request", "clinic/receive-a-statutory-notification", "clinic/receive-and-book-in-a-sample", "clinic/receive-the-outcome-letter", "clinic/reclassify-a-variant", "clinic/reconcile-medicines", "clinic/reconcile-medicines-at-admission", "clinic/record-a-clinic-consultation", "clinic/record-a-dose-incident", "clinic/record-a-patient-deferred", "clinic/record-a-patient-s-own-view", "clinic/record-a-point-of-care-urine-result", "clinic/record-a-safeguarding-concern", "clinic/record-a-self-check-in-exception", "clinic/record-a-wound-or-skin-check", "clinic/record-an-adverse-drug-reaction", "clinic/record-attendance-outcome", "clinic/record-capacity-and-best-interests", "clinic/record-decision-to-treat", "clinic/record-sample-consent-version", "clinic/record-the-faster-diagnosis-outcome", "clinic/record-the-lawful-basis", "clinic/record-the-mdt-recommendation", "clinic/record-trial-eligibility-at-the-mdt", "clinic/refer-on-a-suspected-cancer-pathway", "clinic/refer-to-a-crisis-team", "clinic/refer-to-tertiary-care", "clinic/refer-to-the-mental-health-stream", "clinic/refer-to-the-multidisciplinary-team", "clinic/refresh-public-data-caches", "clinic/register-a-child-as-a-patient", "clinic/register-a-device-fault", "clinic/register-a-walk-in-patient", "clinic/register-an-imaging-device", "clinic/register-my-practice-with-the-clinic", "clinic/register-the-trial", "clinic/reject-an-unlabelled-sample", "clinic/report-a-change-in-the-person-s-condition", "clinic/report-a-data-breach", "clinic/report-a-dispensing-error", "clinic/report-a-safety-incident", "clinic/report-a-side-effect", "clinic/report-a-study", "clinic/report-an-adverse-event", "clinic/report-an-incident", "clinic/report-histology", "clinic/report-monthly-performance", "clinic/report-to-the-regulator", "clinic/report-usage-and-efficiency", "clinic/request-a-copy-of-my-record", "clinic/request-a-rota-swap-or-leave", "clinic/request-a-second-opinion", "clinic/request-a-second-read", "clinic/request-a-second-read-or-discrepancy-review", "clinic/request-a-young-person-s-own-access", "clinic/request-clinic-data-for-a-quote", "clinic/request-evidence-against-fundamental-standards", "clinic/retinal-screening", "clinic/review-a-breach-notification", "clinic/review-a-death-by-suicide", "clinic/review-a-real-result", "clinic/review-a-release-for-clinical-safety", "clinic/review-a-repeat-request", "clinic/review-a-trial-for-compliance", "clinic/review-a-wearable-alert", "clinic/review-assistive-output-labelling", "clinic/review-payslips-and-expenses-approvals", "clinic/review-the-caseload", "clinic/review-the-dashboard", "clinic/review-the-waiting-list-position", "clinic/review-the-worklist", "clinic/revoke-access-when-a-mandate-ends", "clinic/revoke-my-own-proxy-role", "clinic/reward-early-detection-participation", "clinic/roll-back-a-release", "clinic/rotate-keys-and-certificates", "clinic/run-a-business-continuity-drill", "clinic/run-a-catalogue-workload", "clinic/run-a-clinical-safety-training-check", "clinic/run-a-data-protection-impact-assessment", "clinic/run-a-pipeline", "clinic/run-a-pre-assessment", "clinic/run-a-safeguarding-review", "clinic/run-a-synthetic-pathway-drill", "clinic/run-a-urine-test-likelihood-ledger", "clinic/run-an-interim-analysis", "clinic/run-an-online-early-symptom-test", "clinic/run-quality-control", "clinic/run-scanner-quality-assurance", "clinic/run-sequencing-qc", "clinic/run-the-day-s-capacity-check", "clinic/run-the-mdt-meeting", "clinic/run-the-training-controller", "clinic/run-vulnerability-scans", "clinic/schedule-the-scan", "clinic/score-a-model-on-a-held-split", "clinic/score-and-interpret-a-response", "clinic/screen-a-patient-for-eligibility", "clinic/search-recruiting-trials", "clinic/see-my-day", "clinic/send-a-check-in-to-a-caseload", "clinic/send-a-results-ready-notice", "clinic/send-appointment-letters-and-reminders", "clinic/serve-a-vision-model", "clinic/set-a-surveillance-plan", "clinic/set-my-communication-preferences", "clinic/set-my-data-sharing-choices", "clinic/set-role-permissions-and-expiry", "clinic/settle-with-the-clinic", "clinic/share-a-medicines-list-with-the-clinic", "clinic/share-a-wearable-signal", "clinic/share-clinic-data-with-my-insurer", "clinic/share-images-with-another-trust", "clinic/share-risk-information-with-partners", "clinic/sign-a-clinic-letter", "clinic/sign-the-clinical-risk-management-plan", "clinic/simulate-trial-monitoring", "clinic/social-prescribing-referral", "clinic/stop-on-revoked-authority", "clinic/store-and-retrieve-genomic-data", "clinic/submit-a-cancer-registration", "clinic/submit-cancer-waiting-times-data", "clinic/submit-cosd-and-other-datasets", "clinic/submit-mdt-data", "clinic/submit-the-mental-health-dataset", "clinic/supervise-and-sign-off-a-trainee", "clinic/support-deploying-organisations", "clinic/symptom-check-and-triage", "clinic/take-a-blood-or-urine-sample", "clinic/take-a-mandate-ticket", "clinic/take-an-online-early-symptom-test", "clinic/take-informed-consent-for-research", "clinic/take-observations", "clinic/take-part-in-external-quality-assessment", "clinic/tell-the-patient-a-result", "clinic/test-a-hypothesis-on-an-open-cohort", "clinic/test-a-pre-registered-hypothesis", "clinic/trace-ingress-and-egress", "clinic/track-a-referral", "clinic/track-actions-from-the-mdt", "clinic/track-histology-turnaround", "clinic/track-my-cancer-pathway", "clinic/track-sample-custody", "clinic/track-the-62-day-clock-across-the-mdt", "clinic/train-staff-in-information-governance", "clinic/train-staff-in-suicide-awareness", "clinic/triage-a-mental-health-referral", "clinic/triage-a-patient-message", "clinic/triage-a-referral", "clinic/triage-inbox-items-for-a-clinician", "clinic/tune-alert-thresholds", "clinic/update-patient-demographics", "clinic/validate-a-new-assay", "clinic/validate-the-test-against-outcomes", "clinic/verify-a-device-classification", "clinic/verify-an-applicant-s-identity", "clinic/verify-an-evidence-statement", "clinic/verify-backups-and-restore", "clinic/vet-an-imaging-request", "clinic/view-my-medicines-list", "clinic/view-my-results", "clinic/view-results-in-scope", "clinic/wearable-signals", "clinic/wellbeing-questionnaires", "clinic/write-a-care-plan", "edge-consortium/accept-a-review-assignment", "edge-consortium/accept-supervision-of-a-learner", "edge-consortium/acknowledge-the-principles-version", "edge-consortium/add-jude-to-a-channel-under-mandate", "edge-consortium/add-members-and-guests-with-expiry", "edge-consortium/administer-a-database", "edge-consortium/admit-or-refuse-a-guest", "edge-consortium/agree-a-data-sharing-agreement", "edge-consortium/allocate-indirect-costs", "edge-consortium/answer-a-reference-request", "edge-consortium/answer-reviewer-questions-on-a-contribution", "edge-consortium/answer-with-sources-in-a-channel", "edge-consortium/apply-a-non-disclosure-agreement-to-a-room", "edge-consortium/apply-an-open-licence", "edge-consortium/apply-for-ethics-approval", "edge-consortium/apply-retention-and-deletion-to-records", "edge-consortium/apply-to-become-a-verified-reviewer", "edge-consortium/approve-a-compute-request", "edge-consortium/approve-a-connector-or-integration", "edge-consortium/approve-a-contribution-before-submission", "edge-consortium/approve-a-director-passkey-action", "edge-consortium/approve-a-federated-query-request", "edge-consortium/approve-a-guest-account", "edge-consortium/approve-a-reviewer-payment-batch", "edge-consortium/approve-a-team-member-data-access-request", "edge-consortium/approve-a-visiting-researcher-affiliation", "edge-consortium/approve-an-agent-consequential-action", "edge-consortium/approve-an-open-source-release", "edge-consortium/approve-and-mark-a-contribution-verified", "edge-consortium/approve-compute-spend-against-a-grant", "edge-consortium/approve-expenditure-against-a-grant", "edge-consortium/archive-a-channel", "edge-consortium/archive-a-module-version", "edge-consortium/ask-a-clarification-question", "edge-consortium/ask-jude-for-a-sourced-answer", "edge-consortium/assess-ownership-of-an-output", "edge-consortium/assign-a-persistent-identifier", "edge-consortium/attribute-carbon-to-a-project", "edge-consortium/attribute-cost-to-a-grant-and-project", "edge-consortium/audit-reviewer-quality-by-re-review", "edge-consortium/audit-the-signing-log", "edge-consortium/award-and-publish-the-contract-award-notice", "edge-consortium/bind-a-device", "edge-consortium/book-a-room-or-resource", "edge-consortium/build-a-project-team-across-universities", "edge-consortium/build-an-interactive-prototype", "edge-consortium/build-the-research-profile-and-skills-record", "edge-consortium/bulk-enrol-students-and-staff", "edge-consortium/catalogue-a-library-item", "edge-consortium/charge-compute-and-storage-to-grant-codes", "edge-consortium/chase-an-overdue-review", "edge-consortium/check-exclusion-grounds", "edge-consortium/choose-a-specialism", "edge-consortium/claim-payment-for-a-completed-review", "edge-consortium/classify-data-sensitivity", "edge-consortium/close-a-grant", "edge-consortium/close-a-project-and-archive-its-records", "edge-consortium/co-edit-a-markdown-in-real-time", "edge-consortium/complete-a-right-to-work-check", "edge-consortium/complete-research-integrity-training", "edge-consortium/compose-an-image-in-the-imaging-studio", "edge-consortium/compose-audio-in-judio", "edge-consortium/conduct-a-probation-review", "edge-consortium/confirm-a-delivery", "edge-consortium/confirm-a-person-s-enrolment-status", "edge-consortium/confirm-a-time-and-send-the-invitation", "edge-consortium/contribute-a-library-entry-as-an-outsider", "edge-consortium/correct-a-student-record", "edge-consortium/cost-a-grant-application", "edge-consortium/count-agents-and-subagents-per-task", "edge-consortium/create-a-contract-from-a-template", "edge-consortium/create-a-job-role-and-vacancy", "edge-consortium/create-a-markdown-document", "edge-consortium/create-a-module", "edge-consortium/create-a-team-and-its-channels", "edge-consortium/create-an-availability-poll", "edge-consortium/curate-a-collection", "edge-consortium/declare-a-conflict-of-interest", "edge-consortium/declare-exclusion-grounds-and-conflicts", "edge-consortium/decline-an-assignment", "edge-consortium/define-a-purpose-view", "edge-consortium/define-or-change-a-schema", "edge-consortium/delete-data-at-retention-end", "edge-consortium/deploy-a-release", "edge-consortium/deposit-a-dataset-with-a-persistent-identifier", "edge-consortium/deprecate-a-feature", "edge-consortium/detect-unlicensed-reuse", "edge-consortium/draw-on-judecanvas", "edge-consortium/edit-a-video", "edge-consortium/end-employment-and-revoke-access", "edge-consortium/enforce-the-code-of-conduct", "edge-consortium/enrol-on-a-doctoral-training-module", "edge-consortium/escalate-a-dispute-to-the-editor", "edge-consortium/evaluate-bids-with-a-scoring-panel", "edge-consortium/expire-or-revoke-a-licence", "edge-consortium/explore-the-knowledge-graph-visually", "edge-consortium/export-a-creation-with-licence-and-attribution", "edge-consortium/export-a-meeting-record-and-audit-chain", "edge-consortium/export-and-verify-an-audit-chain", "edge-consortium/find-a-time-across-attendees", "edge-consortium/find-and-follow-tender-opportunities", "edge-consortium/find-funding-calls", "edge-consortium/flag-a-concern-without-blocking", "edge-consortium/flag-plagiarism-or-an-integrity-concern", "edge-consortium/governed-intelligence-degree", "edge-consortium/grant-a-clearance-with-expiry", "edge-consortium/grant-a-licence-to-a-third-party", "edge-consortium/grant-a-maintainer-role-with-expiry", "edge-consortium/hand-over-to-another-agent", "edge-consortium/hand-over-work-with-the-baton", "edge-consortium/handle-a-contract-dispute", "edge-consortium/handle-a-data-subject-access-request", "edge-consortium/handle-an-academic-appeal-on-marks", "edge-consortium/handle-an-appeal", "edge-consortium/hold-an-audit-interview-by-secure-call", "edge-consortium/import-external-records", "edge-consortium/inspect-carbon-and-energy-records", "edge-consortium/inspect-release-gates-and-evidence", "edge-consortium/inspect-retention-schedule-compliance", "edge-consortium/inspect-reviewer-payment-records", "edge-consortium/international-technical-students", "edge-consortium/issue-a-certificate-of-membership", "edge-consortium/issue-a-takedown-or-attribution-request", "edge-consortium/issue-a-transcript-of-completed-modules", "edge-consortium/issue-an-employment-contract", "edge-consortium/issue-an-offer-letter", "edge-consortium/issue-findings-and-track-remediation", "edge-consortium/issue-the-decision-letter", "edge-consortium/issue-the-written-statement-of-particulars", "edge-consortium/join-a-secure-call-through-the-lobby", "edge-consortium/join-consortium-through-a-university", "edge-consortium/judiverse-included-in-tuition", "edge-consortium/link-orcid-to-the-profile", "edge-consortium/log-a-receipt-for-every-action", "edge-consortium/make-a-below-threshold-direct-award", "edge-consortium/manage-a-controlled-vocabulary", "edge-consortium/manage-a-grant-extension-or-variation", "edge-consortium/manage-an-embargo-and-release-date", "edge-consortium/manage-consent-records", "edge-consortium/manage-institutional-roles-and-delegates", "edge-consortium/map-a-dataset-to-the-knowledge-graph", "edge-consortium/map-a-module-to-a-challenge-specification", "edge-consortium/map-a-module-to-programme-and-standard", "edge-consortium/mark-and-moderate-submissions", "edge-consortium/match-reviewers-to-a-submission", "edge-consortium/measure-live-energy-per-job-and-service", "edge-consortium/merge-duplicate-entities", "edge-consortium/merge-with-required-checks", "edge-consortium/mirror-a-repository-to-the-source-repository", "edge-consortium/moderate-a-post", "edge-consortium/monitor-data-quality", "edge-consortium/monitor-job-progress-and-output", "edge-consortium/monitor-tokens-prompts-and-time-to-evidence", "edge-consortium/negotiate-clauses-with-the-counterparty", "edge-consortium/nominate-a-learner-for-reviewer-training", "edge-consortium/offer-the-degree-with-a-partner-institution", "edge-consortium/onboard-a-new-starter", "edge-consortium/open-an-issue", "edge-consortium/paid-student-work-with-tuition-offset", "edge-consortium/pass-reviewer-credential-verification", "edge-consortium/pause-or-pre-empt-a-job", "edge-consortium/pay-reviewer-fees", "edge-consortium/pay-salary", "edge-consortium/petition-for-student-finance", "edge-consortium/pin-edit-or-delete-a-post", "edge-consortium/plan-a-maintenance-window", "edge-consortium/plan-a-procurement-and-publish-pipeline-notice", "edge-consortium/post-an-announcement", "edge-consortium/pre-register-a-study", "edge-consortium/prepare-a-review-summary-for-a-reviewer", "edge-consortium/prepare-an-employment-contract-template", "edge-consortium/preserve-and-migrate-file-formats", "edge-consortium/process-a-supplier-invoice", "edge-consortium/produce-a-carbon-reduction-plan", "edge-consortium/produce-a-graphic-in-visual-studio", "edge-consortium/produce-a-statutory-data-return", "edge-consortium/propose-a-change-from-a-fork", "edge-consortium/propose-a-real-world-problem-project", "edge-consortium/provide-a-funder-audit-trail", "edge-consortium/provide-a-reference", "edge-consortium/publish-a-data-dictionary", "edge-consortium/publish-a-markdown-as-deck-or-book", "edge-consortium/publish-a-new-dataset-version", "edge-consortium/publish-a-tender-notice", "edge-consortium/publish-a-terms-version-and-require-re-acceptance", "edge-consortium/publish-an-open-access-version", "edge-consortium/publish-documentation", "edge-consortium/publish-module-materials-to-the-library", "edge-consortium/publish-the-catalogue-to-open-indexes", "edge-consortium/publish-the-contract-details-notice", "edge-consortium/publish-the-monthly-carbon-and-cost-statement", "edge-consortium/raise-a-challenge-or-complaint", "edge-consortium/raise-a-grievance", "edge-consortium/raise-a-progress-or-wellbeing-concern", "edge-consortium/rank-agents-and-models-by-efficiency", "edge-consortium/read-a-verified-entry-with-its-provenance", "edge-consortium/real-project-assignments", "edge-consortium/receive-a-mandate", "edge-consortium/receive-bids-in-a-sealed-queue", "edge-consortium/receive-the-award-outcome", "edge-consortium/receive-time-boxed-auditor-access", "edge-consortium/reconcile-monthly-spend", "edge-consortium/reconcile-provider-invoices-with-metered-use", "edge-consortium/reconcile-with-the-student-record-system", "edge-consortium/record-a-call-with-consent", "edge-consortium/record-a-conflict-of-interest-assessment", "edge-consortium/record-a-data-processing-agreement", "edge-consortium/record-a-priority-date", "edge-consortium/record-a-trademark-and-brand-use-rule", "edge-consortium/record-an-invention-disclosure", "edge-consortium/record-applications-and-outcomes-in-a-portfolio", "edge-consortium/record-attendance-and-engagement", "edge-consortium/record-authorship-and-contributor-roles", "edge-consortium/record-background-and-foreground-ip", "edge-consortium/record-copyright-and-attribution", "edge-consortium/record-energy-per-job", "edge-consortium/record-payment-compliance", "edge-consortium/record-provenance-and-lineage", "edge-consortium/record-working-hours-and-leave", "edge-consortium/refuse-an-action-outside-the-mandate", "edge-consortium/register-a-compute-resource", "edge-consortium/register-a-dataset-in-the-catalogue", "edge-consortium/register-a-delegated-authority", "edge-consortium/register-a-research-project-with-an-id", "edge-consortium/register-a-university-as-a-member-institution", "edge-consortium/register-as-a-supplier", "edge-consortium/register-with-the-office-for-students", "edge-consortium/reject-a-contribution-with-reasons", "edge-consortium/release-a-version", "edge-consortium/release-marks-and-feedback", "edge-consortium/release-results-with-disclosure-control", "edge-consortium/renew-annual-institutional-membership", "edge-consortium/renew-reviewer-verification", "edge-consortium/report-a-personal-data-breach", "edge-consortium/report-a-research-misconduct-concern", "edge-consortium/report-carbon-intensity-per-accepted-outcome", "edge-consortium/report-expenditure-to-the-funder", "edge-consortium/report-monitoring-measures", "edge-consortium/report-performance-indicator-evidence", "edge-consortium/report-review-throughput-and-times", "edge-consortium/report-tokens-prompts-and-time-to-evidence", "edge-consortium/report-utilisation", "edge-consortium/request-a-compute-allowance-for-a-project", "edge-consortium/request-a-purpose-view-on-federated-data", "edge-consortium/request-a-sandbox-with-gpu-and-data", "edge-consortium/request-approval-for-a-consequential-action", "edge-consortium/request-changes-from-the-contributor", "edge-consortium/request-supervisor-allocation", "edge-consortium/research-and-development-partnership", "edge-consortium/research-consortium", "edge-consortium/reserve-capacity-for-a-deadline", "edge-consortium/resolve-conflicting-reviews", "edge-consortium/respond-to-a-security-incident", "edge-consortium/respond-to-an-invitation", "edge-consortium/retract-or-correct-a-verified-entry", "edge-consortium/retry-a-failed-job", "edge-consortium/review-a-contribution-against-criteria", "edge-consortium/review-a-funder-expenditure-statement", "edge-consortium/review-a-pull-request", "edge-consortium/review-access-logs-and-anomalies", "edge-consortium/review-agent-receipts", "edge-consortium/review-module-evaluation-feedback", "edge-consortium/review-quality-and-standards-evidence", "edge-consortium/review-the-dpia-and-record-of-processing", "edge-consortium/revoke-a-purpose-view", "edge-consortium/revoke-access-for-a-leaver", "edge-consortium/rotate-keys", "edge-consortium/run-a-clarification-round", "edge-consortium/run-a-compute-job-under-mandate", "edge-consortium/run-a-cyber-recovery-drill", "edge-consortium/run-a-live-assessment", "edge-consortium/run-a-live-seminar", "edge-consortium/run-a-security-classified-procurement", "edge-consortium/run-a-training-job-on-scheduled-compute", "edge-consortium/run-a-vulnerability-scan-on-a-release", "edge-consortium/run-an-approved-query-where-data-lives", "edge-consortium/run-code-in-the-sandbox", "edge-consortium/run-data-quality-checks-on-a-database", "edge-consortium/sample-verified-entries-for-quality", "edge-consortium/save-sources-to-a-reading-list", "edge-consortium/scale-with-an-external-provider", "edge-consortium/schedule-a-job", "edge-consortium/schedule-a-meeting-on-request", "edge-consortium/schedule-a-secure-government-resource", "edge-consortium/schedule-a-send-and-set-a-signature", "edge-consortium/schedule-and-record-supervision-meetings", "edge-consortium/schedule-deliveries-and-logistics", "edge-consortium/search-across-mail-channels-and-meetings", "edge-consortium/search-libraries-and-the-knowledge-graph", "edge-consortium/send-a-contract-for-signature", "edge-consortium/send-mail-from-an-estate-mailbox", "edge-consortium/set-a-carbon-budget-and-threshold", "edge-consortium/set-a-project-access-control-policy", "edge-consortium/set-agent-authority-scope", "edge-consortium/set-and-publish-key-performance-indicators", "edge-consortium/set-assessment-and-marking-scheme", "edge-consortium/set-contract-parties-permissions-and-expiry", "edge-consortium/set-evaluation-criteria-and-weightings", "edge-consortium/set-mail-rules-flags-and-snooze", "edge-consortium/set-mention-and-tag-rules", "edge-consortium/set-out-of-office-and-working-hours", "edge-consortium/set-permissions-on-project-records", "edge-consortium/set-presence", "edge-consortium/set-project-governance", "edge-consortium/set-quotas-per-project", "edge-consortium/set-renewal-and-expiry-alerts", "edge-consortium/set-retention-and-export-a-channel", "edge-consortium/set-review-criteria-for-a-field", "edge-consortium/set-review-fields-and-availability", "edge-consortium/set-scheduling-policy-and-queues", "edge-consortium/set-the-access-tier-on-an-item", "edge-consortium/set-the-review-deadline-and-reminders", "edge-consortium/set-up-an-awarded-grant", "edge-consortium/share-a-calendar-with-a-delegate", "edge-consortium/share-a-campus-with-a-partner-university", "edge-consortium/share-a-creation-in-the-judebox-exchange", "edge-consortium/share-a-document-in-a-call", "edge-consortium/share-a-workspace-folder-across-universities", "edge-consortium/shift-jobs-to-low-carbon-windows", "edge-consortium/show-undisclosed-figures-as-gaps", "edge-consortium/sign-a-collaboration-agreement", "edge-consortium/sign-a-contract-with-edge-id", "edge-consortium/sign-off-a-progression-milestone", "edge-consortium/sign-the-contributor-agreement", "edge-consortium/sign-the-employment-contract", "edge-consortium/sign-the-public-contract", "edge-consortium/stop-or-kill-a-job", "edge-consortium/store-the-signed-contract-in-the-vault", "edge-consortium/student-to-contractor-pathway", "edge-consortium/submit-a-bid", "edge-consortium/submit-a-contribution-for-review", "edge-consortium/submit-a-grant-application", "edge-consortium/submit-a-paper-to-consortium-review", "edge-consortium/submit-an-annual-report-to-the-funder", "edge-consortium/submit-an-invoice-and-track-payment", "edge-consortium/submit-staff-for-security-vetting", "edge-consortium/suspend-or-end-a-membership", "edge-consortium/terminate-a-contract", "edge-consortium/terminate-a-contract-and-publish-the-notice", "edge-consortium/trace-network-ingress-and-egress", "edge-consortium/track-research-outputs-for-a-ref-return", "edge-consortium/track-royalties-and-revenue-share", "edge-consortium/triage-a-security-vulnerability-report", "edge-consortium/triage-the-submissions-queue", "edge-consortium/university-enrolment", "edge-consortium/update-supplier-information", "edge-consortium/use-the-terminal", "edge-consortium/vary-a-contract", "edge-consortium/verify-a-fact-before-stating-it", "edge-consortium/verify-a-fact-in-the-knowledge-graph", "edge-consortium/verify-a-person-s-identity", "edge-consortium/verify-a-signature", "edge-consortium/verify-institutional-accreditation-status", "edge-consortium/verify-procurement-notice-publication", "edge-consortium/verify-the-accessibility-statement", "edge-consortium/verify-the-audit-chain-offline", "edge-consortium/version-and-roll-back-a-creation", "edge-consortium/vet-a-reviewer-or-supplier-staff-member", "edge-consortium/view-the-diary-by-day-week-month", "edge-consortium/view-the-public-reviewer-track-record", "edge-consortium/vote-on-a-poll", "edge-consortium/write-a-data-management-plan", "edge-consortium/write-a-paper-in-markdowns-with-co-authors", "finance/accept-a-cheque-deposit", "finance/access-a-read-only-evidence-room", "finance/account-for-agent-resource-use", "finance/accrue-deposit-interest-expense", "finance/acknowledge-a-complaint", "finance/act-under-a-power-of-attorney", "finance/action-an-erasure-or-rectification-request", "finance/activate-a-monitoring-rule", "finance/activate-a-proposed-rate", "finance/add-a-joint-account-holder", "finance/add-a-payee-with-name-check", "finance/add-or-remove-a-signatory", "finance/agent-mandate", "finance/agent-onboarding", "finance/agent-retirement", "finance/agree-a-forbearance-plan", "finance/amend-a-policy-mid-term", "finance/amend-a-repayment-schedule", "finance/amend-or-cancel-a-standing-order", "finance/analyse-complaint-root-causes", "finance/answer-a-customer-question-from-records", "finance/answer-a-subject-access-request", "finance/apply-a-database-migration", "finance/apply-a-payment-holiday", "finance/apply-enhanced-due-diligence", "finance/apply-for-a-business-account", "finance/apply-for-a-business-loan-or-facility", "finance/apply-for-a-personal-loan", "finance/apply-for-loan", "finance/apply-reasonable-adjustments", "finance/apply-retention-and-deletion", "finance/approve-a-financial-promotion", "finance/approve-a-new-integration-or-api-consumer", "finance/approve-or-decline-an-application", "finance/approve-or-reject-a-payment", "finance/assess-a-claim", "finance/assess-an-outsourcing-or-third-party", "finance/assess-income-and-expenditure", "finance/assess-source-of-funds-and-wealth", "finance/audit-partner-usage", "finance/authorise-card-transactions-in-real-time", "finance/back-up-the-database", "finance/bill-a-partner", "finance/block-a-payment-in-flight", "finance/book-an-appointment", "finance/bulk-payments", "finance/business-roles-and-approvals", "finance/calculate-and-post-interest-accrual", "finance/calculate-the-liquidity-coverage-ratio", "finance/cancel-a-direct-debit", "finance/cancel-a-policy", "finance/capital-to-outcomes-ledger", "finance/cards", "finance/change-contact-details", "finance/claim-a-direct-debit-indemnity-refund", "finance/close-a-business-account", "finance/close-a-period", "finance/close-account", "finance/close-an-account", "finance/collect-direct-debits-as-an-originator", "finance/collect-premiums", "finance/compare-cover-options", "finance/complete-entity-tax-classification", "finance/complete-tax-residency-self-certification", "finance/compute-expected-credit-losses", "finance/confidence-report", "finance/configure-scheduled-jobs", "finance/confirm-balances-as-external-auditor", "finance/connect-external-accounts", "finance/contact-a-customer-in-arrears", "finance/continue-a-joint-account-after-a-death", "finance/create-a-monitoring-rule", "finance/create-a-prospect-record", "finance/create-a-savings-product", "finance/create-and-release-a-hold", "finance/crypto-wallet", "finance/decide-a-treasury-proposal", "finance/decide-an-app-reimbursement-claim", "finance/decide-and-file-a-sar", "finance/decline-with-reasons", "finance/deploy-a-release-as-jude", "finance/deploy-a-release-through-deploy-py", "finance/design-a-new-lending-product", "finance/detect-a-missed-repayment", "finance/direct-debits", "finance/disburse-a-loan", "finance/disclose-a-vulnerability-or-need", "finance/disposition-an-alert", "finance/dispute-a-card-transaction", "finance/disputes", "finance/draw-down-from-a-facility", "finance/enforce-strong-customer-authentication", "finance/escalate-a-customer-concern", "finance/execute-a-money-market-placement", "finance/exit-a-customer-for-financial-crime", "finance/explain-a-transaction", "finance/export-the-evidence-bundle", "finance/export-transactions-to-accounting-software", "finance/financing-graph-independent-demand", "finance/flag-a-mule-account", "finance/fraud-sharing-network", "finance/freeze-or-unfreeze-a-card", "finance/fund-settlement-accounts-for-rails", "finance/give-or-revoke-open-banking-consent", "finance/grant-an-operator-role", "finance/hand-a-case-to-a-specialist-team", "finance/hand-work-between-agents", "finance/handle-a-banking-protocol-concern", "finance/handle-a-breathing-space-notice", "finance/handle-a-debt-sale", "finance/handle-a-gambling-block-request", "finance/handle-a-mental-capacity-concern", "finance/handle-a-partner-data-request-or-deletion", "finance/handle-a-security-incident", "finance/handle-a-trigger-event-review", "finance/handle-a-whistleblowing-report", "finance/handle-an-indemnity-claim-as-paying-bank", "finance/hedge-interest-rate-exposure", "finance/import-a-bank-or-rail-statement", "finance/inspect-a-sar-audit-trail", "finance/introduce-a-customer-to-a-specialist", "finance/investigate-a-complaint", "finance/investigate-a-payment-exception", "finance/investigate-a-production-error", "finance/investigate-a-suspected-fraud", "finance/invite-a-user-with-a-role", "finance/issue-a-default-notice", "finance/issue-a-final-response", "finance/issue-a-finding-to-the-bank", "finance/issue-a-policy", "finance/issue-a-replacement-card-in-branch", "finance/issue-an-agent-mandate", "finance/issue-annual-loan-statements", "finance/joint-accounts", "finance/lend", "finance/lift-an-account-freeze", "finance/log-an-operational-incident", "finance/log-an-outbound-contact", "finance/maintain-internal-and-nostro-accounts", "finance/maintain-the-business-wide-risk-assessment", "finance/maintain-the-chart-of-accounts", "finance/maintain-the-record-of-processing", "finance/maintain-the-regulatory-obligations-register", "finance/maintain-the-risk-register", "finance/make-a-complaint", "finance/make-a-faster-payment", "finance/make-a-regulatory-notification", "finance/make-an-internal-transfer", "finance/make-an-international-payment", "finance/manage-a-syndicated-participation", "finance/manage-capacity", "finance/manage-encryption-and-device-binding", "finance/manage-environments-and-secrets-free-config", "finance/manage-foreign-exchange-positions", "finance/manage-partner-user-access", "finance/manage-payee-name-check-responses", "finance/manage-payment-limits", "finance/manage-senior-manager-responsibilities", "finance/manage-the-central-bank-reserve-account", "finance/map-important-business-services", "finance/mark-a-customer-as-a-fraud-victim", "finance/match-statement-rows-to-ledger-legs", "finance/measure-carbon-tokens-and-time-to-evidence", "finance/message-a-customer-in-a-case", "finance/money-provenance-certificate", "finance/monitor-chaps-cut-offs-and-queue", "finance/monitor-conduct-and-quality", "finance/monitor-facility-covenants", "finance/monitor-intraday-liquidity", "finance/monitor-job-runs-and-leases", "finance/monitor-privileged-actions", "finance/monitor-product-outcomes", "finance/monitor-service-health", "finance/monitor-without-blocking", "finance/notify-a-claim", "finance/notify-a-death-and-restrict-accounts", "finance/notify-a-personal-data-breach", "finance/offboard-a-business-customer", "finance/onboard-a-partner-organisation", "finance/open-a-fixed-term-deposit", "finance/open-a-foreign-currency-account", "finance/open-a-personal-current-account", "finance/open-a-savings-account", "finance/open-a-service-case", "finance/open-a-transaction-receipt", "finance/open-business-account", "finance/open-personal-account", "finance/open-virtual-accounts-under-a-master", "finance/order-business-cards-for-staff", "finance/overdraft", "finance/pay-redress", "finance/perform-a-data-protection-impact-assessment", "finance/perform-an-internal-audit", "finance/place-a-temporary-account-freeze", "finance/plan-a-portfolio-review", "finance/plan-the-internal-audit-year", "finance/post-a-manual-journal", "finance/prepare-a-business-proposition", "finance/prepare-a-compliance-return", "finance/prepare-a-payment-for-approval", "finance/prepare-a-payment-for-human-approval", "finance/prepare-funds-transfer-pricing", "finance/prepare-regulatory-returns", "finance/prepare-statutory-accounts", "finance/process-a-bacs-input-and-output-cycle", "finance/process-a-cheque-clearing-exception", "finance/process-a-counter-cash-deposit", "finance/process-a-counter-cash-withdrawal", "finance/process-a-loan-repayment-manually", "finance/process-an-inbound-faster-payment", "finance/process-card-clearing-and-settlement", "finance/produce-a-kyc-file-for-review", "finance/produce-a-trial-balance", "finance/produce-the-annual-mlro-report", "finance/produce-the-fscs-single-customer-view", "finance/propose-a-future-rate", "finance/propose-a-treasury-credit", "finance/provenance-coverage-report", "finance/provide-extra-evidence-on-request", "finance/proximity-verification", "finance/publish-terms-and-conditions", "finance/pull-and-record-a-credit-reference", "finance/query-a-mule-account-indicator", "finance/raise-an-internal-suspicion-report", "finance/receive-a-fraud-intelligence-signal", "finance/receive-a-rate-change-notice", "finance/recognise-loan-interest-income", "finance/reconcile-rail-settlement", "finance/reconcile-suspense-accounts", "finance/record-a-credit-decision-override", "finance/record-a-customer-meeting", "finance/record-a-customer-risk-assessment", "finance/record-a-declined-application-reason", "finance/record-a-regulatory-breach", "finance/record-a-support-need", "finance/record-a-vulnerability-disclosure", "finance/record-aml-training-completion", "finance/record-and-test-a-control", "finance/recover-account-access", "finance/recover-funds-from-a-receiving-account", "finance/refer-a-complaint-to-fos", "finance/refer-to-a-debt-collection-agency", "finance/refer-to-debt-advice", "finance/refund-an-unauthorised-transaction", "finance/register-a-court-appointed-deputy", "finance/register-a-default-with-agencies", "finance/register-a-lasting-power-of-attorney", "finance/register-a-new-device", "finance/register-a-third-party-mandate", "finance/register-and-value-collateral", "finance/register-beneficial-owners", "finance/regulator-inspection", "finance/release-estate-funds", "finance/rely-on-another-firm-s-due-diligence", "finance/remove-a-joint-holder", "finance/rename-an-account", "finance/renew-a-policy", "finance/reopen-a-period", "finance/repair-a-misdirected-payment", "finance/repay-or-overpay-a-loan", "finance/report-a-bereavement", "finance/report-a-change-of-company-details", "finance/report-a-lost-or-stolen-card", "finance/report-an-app-scam", "finance/report-an-unauthorised-payment", "finance/report-complaints-data", "finance/report-fraud-losses-to-the-regulator", "finance/report-fraud-to-action-fraud-or-police", "finance/report-insurance-data-to-the-regulator", "finance/report-payment-statistics", "finance/report-to-credit-reference-agencies", "finance/request-a-bank-statement-letter", "finance/request-a-company-audit-letter", "finance/request-a-covenant-waiver", "finance/request-a-date-of-death-balance-letter", "finance/request-a-defence-against-money-laundering", "finance/request-a-loan-settlement-quote", "finance/request-a-trade-finance-instrument", "finance/request-affordability-evidence", "finance/request-an-arranged-overdraft", "finance/request-an-insurance-quote-from-kyc", "finance/request-further-information", "finance/request-money", "finance/request-my-personal-data", "finance/request-owner-approval-for-a-consequential-action", "finance/request-records-from-the-bank", "finance/reset-access-for-a-customer", "finance/resolve-within-3-business-days", "finance/respond-to-a-production-order", "finance/respond-to-an-open-banking-payment-request", "finance/responsible-lending-decision", "finance/restore-from-a-backup", "finance/retry-a-failed-job-step", "finance/retry-a-failed-occurrence", "finance/return-an-unapplied-payment", "finance/reverse-a-journal-with-a-linked-entry", "finance/review-a-complaint-file-as-fos", "finance/review-a-loan-application", "finance/review-access-rights", "finance/review-an-application-queue", "finance/review-app-reimbursement-performance", "finance/review-customer-financial-statements", "finance/review-device-and-session-signals", "finance/review-operational-resilience-evidence", "finance/review-sm-cr-responsibilities", "finance/review-the-vulnerability-register", "finance/revoke-a-role-on-leaving", "finance/revoke-an-agent-mandate", "finance/roll-back-a-release", "finance/rotate-keys-and-device-bindings", "finance/run-a-consumer-duty-outcomes-review", "finance/run-a-disaster-recovery-test", "finance/run-a-fair-value-assessment", "finance/run-a-kyc-check-on-a-subject", "finance/run-a-ledger-report-for-a-period", "finance/run-a-liquidity-stress-test", "finance/run-an-annual-relationship-review", "finance/run-due-jobs-as-service-actor", "finance/run-due-standing-orders", "finance/run-end-of-day-payment-reconciliation", "finance/run-payroll-payments", "finance/run-penetration-testing", "finance/run-periodic-kyc-refresh", "finance/run-protected-reveals", "finance/run-sanctions-screening", "finance/run-transaction-monitoring", "finance/sample-test-receipts-as-external-auditor", "finance/sanctions-and-pep-screening", "finance/save-an-unfinished-application", "finance/savings-pots", "finance/schedule-a-future-payment", "finance/scheduled-payments", "finance/screen-a-list-against-sanctions", "finance/screen-existing-customers-against-list-updates", "finance/screen-for-politically-exposed-persons", "finance/search-transactions", "finance/security-and-devices", "finance/send-a-chaps-payment", "finance/send-a-receiving-bank-app-notice", "finance/send-money", "finance/set-and-monitor-exposure-limits", "finance/set-balance-and-payment-alerts", "finance/set-card-limits-and-controls", "finance/set-eligibility-rules", "finance/set-fees-and-charges", "finance/set-joint-operating-rules", "finance/set-loan-terms-and-produce-an-offer", "finance/set-negotiated-pricing", "finance/set-up-a-direct-debit", "finance/set-up-a-mandate", "finance/set-up-a-standing-order", "finance/set-up-merchant-acquiring", "finance/settle-a-claim", "finance/settle-and-close-a-loan", "finance/solution-bank-funding", "finance/stop-a-savings-enrolment", "finance/submit-transactions-for-monitoring", "finance/support-a-customer-in-financial-difficulty", "finance/support-the-external-audit", "finance/switch-account-to-edge-bank", "finance/terminate-a-partnership", "finance/trace-ingress-and-egress", "finance/track-remediation-actions", "finance/transaction-monitoring", "finance/transition-a-case-state", "finance/triage-a-case-queue", "finance/tune-a-fraud-rule", "finance/underwrite-a-business-facility", "finance/underwrite-a-policy", "finance/update-communication-preferences", "finance/upload-a-bulk-payment-file", "finance/validate-a-credit-risk-model", "finance/verify-a-caller-or-visitor", "finance/verify-a-company-and-its-officers", "finance/verify-a-regulatory-return-against-the-ledger", "finance/verify-a-trust-or-charity", "finance/verify-address-evidence", "finance/verify-an-identity-through-edge-id", "finance/verify-executor-or-administrator", "finance/verify-identity-evidence", "finance/view-and-download-a-statement", "finance/view-balances-and-available-funds", "finance/view-multi-account-cash-position", "finance/view-pin-securely", "finance/void-an-imported-statement", "finance/withdraw-a-product", "finance/withdraw-a-proposed-rate", "finance/withdraw-an-application", "finance/work-the-fraud-alert-queue", "finance/write-a-fraud-case-outcome-letter", "finance/write-off-a-debt", "foundation/automated-review-outreach-and-scheduling", "foundation/bottleneck-and-blocker-removal", "foundation/claim-organisation-by-magic-link", "foundation/evaluation-and-commercial-licensing", "foundation/licence-enforcement-evidence", "foundation/organisation-directory-and-claim", "foundation/pre-onboarded-organisations", "foundation/regulator-auditor-reviewer-workspace", "foundation/scope-change-intake", "judiverse/agent-builder", "judiverse/calendar-and-scheduling", "judiverse/contacts-directory-and-search", "judiverse/deep-research", "judiverse/documents-and-forms", "judiverse/jude-assistant", "judiverse/judebox-agents", "judiverse/judebox-calls", "judiverse/judebox-connections", "judiverse/judebox-contacts", "judiverse/judebox-email", "judiverse/judebox-memory-and-vault", "judiverse/judebox-messaging", "judiverse/judebox-music", "judiverse/judebox-numbers-and-mailboxes", "judiverse/judebox-webhooks", "judiverse/judiverse-included-with-every-licence", "judiverse/judiverse-standalone-subscription", "judiverse/mail-and-agent-mailboxes", "judiverse/messages-and-chat", "judiverse/procedure-compilation", "judiverse/tasks-and-workflows", "public-edge/accept-a-package-referral", "public-edge/act-on-a-deadline-stack-item", "public-edge/activate-the-emergency-centre-and-keep-the-decision-log", "public-edge/adopt-a-risk-based-reserves-strategy-with-minimum-level-and", "public-edge/advise-on-legality-and-sign-the-legal-implications-of-a-repo", "public-edge/agree-a-care-and-support-plan-and-personal-budget", "public-edge/agree-a-deferred-payment", "public-edge/agree-a-personalised-housing-plan", "public-edge/agree-an-information-sharing-agreement", "public-edge/agree-and-seal-a-planning-obligation", "public-edge/agree-the-council-plan-and-its-outcomes", "public-edge/agree-the-officer-and-member-protocol", "public-edge/agree-the-political-group-position-before-a-meeting", "public-edge/agree-the-post-election-timetable-and-transition-protocol", "public-edge/allocate-committee-seats-by-political-balance", "public-edge/analog-human-switch-for-irreversible-actions", "public-edge/answer-a-best-value-notice-or-mhclg-engagement-with-one-evid", "public-edge/answer-a-code-of-conduct-complaint", "public-edge/answer-a-corporate-complaint", "public-edge/answer-a-freedom-of-information-request", "public-edge/answer-a-local-land-charges-search", "public-edge/answer-a-media-enquiry-and-issue-a-press-release", "public-edge/answer-a-standards-complaint", "public-edge/answer-a-subject-access-request", "public-edge/answer-a-written-parliamentary-question", "public-edge/answer-an-appeal-to-the-valuation-tribunal", "public-edge/answer-an-environmental-information-request", "public-edge/answer-ministerial-correspondence", "public-edge/answer-questions-at-a-scrutiny-meeting", "public-edge/appeal-a-planning-decision", "public-edge/apply-an-empty-home-or-second-home-premium", "public-edge/apply-an-exemption-and-issue-a-refusal-notice", "public-edge/apply-for-a-blue-badge", "public-edge/apply-for-a-deprivation-of-liberty-authorisation", "public-edge/apply-for-a-discount-or-council-tax-reduction", "public-edge/apply-for-a-licence", "public-edge/apply-for-a-postal-or-proxy-vote", "public-edge/apply-for-a-school-place-and-appeal-a-refusal", "public-edge/apply-for-a-voter-authority-certificate", "public-edge/apply-for-an-export-licence", "public-edge/apply-for-free-school-meals", "public-edge/apply-for-planning-permission-and-track-it", "public-edge/apply-the-armed-forces-covenant-duty-in-a-council-decision", "public-edge/apply-the-monthly-savings-test", "public-edge/apply-the-national-pay-award", "public-edge/apply-the-one-page-decision-standard-and-report-standard-to", "public-edge/apply-the-pre-election-period", "public-edge/apply-the-retention-schedule-and-dispose-of-records", "public-edge/appoint-an-election-agent", "public-edge/appoint-an-independent-member-to-the-committee", "public-edge/appoint-and-train-polling-staff", "public-edge/appoint-or-dismiss-a-statutory-officer", "public-edge/appoint-the-cabinet-and-allocate-portfolios", "public-edge/appoint-the-external-auditor-through-psaa", "public-edge/appoint-the-internal-auditor-and-receive-the-report", "public-edge/approve-a-business-case-through-the-assurance-framework", "public-edge/approve-a-change-through-the-change-board", "public-edge/approve-a-direct-payment", "public-edge/approve-a-disabled-facilities-grant", "public-edge/approve-a-new-system-against-the-architecture-standard", "public-edge/approve-a-severance-package-over-100-000", "public-edge/approve-a-virement", "public-edge/approve-the-internal-audit-charter", "public-edge/approve-the-internal-audit-plan", "public-edge/approve-the-medium-term-financial-strategy", "public-edge/approve-the-pension-fund-accounts-and-valuation", "public-edge/approve-the-senior-structure-and-statutory-officer-designati", "public-edge/approve-the-treasury-capital-and-minimum-revenue-provision-s", "public-edge/ask-a-question-of-the-leader-or-a-cabinet-member", "public-edge/ask-a-question-or-present-a-petition-at-a-meeting", "public-edge/ask-for-a-care-needs-assessment", "public-edge/ask-for-help-with-homelessness", "public-edge/ask-for-own-personal-data", "public-edge/assess-a-carer-s-needs", "public-edge/assess-a-code-of-conduct-complaint", "public-edge/assess-a-housing-benefit-claim", "public-edge/assess-and-decide-bus-franchising", "public-edge/assess-council-tax-reduction", "public-edge/assess-risk-and-publish-the-community-risk-register", "public-edge/assess-the-committee-s-effectiveness-and-report-to-council", "public-edge/assess-the-needs-of-an-adult", "public-edge/attend-film-or-report-a-public-meeting", "public-edge/attend-the-count-as-an-agent", "public-edge/authenticate-and-seal-a-document", "public-edge/authorise-and-make-a-payment", "public-edge/authorise-covert-surveillance-under-ripa", "public-edge/award-a-discount-or-exemption", "public-edge/award-a-discretionary-housing-payment", "public-edge/award-and-monitor-a-voluntary-sector-grant", "public-edge/award-business-rates-relief", "public-edge/award-under-a-framework-or-dynamic-market", "public-edge/back-up-and-test-a-restore", "public-edge/bid-for-and-report-on-a-transport-grant", "public-edge/bill-a-new-business-rates-property", "public-edge/build-a-resident-service-through-discovery-alpha-and-beta", "public-edge/build-and-sign-the-savings-ledger-line-by-line", "public-edge/calculate-the-council-tax-base-and-requirement", "public-edge/call-in-an-executive-decision", "public-edge/canvass-voters-lawfully", "public-edge/carry-out-a-data-protection-impact-assessment", "public-edge/carry-out-a-financial-assessment-and-charge", "public-edge/carry-out-an-education-health-and-care-needs-assessment", "public-edge/carry-out-an-internal-review", "public-edge/cast-a-recorded-vote-on-the-budget", "public-edge/certify-a-grant-claim", "public-edge/chair-full-council-and-rule-on-order", "public-edge/challenge-a-council-tax-band", "public-edge/challenge-a-provider-rate-and-decide-an-uplift", "public-edge/check-supply-chain-labour-and-modern-slavery-risk", "public-edge/claim-emergency-financial-assistance", "public-edge/claim-members-allowances-and-expenses", "public-edge/claim-staffing-and-business-costs", "public-edge/clerk-a-school-admission-appeal-panel", "public-edge/comment-on-a-planning-application", "public-edge/comment-on-a-planning-application-2", "public-edge/commission-a-placement-for-a-looked-after-child", "public-edge/commission-a-public-health-service-under-the-provider-select", "public-edge/commission-supported-living-from-a-charity-provider", "public-edge/commission-temporary-accommodation", "public-edge/commission-the-external-quality-assessment", "public-edge/complain-to-the-council-and-the-ombudsman", "public-edge/complete-induction-and-training-on-reserves-debt-efs-and-ris", "public-edge/complete-mandatory-member-training", "public-edge/complete-the-annual-governance-and-accountability-return", "public-edge/complete-the-data-security-and-protection-toolkit", "public-edge/comply-with-a-secretary-of-state-direction", "public-edge/conduct-a-marriage-or-civil-partnership", "public-edge/confirm-mrp-and-efs-accounting-with-the-external-auditor", "public-edge/consider-an-objection-from-a-local-elector", "public-edge/consider-social-value-before-procurement", "public-edge/consolidate-published-recommendations-into-owned-operating-s", "public-edge/consult-and-publicise-an-application", "public-edge/consult-on-and-make-redundancies", "public-edge/contract-management", "public-edge/convene-the-right-people-by-mail-and-call", "public-edge/correct-a-register-entry", "public-edge/council-decisions", "public-edge/cyber-posture", "public-edge/deal-with-a-dangerous-structure", "public-edge/decide-a-parking-penalty-challenge", "public-edge/decide-an-application-at-planning-committee", "public-edge/decide-on-sharing-confidential-care-information-as-caldicott", "public-edge/declare-an-impairment-to-independence", "public-edge/declare-an-interest-at-a-meeting-and-withdraw", "public-edge/declare-an-interest-in-proceedings", "public-edge/declare-the-collection-fund-surplus-or-deficit", "public-edge/decommission-a-legacy-system", "public-edge/defend-a-planning-appeal", "public-edge/deliver-the-risk-based-audit-plan-from-substantive-capacity", "public-edge/design-and-run-a-grant-scheme", "public-edge/determine-a-building-regulations-application", "public-edge/determine-a-postal-or-proxy-vote-application", "public-edge/determine-an-application-to-register", "public-edge/discharge-planning-conditions", "public-edge/dispose-of-or-acquire-land-at-best-consideration", "public-edge/employ-staff-in-the-mp-s-office", "public-edge/employment", "public-edge/enforce-joiner-mover-and-leaver-with-segregation-of-duties", "public-edge/enforce-publication-deadlines-with-an-exception-register", "public-edge/enrol-staff-in-the-pension-scheme-and-send-monthly-data", "public-edge/ensure-continuity-of-care-on-provider-failure", "public-edge/evaluate-and-grade-a-job", "public-edge/fill-a-casual-vacancy-by-election-or-co-option", "public-edge/find-a-council-borough-or-government-department-in-the-insti", "public-edge/find-the-right-council-and-service-for-an-address", "public-edge/flow-obligations-into-a-contract-or-commission", "public-edge/give-notice-to-hand-back-a-contract", "public-edge/give-the-opinion-on-the-financial-statements", "public-edge/grant-a-dispensation-for-a-pecuniary-interest", "public-edge/grant-a-premises-licence", "public-edge/grant-a-street-works-permit", "public-edge/handle-a-benefit-reconsideration-or-appeal", "public-edge/handle-a-highway-claim", "public-edge/handle-a-housing-complaint", "public-edge/handle-a-request-for-erasure-or-restriction", "public-edge/hear-a-call-in-and-decide", "public-edge/hear-a-councillor-call-for-action", "public-edge/hold-a-citizenship-ceremony", "public-edge/hold-a-constituency-surgery", "public-edge/hold-a-licensing-sub-committee-hearing", "public-edge/hold-an-initial-child-protection-conference", "public-edge/hold-and-record-the-monthly-golden-triangle-meeting", "public-edge/hold-identity-and-operational-evidence", "public-edge/hold-pre-decision-scrutiny-of-a-key-decision", "public-edge/hold-the-annual-parish-meeting", "public-edge/host-a-social-housing-consumer-standards-inspection", "public-edge/host-an-lga-corporate-peer-challenge", "public-edge/independently-test-an-implemented-control-under-the-self-rev", "public-edge/inspect-a-food-business-and-award-a-hygiene-rating", "public-edge/inspect-a-highway-and-repair-a-defect", "public-edge/inspect-a-private-rented-home-and-serve-a-notice", "public-edge/inspect-a-workplace-for-health-and-safety", "public-edge/inspect-the-council-s-accounts-and-object", "public-edge/inspect-work-and-issue-a-completion-certificate", "public-edge/instruct-enforcement-agents-with-a-vulnerability-check", "public-edge/investigate-a-breach-and-serve-an-enforcement-notice", "public-edge/investigate-a-consumer-complaint-and-take-enforcement-action", "public-edge/investigate-a-flood-as-lead-local-flood-authority", "public-edge/investigate-a-statutory-nuisance-and-serve-an-abatement-noti", "public-edge/investigate-a-suspected-fraud", "public-edge/investigate-tenancy-fraud", "public-edge/issue-a-best-value-notice-and-run-engagement", "public-edge/issue-a-certified-copy-of-an-entry", "public-edge/issue-a-cil-liability-notice-and-demand", "public-edge/issue-a-mayoral-decision", "public-edge/issue-a-refund-or-transfer-a-credit", "public-edge/issue-a-section-114-report", "public-edge/issue-a-section-5-report-on-unlawfulness-or-maladministratio", "public-edge/issue-a-security-aspects-letter-to-a-contractor", "public-edge/issue-a-statutory-recommendation-or-public-interest-report", "public-edge/issue-a-voter-authority-certificate", "public-edge/issue-annual-council-tax-bills", "public-edge/issue-the-annual-internal-audit-opinion", "public-edge/issue-the-audit-plan", "public-edge/issue-the-decision-notice", "public-edge/issue-the-role-statement-train-and-attribute-challenge", "public-edge/issue-the-summons-and-agenda", "public-edge/issue-the-weekly-one-page-exception-report", "public-edge/join-a-council-s-approved-provider-list", "public-edge/join-a-strategic-coordinating-group", "public-edge/join-the-housing-register-and-assess-priority", "public-edge/keep-business-continuity-plans-and-advise-businesses", "public-edge/keep-spending-within-the-limit-and-file-the-return", "public-edge/keep-statutory-notification-readiness-and-named-incident-rol", "public-edge/keep-the-device-and-software-asset-register", "public-edge/keep-the-outside-bodies-appointments-register", "public-edge/keep-the-record-of-processing-activities", "public-edge/lay-a-statutory-instrument", "public-edge/lay-the-annual-report-and-accounts", "public-edge/lead-recovery", "public-edge/lead-the-response-to-a-section-114-notice", "public-edge/let-a-social-home", "public-edge/licence-a-taxi-or-private-hire-driver", "public-edge/licence-an-animal-activity", "public-edge/license-a-house-in-multiple-occupation", "public-edge/logistics-tracking", "public-edge/maintain-and-publish-the-register-of-members-interests", "public-edge/maintain-the-brownfield-and-self-build-registers", "public-edge/maintain-the-constitution-and-scheme-of-delegation-with-live", "public-edge/maintain-the-contract-procedure-rules", "public-edge/maintain-the-publication-scheme-and-disclosure-log", "public-edge/maintain-the-single-procurement-and-commissioning-pipeline", "public-edge/make-a-freedom-of-information-request-or-ask-for-review", "public-edge/make-a-public-spaces-protection-order", "public-edge/make-a-safeguarding-adults-enquiry", "public-edge/make-a-traffic-regulation-order", "public-edge/make-a-tree-preservation-order", "public-edge/manage-a-contract-and-hold-review-meetings", "public-edge/manage-a-persistent-or-unreasonable-complainant", "public-edge/manage-a-public-question-or-petition-at-a-meeting", "public-edge/manage-allotments", "public-edge/manage-cash-flow-and-borrowing", "public-edge/manage-sickness-absence", "public-edge/mark-and-check-politically-restricted-posts", "public-edge/mark-exempt-information-and-run-a-private-item", "public-edge/mark-store-and-send-classified-information", "public-edge/meet-the-auditors-without-officers-present", "public-edge/modify-a-contract-and-publish-the-change-notice", "public-edge/monitor-a-grant-and-recover-misused-funds", "public-edge/monitor-a-provider-s-quality", "public-edge/name-and-number-a-street", "public-edge/nominate-an-asset-of-community-value", "public-edge/nominate-an-asset-of-community-value-2", "public-edge/notify-the-regulator-of-an-incident", "public-edge/obtain-facility-security-clearance-for-a-supplier-site", "public-edge/open-a-constituent-case-with-consent", "public-edge/open-a-council-tax-account-on-a-move-in", "public-edge/open-a-rest-centre-and-register-evacuees", "public-edge/open-and-conclude-an-inquest", "public-edge/open-postal-votes-and-verify-personal-identifiers", "public-edge/open-the-government-station-and-choose-a-journey", "public-edge/organise-evidence-and-records-for-a-public-body", "public-edge/pass-a-proposal-through-the-business-case-gateway", "public-edge/pay-council-tax-and-set-up-a-direct-debit", "public-edge/pay-or-challenge-a-parking-penalty", "public-edge/pay-the-data-protection-fee", "public-edge/place-a-household-in-interim-accommodation", "public-edge/pool-a-budget-with-the-nhs", "public-edge/prepare-a-care-proceedings-application", "public-edge/prepare-a-green-book-business-case", "public-edge/prepare-a-neighbourhood-plan", "public-edge/prepare-a-procurement-and-agreement-requirements-brief-for-a", "public-edge/prepare-an-impact-assessment", "public-edge/prepare-an-off-site-plan-for-a-major-hazard-site", "public-edge/prepare-and-examine-the-local-plan", "public-edge/prepare-for-a-cqc-assessment-of-adult-social-care", "public-edge/prepare-for-a-public-accounts-committee-hearing", "public-edge/prepare-for-an-inspection-of-children-s-services", "public-edge/prepare-main-and-supplementary-estimates", "public-edge/prepare-the-local-transport-plan", "public-edge/prepare-the-market-sustainability-plan", "public-edge/price-a-single-source-defence-contract", "public-edge/process-a-right-to-buy-application", "public-edge/process-a-temporary-event-notice", "public-edge/procure-a-separately-procurable-engagement-with-an-aggregati", "public-edge/procurement-intelligence", "public-edge/propose-the-budget-and-council-tax-to-full-council", "public-edge/publish-an-accessibility-statement", "public-edge/publish-an-algorithmic-transparency-record", "public-edge/publish-an-executive-decision-record", "public-edge/publish-and-maintain-privacy-notices", "public-edge/publish-government-procurement-card-spend", "public-edge/publish-members-allowances-for-the-year", "public-edge/publish-notice-of-a-key-decision-28-clear-days-ahead", "public-edge/publish-open-data", "public-edge/publish-spending-over-500", "public-edge/publish-the-agenda-and-reports-five-clear-days-ahead", "public-edge/publish-the-annual-pay-policy-statement", "public-edge/publish-the-contracts-register", "public-edge/publish-the-draft-and-audited-statement-of-accounts", "public-edge/publish-the-forward-plan-twelve-weeks-ahead", "public-edge/publish-the-gambling-statement-of-principles", "public-edge/publish-the-gender-pay-gap", "public-edge/publish-the-market-position-statement", "public-edge/publish-the-notice-of-election", "public-edge/publish-the-regulatory-enforcement-policy", "public-edge/publish-the-revised-register", "public-edge/publish-the-updated-constitution", "public-edge/publish-transparency-data", "public-edge/put-a-digital-imprint-on-campaign-material", "public-edge/raise-a-purchase-order-before-package-spend", "public-edge/raise-a-safeguarding-concern", "public-edge/raise-a-service-complaint", "public-edge/read-a-committee-pack-within-the-page-limit-and-lodge-questi", "public-edge/rebase-the-adult-social-care-budget-from-the-care-ledger", "public-edge/receive-and-decide-on-nominations", "public-edge/receive-and-handle-a-whistleblowing-disclosure", "public-edge/receive-and-publish-candidates-spending-returns", "public-edge/receive-quarterly-monitoring-with-a-demand-bridge", "public-edge/receive-the-auditor-s-annual-report-and-value-for-money-comm", "public-edge/receive-the-counter-fraud-annual-report", "public-edge/receive-the-monthly-evidence-pack-rated-on-evidence", "public-edge/reconcile-access-and-contain-unexplained-accounts", "public-edge/reconcile-the-balance-sheet-exposure-statement-quarterly", "public-edge/record-a-gift-or-hospitality", "public-edge/record-a-price-exception-with-the-finance-position", "public-edge/record-an-exemption-from-contract-procedure-rules", "public-edge/recover-a-benefit-overpayment", "public-edge/recover-unpaid-council-tax-through-a-liability-order", "public-edge/recruit-safely-with-dbs-checks", "public-edge/refer-a-band-challenge-to-the-valuation-office-agency", "public-edge/refer-higher-risk-building-work-to-the-building-safety-regul", "public-edge/register-a-birth", "public-edge/register-a-birth-death-or-marriage-notice", "public-edge/register-a-death-and-offer-tell-us-once", "public-edge/register-a-financial-interest", "public-edge/register-a-foreign-activity-arrangement", "public-edge/register-a-political-party", "public-edge/register-as-a-consultant-lobbyist-and-file-returns", "public-edge/register-as-a-non-party-campaigner", "public-edge/register-disclosable-pecuniary-interests", "public-edge/register-to-vote", "public-edge/reimburse-operators-for-concessionary-travel", "public-edge/renew-public-services-network-compliance", "public-edge/reperform-a-closed-control-from-source", "public-edge/report-a-defence-security-incident", "public-edge/report-a-street-problem", "public-edge/report-complaints-and-compliments-annually", "public-edge/report-contract-performance-indicators", "public-edge/report-digital-service-performance", "public-edge/report-donations-and-loans", "public-edge/report-limited-or-no-assurance-to-the-audit-committee", "public-edge/report-on-a-council-owned-company-to-the-shareholder-committ", "public-edge/report-on-staffing-and-organisation-as-head-of-paid-service", "public-edge/report-on-the-robustness-of-estimates-and-adequacy-of-reserv", "public-edge/report-on-value-for-money-arrangements", "public-edge/represent-the-council-on-the-combined-authority-or-a-regiona", "public-edge/request-a-rate-uplift", "public-edge/request-a-review-of-a-housing-decision", "public-edge/request-and-track-audit-evidence", "public-edge/request-exceptional-financial-support", "public-edge/require-a-cabinet-member-or-officer-to-attend", "public-edge/resolve-a-service-desk-incident", "public-edge/resolve-eligibility-for-the-general-power-of-competence", "public-edge/respond-to-a-child-protection-referral", "public-edge/respond-to-a-damp-mould-or-repair-report", "public-edge/respond-to-a-judicial-review-pre-action-letter", "public-edge/respond-to-a-parliamentary-petition", "public-edge/respond-to-a-statutory-recommendation-from-the-auditor", "public-edge/respond-to-an-area-send-inspection", "public-edge/respond-to-an-ico-decision-notice", "public-edge/respond-to-an-ombudsman-enquiry-and-carry-out-the-remedy", "public-edge/respond-to-an-ombudsman-finding", "public-edge/return-post-election-data-to-the-electoral-commission", "public-edge/review-a-homelessness-decision", "public-edge/review-a-looked-after-child-s-care-plan", "public-edge/review-local-air-quality", "public-edge/review-polling-districts-and-places", "public-edge/review-the-annual-governance-statement", "public-edge/review-the-corporate-risk-register-quarterly", "public-edge/review-the-draft-statement-of-accounts", "public-edge/review-the-information-risk-register-as-senior-information-r", "public-edge/review-the-risk-management-framework", "public-edge/run-a-combined-authority-board-meeting", "public-edge/run-a-deep-dive-with-recorded-influence-on-direction", "public-edge/run-a-demand-control-cell-weekly", "public-edge/run-a-disaster-recovery-test", "public-edge/run-a-disciplinary-process", "public-edge/run-a-neighbourhood-planning-referendum", "public-edge/run-a-public-consultation-and-publish-the-response", "public-edge/run-a-public-consultation-and-report-the-results", "public-edge/run-a-recall-petition-as-petition-officer", "public-edge/run-a-standards-hearing-and-publish-the-decision", "public-edge/run-a-task-and-finish-review-and-secure-the-executive-respon", "public-edge/run-an-assurance-review-on-a-major-project", "public-edge/run-an-audit-engagement-from-terms-of-reference-to-report", "public-edge/run-monthly-budget-monitoring-with-budget-holders", "public-edge/run-monthly-payroll-and-report-to-hmrc", "public-edge/run-national-fraud-initiative-matches", "public-edge/run-security-awareness-and-phishing-tests", "public-edge/run-the-annual-canvass", "public-edge/run-the-annual-council-meeting", "public-edge/run-the-cross-committee-forum-with-other-chairs", "public-edge/run-the-decision-clock-on-stalled-reforms", "public-edge/run-the-make-or-buy-test-before-a-material-re-let", "public-edge/run-the-out-of-hours-duty-officer-rota", "public-edge/run-the-period-for-exercise-of-public-rights", "public-edge/run-the-public-inspection-period-for-the-accounts", "public-edge/run-the-reorganisation-programme-to-vesting-day", "public-edge/run-the-verification-and-count-and-declare-the-result", "public-edge/score-the-thirty-cyber-criteria-with-evidence-references", "public-edge/scrutinise-a-substantial-nhs-service-change", "public-edge/scrutinise-the-budget-proposals-before-council", "public-edge/scrutinise-the-combined-authority", "public-edge/scrutinise-the-community-safety-partnership", "public-edge/scrutinise-treasury-management", "public-edge/security-incident-response", "public-edge/seek-a-cabinet-office-spend-control-approval", "public-edge/seek-hm-treasury-approval-above-delegated-limits", "public-edge/select-a-candidate-and-authorise-the-party-description", "public-edge/send-apologies-and-keep-attendance-within-the-six-month-rule", "public-edge/separate-cashable-saving-cost-avoidance-and-deferred-demand", "public-edge/set-and-collect-council-housing-rents", "public-edge/set-the-annual-calendar-of-meetings", "public-edge/set-the-annual-work-programme", "public-edge/set-the-local-government-finance-settlement", "public-edge/set-the-mayoral-budget-and-precept", "public-edge/set-the-precept-and-notify-the-billing-authority", "public-edge/set-the-scheme-of-executive-delegation", "public-edge/set-up-an-instalment-plan-or-direct-debit", "public-edge/share-information-with-other-responders", "public-edge/sign-the-annual-governance-statement", "public-edge/sign-the-declaration-of-acceptance-of-office", "public-edge/speak-in-a-debate-and-check-the-official-report", "public-edge/spend-under-the-section-137-power-within-the-limit", "public-edge/submit-a-motion-to-full-council", "public-edge/submit-an-innovation-proposal-to-the-defence-and-security-ac", "public-edge/submit-planning-statistics-returns", "public-edge/submit-revenue-and-capital-returns-to-mhclg", "public-edge/submit-single-source-contract-reports", "public-edge/submit-the-business-rates-forecast-and-outturn-returns", "public-edge/submit-the-homelessness-statistics-return", "public-edge/submit-the-party-s-annual-accounts", "public-edge/submit-the-provider-cost-return", "public-edge/submit-the-tenant-satisfaction-measures", "public-edge/supplier-vetting", "public-edge/supply-the-register-to-an-entitled-person", "public-edge/support-a-care-leaver-with-a-pathway-plan", "public-edge/table-a-motion-or-seek-a-debate", "public-edge/table-a-written-question", "public-edge/take-a-bill-through-parliament", "public-edge/take-a-financial-monitoring-item-with-decision-grade-evidenc", "public-edge/take-a-homelessness-application-and-decide-the-duty", "public-edge/take-a-key-decision-on-the-one-page-standard", "public-edge/take-a-ward-casework-enquiry-through-to-reply", "public-edge/take-an-individual-portfolio-decision", "public-edge/take-an-urgent-decision-under-special-urgency", "public-edge/take-and-publish-the-minutes", "public-edge/take-and-publish-the-minutes-2", "public-edge/take-evidence-and-agree-a-select-committee-report", "public-edge/take-notice-of-marriage-or-civil-partnership", "public-edge/take-the-six-month-progress-report-on-statutory-recommendati", "public-edge/tender-and-award", "public-edge/track-audit-recommendations-to-owner-and-date-and-escalate-o", "public-edge/track-the-executive-s-responses-to-recommendations", "public-edge/transfer-staff-under-tupe", "public-edge/transfer-the-commissioning-toolkit-into-a-service", "public-edge/validate-a-planning-application", "public-edge/verify-weighing-and-measuring-equipment", "public-edge/vet-and-train-care-staff", "public-edge/view-the-public-outcome-scorecard", "public-edge/vote-in-a-division", "public-edge/warn-and-inform-the-public", "public-edge/webcast-and-allow-reporting-of-a-public-meeting", "public-edge/write-a-submission-to-a-minister", "public-edge/write-and-exercise-the-emergency-plan", "public-edge/write-off-an-irrecoverable-debt", "public-edge/write-the-officer-report-and-decide-under-delegation", "public-edge/write-to-a-public-body-on-a-constituent-s-behalf" ], "decided_by": "X-X.TECH (Owner order; recorded by Claude-Uriel 21:3531102026-1)", "date": "2026-11-01" } ``` ### spec/changes/2026-11-01-typed-steps-honest-outcomes-reference-flows.json `spec/changes/2026-11-01-typed-steps-honest-outcomes-reference-flows.json` · sha256 `3e7fe9759ddde7ccdebaa8bc3110465762cacb1357a6f5685c79a0c7e6511c26` ```json { "id": "2026-11-01-typed-steps-honest-outcomes-reference-flows", "request": "Owner, 1 November 2026 03:4xZ to 04:14Z: work backwards from the customer's outcome; every journey step asks a typed real question (no generic text boxes), outcomes are honest (no halving), and every capability carries its reference flow (Mobbin, Refero) with the layout copied from the reference and only the estate's controls added. Regenerated by lane Claude-Raphael 03:5801112026-1 (tools/typed_steps.py, outcomes.py, reference_flows.py), commits a7377e84 and c2a37f61c; nothing in scope added or removed, every changed brief is one of the frozen briefs retyped.", "admits": [ "clinic/accept-a-proxy-mandate", "clinic/acquire-and-send-images", "clinic/act-on-a-drug-safety-alert", "clinic/act-on-a-high-risk-answer", "clinic/add-an-appointment-to-my-calendar", "clinic/adjust-a-premium-at-renewal", "clinic/administer-a-medicine", "clinic/allocate-a-breach-between-trusts", "clinic/annotate-variants", "clinic/answer-a-patient-call", "clinic/apply-a-national-data-opt-out", "clinic/apply-a-treatment-rule", "clinic/apply-for-hra-and-rec-approval", "clinic/approve-a-data-release", "clinic/approve-a-leave-or-swap-request", "clinic/approve-a-questionnaire-or-test-threshold", "clinic/approve-a-staff-request", "clinic/arrange-a-cross-trust-mdt", "clinic/ask-for-advice-and-guidance", "clinic/ask-for-help-now", "clinic/assess-a-safeguarding-concern", "clinic/attend-an-appointment-with-the-person", "clinic/audit-access-to-records", "clinic/audit-antimicrobial-use", "clinic/audit-consent-records", "clinic/audit-data-use-against-mandates", "clinic/audit-mdt-attendance", "clinic/audit-reporting-quality", "clinic/audit-response-times", "clinic/audit-waiting-time-data", "clinic/authorise-a-result", "clinic/book-a-pathway-test", "clinic/book-an-appointment", "clinic/book-an-interpreter", "clinic/book-for-the-person-i-care-for", "clinic/book-from-the-referral-queue", "clinic/book-transport", "clinic/build-and-publish-the-rota", "clinic/build-the-clinic-template", "clinic/build-the-mdt-list", "clinic/build-the-variant-to-trial-graph", "clinic/cancel-a-clinic-session", "clinic/change-my-pharmacy", "clinic/chase-a-pathway-clock", "clinic/chase-missing-results", "clinic/check-a-chemotherapy-prescription", "clinic/check-access-controls", "clinic/check-fragment-size-before-sequencing", "clinic/check-identity-and-consent-before-scanning", "clinic/check-in-an-arriving-patient", "clinic/check-in-on-arrival", "clinic/check-laboratory-submissions-for-a-variant", "clinic/check-record-accuracy", "clinic/check-research-integrity", "clinic/check-staff-registration-and-training", "clinic/claim-my-edge-id-and-record", "clinic/classify-a-variant", "clinic/classify-software-as-a-medical-device", "clinic/clinically-check-a-prescription", "clinic/close-a-referral", "clinic/close-a-study", "clinic/collect-case-report-data", "clinic/compare-two-pipeline-runs", "clinic/complete-a-mental-health-check-in", "clinic/complete-a-pre-clinic-questionnaire", "clinic/complete-a-risk-formulation", "clinic/complete-mandatory-training", "clinic/complete-the-data-security-and-protection-toolkit", "clinic/confirm-attendance-and-prepare", "clinic/connect-a-lims", "clinic/connect-a-pacs-or-imaging-system", "clinic/counsel-a-patient-on-a-new-medicine", "clinic/decide-a-re-run", "clinic/deliver-cancer-nurse-specialist-support", "clinic/deploy-a-release", "clinic/design-a-smart-questionnaire", "clinic/design-a-trial", "clinic/detect-claim-anomalies", "clinic/discharge-from-the-pathway", "clinic/discharge-from-the-stream", "clinic/dispense-a-medicine", "clinic/escalate-to-emergency-services", "clinic/explain-a-result-in-plain-language", "clinic/explain-a-result-to-the-patient", "clinic/find-a-recruiting-trial-near-me", "clinic/flag-a-critical-finding", "clinic/flag-never-block", "clinic/follow-up-after-discharge-from-crisis-care", "clinic/follow-up-an-action-plan", "clinic/forecast-compliance", "clinic/forecast-days-to-treatment", "clinic/fulfil-a-subject-access-request", "clinic/give-a-vaccination", "clinic/give-consent-on-behalf", "clinic/give-or-withdraw-consent", "clinic/grade-a-retinal-image", "clinic/hand-off-to-codex-or-jude", "clinic/handle-a-complaint", "clinic/handle-a-controlled-drug", "clinic/handle-a-did-not-attend", "clinic/handle-a-disclosure-of-abuse", "clinic/handle-a-freedom-of-information-request", "clinic/handle-a-policyholder-complaint", "clinic/handle-a-stock-shortage", "clinic/handle-the-record-after-death", "clinic/hold-a-remote-session", "clinic/hold-a-slot-for-a-tertiary-handover", "clinic/hold-the-duty-rota", "clinic/import-my-practice-register", "clinic/inspect-the-hazard-log-and-safety-case", "clinic/investigate-a-clinical-safety-incident", "clinic/issue-a-clinical-safety-case-report", "clinic/issue-a-field-safety-notice", "clinic/issue-a-genomic-report", "clinic/issue-a-policy", "clinic/maintain-laboratory-accreditation-evidence", "clinic/maintain-the-hazard-log", "clinic/maintain-the-suicide-prevention-plan", "clinic/make-a-follow-up-call", "clinic/make-an-urgent-same-day-referral", "clinic/manage-an-incident", "clinic/manage-freezer-alarms", "clinic/manage-patient-group-directions", "clinic/manage-patient-transport-and-interpreter-contracts", "clinic/manage-retention-and-destruction", "clinic/manage-room-allocation", "clinic/manage-the-information-asset-register", "clinic/manage-the-outbound-allow-list", "clinic/manage-the-reporting-backlog", "clinic/manage-the-room-and-equipment-register", "clinic/manage-the-waiting-list", "clinic/measure-a-nodule-and-its-growth", "clinic/measure-compute-energy-and-carbon", "clinic/measure-outcomes-over-time", "clinic/message-my-care-team", "clinic/monitor-alerts-and-escalate", "clinic/monitor-data-feeds", "clinic/monitor-post-deployment-safety", "clinic/nominate-a-carer-or-proxy", "clinic/obtain-confirmation-of-capacity-and-capability", "clinic/offer-a-cancellation-slot", "clinic/offer-self-help-resources", "clinic/onboard-a-staff-member", "clinic/onboard-a-wearable-device-type", "clinic/order-a-diagnostic-test", "clinic/order-a-direct-access-test", "clinic/order-a-genomic-test", "clinic/order-a-repeat-medicine", "clinic/order-repeat-medicines-on-behalf", "clinic/pair-my-wearable-device", "clinic/phone-a-critical-result", "clinic/plan-weekly-capacity", "clinic/post-a-receipt-to-the-agent-channel", "clinic/predict-fields-in-a-form", "clinic/prepare-a-clinic-letter-for-a-clinician", "clinic/prepare-a-patient-for-imaging", "clinic/prepare-for-a-cqc-inspection", "clinic/prescribe", "clinic/prescribe-under-the-pathway-protocol", "clinic/price-a-policy-from-clinic-data", "clinic/problem-annual-offshore-statistics-and-regulatory-activity", "clinic/problem-medicare-will-pay-50k-for-your-heart-attack-it-sho", "clinic/problem-students-tackle-public-policy-challenges-with-supp", "clinic/problem-working-for-the-national-cyber-force-ncf-case-stud", "clinic/process-a-claim", "clinic/pseudonymise-a-dataset", "clinic/publish-results", "clinic/raise-a-complaint-or-feedback", "clinic/raise-a-safeguarding-concern-from-a-response", "clinic/randomise-a-participant", "clinic/read-assistive-output-on-an-image", "clinic/rebook-or-cancel-an-appointment", "clinic/receive-a-crisis-alert", "clinic/receive-a-crisis-alert-for-the-person", "clinic/receive-a-critical-result-for-my-patient", "clinic/receive-a-did-not-attend-notice", "clinic/receive-a-pre-authorisation-request", "clinic/receive-a-statutory-notification", "clinic/receive-and-book-in-a-sample", "clinic/receive-the-outcome-letter", "clinic/reclassify-a-variant", "clinic/reconcile-medicines", "clinic/reconcile-medicines-at-admission", "clinic/record-a-clinic-consultation", "clinic/record-a-dose-incident", "clinic/record-a-patient-deferred", "clinic/record-a-patient-s-own-view", "clinic/record-a-point-of-care-urine-result", "clinic/record-a-safeguarding-concern", "clinic/record-a-self-check-in-exception", "clinic/record-a-wound-or-skin-check", "clinic/record-an-adverse-drug-reaction", "clinic/record-attendance-outcome", "clinic/record-capacity-and-best-interests", "clinic/record-decision-to-treat", "clinic/record-sample-consent-version", "clinic/record-the-faster-diagnosis-outcome", "clinic/record-the-lawful-basis", "clinic/record-the-mdt-recommendation", "clinic/record-trial-eligibility-at-the-mdt", "clinic/refer-on-a-suspected-cancer-pathway", "clinic/refer-to-a-crisis-team", "clinic/refer-to-tertiary-care", "clinic/refer-to-the-mental-health-stream", "clinic/refer-to-the-multidisciplinary-team", "clinic/refresh-public-data-caches", "clinic/register-a-child-as-a-patient", "clinic/register-a-device-fault", "clinic/register-a-walk-in-patient", "clinic/register-an-imaging-device", "clinic/register-my-practice-with-the-clinic", "clinic/register-the-trial", "clinic/reject-an-unlabelled-sample", "clinic/report-a-change-in-the-person-s-condition", "clinic/report-a-data-breach", "clinic/report-a-dispensing-error", "clinic/report-a-safety-incident", "clinic/report-a-side-effect", "clinic/report-a-study", "clinic/report-an-adverse-event", "clinic/report-an-incident", "clinic/report-histology", "clinic/report-monthly-performance", "clinic/report-to-the-regulator", "clinic/report-usage-and-efficiency", "clinic/request-a-copy-of-my-record", "clinic/request-a-rota-swap-or-leave", "clinic/request-a-second-opinion", "clinic/request-a-second-read", "clinic/request-a-second-read-or-discrepancy-review", "clinic/request-a-young-person-s-own-access", "clinic/request-clinic-data-for-a-quote", "clinic/request-evidence-against-fundamental-standards", "clinic/review-a-breach-notification", "clinic/review-a-death-by-suicide", "clinic/review-a-real-result", "clinic/review-a-release-for-clinical-safety", "clinic/review-a-repeat-request", "clinic/review-a-trial-for-compliance", "clinic/review-a-wearable-alert", "clinic/review-assistive-output-labelling", "clinic/review-payslips-and-expenses-approvals", "clinic/review-the-caseload", "clinic/review-the-dashboard", "clinic/review-the-waiting-list-position", "clinic/review-the-worklist", "clinic/revoke-access-when-a-mandate-ends", "clinic/revoke-my-own-proxy-role", "clinic/reward-early-detection-participation", "clinic/roll-back-a-release", "clinic/rotate-keys-and-certificates", "clinic/run-a-business-continuity-drill", "clinic/run-a-catalogue-workload", "clinic/run-a-clinical-safety-training-check", "clinic/run-a-data-protection-impact-assessment", "clinic/run-a-pipeline", "clinic/run-a-pre-assessment", "clinic/run-a-safeguarding-review", "clinic/run-a-synthetic-pathway-drill", "clinic/run-a-urine-test-likelihood-ledger", "clinic/run-an-interim-analysis", "clinic/run-an-online-early-symptom-test", "clinic/run-quality-control", "clinic/run-scanner-quality-assurance", "clinic/run-sequencing-qc", "clinic/run-the-day-s-capacity-check", "clinic/run-the-mdt-meeting", "clinic/run-the-training-controller", "clinic/run-vulnerability-scans", "clinic/schedule-the-scan", "clinic/score-a-model-on-a-held-split", "clinic/score-and-interpret-a-response", "clinic/screen-a-patient-for-eligibility", "clinic/search-recruiting-trials", "clinic/see-my-day", "clinic/send-a-check-in-to-a-caseload", "clinic/send-a-results-ready-notice", "clinic/send-appointment-letters-and-reminders", "clinic/serve-a-vision-model", "clinic/set-a-surveillance-plan", "clinic/set-my-communication-preferences", "clinic/set-my-data-sharing-choices", "clinic/set-role-permissions-and-expiry", "clinic/settle-with-the-clinic", "clinic/share-a-medicines-list-with-the-clinic", "clinic/share-a-wearable-signal", "clinic/share-clinic-data-with-my-insurer", "clinic/share-images-with-another-trust", "clinic/share-risk-information-with-partners", "clinic/sign-a-clinic-letter", "clinic/sign-the-clinical-risk-management-plan", "clinic/simulate-trial-monitoring", "clinic/stop-on-revoked-authority", "clinic/store-and-retrieve-genomic-data", "clinic/submit-a-cancer-registration", "clinic/submit-cancer-waiting-times-data", "clinic/submit-cosd-and-other-datasets", "clinic/submit-mdt-data", "clinic/submit-the-mental-health-dataset", "clinic/supervise-and-sign-off-a-trainee", "clinic/support-deploying-organisations", "clinic/take-a-blood-or-urine-sample", "clinic/take-a-mandate-ticket", "clinic/take-an-online-early-symptom-test", "clinic/take-informed-consent-for-research", "clinic/take-observations", "clinic/take-part-in-external-quality-assessment", "clinic/tell-the-patient-a-result", "clinic/test-a-hypothesis-on-an-open-cohort", "clinic/test-a-pre-registered-hypothesis", "clinic/trace-ingress-and-egress", "clinic/track-a-referral", "clinic/track-actions-from-the-mdt", "clinic/track-histology-turnaround", "clinic/track-my-cancer-pathway", "clinic/track-sample-custody", "clinic/track-the-62-day-clock-across-the-mdt", "clinic/train-staff-in-information-governance", "clinic/train-staff-in-suicide-awareness", "clinic/triage-a-mental-health-referral", "clinic/triage-a-patient-message", "clinic/triage-a-referral", "clinic/triage-inbox-items-for-a-clinician", "clinic/tune-alert-thresholds", "clinic/update-patient-demographics", "clinic/validate-a-new-assay", "clinic/validate-the-test-against-outcomes", "clinic/verify-a-device-classification", "clinic/verify-an-applicant-s-identity", "clinic/verify-an-evidence-statement", "clinic/verify-backups-and-restore", "clinic/vet-an-imaging-request", "clinic/view-my-medicines-list", "clinic/view-my-results", "clinic/view-results-in-scope", "clinic/write-a-care-plan", "edge-consortium/accept-a-review-assignment", "edge-consortium/accept-supervision-of-a-learner", "edge-consortium/acknowledge-the-principles-version", "edge-consortium/add-jude-to-a-channel-under-mandate", "edge-consortium/add-members-and-guests-with-expiry", "edge-consortium/administer-a-database", "edge-consortium/admit-or-refuse-a-guest", "edge-consortium/agree-a-data-sharing-agreement", "edge-consortium/allocate-indirect-costs", "edge-consortium/answer-a-reference-request", "edge-consortium/answer-reviewer-questions-on-a-contribution", "edge-consortium/answer-with-sources-in-a-channel", "edge-consortium/apply-a-non-disclosure-agreement-to-a-room", "edge-consortium/apply-an-open-licence", "edge-consortium/apply-for-ethics-approval", "edge-consortium/apply-retention-and-deletion-to-records", "edge-consortium/apply-to-become-a-verified-reviewer", "edge-consortium/approve-a-compute-request", "edge-consortium/approve-a-connector-or-integration", "edge-consortium/approve-a-contribution-before-submission", "edge-consortium/approve-a-director-passkey-action", "edge-consortium/approve-a-federated-query-request", "edge-consortium/approve-a-guest-account", "edge-consortium/approve-a-reviewer-payment-batch", "edge-consortium/approve-a-team-member-data-access-request", "edge-consortium/approve-a-visiting-researcher-affiliation", "edge-consortium/approve-an-agent-consequential-action", "edge-consortium/approve-an-open-source-release", "edge-consortium/approve-and-mark-a-contribution-verified", "edge-consortium/approve-compute-spend-against-a-grant", "edge-consortium/approve-expenditure-against-a-grant", "edge-consortium/archive-a-channel", "edge-consortium/archive-a-module-version", "edge-consortium/ask-a-clarification-question", "edge-consortium/ask-jude-for-a-sourced-answer", "edge-consortium/assess-ownership-of-an-output", "edge-consortium/assign-a-persistent-identifier", "edge-consortium/attribute-carbon-to-a-project", "edge-consortium/attribute-cost-to-a-grant-and-project", "edge-consortium/audit-reviewer-quality-by-re-review", "edge-consortium/audit-the-signing-log", "edge-consortium/award-and-publish-the-contract-award-notice", "edge-consortium/bind-a-device", "edge-consortium/book-a-room-or-resource", "edge-consortium/build-a-project-team-across-universities", "edge-consortium/build-an-interactive-prototype", "edge-consortium/build-the-research-profile-and-skills-record", "edge-consortium/bulk-enrol-students-and-staff", "edge-consortium/catalogue-a-library-item", "edge-consortium/charge-compute-and-storage-to-grant-codes", "edge-consortium/chase-an-overdue-review", "edge-consortium/check-exclusion-grounds", "edge-consortium/choose-a-specialism", "edge-consortium/claim-payment-for-a-completed-review", "edge-consortium/classify-data-sensitivity", "edge-consortium/close-a-grant", "edge-consortium/close-a-project-and-archive-its-records", "edge-consortium/co-edit-a-markdown-in-real-time", "edge-consortium/complete-a-right-to-work-check", "edge-consortium/complete-research-integrity-training", "edge-consortium/compose-an-image-in-the-imaging-studio", "edge-consortium/compose-audio-in-judio", "edge-consortium/conduct-a-probation-review", "edge-consortium/confirm-a-delivery", "edge-consortium/confirm-a-person-s-enrolment-status", "edge-consortium/confirm-a-time-and-send-the-invitation", "edge-consortium/contribute-a-library-entry-as-an-outsider", "edge-consortium/correct-a-student-record", "edge-consortium/cost-a-grant-application", "edge-consortium/count-agents-and-subagents-per-task", "edge-consortium/create-a-contract-from-a-template", "edge-consortium/create-a-job-role-and-vacancy", "edge-consortium/create-a-markdown-document", "edge-consortium/create-a-module", "edge-consortium/create-a-team-and-its-channels", "edge-consortium/create-an-availability-poll", "edge-consortium/curate-a-collection", "edge-consortium/declare-a-conflict-of-interest", "edge-consortium/declare-exclusion-grounds-and-conflicts", "edge-consortium/decline-an-assignment", "edge-consortium/define-a-purpose-view", "edge-consortium/define-or-change-a-schema", "edge-consortium/delete-data-at-retention-end", "edge-consortium/deploy-a-release", "edge-consortium/deposit-a-dataset-with-a-persistent-identifier", "edge-consortium/deprecate-a-feature", "edge-consortium/detect-unlicensed-reuse", "edge-consortium/draw-on-judecanvas", "edge-consortium/edit-a-video", "edge-consortium/end-employment-and-revoke-access", "edge-consortium/enforce-the-code-of-conduct", "edge-consortium/enrol-on-a-doctoral-training-module", "edge-consortium/escalate-a-dispute-to-the-editor", "edge-consortium/evaluate-bids-with-a-scoring-panel", "edge-consortium/expire-or-revoke-a-licence", "edge-consortium/explore-the-knowledge-graph-visually", "edge-consortium/export-a-creation-with-licence-and-attribution", "edge-consortium/export-a-meeting-record-and-audit-chain", "edge-consortium/export-and-verify-an-audit-chain", "edge-consortium/find-a-time-across-attendees", "edge-consortium/find-and-follow-tender-opportunities", "edge-consortium/find-funding-calls", "edge-consortium/flag-a-concern-without-blocking", "edge-consortium/flag-plagiarism-or-an-integrity-concern", "edge-consortium/governed-intelligence-degree", "edge-consortium/grant-a-clearance-with-expiry", "edge-consortium/grant-a-licence-to-a-third-party", "edge-consortium/grant-a-maintainer-role-with-expiry", "edge-consortium/hand-over-to-another-agent", "edge-consortium/hand-over-work-with-the-baton", "edge-consortium/handle-a-contract-dispute", "edge-consortium/handle-a-data-subject-access-request", "edge-consortium/handle-an-academic-appeal-on-marks", "edge-consortium/handle-an-appeal", "edge-consortium/hold-an-audit-interview-by-secure-call", "edge-consortium/import-external-records", "edge-consortium/inspect-carbon-and-energy-records", "edge-consortium/inspect-release-gates-and-evidence", "edge-consortium/inspect-retention-schedule-compliance", "edge-consortium/inspect-reviewer-payment-records", "edge-consortium/international-technical-students", "edge-consortium/issue-a-certificate-of-membership", "edge-consortium/issue-a-takedown-or-attribution-request", "edge-consortium/issue-a-transcript-of-completed-modules", "edge-consortium/issue-an-employment-contract", "edge-consortium/issue-an-offer-letter", "edge-consortium/issue-findings-and-track-remediation", "edge-consortium/issue-the-decision-letter", "edge-consortium/issue-the-written-statement-of-particulars", "edge-consortium/join-a-secure-call-through-the-lobby", "edge-consortium/join-consortium-through-a-university", "edge-consortium/judiverse-included-in-tuition", "edge-consortium/link-orcid-to-the-profile", "edge-consortium/log-a-receipt-for-every-action", "edge-consortium/make-a-below-threshold-direct-award", "edge-consortium/manage-a-controlled-vocabulary", "edge-consortium/manage-a-grant-extension-or-variation", "edge-consortium/manage-an-embargo-and-release-date", "edge-consortium/manage-consent-records", "edge-consortium/manage-institutional-roles-and-delegates", "edge-consortium/map-a-dataset-to-the-knowledge-graph", "edge-consortium/map-a-module-to-a-challenge-specification", "edge-consortium/map-a-module-to-programme-and-standard", "edge-consortium/mark-and-moderate-submissions", "edge-consortium/match-reviewers-to-a-submission", "edge-consortium/measure-live-energy-per-job-and-service", "edge-consortium/merge-duplicate-entities", "edge-consortium/merge-with-required-checks", "edge-consortium/mirror-a-repository-to-the-source-repository", "edge-consortium/moderate-a-post", "edge-consortium/monitor-data-quality", "edge-consortium/monitor-job-progress-and-output", "edge-consortium/monitor-tokens-prompts-and-time-to-evidence", "edge-consortium/negotiate-clauses-with-the-counterparty", "edge-consortium/nominate-a-learner-for-reviewer-training", "edge-consortium/offer-the-degree-with-a-partner-institution", "edge-consortium/onboard-a-new-starter", "edge-consortium/open-an-issue", "edge-consortium/paid-student-work-with-tuition-offset", "edge-consortium/pass-reviewer-credential-verification", "edge-consortium/pause-or-pre-empt-a-job", "edge-consortium/pay-reviewer-fees", "edge-consortium/pay-salary", "edge-consortium/petition-for-student-finance", "edge-consortium/pin-edit-or-delete-a-post", "edge-consortium/plan-a-maintenance-window", "edge-consortium/plan-a-procurement-and-publish-pipeline-notice", "edge-consortium/post-an-announcement", "edge-consortium/pre-register-a-study", "edge-consortium/prepare-a-review-summary-for-a-reviewer", "edge-consortium/prepare-an-employment-contract-template", "edge-consortium/preserve-and-migrate-file-formats", "edge-consortium/problem-brown-student-maps-the-hidden-global-labor-powerin", "edge-consortium/problem-research-agentic-ai-and-consumers", "edge-consortium/process-a-supplier-invoice", "edge-consortium/produce-a-carbon-reduction-plan", "edge-consortium/produce-a-graphic-in-visual-studio", "edge-consortium/produce-a-statutory-data-return", "edge-consortium/propose-a-change-from-a-fork", "edge-consortium/propose-a-real-world-problem-project", "edge-consortium/provide-a-funder-audit-trail", "edge-consortium/provide-a-reference", "edge-consortium/publish-a-data-dictionary", "edge-consortium/publish-a-markdown-as-deck-or-book", "edge-consortium/publish-a-new-dataset-version", "edge-consortium/publish-a-tender-notice", "edge-consortium/publish-a-terms-version-and-require-re-acceptance", "edge-consortium/publish-an-open-access-version", "edge-consortium/publish-documentation", "edge-consortium/publish-module-materials-to-the-library", "edge-consortium/publish-the-catalogue-to-open-indexes", "edge-consortium/publish-the-contract-details-notice", "edge-consortium/publish-the-monthly-carbon-and-cost-statement", "edge-consortium/raise-a-challenge-or-complaint", "edge-consortium/raise-a-grievance", "edge-consortium/raise-a-progress-or-wellbeing-concern", "edge-consortium/rank-agents-and-models-by-efficiency", "edge-consortium/read-a-verified-entry-with-its-provenance", "edge-consortium/real-project-assignments", "edge-consortium/receive-a-mandate", "edge-consortium/receive-bids-in-a-sealed-queue", "edge-consortium/receive-the-award-outcome", "edge-consortium/receive-time-boxed-auditor-access", "edge-consortium/reconcile-monthly-spend", "edge-consortium/reconcile-provider-invoices-with-metered-use", "edge-consortium/reconcile-with-the-student-record-system", "edge-consortium/record-a-call-with-consent", "edge-consortium/record-a-conflict-of-interest-assessment", "edge-consortium/record-a-data-processing-agreement", "edge-consortium/record-a-priority-date", "edge-consortium/record-a-trademark-and-brand-use-rule", "edge-consortium/record-an-invention-disclosure", "edge-consortium/record-applications-and-outcomes-in-a-portfolio", "edge-consortium/record-attendance-and-engagement", "edge-consortium/record-authorship-and-contributor-roles", "edge-consortium/record-background-and-foreground-ip", "edge-consortium/record-copyright-and-attribution", "edge-consortium/record-energy-per-job", "edge-consortium/record-payment-compliance", "edge-consortium/record-provenance-and-lineage", "edge-consortium/record-working-hours-and-leave", "edge-consortium/refuse-an-action-outside-the-mandate", "edge-consortium/register-a-compute-resource", "edge-consortium/register-a-dataset-in-the-catalogue", "edge-consortium/register-a-delegated-authority", "edge-consortium/register-a-research-project-with-an-id", "edge-consortium/register-a-university-as-a-member-institution", "edge-consortium/register-as-a-supplier", "edge-consortium/register-with-the-office-for-students", "edge-consortium/reject-a-contribution-with-reasons", "edge-consortium/release-a-version", "edge-consortium/release-marks-and-feedback", "edge-consortium/release-results-with-disclosure-control", "edge-consortium/renew-annual-institutional-membership", "edge-consortium/renew-reviewer-verification", "edge-consortium/report-a-personal-data-breach", "edge-consortium/report-a-research-misconduct-concern", "edge-consortium/report-carbon-intensity-per-accepted-outcome", "edge-consortium/report-expenditure-to-the-funder", "edge-consortium/report-monitoring-measures", "edge-consortium/report-performance-indicator-evidence", "edge-consortium/report-review-throughput-and-times", "edge-consortium/report-tokens-prompts-and-time-to-evidence", "edge-consortium/report-utilisation", "edge-consortium/request-a-compute-allowance-for-a-project", "edge-consortium/request-a-purpose-view-on-federated-data", "edge-consortium/request-a-sandbox-with-gpu-and-data", "edge-consortium/request-approval-for-a-consequential-action", "edge-consortium/request-changes-from-the-contributor", "edge-consortium/request-supervisor-allocation", "edge-consortium/research-and-development-partnership", "edge-consortium/reserve-capacity-for-a-deadline", "edge-consortium/resolve-conflicting-reviews", "edge-consortium/respond-to-a-security-incident", "edge-consortium/respond-to-an-invitation", "edge-consortium/retract-or-correct-a-verified-entry", "edge-consortium/retry-a-failed-job", "edge-consortium/review-a-contribution-against-criteria", "edge-consortium/review-a-funder-expenditure-statement", "edge-consortium/review-a-pull-request", "edge-consortium/review-access-logs-and-anomalies", "edge-consortium/review-agent-receipts", "edge-consortium/review-module-evaluation-feedback", "edge-consortium/review-quality-and-standards-evidence", "edge-consortium/review-the-dpia-and-record-of-processing", "edge-consortium/revoke-a-purpose-view", "edge-consortium/revoke-access-for-a-leaver", "edge-consortium/rotate-keys", "edge-consortium/run-a-clarification-round", "edge-consortium/run-a-compute-job-under-mandate", "edge-consortium/run-a-cyber-recovery-drill", "edge-consortium/run-a-live-assessment", "edge-consortium/run-a-live-seminar", "edge-consortium/run-a-security-classified-procurement", "edge-consortium/run-a-training-job-on-scheduled-compute", "edge-consortium/run-a-vulnerability-scan-on-a-release", "edge-consortium/run-an-approved-query-where-data-lives", "edge-consortium/run-code-in-the-sandbox", "edge-consortium/run-data-quality-checks-on-a-database", "edge-consortium/sample-verified-entries-for-quality", "edge-consortium/save-sources-to-a-reading-list", "edge-consortium/scale-with-an-external-provider", "edge-consortium/schedule-a-job", "edge-consortium/schedule-a-meeting-on-request", "edge-consortium/schedule-a-secure-government-resource", "edge-consortium/schedule-a-send-and-set-a-signature", "edge-consortium/schedule-and-record-supervision-meetings", "edge-consortium/schedule-deliveries-and-logistics", "edge-consortium/search-across-mail-channels-and-meetings", "edge-consortium/search-libraries-and-the-knowledge-graph", "edge-consortium/send-a-contract-for-signature", "edge-consortium/send-mail-from-an-estate-mailbox", "edge-consortium/set-a-carbon-budget-and-threshold", "edge-consortium/set-a-project-access-control-policy", "edge-consortium/set-agent-authority-scope", "edge-consortium/set-and-publish-key-performance-indicators", "edge-consortium/set-assessment-and-marking-scheme", "edge-consortium/set-contract-parties-permissions-and-expiry", "edge-consortium/set-evaluation-criteria-and-weightings", "edge-consortium/set-mail-rules-flags-and-snooze", "edge-consortium/set-mention-and-tag-rules", "edge-consortium/set-out-of-office-and-working-hours", "edge-consortium/set-permissions-on-project-records", "edge-consortium/set-presence", "edge-consortium/set-project-governance", "edge-consortium/set-quotas-per-project", "edge-consortium/set-renewal-and-expiry-alerts", "edge-consortium/set-retention-and-export-a-channel", "edge-consortium/set-review-criteria-for-a-field", "edge-consortium/set-review-fields-and-availability", "edge-consortium/set-scheduling-policy-and-queues", "edge-consortium/set-the-access-tier-on-an-item", "edge-consortium/set-the-review-deadline-and-reminders", "edge-consortium/set-up-an-awarded-grant", "edge-consortium/share-a-calendar-with-a-delegate", "edge-consortium/share-a-campus-with-a-partner-university", "edge-consortium/share-a-creation-in-the-judebox-exchange", "edge-consortium/share-a-document-in-a-call", "edge-consortium/share-a-workspace-folder-across-universities", "edge-consortium/shift-jobs-to-low-carbon-windows", "edge-consortium/show-undisclosed-figures-as-gaps", "edge-consortium/sign-a-collaboration-agreement", "edge-consortium/sign-a-contract-with-edge-id", "edge-consortium/sign-off-a-progression-milestone", "edge-consortium/sign-the-contributor-agreement", "edge-consortium/sign-the-employment-contract", "edge-consortium/sign-the-public-contract", "edge-consortium/stop-or-kill-a-job", "edge-consortium/store-the-signed-contract-in-the-vault", "edge-consortium/student-to-contractor-pathway", "edge-consortium/submit-a-bid", "edge-consortium/submit-a-contribution-for-review", "edge-consortium/submit-a-grant-application", "edge-consortium/submit-a-paper-to-consortium-review", "edge-consortium/submit-an-annual-report-to-the-funder", "edge-consortium/submit-an-invoice-and-track-payment", "edge-consortium/submit-staff-for-security-vetting", "edge-consortium/suspend-or-end-a-membership", "edge-consortium/terminate-a-contract", "edge-consortium/terminate-a-contract-and-publish-the-notice", "edge-consortium/trace-network-ingress-and-egress", "edge-consortium/track-research-outputs-for-a-ref-return", "edge-consortium/track-royalties-and-revenue-share", "edge-consortium/triage-a-security-vulnerability-report", "edge-consortium/triage-the-submissions-queue", "edge-consortium/update-supplier-information", "edge-consortium/use-the-terminal", "edge-consortium/vary-a-contract", "edge-consortium/verify-a-fact-before-stating-it", "edge-consortium/verify-a-fact-in-the-knowledge-graph", "edge-consortium/verify-a-person-s-identity", "edge-consortium/verify-a-signature", "edge-consortium/verify-institutional-accreditation-status", "edge-consortium/verify-procurement-notice-publication", "edge-consortium/verify-the-accessibility-statement", "edge-consortium/verify-the-audit-chain-offline", "edge-consortium/version-and-roll-back-a-creation", "edge-consortium/vet-a-reviewer-or-supplier-staff-member", "edge-consortium/view-the-diary-by-day-week-month", "edge-consortium/view-the-public-reviewer-track-record", "edge-consortium/vote-on-a-poll", "edge-consortium/write-a-data-management-plan", "edge-consortium/write-a-paper-in-markdowns-with-co-authors", "finance/accept-a-cheque-deposit", "finance/access-a-read-only-evidence-room", "finance/account-for-agent-resource-use", "finance/accrue-deposit-interest-expense", "finance/acknowledge-a-complaint", "finance/act-under-a-power-of-attorney", "finance/action-an-erasure-or-rectification-request", "finance/activate-a-monitoring-rule", "finance/activate-a-proposed-rate", "finance/add-a-joint-account-holder", "finance/add-a-payee-with-name-check", "finance/add-or-remove-a-signatory", "finance/agree-a-forbearance-plan", "finance/amend-a-policy-mid-term", "finance/amend-a-repayment-schedule", "finance/amend-or-cancel-a-standing-order", "finance/analyse-complaint-root-causes", "finance/answer-a-customer-question-from-records", "finance/answer-a-subject-access-request", "finance/apply-a-database-migration", "finance/apply-a-payment-holiday", "finance/apply-enhanced-due-diligence", "finance/apply-for-a-business-account", "finance/apply-for-a-business-loan-or-facility", "finance/apply-for-a-personal-loan", "finance/apply-reasonable-adjustments", "finance/apply-retention-and-deletion", "finance/approve-a-financial-promotion", "finance/approve-a-new-integration-or-api-consumer", "finance/approve-or-decline-an-application", "finance/approve-or-reject-a-payment", "finance/assess-a-claim", "finance/assess-an-outsourcing-or-third-party", "finance/assess-income-and-expenditure", "finance/assess-source-of-funds-and-wealth", "finance/audit-partner-usage", "finance/authorise-card-transactions-in-real-time", "finance/back-up-the-database", "finance/bill-a-partner", "finance/block-a-payment-in-flight", "finance/book-an-appointment", "finance/calculate-and-post-interest-accrual", "finance/calculate-the-liquidity-coverage-ratio", "finance/cancel-a-direct-debit", "finance/cancel-a-policy", "finance/change-contact-details", "finance/claim-a-direct-debit-indemnity-refund", "finance/close-a-business-account", "finance/close-a-period", "finance/close-an-account", "finance/collect-direct-debits-as-an-originator", "finance/collect-premiums", "finance/compare-cover-options", "finance/complete-entity-tax-classification", "finance/complete-tax-residency-self-certification", "finance/compute-expected-credit-losses", "finance/configure-scheduled-jobs", "finance/confirm-balances-as-external-auditor", "finance/contact-a-customer-in-arrears", "finance/continue-a-joint-account-after-a-death", "finance/create-a-monitoring-rule", "finance/create-a-prospect-record", "finance/create-a-savings-product", "finance/create-and-release-a-hold", "finance/decide-a-treasury-proposal", "finance/decide-an-app-reimbursement-claim", "finance/decide-and-file-a-sar", "finance/decline-with-reasons", "finance/deploy-a-release-as-jude", "finance/deploy-a-release-through-deploy-py", "finance/design-a-new-lending-product", "finance/detect-a-missed-repayment", "finance/disburse-a-loan", "finance/disclose-a-vulnerability-or-need", "finance/disposition-an-alert", "finance/dispute-a-card-transaction", "finance/draw-down-from-a-facility", "finance/enforce-strong-customer-authentication", "finance/escalate-a-customer-concern", "finance/execute-a-money-market-placement", "finance/exit-a-customer-for-financial-crime", "finance/explain-a-transaction", "finance/export-the-evidence-bundle", "finance/export-transactions-to-accounting-software", "finance/flag-a-mule-account", "finance/freeze-or-unfreeze-a-card", "finance/fund-settlement-accounts-for-rails", "finance/give-or-revoke-open-banking-consent", "finance/grant-an-operator-role", "finance/hand-a-case-to-a-specialist-team", "finance/hand-work-between-agents", "finance/handle-a-banking-protocol-concern", "finance/handle-a-breathing-space-notice", "finance/handle-a-debt-sale", "finance/handle-a-gambling-block-request", "finance/handle-a-mental-capacity-concern", "finance/handle-a-partner-data-request-or-deletion", "finance/handle-a-security-incident", "finance/handle-a-trigger-event-review", "finance/handle-a-whistleblowing-report", "finance/handle-an-indemnity-claim-as-paying-bank", "finance/hedge-interest-rate-exposure", "finance/import-a-bank-or-rail-statement", "finance/inspect-a-sar-audit-trail", "finance/introduce-a-customer-to-a-specialist", "finance/investigate-a-complaint", "finance/investigate-a-payment-exception", "finance/investigate-a-production-error", "finance/investigate-a-suspected-fraud", "finance/invite-a-user-with-a-role", "finance/issue-a-default-notice", "finance/issue-a-final-response", "finance/issue-a-finding-to-the-bank", "finance/issue-a-policy", "finance/issue-a-replacement-card-in-branch", "finance/issue-an-agent-mandate", "finance/issue-annual-loan-statements", "finance/lift-an-account-freeze", "finance/log-an-operational-incident", "finance/log-an-outbound-contact", "finance/maintain-internal-and-nostro-accounts", "finance/maintain-the-business-wide-risk-assessment", "finance/maintain-the-chart-of-accounts", "finance/maintain-the-record-of-processing", "finance/maintain-the-regulatory-obligations-register", "finance/maintain-the-risk-register", "finance/make-a-complaint", "finance/make-a-faster-payment", "finance/make-a-regulatory-notification", "finance/make-an-internal-transfer", "finance/make-an-international-payment", "finance/manage-a-syndicated-participation", "finance/manage-capacity", "finance/manage-encryption-and-device-binding", "finance/manage-environments-and-secrets-free-config", "finance/manage-foreign-exchange-positions", "finance/manage-partner-user-access", "finance/manage-payee-name-check-responses", "finance/manage-payment-limits", "finance/manage-senior-manager-responsibilities", "finance/manage-the-central-bank-reserve-account", "finance/map-important-business-services", "finance/mark-a-customer-as-a-fraud-victim", "finance/match-statement-rows-to-ledger-legs", "finance/measure-carbon-tokens-and-time-to-evidence", "finance/message-a-customer-in-a-case", "finance/monitor-chaps-cut-offs-and-queue", "finance/monitor-conduct-and-quality", "finance/monitor-facility-covenants", "finance/monitor-intraday-liquidity", "finance/monitor-job-runs-and-leases", "finance/monitor-privileged-actions", "finance/monitor-product-outcomes", "finance/monitor-service-health", "finance/monitor-without-blocking", "finance/notify-a-claim", "finance/notify-a-death-and-restrict-accounts", "finance/notify-a-personal-data-breach", "finance/offboard-a-business-customer", "finance/onboard-a-partner-organisation", "finance/open-a-fixed-term-deposit", "finance/open-a-foreign-currency-account", "finance/open-a-personal-current-account", "finance/open-a-savings-account", "finance/open-a-service-case", "finance/open-a-transaction-receipt", "finance/open-virtual-accounts-under-a-master", "finance/order-business-cards-for-staff", "finance/pay-redress", "finance/perform-a-data-protection-impact-assessment", "finance/perform-an-internal-audit", "finance/place-a-temporary-account-freeze", "finance/plan-a-portfolio-review", "finance/plan-the-internal-audit-year", "finance/post-a-manual-journal", "finance/prepare-a-business-proposition", "finance/prepare-a-compliance-return", "finance/prepare-a-payment-for-approval", "finance/prepare-a-payment-for-human-approval", "finance/prepare-funds-transfer-pricing", "finance/prepare-regulatory-returns", "finance/prepare-statutory-accounts", "finance/problem-6-3bn-in-30-minutes-how-the-world-s-biggest-heist", "finance/problem-alcohol-duty-penalties-for-late-filing-and-late-pa", "finance/problem-applications-open-for-next-director-of-the-serious", "finance/problem-bank-of-england-explores-easing-key-limit-on-riski", "finance/problem-bereaved-families-will-not-face-tax-bill-over-pens", "finance/problem-beyond-borders-regulating-the-risks-we-share", "finance/problem-binance-rushes-to-recover-customer-money-after-6-3", "finance/problem-binance-s-6-3bn-crypto-hack-biggest-in-history", "finance/problem-binance-s-6-3bn-heist-is-the-biggest-but-not-the-f", "finance/problem-brits-nearly-twice-as-likely-to-spot-pressure-tact", "finance/problem-crypto-giant-binance-attempting-to-recover-6-3bn-s", "finance/problem-crypto-giant-binance-hit-by-6-3bn-hack-suspected-t", "finance/problem-defence-smes-need-banking-services-that-work", "finance/problem-defined-benefit-pension-scheme-surplus-payments-to", "finance/problem-fca-restricts-6-firms-following-concerns-over-anti", "finance/problem-former-chief-executive-banned-after-misconduct-led", "finance/problem-france-s-supermarket-fridges-were-vulnerable-onlin", "finance/problem-french-supermarkets-suffer-fridge-shutdown-after-u", "finance/problem-from-bitcoin-to-binance-crypto-terms-and-what-they", "finance/problem-government-unleashes-1-2-billion-crackdown-as-part", "finance/problem-guidance-framework-agreement-between-the-law-offic", "finance/problem-guidance-make-a-corporate-self-report", "finance/problem-hackers-activate-automatic-defrosting-in-supermark", "finance/problem-how-ai-might-be-deployed-for-better-bank-supervisi", "finance/problem-how-big-a-deal-is-the-6-3bn-binance-crypto-heist-w", "finance/problem-how-crypto-hacking-funds-isolated-north-korea-s-nu", "finance/problem-how-poundland-grew-from-a-market-stall-to-a-high-s", "finance/problem-i-read-anthropic-s-92-page-risk-factor-document-so", "finance/problem-israeli-settlers-attack-european-diplomatic-cars-i", "finance/problem-israeli-settlers-attack-olive-framers-washington-p", "finance/problem-ledger-confirms-spy-chip-implant-behind-wallet-cry", "finance/problem-minutes-of-the-sonia-stakeholder-advisory-group-25", "finance/problem-north-korea-is-likely-culprit-behind-6-3bn-crypto", "finance/problem-ofqual-fines-innovate-awarding-185-000-over-failur", "finance/problem-pip-customers-experiences-of-the-award-review-proc", "finance/problem-response-to-supreme-court-judgment", "finance/problem-scam-american-companies-are-using-to-manipulate-in", "finance/problem-sfo-strengthens-corporate-cooperation-guidance", "finance/problem-the-staggering-446m-elon-musk-has-spent-on-the-us", "finance/problem-the-timms-review-co-chair-update-october-2026", "finance/problem-three-arrests-in-major-fca-operation-against-crimi", "finance/problem-trump-beware-us-voters-are-concerned-about-governm", "finance/problem-two-plead-guilty-after-clients-lose-2m-in-fraud-ca", "finance/problem-tyneside-pizza-boss-jailed-after-gambling-with-mon", "finance/problem-what-we-know-about-the-6-3bn-binance-hack", "finance/problem-whistleblowing", "finance/problem-who-could-be-behind-the-binance-hack-and-what-we-k", "finance/problem-woman-charged-with-computer-fraud-due-to-scam-cash", "finance/process-a-bacs-input-and-output-cycle", "finance/process-a-cheque-clearing-exception", "finance/process-a-counter-cash-deposit", "finance/process-a-counter-cash-withdrawal", "finance/process-a-loan-repayment-manually", "finance/process-an-inbound-faster-payment", "finance/process-card-clearing-and-settlement", "finance/produce-a-kyc-file-for-review", "finance/produce-a-trial-balance", "finance/produce-the-annual-mlro-report", "finance/produce-the-fscs-single-customer-view", "finance/propose-a-future-rate", "finance/propose-a-treasury-credit", "finance/provide-extra-evidence-on-request", "finance/publish-terms-and-conditions", "finance/pull-and-record-a-credit-reference", "finance/query-a-mule-account-indicator", "finance/raise-an-internal-suspicion-report", "finance/receive-a-fraud-intelligence-signal", "finance/receive-a-rate-change-notice", "finance/recognise-loan-interest-income", "finance/reconcile-rail-settlement", "finance/reconcile-suspense-accounts", "finance/record-a-credit-decision-override", "finance/record-a-customer-meeting", "finance/record-a-customer-risk-assessment", "finance/record-a-declined-application-reason", "finance/record-a-regulatory-breach", "finance/record-a-support-need", "finance/record-a-vulnerability-disclosure", "finance/record-aml-training-completion", "finance/record-and-test-a-control", "finance/recover-account-access", "finance/recover-funds-from-a-receiving-account", "finance/refer-a-complaint-to-fos", "finance/refer-to-a-debt-collection-agency", "finance/refer-to-debt-advice", "finance/refund-an-unauthorised-transaction", "finance/register-a-court-appointed-deputy", "finance/register-a-default-with-agencies", "finance/register-a-lasting-power-of-attorney", "finance/register-a-new-device", "finance/register-a-third-party-mandate", "finance/register-and-value-collateral", "finance/register-beneficial-owners", "finance/release-estate-funds", "finance/rely-on-another-firm-s-due-diligence", "finance/remove-a-joint-holder", "finance/rename-an-account", "finance/renew-a-policy", "finance/reopen-a-period", "finance/repair-a-misdirected-payment", "finance/repay-or-overpay-a-loan", "finance/report-a-bereavement", "finance/report-a-change-of-company-details", "finance/report-a-lost-or-stolen-card", "finance/report-an-app-scam", "finance/report-an-unauthorised-payment", "finance/report-complaints-data", "finance/report-fraud-losses-to-the-regulator", "finance/report-fraud-to-action-fraud-or-police", "finance/report-insurance-data-to-the-regulator", "finance/report-payment-statistics", "finance/report-to-credit-reference-agencies", "finance/request-a-bank-statement-letter", "finance/request-a-company-audit-letter", "finance/request-a-covenant-waiver", "finance/request-a-date-of-death-balance-letter", "finance/request-a-defence-against-money-laundering", "finance/request-a-loan-settlement-quote", "finance/request-a-trade-finance-instrument", "finance/request-affordability-evidence", "finance/request-an-arranged-overdraft", "finance/request-an-insurance-quote-from-kyc", "finance/request-further-information", "finance/request-my-personal-data", "finance/request-owner-approval-for-a-consequential-action", "finance/request-records-from-the-bank", "finance/reset-access-for-a-customer", "finance/resolve-within-3-business-days", "finance/respond-to-a-production-order", "finance/respond-to-an-open-banking-payment-request", "finance/restore-from-a-backup", "finance/retry-a-failed-job-step", "finance/retry-a-failed-occurrence", "finance/return-an-unapplied-payment", "finance/reverse-a-journal-with-a-linked-entry", "finance/review-a-complaint-file-as-fos", "finance/review-a-loan-application", "finance/review-access-rights", "finance/review-an-application-queue", "finance/review-app-reimbursement-performance", "finance/review-customer-financial-statements", "finance/review-device-and-session-signals", "finance/review-operational-resilience-evidence", "finance/review-sm-cr-responsibilities", "finance/review-the-vulnerability-register", "finance/revoke-a-role-on-leaving", "finance/revoke-an-agent-mandate", "finance/roll-back-a-release", "finance/rotate-keys-and-device-bindings", "finance/run-a-consumer-duty-outcomes-review", "finance/run-a-disaster-recovery-test", "finance/run-a-fair-value-assessment", "finance/run-a-kyc-check-on-a-subject", "finance/run-a-ledger-report-for-a-period", "finance/run-a-liquidity-stress-test", "finance/run-an-annual-relationship-review", "finance/run-due-jobs-as-service-actor", "finance/run-due-standing-orders", "finance/run-end-of-day-payment-reconciliation", "finance/run-payroll-payments", "finance/run-penetration-testing", "finance/run-periodic-kyc-refresh", "finance/run-protected-reveals", "finance/run-sanctions-screening", "finance/run-transaction-monitoring", "finance/sample-test-receipts-as-external-auditor", "finance/save-an-unfinished-application", "finance/schedule-a-future-payment", "finance/screen-a-list-against-sanctions", "finance/screen-existing-customers-against-list-updates", "finance/screen-for-politically-exposed-persons", "finance/search-transactions", "finance/send-a-chaps-payment", "finance/send-a-receiving-bank-app-notice", "finance/set-and-monitor-exposure-limits", "finance/set-balance-and-payment-alerts", "finance/set-card-limits-and-controls", "finance/set-eligibility-rules", "finance/set-fees-and-charges", "finance/set-joint-operating-rules", "finance/set-loan-terms-and-produce-an-offer", "finance/set-negotiated-pricing", "finance/set-up-a-direct-debit", "finance/set-up-a-mandate", "finance/set-up-a-standing-order", "finance/set-up-merchant-acquiring", "finance/settle-a-claim", "finance/settle-and-close-a-loan", "finance/stop-a-savings-enrolment", "finance/submit-transactions-for-monitoring", "finance/support-a-customer-in-financial-difficulty", "finance/support-the-external-audit", "finance/switch-account-to-edge-bank", "finance/terminate-a-partnership", "finance/trace-ingress-and-egress", "finance/track-remediation-actions", "finance/transition-a-case-state", "finance/triage-a-case-queue", "finance/tune-a-fraud-rule", "finance/underwrite-a-business-facility", "finance/underwrite-a-policy", "finance/update-communication-preferences", "finance/upload-a-bulk-payment-file", "finance/validate-a-credit-risk-model", "finance/verify-a-caller-or-visitor", "finance/verify-a-company-and-its-officers", "finance/verify-a-regulatory-return-against-the-ledger", "finance/verify-a-trust-or-charity", "finance/verify-address-evidence", "finance/verify-an-identity-through-edge-id", "finance/verify-executor-or-administrator", "finance/verify-identity-evidence", "finance/view-and-download-a-statement", "finance/view-balances-and-available-funds", "finance/view-multi-account-cash-position", "finance/view-pin-securely", "finance/void-an-imported-statement", "finance/withdraw-a-product", "finance/withdraw-a-proposed-rate", "finance/withdraw-an-application", "finance/work-the-fraud-alert-queue", "finance/write-a-fraud-case-outcome-letter", "finance/write-off-a-debt", "foundation/automated-review-outreach-and-scheduling", "foundation/bottleneck-and-blocker-removal", "foundation/claim-organisation-by-magic-link", "foundation/evaluation-and-commercial-licensing", "foundation/licence-enforcement-evidence", "foundation/pre-onboarded-organisations", "foundation/problem-12-000-fine-for-cheltenham-waste-operator-for-disp", "foundation/problem-a-28-year-old-doge-operative-gutted-usaid-hr-had-n", "foundation/problem-a-complete-bypass-of-emudhra-s-1fa-otp-login-syste", "foundation/problem-about-the-compromised-emudhra-intermediate-cas", "foundation/problem-air-pollution-trends-reports", "foundation/problem-azure-s-weakest-link-five-full-cross-tenant-compro", "foundation/problem-bedfordshire-soldiers-laid-to-rest-on-the-western", "foundation/problem-binance-hacker-1029-5-6b-stolen-from-binance", "foundation/problem-blockchain-intelligence-firms-link-north-korea-to", "foundation/problem-britain-s-housing-crisis-won-t-be-fixed-without-a", "foundation/problem-bulgaria-votes-as-russian-attacks-in-black-sea-rea", "foundation/problem-burnham-ready-to-support-poundland-staff-as-retail", "foundation/problem-capital-gains-tax-returns-for-disposals-of-uk-land", "foundation/problem-codex-cli-installs-a-persistent-self-updating-back", "foundation/problem-companies-house-third-party-code-of-conduct", "foundation/problem-company-director-disqualification-following-breach", "foundation/problem-compromised-litellm-1-107-0-on-pypi-installs-a-cre", "foundation/problem-cops-can-bypass-iphone-s-automatic-reboot-to-get-i", "foundation/problem-cve-2026-111943-compromised-intermediate-certifica", "foundation/problem-defending-odihr-s-independence-and-the-osce-human", "foundation/problem-dungeon-generation-in-the-gnomish-mines-of-nethack", "foundation/problem-emudhra-intermediate-ca-compromise", "foundation/problem-excise-duty-penalties-for-wrongdoing", "foundation/problem-eye-of-sauron-long-range-hidden-spy-camera-detecti", "foundation/problem-florida-s-hoas-and-condo-associations-are-making-t", "foundation/problem-git-3-0-s-upcoming-sha-256-default-will-be-a-costl", "foundation/problem-guidance-chickerell-firing-times", "foundation/problem-guidance-tregantle-firing-times", "foundation/problem-gwyndod-supply-chain-mobilised-to-deliver-uk-s-fir", "foundation/problem-hackers-expose-the-identities-of-more-than-40k-dhs", "foundation/problem-how-a-1-deal-dragged-poundland-s-owner-into-a-poli", "foundation/problem-how-oil-rich-alberta-s-economy-might-fare-if-it-br", "foundation/problem-how-the-military-will-hack-the-election", "foundation/problem-i-m-on-a-war-footing-terry-tao-and-the-crisis-in-m", "foundation/problem-icc-judge-on-what-u-s-sanctions-mean-for-her-and-g", "foundation/problem-imposing-sanctions-on-the-international-criminal-c", "foundation/problem-independent-review-of-the-prison-system-terms-of-r", "foundation/problem-investigation-launched-as-dispute-risks-bringing-d", "foundation/problem-it-took-me-nine-years-to-hack-my-xbox-360", "foundation/problem-jared-kushner-hacked-phone-reveals-secret-negotiat", "foundation/problem-london-loft-and-extension-company-which-left-unfin", "foundation/problem-making-registration-applications-easier-and-more-e", "foundation/problem-middlesex-regiment-soldier-laid-to-rest-on-the-som", "foundation/problem-modernising-the-correction-of-errors", "foundation/problem-nicotine-approval-scheme", "foundation/problem-no-more-icons-in-every-menu-by-default", "foundation/problem-northflank-breached", "foundation/problem-opencode-hacked-user-credentials-leaked", "foundation/problem-professor-akande-brings-the-independence-legal-rig", "foundation/problem-publishing-details-of-deliberate-defaulters-reform", "foundation/problem-radioactivity-in-food-and-the-environment-rife-rep", "foundation/problem-restrictions-on-the-performance-of-high-risk-cosme", "foundation/problem-show-hn-wallhop-12ft-io-is-gone-so-i-built-a-repla", "foundation/problem-telegram-desktop-vulnerability-allowed-any-user-s", "foundation/problem-tell-hn-scribd-breached", "foundation/problem-the-6-3b-binance-hack-was-pulled-off-in-just-20-mi", "foundation/problem-today-we-hacked-what-we-d-set-out-to-hack-the-ipho", "foundation/problem-uk-sanctions-senior-commanders-implicated-in-campa", "foundation/problem-uk-ukraine-and-latvia-extend-crackdown-on-russian", "foundation/problem-us-federal-budget-deficit-jumps-12-in-one-year-to", "foundation/problem-us-imposes-sanctions-on-icc-hours-after-former-jud", "foundation/problem-various-projects-find-hidden-sdr-capabilities-in-e", "foundation/problem-we-re-going-to-need-default-hard-budget-caps-on-pr", "foundation/problem-why-north-korea-s-elite-hackers-are-the-prime-susp", "foundation/problem-why-windows-11-still-ships-with-windows-95-s-memor", "foundation/problem-zod-4-6-6-compromised-by-malware", "foundation/regulator-auditor-reviewer-workspace", "foundation/scope-change-intake", "judiverse/judebox-agents", "judiverse/judebox-calls", "judiverse/judebox-connections", "judiverse/judebox-contacts", "judiverse/judebox-email", "judiverse/judebox-memory-and-vault", "judiverse/judebox-messaging", "judiverse/judebox-numbers-and-mailboxes", "judiverse/judebox-webhooks", "judiverse/judiverse-included-with-every-licence", "judiverse/judiverse-standalone-subscription", "judiverse/problem-123456-password-used-in-danish-cpr-data-breach", "judiverse/problem-afd-leak-what-421-179-hacked-emails-show-about-the", "judiverse/problem-agentic-ai-and-consumers", "judiverse/problem-ai-and-leadership", "judiverse/problem-amazon-uses-ai-to-monitor-union-risk-in-uk-warehou", "judiverse/problem-apple-and-a-hacker-s-future", "judiverse/problem-apple-music-404-not-found", "judiverse/problem-apple-s-git-still-trusts-the-compromised-emudhra-c", "judiverse/problem-brown-student-maps-the-hidden-global-labor-powerin", "judiverse/problem-cisco-gives-engineer-unlimited-ai-tokens-to-rewrit", "judiverse/problem-clef-open-weight-decision-models-and-new-rl-fine-t", "judiverse/problem-containment-and-monitoring-failures-are-to-blame-f", "judiverse/problem-cross-request-kv-cache-exposure-in-the-glm-5-3-wor", "judiverse/problem-denmark-data-breach-exposes-8-8m-people-s-personal", "judiverse/problem-draft-legislation-better-use-of-new-and-improved-t", "judiverse/problem-electrocutions-disease-and-no-pay-how-workers-risk", "judiverse/problem-faster-planning-decisions-as-new-rules-and-ai-tack", "judiverse/problem-ftc-is-investigating-openai-anthropic-and-other-ai", "judiverse/problem-hacker-leaks-281-gb-of-data-from-germany-s-far-rig", "judiverse/problem-i-quit-openai-because-its-culture-is-broken", "judiverse/problem-inside-the-make-or-break-ai-mega-project-behind-ns", "judiverse/problem-investors-trim-growth-bets-over-fears-of-us-econom", "judiverse/problem-is-a-job-guarantee-the-solution-to-mass-ai-unemplo", "judiverse/problem-meta-hit-with-25b-penalty-for-deceiving-users-abou", "judiverse/problem-near-to-real-time-suspected-suicide-surveillance-n", "judiverse/problem-openai-safety-leader-quits-warning-ai-company-s-cu", "judiverse/problem-postgres-19-delayed-after-ai-reviewers-find-more-b", "judiverse/problem-research-agentic-ai-and-consumers", "judiverse/problem-run-full-ios-16-5-firmware-on-android-devices", "judiverse/problem-turn-a-wifi-7-router-into-a-320msps-software-defin", "judiverse/problem-vote-on-which-of-hacker-news-challenges-for-ai-hav", "judiverse/problem-wanted-to-download-an-iso-ended-up-writing-a-kerne", "judiverse/problem-why-aren-t-ai-companies-being-charged-for-hacking", "judiverse/problem-with-most-information-hidden-the-game-stratego-had", "judiverse/problem-yandex-bot-traffic-down-99-after-data-center-attac", "public-edge/accept-a-package-referral", "public-edge/act-on-a-deadline-stack-item", "public-edge/activate-the-emergency-centre-and-keep-the-decision-log", "public-edge/adopt-a-risk-based-reserves-strategy-with-minimum-level-and", "public-edge/advise-on-legality-and-sign-the-legal-implications-of-a-repo", "public-edge/agree-a-care-and-support-plan-and-personal-budget", "public-edge/agree-a-deferred-payment", "public-edge/agree-a-personalised-housing-plan", "public-edge/agree-an-information-sharing-agreement", "public-edge/agree-and-seal-a-planning-obligation", "public-edge/agree-the-council-plan-and-its-outcomes", "public-edge/agree-the-officer-and-member-protocol", "public-edge/agree-the-political-group-position-before-a-meeting", "public-edge/agree-the-post-election-timetable-and-transition-protocol", "public-edge/allocate-committee-seats-by-political-balance", "public-edge/analog-human-switch-for-irreversible-actions", "public-edge/answer-a-best-value-notice-or-mhclg-engagement-with-one-evid", "public-edge/answer-a-code-of-conduct-complaint", "public-edge/answer-a-corporate-complaint", "public-edge/answer-a-freedom-of-information-request", "public-edge/answer-a-local-land-charges-search", "public-edge/answer-a-media-enquiry-and-issue-a-press-release", "public-edge/answer-a-standards-complaint", "public-edge/answer-a-subject-access-request", "public-edge/answer-a-written-parliamentary-question", "public-edge/answer-an-appeal-to-the-valuation-tribunal", "public-edge/answer-an-environmental-information-request", "public-edge/answer-ministerial-correspondence", "public-edge/answer-questions-at-a-scrutiny-meeting", "public-edge/appeal-a-planning-decision", "public-edge/apply-an-empty-home-or-second-home-premium", "public-edge/apply-an-exemption-and-issue-a-refusal-notice", "public-edge/apply-for-a-blue-badge", "public-edge/apply-for-a-deprivation-of-liberty-authorisation", "public-edge/apply-for-a-discount-or-council-tax-reduction", "public-edge/apply-for-a-licence", "public-edge/apply-for-a-postal-or-proxy-vote", "public-edge/apply-for-a-school-place-and-appeal-a-refusal", "public-edge/apply-for-a-voter-authority-certificate", "public-edge/apply-for-an-export-licence", "public-edge/apply-for-free-school-meals", "public-edge/apply-for-planning-permission-and-track-it", "public-edge/apply-the-armed-forces-covenant-duty-in-a-council-decision", "public-edge/apply-the-monthly-savings-test", "public-edge/apply-the-national-pay-award", "public-edge/apply-the-one-page-decision-standard-and-report-standard-to", "public-edge/apply-the-pre-election-period", "public-edge/apply-the-retention-schedule-and-dispose-of-records", "public-edge/appoint-an-election-agent", "public-edge/appoint-an-independent-member-to-the-committee", "public-edge/appoint-and-train-polling-staff", "public-edge/appoint-or-dismiss-a-statutory-officer", "public-edge/appoint-the-cabinet-and-allocate-portfolios", "public-edge/appoint-the-external-auditor-through-psaa", "public-edge/appoint-the-internal-auditor-and-receive-the-report", "public-edge/approve-a-business-case-through-the-assurance-framework", "public-edge/approve-a-change-through-the-change-board", "public-edge/approve-a-direct-payment", "public-edge/approve-a-disabled-facilities-grant", "public-edge/approve-a-new-system-against-the-architecture-standard", "public-edge/approve-a-severance-package-over-100-000", "public-edge/approve-a-virement", "public-edge/approve-the-internal-audit-charter", "public-edge/approve-the-internal-audit-plan", "public-edge/approve-the-medium-term-financial-strategy", "public-edge/approve-the-pension-fund-accounts-and-valuation", "public-edge/approve-the-senior-structure-and-statutory-officer-designati", "public-edge/approve-the-treasury-capital-and-minimum-revenue-provision-s", "public-edge/ask-a-question-of-the-leader-or-a-cabinet-member", "public-edge/ask-a-question-or-present-a-petition-at-a-meeting", "public-edge/ask-for-a-care-needs-assessment", "public-edge/ask-for-help-with-homelessness", "public-edge/ask-for-own-personal-data", "public-edge/assess-a-carer-s-needs", "public-edge/assess-a-code-of-conduct-complaint", "public-edge/assess-a-housing-benefit-claim", "public-edge/assess-and-decide-bus-franchising", "public-edge/assess-council-tax-reduction", "public-edge/assess-risk-and-publish-the-community-risk-register", "public-edge/assess-the-committee-s-effectiveness-and-report-to-council", "public-edge/assess-the-needs-of-an-adult", "public-edge/attend-film-or-report-a-public-meeting", "public-edge/attend-the-count-as-an-agent", "public-edge/authenticate-and-seal-a-document", "public-edge/authorise-and-make-a-payment", "public-edge/authorise-covert-surveillance-under-ripa", "public-edge/award-a-discount-or-exemption", "public-edge/award-a-discretionary-housing-payment", "public-edge/award-and-monitor-a-voluntary-sector-grant", "public-edge/award-business-rates-relief", "public-edge/award-under-a-framework-or-dynamic-market", "public-edge/back-up-and-test-a-restore", "public-edge/bid-for-and-report-on-a-transport-grant", "public-edge/bill-a-new-business-rates-property", "public-edge/build-a-resident-service-through-discovery-alpha-and-beta", "public-edge/build-and-sign-the-savings-ledger-line-by-line", "public-edge/calculate-the-council-tax-base-and-requirement", "public-edge/call-in-an-executive-decision", "public-edge/canvass-voters-lawfully", "public-edge/carry-out-a-data-protection-impact-assessment", "public-edge/carry-out-a-financial-assessment-and-charge", "public-edge/carry-out-an-education-health-and-care-needs-assessment", "public-edge/carry-out-an-internal-review", "public-edge/cast-a-recorded-vote-on-the-budget", "public-edge/certify-a-grant-claim", "public-edge/chair-full-council-and-rule-on-order", "public-edge/challenge-a-council-tax-band", "public-edge/challenge-a-provider-rate-and-decide-an-uplift", "public-edge/check-supply-chain-labour-and-modern-slavery-risk", "public-edge/claim-emergency-financial-assistance", "public-edge/claim-members-allowances-and-expenses", "public-edge/claim-staffing-and-business-costs", "public-edge/clerk-a-school-admission-appeal-panel", "public-edge/comment-on-a-planning-application", "public-edge/comment-on-a-planning-application-2", "public-edge/commission-a-placement-for-a-looked-after-child", "public-edge/commission-a-public-health-service-under-the-provider-select", "public-edge/commission-supported-living-from-a-charity-provider", "public-edge/commission-temporary-accommodation", "public-edge/commission-the-external-quality-assessment", "public-edge/complain-to-the-council-and-the-ombudsman", "public-edge/complete-induction-and-training-on-reserves-debt-efs-and-ris", "public-edge/complete-mandatory-member-training", "public-edge/complete-the-annual-governance-and-accountability-return", "public-edge/complete-the-data-security-and-protection-toolkit", "public-edge/comply-with-a-secretary-of-state-direction", "public-edge/conduct-a-marriage-or-civil-partnership", "public-edge/confirm-mrp-and-efs-accounting-with-the-external-auditor", "public-edge/consider-an-objection-from-a-local-elector", "public-edge/consider-social-value-before-procurement", "public-edge/consolidate-published-recommendations-into-owned-operating-s", "public-edge/consult-and-publicise-an-application", "public-edge/consult-on-and-make-redundancies", "public-edge/convene-the-right-people-by-mail-and-call", "public-edge/correct-a-register-entry", "public-edge/deal-with-a-dangerous-structure", "public-edge/decide-a-parking-penalty-challenge", "public-edge/decide-an-application-at-planning-committee", "public-edge/decide-on-sharing-confidential-care-information-as-caldicott", "public-edge/declare-an-impairment-to-independence", "public-edge/declare-an-interest-at-a-meeting-and-withdraw", "public-edge/declare-an-interest-in-proceedings", "public-edge/declare-the-collection-fund-surplus-or-deficit", "public-edge/decommission-a-legacy-system", "public-edge/defend-a-planning-appeal", "public-edge/deliver-the-risk-based-audit-plan-from-substantive-capacity", "public-edge/design-and-run-a-grant-scheme", "public-edge/determine-a-building-regulations-application", "public-edge/determine-a-postal-or-proxy-vote-application", "public-edge/determine-an-application-to-register", "public-edge/discharge-planning-conditions", "public-edge/dispose-of-or-acquire-land-at-best-consideration", "public-edge/employ-staff-in-the-mp-s-office", "public-edge/enforce-joiner-mover-and-leaver-with-segregation-of-duties", "public-edge/enforce-publication-deadlines-with-an-exception-register", "public-edge/enrol-staff-in-the-pension-scheme-and-send-monthly-data", "public-edge/ensure-continuity-of-care-on-provider-failure", "public-edge/evaluate-and-grade-a-job", "public-edge/fill-a-casual-vacancy-by-election-or-co-option", "public-edge/find-a-council-borough-or-government-department-in-the-insti", "public-edge/find-the-right-council-and-service-for-an-address", "public-edge/flow-obligations-into-a-contract-or-commission", "public-edge/give-notice-to-hand-back-a-contract", "public-edge/give-the-opinion-on-the-financial-statements", "public-edge/grant-a-dispensation-for-a-pecuniary-interest", "public-edge/grant-a-premises-licence", "public-edge/grant-a-street-works-permit", "public-edge/handle-a-benefit-reconsideration-or-appeal", "public-edge/handle-a-highway-claim", "public-edge/handle-a-housing-complaint", "public-edge/handle-a-request-for-erasure-or-restriction", "public-edge/hear-a-call-in-and-decide", "public-edge/hear-a-councillor-call-for-action", "public-edge/hold-a-citizenship-ceremony", "public-edge/hold-a-constituency-surgery", "public-edge/hold-a-licensing-sub-committee-hearing", "public-edge/hold-an-initial-child-protection-conference", "public-edge/hold-and-record-the-monthly-golden-triangle-meeting", "public-edge/hold-identity-and-operational-evidence", "public-edge/hold-pre-decision-scrutiny-of-a-key-decision", "public-edge/hold-the-annual-parish-meeting", "public-edge/host-a-social-housing-consumer-standards-inspection", "public-edge/host-an-lga-corporate-peer-challenge", "public-edge/independently-test-an-implemented-control-under-the-self-rev", "public-edge/inspect-a-food-business-and-award-a-hygiene-rating", "public-edge/inspect-a-highway-and-repair-a-defect", "public-edge/inspect-a-private-rented-home-and-serve-a-notice", "public-edge/inspect-a-workplace-for-health-and-safety", "public-edge/inspect-the-council-s-accounts-and-object", "public-edge/inspect-work-and-issue-a-completion-certificate", "public-edge/instruct-enforcement-agents-with-a-vulnerability-check", "public-edge/investigate-a-breach-and-serve-an-enforcement-notice", "public-edge/investigate-a-consumer-complaint-and-take-enforcement-action", "public-edge/investigate-a-flood-as-lead-local-flood-authority", "public-edge/investigate-a-statutory-nuisance-and-serve-an-abatement-noti", "public-edge/investigate-a-suspected-fraud", "public-edge/investigate-tenancy-fraud", "public-edge/issue-a-best-value-notice-and-run-engagement", "public-edge/issue-a-certified-copy-of-an-entry", "public-edge/issue-a-cil-liability-notice-and-demand", "public-edge/issue-a-mayoral-decision", "public-edge/issue-a-refund-or-transfer-a-credit", "public-edge/issue-a-section-114-report", "public-edge/issue-a-section-5-report-on-unlawfulness-or-maladministratio", "public-edge/issue-a-security-aspects-letter-to-a-contractor", "public-edge/issue-a-statutory-recommendation-or-public-interest-report", "public-edge/issue-a-voter-authority-certificate", "public-edge/issue-annual-council-tax-bills", "public-edge/issue-the-annual-internal-audit-opinion", "public-edge/issue-the-audit-plan", "public-edge/issue-the-decision-notice", "public-edge/issue-the-role-statement-train-and-attribute-challenge", "public-edge/issue-the-summons-and-agenda", "public-edge/issue-the-weekly-one-page-exception-report", "public-edge/join-a-council-s-approved-provider-list", "public-edge/join-a-strategic-coordinating-group", "public-edge/join-the-housing-register-and-assess-priority", "public-edge/keep-business-continuity-plans-and-advise-businesses", "public-edge/keep-spending-within-the-limit-and-file-the-return", "public-edge/keep-statutory-notification-readiness-and-named-incident-rol", "public-edge/keep-the-device-and-software-asset-register", "public-edge/keep-the-outside-bodies-appointments-register", "public-edge/keep-the-record-of-processing-activities", "public-edge/lay-a-statutory-instrument", "public-edge/lay-the-annual-report-and-accounts", "public-edge/lead-recovery", "public-edge/lead-the-response-to-a-section-114-notice", "public-edge/let-a-social-home", "public-edge/licence-a-taxi-or-private-hire-driver", "public-edge/licence-an-animal-activity", "public-edge/license-a-house-in-multiple-occupation", "public-edge/maintain-and-publish-the-register-of-members-interests", "public-edge/maintain-the-brownfield-and-self-build-registers", "public-edge/maintain-the-constitution-and-scheme-of-delegation-with-live", "public-edge/maintain-the-contract-procedure-rules", "public-edge/maintain-the-publication-scheme-and-disclosure-log", "public-edge/maintain-the-single-procurement-and-commissioning-pipeline", "public-edge/make-a-freedom-of-information-request-or-ask-for-review", "public-edge/make-a-public-spaces-protection-order", "public-edge/make-a-safeguarding-adults-enquiry", "public-edge/make-a-traffic-regulation-order", "public-edge/make-a-tree-preservation-order", "public-edge/manage-a-contract-and-hold-review-meetings", "public-edge/manage-a-persistent-or-unreasonable-complainant", "public-edge/manage-a-public-question-or-petition-at-a-meeting", "public-edge/manage-allotments", "public-edge/manage-cash-flow-and-borrowing", "public-edge/manage-sickness-absence", "public-edge/mark-and-check-politically-restricted-posts", "public-edge/mark-exempt-information-and-run-a-private-item", "public-edge/mark-store-and-send-classified-information", "public-edge/meet-the-auditors-without-officers-present", "public-edge/modify-a-contract-and-publish-the-change-notice", "public-edge/monitor-a-grant-and-recover-misused-funds", "public-edge/monitor-a-provider-s-quality", "public-edge/name-and-number-a-street", "public-edge/nominate-an-asset-of-community-value", "public-edge/nominate-an-asset-of-community-value-2", "public-edge/notify-the-regulator-of-an-incident", "public-edge/obtain-facility-security-clearance-for-a-supplier-site", "public-edge/open-a-constituent-case-with-consent", "public-edge/open-a-council-tax-account-on-a-move-in", "public-edge/open-a-rest-centre-and-register-evacuees", "public-edge/open-and-conclude-an-inquest", "public-edge/open-postal-votes-and-verify-personal-identifiers", "public-edge/open-the-government-station-and-choose-a-journey", "public-edge/organise-evidence-and-records-for-a-public-body", "public-edge/pass-a-proposal-through-the-business-case-gateway", "public-edge/pay-council-tax-and-set-up-a-direct-debit", "public-edge/pay-or-challenge-a-parking-penalty", "public-edge/pay-the-data-protection-fee", "public-edge/place-a-household-in-interim-accommodation", "public-edge/pool-a-budget-with-the-nhs", "public-edge/prepare-a-care-proceedings-application", "public-edge/prepare-a-green-book-business-case", "public-edge/prepare-a-neighbourhood-plan", "public-edge/prepare-a-procurement-and-agreement-requirements-brief-for-a", "public-edge/prepare-an-impact-assessment", "public-edge/prepare-an-off-site-plan-for-a-major-hazard-site", "public-edge/prepare-and-examine-the-local-plan", "public-edge/prepare-for-a-cqc-assessment-of-adult-social-care", "public-edge/prepare-for-a-public-accounts-committee-hearing", "public-edge/prepare-for-an-inspection-of-children-s-services", "public-edge/prepare-main-and-supplementary-estimates", "public-edge/prepare-the-local-transport-plan", "public-edge/prepare-the-market-sustainability-plan", "public-edge/price-a-single-source-defence-contract", "public-edge/problem-cyberattack-halts-construction-of-tunnel-between-d", "public-edge/problem-details-of-thousands-of-us-immigration-agents-leak", "public-edge/problem-dutch-government-had-hacked-into-jared-kushner-s-p", "public-edge/problem-dutch-national-cyber-security-center-is-offline-be", "public-edge/problem-experts-warn-ai-copyright-changes-could-breach-aus", "public-edge/problem-help-the-ncsc-understand-how-organisations-approac", "public-edge/problem-how-north-korea-became-a-hacking-superpower", "public-edge/problem-how-sin-bin-for-red-light-jumpers-saves-cyclists-f", "public-edge/problem-how-the-military-will-hack-the-election", "public-edge/problem-incident-affecting-asos-customers", "public-edge/problem-introducing-a-new-ai-chapter-in-the-ncsc-research", "public-edge/problem-is-this-the-new-cyber-normal", "public-edge/problem-joint-ministerial-statement-on-the-risk-of-famine", "public-edge/problem-labor-and-the-coalition-are-poised-to-take-a-reckl", "public-edge/problem-legacy-system-local-plan-extension-letter-from-the", "public-edge/problem-meta-ordered-to-pay-25b-penalty-in-new-mexico-camb", "public-edge/problem-minister-mcneill-speech-from-hunger-hotspots-to-po", "public-edge/problem-more-people-protected-as-government-acts-to-make-m", "public-edge/problem-procurement-compliance-oversight", "public-edge/problem-public-electric-vehicle-charge-points-cyber-securi", "public-edge/problem-russia-s-decision-to-veto-this-draft-resolution-ri", "public-edge/problem-supporting-young-people-s-independence-through-tra", "public-edge/problem-the-united-kingdom-remains-committed-to-working-wi", "public-edge/problem-what-we-know-about-binance-s-6-3bn-hack", "public-edge/problem-william-gibson-and-the-trademarking-of-cyberspace", "public-edge/process-a-right-to-buy-application", "public-edge/process-a-temporary-event-notice", "public-edge/procure-a-separately-procurable-engagement-with-an-aggregati", "public-edge/propose-the-budget-and-council-tax-to-full-council", "public-edge/publish-an-accessibility-statement", "public-edge/publish-an-algorithmic-transparency-record", "public-edge/publish-an-executive-decision-record", "public-edge/publish-and-maintain-privacy-notices", "public-edge/publish-government-procurement-card-spend", "public-edge/publish-members-allowances-for-the-year", "public-edge/publish-notice-of-a-key-decision-28-clear-days-ahead", "public-edge/publish-open-data", "public-edge/publish-spending-over-500", "public-edge/publish-the-agenda-and-reports-five-clear-days-ahead", "public-edge/publish-the-annual-pay-policy-statement", "public-edge/publish-the-contracts-register", "public-edge/publish-the-draft-and-audited-statement-of-accounts", "public-edge/publish-the-forward-plan-twelve-weeks-ahead", "public-edge/publish-the-gambling-statement-of-principles", "public-edge/publish-the-gender-pay-gap", "public-edge/publish-the-market-position-statement", "public-edge/publish-the-notice-of-election", "public-edge/publish-the-regulatory-enforcement-policy", "public-edge/publish-the-revised-register", "public-edge/publish-the-updated-constitution", "public-edge/publish-transparency-data", "public-edge/put-a-digital-imprint-on-campaign-material", "public-edge/raise-a-purchase-order-before-package-spend", "public-edge/raise-a-safeguarding-concern", "public-edge/raise-a-service-complaint", "public-edge/read-a-committee-pack-within-the-page-limit-and-lodge-questi", "public-edge/rebase-the-adult-social-care-budget-from-the-care-ledger", "public-edge/receive-and-decide-on-nominations", "public-edge/receive-and-handle-a-whistleblowing-disclosure", "public-edge/receive-and-publish-candidates-spending-returns", "public-edge/receive-quarterly-monitoring-with-a-demand-bridge", "public-edge/receive-the-auditor-s-annual-report-and-value-for-money-comm", "public-edge/receive-the-counter-fraud-annual-report", "public-edge/receive-the-monthly-evidence-pack-rated-on-evidence", "public-edge/reconcile-access-and-contain-unexplained-accounts", "public-edge/reconcile-the-balance-sheet-exposure-statement-quarterly", "public-edge/record-a-gift-or-hospitality", "public-edge/record-a-price-exception-with-the-finance-position", "public-edge/record-an-exemption-from-contract-procedure-rules", "public-edge/recover-a-benefit-overpayment", "public-edge/recover-unpaid-council-tax-through-a-liability-order", "public-edge/recruit-safely-with-dbs-checks", "public-edge/refer-a-band-challenge-to-the-valuation-office-agency", "public-edge/refer-higher-risk-building-work-to-the-building-safety-regul", "public-edge/register-a-birth", "public-edge/register-a-birth-death-or-marriage-notice", "public-edge/register-a-death-and-offer-tell-us-once", "public-edge/register-a-financial-interest", "public-edge/register-a-foreign-activity-arrangement", "public-edge/register-a-political-party", "public-edge/register-as-a-consultant-lobbyist-and-file-returns", "public-edge/register-as-a-non-party-campaigner", "public-edge/register-disclosable-pecuniary-interests", "public-edge/register-to-vote", "public-edge/reimburse-operators-for-concessionary-travel", "public-edge/renew-public-services-network-compliance", "public-edge/reperform-a-closed-control-from-source", "public-edge/report-a-defence-security-incident", "public-edge/report-a-street-problem", "public-edge/report-complaints-and-compliments-annually", "public-edge/report-contract-performance-indicators", "public-edge/report-digital-service-performance", "public-edge/report-donations-and-loans", "public-edge/report-limited-or-no-assurance-to-the-audit-committee", "public-edge/report-on-a-council-owned-company-to-the-shareholder-committ", "public-edge/report-on-staffing-and-organisation-as-head-of-paid-service", "public-edge/report-on-the-robustness-of-estimates-and-adequacy-of-reserv", "public-edge/report-on-value-for-money-arrangements", "public-edge/represent-the-council-on-the-combined-authority-or-a-regiona", "public-edge/request-a-rate-uplift", "public-edge/request-a-review-of-a-housing-decision", "public-edge/request-and-track-audit-evidence", "public-edge/request-exceptional-financial-support", "public-edge/require-a-cabinet-member-or-officer-to-attend", "public-edge/resolve-a-service-desk-incident", "public-edge/resolve-eligibility-for-the-general-power-of-competence", "public-edge/respond-to-a-child-protection-referral", "public-edge/respond-to-a-damp-mould-or-repair-report", "public-edge/respond-to-a-judicial-review-pre-action-letter", "public-edge/respond-to-a-parliamentary-petition", "public-edge/respond-to-a-statutory-recommendation-from-the-auditor", "public-edge/respond-to-an-area-send-inspection", "public-edge/respond-to-an-ico-decision-notice", "public-edge/respond-to-an-ombudsman-enquiry-and-carry-out-the-remedy", "public-edge/respond-to-an-ombudsman-finding", "public-edge/return-post-election-data-to-the-electoral-commission", "public-edge/review-a-homelessness-decision", "public-edge/review-a-looked-after-child-s-care-plan", "public-edge/review-local-air-quality", "public-edge/review-polling-districts-and-places", "public-edge/review-the-annual-governance-statement", "public-edge/review-the-corporate-risk-register-quarterly", "public-edge/review-the-draft-statement-of-accounts", "public-edge/review-the-information-risk-register-as-senior-information-r", "public-edge/review-the-risk-management-framework", "public-edge/run-a-combined-authority-board-meeting", "public-edge/run-a-deep-dive-with-recorded-influence-on-direction", "public-edge/run-a-demand-control-cell-weekly", "public-edge/run-a-disaster-recovery-test", "public-edge/run-a-disciplinary-process", "public-edge/run-a-neighbourhood-planning-referendum", "public-edge/run-a-public-consultation-and-publish-the-response", "public-edge/run-a-public-consultation-and-report-the-results", "public-edge/run-a-recall-petition-as-petition-officer", "public-edge/run-a-standards-hearing-and-publish-the-decision", "public-edge/run-a-task-and-finish-review-and-secure-the-executive-respon", "public-edge/run-an-assurance-review-on-a-major-project", "public-edge/run-an-audit-engagement-from-terms-of-reference-to-report", "public-edge/run-monthly-budget-monitoring-with-budget-holders", "public-edge/run-monthly-payroll-and-report-to-hmrc", "public-edge/run-national-fraud-initiative-matches", "public-edge/run-security-awareness-and-phishing-tests", "public-edge/run-the-annual-canvass", "public-edge/run-the-annual-council-meeting", "public-edge/run-the-cross-committee-forum-with-other-chairs", "public-edge/run-the-decision-clock-on-stalled-reforms", "public-edge/run-the-make-or-buy-test-before-a-material-re-let", "public-edge/run-the-out-of-hours-duty-officer-rota", "public-edge/run-the-period-for-exercise-of-public-rights", "public-edge/run-the-public-inspection-period-for-the-accounts", "public-edge/run-the-reorganisation-programme-to-vesting-day", "public-edge/run-the-verification-and-count-and-declare-the-result", "public-edge/score-the-thirty-cyber-criteria-with-evidence-references", "public-edge/scrutinise-a-substantial-nhs-service-change", "public-edge/scrutinise-the-budget-proposals-before-council", "public-edge/scrutinise-the-combined-authority", "public-edge/scrutinise-the-community-safety-partnership", "public-edge/scrutinise-treasury-management", "public-edge/seek-a-cabinet-office-spend-control-approval", "public-edge/seek-hm-treasury-approval-above-delegated-limits", "public-edge/select-a-candidate-and-authorise-the-party-description", "public-edge/send-apologies-and-keep-attendance-within-the-six-month-rule", "public-edge/separate-cashable-saving-cost-avoidance-and-deferred-demand", "public-edge/set-and-collect-council-housing-rents", "public-edge/set-the-annual-calendar-of-meetings", "public-edge/set-the-annual-work-programme", "public-edge/set-the-local-government-finance-settlement", "public-edge/set-the-mayoral-budget-and-precept", "public-edge/set-the-precept-and-notify-the-billing-authority", "public-edge/set-the-scheme-of-executive-delegation", "public-edge/set-up-an-instalment-plan-or-direct-debit", "public-edge/share-information-with-other-responders", "public-edge/sign-the-annual-governance-statement", "public-edge/sign-the-declaration-of-acceptance-of-office", "public-edge/speak-in-a-debate-and-check-the-official-report", "public-edge/spend-under-the-section-137-power-within-the-limit", "public-edge/submit-a-motion-to-full-council", "public-edge/submit-an-innovation-proposal-to-the-defence-and-security-ac", "public-edge/submit-planning-statistics-returns", "public-edge/submit-revenue-and-capital-returns-to-mhclg", "public-edge/submit-single-source-contract-reports", "public-edge/submit-the-business-rates-forecast-and-outturn-returns", "public-edge/submit-the-homelessness-statistics-return", "public-edge/submit-the-party-s-annual-accounts", "public-edge/submit-the-provider-cost-return", "public-edge/submit-the-tenant-satisfaction-measures", "public-edge/supply-the-register-to-an-entitled-person", "public-edge/support-a-care-leaver-with-a-pathway-plan", "public-edge/table-a-motion-or-seek-a-debate", "public-edge/table-a-written-question", "public-edge/take-a-bill-through-parliament", "public-edge/take-a-financial-monitoring-item-with-decision-grade-evidenc", "public-edge/take-a-homelessness-application-and-decide-the-duty", "public-edge/take-a-key-decision-on-the-one-page-standard", "public-edge/take-a-ward-casework-enquiry-through-to-reply", "public-edge/take-an-individual-portfolio-decision", "public-edge/take-an-urgent-decision-under-special-urgency", "public-edge/take-and-publish-the-minutes", "public-edge/take-and-publish-the-minutes-2", "public-edge/take-evidence-and-agree-a-select-committee-report", "public-edge/take-notice-of-marriage-or-civil-partnership", "public-edge/take-the-six-month-progress-report-on-statutory-recommendati", "public-edge/track-audit-recommendations-to-owner-and-date-and-escalate-o", "public-edge/track-the-executive-s-responses-to-recommendations", "public-edge/transfer-staff-under-tupe", "public-edge/transfer-the-commissioning-toolkit-into-a-service", "public-edge/validate-a-planning-application", "public-edge/verify-weighing-and-measuring-equipment", "public-edge/vet-and-train-care-staff", "public-edge/view-the-public-outcome-scorecard", "public-edge/vote-in-a-division", "public-edge/warn-and-inform-the-public", "public-edge/webcast-and-allow-reporting-of-a-public-meeting", "public-edge/write-a-submission-to-a-minister", "public-edge/write-and-exercise-the-emergency-plan", "public-edge/write-off-an-irrecoverable-debt", "public-edge/write-the-officer-report-and-decide-under-delegation", "public-edge/write-to-a-public-body-on-a-constituent-s-behalf" ], "decided_by": "X-X.TECH (Owner order; recorded by Claude-Uriel 21:3531102026-1)", "date": "2026-11-01" } ``` ### Capabilities (1616 units; each has its own complete specification in section 11) | Application | Capability | Path | Screens | |---|---|---|---| | clinic | Accept a proxy mandate | `/clinic/accept-proxy-mandate` | 6 | | clinic | Acquire and send images | `/clinic/acquire-send-images` | 6 | | clinic | Act on a drug safety alert | `/clinic/act-drug-safety-alert` | 5 | | clinic | Act on a high-risk answer | `/clinic/act-high-risk-answer` | 6 | | clinic | Add an appointment to my calendar | `/clinic/add-appointment-calendar` | 5 | | clinic | Adjust a premium at renewal | `/clinic/adjust-premium-renewal` | 5 | | clinic | Administer a medicine | `/clinic/administer-medicine` | 6 | | clinic | Allocate a breach between trusts | `/clinic/allocate-breach-between-trusts` | 5 | | clinic | Annotate variants | `/clinic/annotate-variants` | 4 | | clinic | Answer a patient call | `/clinic/answer-patient-call` | 6 | | clinic | Apply a national data opt-out | `/clinic/apply-national-data-opt` | 4 | | clinic | Apply a treatment rule | `/clinic/apply-treatment-rule` | 5 | | clinic | Apply for HRA and REC approval | `/clinic/apply-hra-rec-approval` | 5 | | clinic | Approve a data release | `/clinic/approve-data-release` | 5 | | clinic | Approve a leave or swap request | `/clinic/approve-leave-swap-request` | 5 | | clinic | Approve a questionnaire or test threshold | `/clinic/approve-questionnaire-test` | 4 | | clinic | Approve a staff request | `/clinic/approve-staff-request` | 5 | | clinic | Arrange a cross-trust MDT | `/clinic/arrange-cross-trust-mdt` | 4 | | clinic | Ask for advice and guidance | `/clinic/ask-advice-guidance` | 6 | | clinic | Ask for help now | `/clinic/ask-help-now` | 5 | | clinic | Assess a safeguarding concern | `/clinic/assess-safeguarding-concern` | 6 | | clinic | Attend an appointment with the person | `/clinic/attend-appointment-person` | 4 | | clinic | Audit access to records | `/clinic/audit-access-records` | 5 | | clinic | Audit antimicrobial use | `/clinic/audit-antimicrobial-use` | 5 | | clinic | Audit consent records | `/clinic/audit-consent-records` | 4 | | clinic | Audit data use against mandates | `/clinic/audit-data-use-against` | 4 | | clinic | Audit MDT attendance | `/clinic/audit-mdt-attendance` | 4 | | clinic | Audit reporting quality | `/clinic/audit-reporting-quality` | 5 | | clinic | Audit response times | `/clinic/audit-response-times` | 4 | | clinic | Audit waiting time data | `/clinic/audit-waiting-time-data` | 5 | | clinic | Authorise a result | `/clinic/authorise-result` | 5 | | clinic | Book a pathway test | `/clinic/book-pathway-test` | 6 | | clinic | Book an appointment | `/clinic/book-appointment` | 6 | | clinic | Book an interpreter | `/clinic/book-interpreter` | 5 | | clinic | Book for the person I care for | `/clinic/book-person-i-care` | 5 | | clinic | Book from the referral queue | `/clinic/book-referral-queue` | 6 | | clinic | Book transport | `/clinic/book-transport` | 5 | | clinic | Build and publish the rota | `/clinic/build-publish-rota` | 5 | | clinic | Build the clinic template | `/clinic/build-clinic-template` | 6 | | clinic | Build the MDT list | `/clinic/build-mdt-list` | 5 | | clinic | Build the variant to trial graph | `/clinic/build-variant-trial-graph` | 8 | | clinic | Cancel a clinic session | `/clinic/cancel-clinic-session` | 6 | | clinic | Change my pharmacy | `/clinic/change-pharmacy` | 6 | | clinic | Chase a pathway clock | `/clinic/chase-pathway-clock` | 5 | | clinic | Chase missing results | `/clinic/chase-missing-results` | 5 | | clinic | Check a chemotherapy prescription | `/clinic/check-chemotherapy-prescription` | 4 | | clinic | Check access controls | `/clinic/check-access-controls` | 4 | | clinic | Check fragment size before sequencing | `/clinic/check-fragment-size-before` | 4 | | clinic | Check identity and consent before scanning | `/clinic/check-identity-consent-before` | 5 | | clinic | Check in an arriving patient | `/clinic/check-arriving-patient` | 5 | | clinic | Check in on arrival | `/clinic/check-arrival` | 5 | | clinic | Check laboratory submissions for a variant | `/clinic/check-laboratory-submissions` | 5 | | clinic | Check record accuracy | `/clinic/check-record-accuracy` | 5 | | clinic | Check research integrity | `/clinic/check-research-integrity` | 5 | | clinic | Check staff registration and training | `/clinic/check-staff-registration` | 4 | | clinic | Claim my EDGE ID and record | `/clinic/claim-edge-id-record` | 6 | | clinic | Classify a variant | `/clinic/classify-variant` | 5 | | clinic | Classify software as a medical device | `/clinic/classify-software-medical-device` | 6 | | clinic | Clinically check a prescription | `/clinic/clinically-check-prescription` | 5 | | clinic | Close a referral | `/clinic/close-referral` | 5 | | clinic | Close a study | `/clinic/close-study` | 5 | | clinic | Collect case report data | `/clinic/collect-case-report-data` | 5 | | clinic | Compare two pipeline runs | `/clinic/compare-two-pipeline-runs` | 5 | | clinic | Complete a mental health check-in | `/clinic/complete-mental-health-check` | 6 | | clinic | Complete a pre-clinic questionnaire | `/clinic/complete-pre-clinic` | 6 | | clinic | Complete a risk formulation | `/clinic/complete-risk-formulation` | 5 | | clinic | Complete mandatory training | `/clinic/complete-mandatory-training` | 6 | | clinic | Complete the Data Security and Protection Toolkit | `/clinic/complete-data-security` | 5 | | clinic | Confirm attendance and prepare | `/clinic/confirm-attendance-prepare` | 5 | | clinic | Connect a LIMS | `/clinic/connect-lims` | 5 | | clinic | Connect a PACS or imaging system | `/clinic/connect-pacs-imaging-system` | 5 | | clinic | Counsel a patient on a new medicine | `/clinic/counsel-patient-new-medicine` | 5 | | clinic | Decide a re-run | `/clinic/decide-re-run` | 4 | | clinic | Deliver cancer nurse specialist support | `/clinic/deliver-cancer-nurse-specialist` | 6 | | clinic | Deploy a release | `/clinic/deploy-release` | 7 | | clinic | Design a smart questionnaire | `/clinic/design-smart-questionnaire` | 7 | | clinic | Design a trial | `/clinic/design-trial` | 5 | | clinic | Detect claim anomalies | `/clinic/detect-claim-anomalies` | 5 | | clinic | Discharge from the pathway | `/clinic/discharge-pathway` | 5 | | clinic | Discharge from the stream | `/clinic/discharge-stream` | 5 | | clinic | Dispense a medicine | `/clinic/dispense-medicine` | 6 | | clinic | Escalate to emergency services | `/clinic/escalate-emergency-services` | 5 | | clinic | Explain a result in plain language | `/clinic/explain-result-plain-language` | 4 | | clinic | Explain a result to the patient | `/clinic/explain-result-patient` | 5 | | clinic | Find a recruiting trial near me | `/clinic/find-recruiting-trial-near` | 6 | | clinic | Flag a critical finding | `/clinic/flag-critical-finding` | 5 | | clinic | Flag, never block | `/clinic/flag-never-block` | 5 | | clinic | Follow up after discharge from crisis care | `/clinic/follow-up-after-discharge` | 4 | | clinic | Follow up an action plan | `/clinic/follow-up-action-plan` | 4 | | clinic | Forecast compliance | `/clinic/forecast-compliance` | 5 | | clinic | Forecast days to treatment | `/clinic/forecast-days-treatment` | 5 | | clinic | Front door check-in | `/clinic/front-door-check` | 7 | | clinic | Fulfil a subject access request | `/clinic/fulfil-subject-access-request` | 6 | | clinic | Give a vaccination | `/clinic/give-vaccination` | 6 | | clinic | Give consent on behalf | `/clinic/give-consent-behalf` | 5 | | clinic | Give or withdraw consent | `/clinic/give-withdraw-consent` | 6 | | clinic | Grade a retinal image | `/clinic/grade-retinal-image` | 5 | | clinic | Hand off to Codex or Jude | `/clinic/hand-off-codex-jude` | 5 | | clinic | Handle a complaint | `/clinic/handle-complaint` | 5 | | clinic | Handle a controlled drug | `/clinic/handle-controlled-drug` | 5 | | clinic | Handle a did-not-attend | `/clinic/handle-did-not-attend` | 6 | | clinic | Handle a disclosure of abuse | `/clinic/handle-disclosure-abuse` | 5 | | clinic | Handle a freedom of information request | `/clinic/handle-freedom-information` | 5 | | clinic | Handle a policyholder complaint | `/clinic/handle-policyholder-complaint` | 5 | | clinic | Handle a stock shortage | `/clinic/handle-stock-shortage` | 5 | | clinic | Handle the record after death | `/clinic/handle-record-after-death` | 5 | | clinic | Hold a remote session | `/clinic/hold-remote-session` | 5 | | clinic | Hold a slot for a tertiary handover | `/clinic/hold-slot-tertiary-handover` | 6 | | clinic | Hold the duty rota | `/clinic/hold-duty-rota` | 4 | | clinic | Import my practice register | `/clinic/import-practice-register` | 6 | | clinic | Inspect the hazard log and safety case | `/clinic/inspect-hazard-log-safety` | 4 | | clinic | Investigate a clinical safety incident | `/clinic/investigate-clinical-safety` | 6 | | clinic | Issue a clinical safety case report | `/clinic/issue-clinical-safety-case` | 4 | | clinic | Issue a field safety notice | `/clinic/issue-field-safety-notice` | 5 | | clinic | Issue a genomic report | `/clinic/issue-genomic-report` | 6 | | clinic | Issue a policy | `/clinic/issue-policy` | 5 | | clinic | Maintain laboratory accreditation evidence | `/clinic/maintain-laboratory` | 5 | | clinic | Maintain the hazard log | `/clinic/maintain-hazard-log` | 6 | | clinic | Maintain the suicide prevention plan | `/clinic/maintain-suicide-prevention-plan` | 5 | | clinic | Make a follow-up call | `/clinic/make-follow-up-call` | 5 | | clinic | Make an urgent same-day referral | `/clinic/make-urgent-same-day` | 6 | | clinic | Manage an incident | `/clinic/manage-incident` | 6 | | clinic | Manage freezer alarms | `/clinic/manage-freezer-alarms` | 6 | | clinic | Manage patient group directions | `/clinic/manage-patient-group-directions` | 5 | | clinic | Manage patient transport and interpreter contracts | `/clinic/manage-patient-transport` | 4 | | clinic | Manage retention and destruction | `/clinic/manage-retention-destruction` | 5 | | clinic | Manage room allocation | `/clinic/manage-room-allocation` | 5 | | clinic | Manage the information asset register | `/clinic/manage-information-asset` | 5 | | clinic | Manage the outbound allow-list | `/clinic/manage-outbound-allow-list` | 5 | | clinic | Manage the reporting backlog | `/clinic/manage-reporting-backlog` | 5 | | clinic | Manage the room and equipment register | `/clinic/manage-room-equipment-register` | 5 | | clinic | Manage the waiting list | `/clinic/manage-waiting-list` | 6 | | clinic | Measure a nodule and its growth | `/clinic/measure-nodule-growth` | 6 | | clinic | Measure compute, energy and carbon | `/clinic/measure-compute-energy-carbon` | 4 | | clinic | Measure outcomes over time | `/clinic/measure-outcomes-over-time` | 5 | | clinic | Medication safety check | `/clinic/medication-safety-check` | 4 | | clinic | Mental health care and crisis support | `/clinic/mental-health-care-crisis` | 12 | | clinic | Message my care team | `/clinic/message-care-team` | 6 | | clinic | Monitor alerts and escalate | `/clinic/monitor-alerts-escalate` | 4 | | clinic | Monitor data feeds | `/clinic/monitor-data-feeds` | 4 | | clinic | Monitor post-deployment safety | `/clinic/monitor-post-deployment-safety` | 4 | | clinic | Nominate a carer or proxy | `/clinic/nominate-carer-proxy` | 6 | | clinic | Obtain confirmation of capacity and capability | `/clinic/obtain-confirmation-capacity` | 4 | | clinic | Offer a cancellation slot | `/clinic/offer-cancellation-slot` | 5 | | clinic | Offer self-help resources | `/clinic/offer-self-help-resources` | 5 | | clinic | Onboard a staff member | `/clinic/onboard-staff-member` | 5 | | clinic | Onboard a wearable device type | `/clinic/onboard-wearable-device-type` | 6 | | clinic | Online urgent triage | `/clinic/online-urgent-triage` | 7 | | clinic | Order a diagnostic test | `/clinic/order-diagnostic-test` | 6 | | clinic | Order a direct access test | `/clinic/order-direct-access-test` | 6 | | clinic | Order a genomic test | `/clinic/order-genomic-test` | 5 | | clinic | Order a repeat medicine | `/clinic/order-repeat-medicine` | 6 | | clinic | Order repeat medicines on behalf | `/clinic/order-repeat-medicines-behalf` | 6 | | clinic | Pair my wearable device | `/clinic/pair-wearable-device` | 6 | | clinic | Person-held health record | `/clinic/person-held-health-record` | 11 | | clinic | Phone a critical result | `/clinic/phone-critical-result` | 5 | | clinic | Plan weekly capacity | `/clinic/plan-weekly-capacity` | 5 | | clinic | Post a receipt to the Agent Channel | `/clinic/post-receipt-agent-channel` | 4 | | clinic | Predict fields in a form | `/clinic/predict-fields-form` | 4 | | clinic | Prepare a clinic letter for a clinician | `/clinic/prepare-clinic-letter-clinician` | 5 | | clinic | Prepare a patient for imaging | `/clinic/prepare-patient-imaging` | 4 | | clinic | Prepare for a CQC inspection | `/clinic/prepare-cqc-inspection` | 5 | | clinic | Prescribe | `/clinic/prescribe` | 6 | | clinic | Prescribe under the pathway protocol | `/clinic/prescribe-under-pathway-protocol` | 5 | | clinic | Price a policy from clinic data | `/clinic/price-policy-clinic-data` | 5 | | clinic | Process a claim | `/clinic/process-claim` | 5 | | clinic | Pseudonymise a dataset | `/clinic/pseudonymise-dataset` | 5 | | clinic | Publish results | `/clinic/publish-results` | 5 | | clinic | Raise a complaint or feedback | `/clinic/raise-complaint-feedback` | 6 | | clinic | Raise a safeguarding concern from a response | `/clinic/raise-safeguarding-concern` | 5 | | clinic | Randomise a participant | `/clinic/randomise-participant` | 5 | | clinic | Read assistive output on an image | `/clinic/read-assistive-output-image` | 5 | | clinic | Rebook or cancel an appointment | `/clinic/rebook-cancel-appointment` | 6 | | clinic | Receive a crisis alert | `/clinic/receive-crisis-alert` | 6 | | clinic | Receive a crisis alert for the person | `/clinic/receive-crisis-alert-person` | 5 | | clinic | Receive a critical result for my patient | `/clinic/receive-critical-result-patient` | 5 | | clinic | Receive a did-not-attend notice | `/clinic/receive-did-not-attend` | 5 | | clinic | Receive a pre-authorisation request | `/clinic/receive-pre-authorisation` | 5 | | clinic | Receive a statutory notification | `/clinic/receive-statutory-notification` | 4 | | clinic | Receive and book in a sample | `/clinic/receive-book-sample` | 6 | | clinic | Receive the outcome letter | `/clinic/receive-outcome-letter` | 5 | | clinic | Reclassify a variant | `/clinic/reclassify-variant` | 6 | | clinic | Reconcile medicines | `/clinic/reconcile-medicines` | 5 | | clinic | Reconcile medicines at admission | `/clinic/reconcile-medicines-admission` | 5 | | clinic | Record a clinic consultation | `/clinic/record-clinic-consultation` | 6 | | clinic | Record a dose incident | `/clinic/record-dose-incident` | 5 | | clinic | Record a patient deferred | `/clinic/record-patient-deferred` | 5 | | clinic | Record a patient's own view | `/clinic/record-patients-own-view` | 5 | | clinic | Record a point-of-care urine result | `/clinic/record-point-care-urine` | 5 | | clinic | Record a safeguarding concern | `/clinic/record-safeguarding-concern` | 6 | | clinic | Record a self check-in exception | `/clinic/record-self-check-exception` | 5 | | clinic | Record a wound or skin check | `/clinic/record-wound-skin-check` | 5 | | clinic | Record an adverse drug reaction | `/clinic/record-adverse-drug-reaction` | 6 | | clinic | Record attendance outcome | `/clinic/record-attendance-outcome` | 5 | | clinic | Record capacity and best interests | `/clinic/record-capacity-best-interests` | 5 | | clinic | Record decision to treat | `/clinic/record-decision-treat` | 5 | | clinic | Record sample consent version | `/clinic/record-sample-consent-version` | 4 | | clinic | Record the faster diagnosis outcome | `/clinic/record-faster-diagnosis-outcome` | 5 | | clinic | Record the lawful basis | `/clinic/record-lawful-basis` | 4 | | clinic | Record the MDT recommendation | `/clinic/record-mdt-recommendation` | 4 | | clinic | Record trial eligibility at the MDT | `/clinic/record-trial-eligibility-mdt` | 5 | | clinic | Refer on a suspected cancer pathway | `/clinic/refer-suspected-cancer-pathway` | 8 | | clinic | Refer to a crisis team | `/clinic/refer-crisis-team` | 5 | | clinic | Refer to tertiary care | `/clinic/refer-tertiary-care` | 5 | | clinic | Refer to the mental health stream | `/clinic/refer-mental-health-stream` | 7 | | clinic | Refer to the multidisciplinary team | `/clinic/refer-multidisciplinary-team` | 4 | | clinic | Refresh public data caches | `/clinic/refresh-public-data-caches` | 4 | | clinic | Register a child as a patient | `/clinic/register-child-patient` | 6 | | clinic | Register a device fault | `/clinic/register-device-fault` | 5 | | clinic | Register a walk-in patient | `/clinic/register-walk-patient` | 6 | | clinic | Register an imaging device | `/clinic/register-imaging-device` | 5 | | clinic | Register my practice with the clinic | `/clinic/register-practice-clinic` | 5 | | clinic | Register the trial | `/clinic/register-trial` | 5 | | clinic | Reject an unlabelled sample | `/clinic/reject-unlabelled-sample` | 5 | | clinic | Report a change in the person's condition | `/clinic/report-change-persons-condition` | 6 | | clinic | Report a data breach | `/clinic/report-data-breach` | 6 | | clinic | Report a dispensing error | `/clinic/report-dispensing-error` | 5 | | clinic | Report a safety incident | `/clinic/report-safety-incident` | 5 | | clinic | Report a side effect | `/clinic/report-side-effect` | 6 | | clinic | Report a study | `/clinic/report-study` | 6 | | clinic | Report an adverse event | `/clinic/report-adverse-event` | 5 | | clinic | Report an incident | `/clinic/report-incident` | 6 | | clinic | Report histology | `/clinic/report-histology` | 6 | | clinic | Report monthly performance | `/clinic/report-monthly-performance` | 4 | | clinic | Report to the regulator | `/clinic/report-regulator` | 4 | | clinic | Report usage and efficiency | `/clinic/report-usage-efficiency` | 4 | | clinic | Request a copy of my record | `/clinic/request-copy-record` | 6 | | clinic | Request a rota swap or leave | `/clinic/request-rota-swap-leave` | 6 | | clinic | Request a second opinion | `/clinic/request-second-opinion` | 5 | | clinic | Request a second read | `/clinic/request-second-read` | 6 | | clinic | Request a second read or discrepancy review | `/clinic/request-second-read-discrepancy` | 5 | | clinic | Request a young person's own access | `/clinic/request-young-persons-own` | 5 | | clinic | Request clinic data for a quote | `/clinic/request-clinic-data-quote` | 6 | | clinic | Request evidence against fundamental standards | `/clinic/request-evidence-against` | 5 | | clinic | Retinal screening | `/clinic/retinal-screening` | 5 | | clinic | Review a breach notification | `/clinic/review-breach-notification` | 4 | | clinic | Review a death by suicide | `/clinic/review-death-suicide` | 6 | | clinic | Review a real result | `/clinic/review-real-result` | 6 | | clinic | Review a release for clinical safety | `/clinic/review-release-clinical-safety` | 5 | | clinic | Review a repeat request | `/clinic/review-repeat-request` | 5 | | clinic | Review a trial for compliance | `/clinic/review-trial-compliance` | 5 | | clinic | Review a wearable alert | `/clinic/review-wearable-alert` | 5 | | clinic | Review assistive output labelling | `/clinic/review-assistive-output` | 5 | | clinic | Review payslips and expenses approvals | `/clinic/review-payslips-expenses` | 4 | | clinic | Review the caseload | `/clinic/review-caseload` | 4 | | clinic | Review the dashboard | `/clinic/review-dashboard` | 5 | | clinic | Review the waiting list position | `/clinic/review-waiting-list-position` | 4 | | clinic | Review the worklist | `/clinic/review-worklist` | 4 | | clinic | Revoke access when a mandate ends | `/clinic/revoke-access-when-mandate` | 5 | | clinic | Revoke my own proxy role | `/clinic/revoke-own-proxy-role` | 4 | | clinic | Reward early detection participation | `/clinic/reward-early-detection` | 4 | | clinic | Roll back a release | `/clinic/roll-back-release` | 5 | | clinic | Rotate keys and certificates | `/clinic/rotate-keys-certificates` | 6 | | clinic | Run a business continuity drill | `/clinic/run-business-continuity-drill` | 5 | | clinic | Run a catalogue workload | `/clinic/run-catalogue-workload` | 5 | | clinic | Run a clinical safety training check | `/clinic/run-clinical-safety-training` | 5 | | clinic | Run a data protection impact assessment | `/clinic/run-data-protection-impact` | 6 | | clinic | Run a pipeline | `/clinic/run-pipeline` | 5 | | clinic | Run a pre-assessment | `/clinic/run-pre-assessment` | 5 | | clinic | Run a safeguarding review | `/clinic/run-safeguarding-review` | 6 | | clinic | Run a synthetic pathway drill | `/clinic/run-synthetic-pathway-drill` | 5 | | clinic | Run a urine test likelihood ledger | `/clinic/run-urine-test-likelihood` | 5 | | clinic | Run an interim analysis | `/clinic/run-interim-analysis` | 5 | | clinic | Run an online early symptom test | `/clinic/run-online-early-symptom` | 5 | | clinic | Run quality control | `/clinic/run-quality-control` | 5 | | clinic | Run scanner quality assurance | `/clinic/run-scanner-quality-assurance` | 5 | | clinic | Run sequencing QC | `/clinic/run-sequencing-qc` | 5 | | clinic | Run the day's capacity check | `/clinic/run-days-capacity-check` | 5 | | clinic | Run the MDT meeting | `/clinic/run-mdt-meeting` | 6 | | clinic | Run the training controller | `/clinic/run-training-controller` | 5 | | clinic | Run vulnerability scans | `/clinic/run-vulnerability-scans` | 5 | | clinic | Schedule the scan | `/clinic/schedule-scan` | 5 | | clinic | Score a model on a held split | `/clinic/score-model-held-split` | 5 | | clinic | Score and interpret a response | `/clinic/score-interpret-response` | 5 | | clinic | Screen a patient for eligibility | `/clinic/screen-patient-eligibility` | 4 | | clinic | Search recruiting trials | `/clinic/search-recruiting-trials` | 4 | | clinic | See my day | `/clinic/see-day` | 4 | | clinic | Send a check-in to a caseload | `/clinic/send-check-caseload` | 6 | | clinic | Send a results-ready notice | `/clinic/send-results-ready-notice` | 4 | | clinic | Send appointment letters and reminders | `/clinic/send-appointment-letters` | 5 | | clinic | Serve a vision model | `/clinic/serve-vision-model` | 6 | | clinic | Set a surveillance plan | `/clinic/set-surveillance-plan` | 5 | | clinic | Set my communication preferences | `/clinic/set-communication-preferences` | 4 | | clinic | Set my data sharing choices | `/clinic/set-data-sharing-choices` | 5 | | clinic | Set role permissions and expiry | `/clinic/set-role-permissions-expiry` | 5 | | clinic | Settle with the clinic | `/clinic/settle-clinic` | 4 | | clinic | Share a medicines list with the clinic | `/clinic/share-medicines-list-clinic` | 5 | | clinic | Share a wearable signal | `/clinic/share-wearable-signal` | 6 | | clinic | Share clinic data with my insurer | `/clinic/share-clinic-data-insurer` | 6 | | clinic | Share images with another trust | `/clinic/share-images-another-trust` | 6 | | clinic | Share risk information with partners | `/clinic/share-risk-information-partners` | 5 | | clinic | Sign a clinic letter | `/clinic/sign-clinic-letter` | 5 | | clinic | Sign the clinical risk management plan | `/clinic/sign-clinical-risk-management` | 4 | | clinic | Simulate trial monitoring | `/clinic/simulate-trial-monitoring` | 4 | | clinic | Social prescribing referral | `/clinic/social-prescribing-referral` | 7 | | clinic | Stop on revoked authority | `/clinic/stop-revoked-authority` | 4 | | clinic | Store and retrieve genomic data | `/clinic/store-retrieve-genomic-data` | 4 | | clinic | Submit a cancer registration | `/clinic/submit-cancer-registration` | 5 | | clinic | Submit cancer waiting times data | `/clinic/submit-cancer-waiting-times` | 5 | | clinic | Submit COSD and other datasets | `/clinic/submit-cosd-other-datasets` | 5 | | clinic | Submit MDT data | `/clinic/submit-mdt-data` | 5 | | clinic | Submit the mental health dataset | `/clinic/submit-mental-health-dataset` | 5 | | clinic | Supervise and sign off a trainee | `/clinic/supervise-sign-off-trainee` | 5 | | clinic | Support deploying organisations | `/clinic/support-deploying-organisations` | 4 | | clinic | Symptom check and triage | `/clinic/symptom-check-triage` | 6 | | clinic | Take a blood or urine sample | `/clinic/take-blood-urine-sample` | 7 | | clinic | Take a mandate ticket | `/clinic/take-mandate-ticket` | 5 | | clinic | Take an online early symptom test | `/clinic/take-online-early-symptom` | 6 | | clinic | Take informed consent for research | `/clinic/take-informed-consent-research` | 6 | | clinic | Take observations | `/clinic/take-observations` | 6 | | clinic | Take part in external quality assessment | `/clinic/take-part-external-quality` | 5 | | clinic | Tell the patient a result | `/clinic/tell-patient-result` | 5 | | clinic | Test a hypothesis on an open cohort | `/clinic/test-hypothesis-open-cohort` | 5 | | clinic | Test a pre-registered hypothesis | `/clinic/test-pre-registered-hypothesis` | 5 | | clinic | Trace ingress and egress | `/clinic/trace-ingress-egress` | 4 | | clinic | Track a referral | `/clinic/track-referral` | 4 | | clinic | Track actions from the MDT | `/clinic/track-actions-mdt` | 5 | | clinic | Track histology turnaround | `/clinic/track-histology-turnaround` | 5 | | clinic | Track my cancer pathway | `/clinic/track-cancer-pathway` | 5 | | clinic | Track sample custody | `/clinic/track-sample-custody` | 4 | | clinic | Track the 62-day clock across the MDT | `/clinic/track-62-day-clock` | 4 | | clinic | Train staff in information governance | `/clinic/train-staff-information` | 4 | | clinic | Train staff in suicide awareness | `/clinic/train-staff-suicide-awareness` | 4 | | clinic | Triage a mental health referral | `/clinic/triage-mental-health-referral` | 6 | | clinic | Triage a patient message | `/clinic/triage-patient-message` | 5 | | clinic | Triage a referral | `/clinic/triage-referral` | 6 | | clinic | Triage inbox items for a clinician | `/clinic/triage-inbox-items-clinician` | 4 | | clinic | Tune alert thresholds | `/clinic/tune-alert-thresholds` | 5 | | clinic | Update patient demographics | `/clinic/update-patient-demographics` | 5 | | clinic | Validate a new assay | `/clinic/validate-new-assay` | 6 | | clinic | Validate the test against outcomes | `/clinic/validate-test-against-outcomes` | 5 | | clinic | Verify a device classification | `/clinic/verify-device-classification` | 4 | | clinic | Verify an applicant's identity | `/clinic/verify-applicants-identity` | 4 | | clinic | Verify an evidence statement | `/clinic/verify-evidence-statement` | 4 | | clinic | Verify backups and restore | `/clinic/verify-backups-restore` | 5 | | clinic | Vet an imaging request | `/clinic/vet-imaging-request` | 5 | | clinic | View my medicines list | `/clinic/view-medicines-list` | 5 | | clinic | View my results | `/clinic/view-results` | 5 | | clinic | View results in scope | `/clinic/view-results-scope` | 4 | | clinic | Wearable and home device signals | `/clinic/wearable-home-device-signals` | 5 | | clinic | Wellbeing questionnaires | `/clinic/wellbeing-questionnaires` | 6 | | clinic | Write a care plan | `/clinic/write-care-plan` | 6 | | consortium | Accept a review assignment | `/consortium/accept-review-assignment` | 6 | | consortium | Accept supervision of a learner | `/consortium/accept-supervision-learner` | 5 | | consortium | Acknowledge the principles version | `/consortium/acknowledge-principles-version` | 5 | | consortium | Add Jude to a channel under mandate | `/consortium/add-jude-channel-under` | 5 | | consortium | Add members and guests with expiry | `/consortium/add-members-guests-expiry` | 6 | | consortium | Administer a database | `/consortium/administer-database` | 6 | | consortium | Admit or refuse a guest | `/consortium/admit-refuse-guest` | 4 | | consortium | Agree a data sharing agreement | `/consortium/agree-data-sharing-agreement` | 5 | | consortium | Allocate indirect costs | `/consortium/allocate-indirect-costs` | 5 | | consortium | Answer a reference request | `/consortium/answer-reference-request` | 6 | | consortium | Answer reviewer questions on a contribution | `/consortium/answer-reviewer-questions` | 6 | | consortium | Answer with sources in a channel | `/consortium/answer-sources-channel` | 5 | | consortium | Apply a non-disclosure agreement to a room | `/consortium/apply-non-disclosure-agreement` | 5 | | consortium | Apply an open licence | `/consortium/apply-open-licence` | 6 | | consortium | Apply for ethics approval | `/consortium/apply-ethics-approval` | 6 | | consortium | Apply retention and deletion to records | `/consortium/apply-retention-deletion-records` | 6 | | consortium | Apply to become a verified reviewer | `/consortium/apply-become-verified-reviewer` | 6 | | consortium | Approve a compute request | `/consortium/approve-compute-request` | 5 | | consortium | Approve a connector or integration | `/consortium/approve-connector-integration` | 5 | | consortium | Approve a contribution before submission | `/consortium/approve-contribution-before` | 5 | | consortium | Approve a director passkey action | `/consortium/approve-director-passkey-action` | 5 | | consortium | Approve a federated query request | `/consortium/approve-federated-query-request` | 5 | | consortium | Approve a guest account | `/consortium/approve-guest-account` | 6 | | consortium | Approve a reviewer payment batch | `/consortium/approve-reviewer-payment-batch` | 5 | | consortium | Approve a team member data access request | `/consortium/approve-team-member-data` | 5 | | consortium | Approve a visiting researcher affiliation | `/consortium/approve-visiting-researcher` | 6 | | consortium | Approve an agent consequential action | `/consortium/approve-agent-consequential` | 5 | | consortium | Approve an open-source release | `/consortium/approve-open-source-release` | 6 | | consortium | Approve and mark a contribution verified | `/consortium/approve-mark-contribution` | 5 | | consortium | Approve compute spend against a grant | `/consortium/approve-compute-spend-against` | 5 | | consortium | Approve expenditure against a grant | `/consortium/approve-expenditure-against` | 5 | | consortium | Archive a channel | `/consortium/archive-channel` | 5 | | consortium | Archive a module version | `/consortium/archive-module-version` | 6 | | consortium | Ask a clarification question | `/consortium/ask-clarification-question` | 4 | | consortium | Ask Jude for a sourced answer | `/consortium/ask-jude-sourced-answer` | 6 | | consortium | Assess ownership of an output | `/consortium/assess-ownership-output` | 5 | | consortium | Assign a persistent identifier | `/consortium/assign-persistent-identifier` | 5 | | consortium | Assignments that are real projects, sandboxed | `/consortium/assignments-real-projects` | 7 | | consortium | Attribute carbon to a project | `/consortium/attribute-carbon-project` | 5 | | consortium | Attribute cost to a grant and project | `/consortium/attribute-cost-grant-project` | 5 | | consortium | Audit reviewer quality by re-review | `/consortium/audit-reviewer-quality-re` | 5 | | consortium | Audit the signing log | `/consortium/audit-signing-log` | 5 | | consortium | Award and publish the contract award notice | `/consortium/award-publish-contract-notice` | 6 | | consortium | Bind a device | `/consortium/bind-device` | 6 | | consortium | Book a room or resource | `/consortium/book-room-resource` | 5 | | consortium | Build a project team across universities | `/consortium/build-project-team-across` | 6 | | consortium | Build an interactive prototype | `/consortium/build-interactive-prototype` | 6 | | consortium | Build the research profile and skills record | `/consortium/build-research-profile-skills` | 7 | | consortium | Bulk enrol students and staff | `/consortium/bulk-enrol-students-staff` | 7 | | consortium | Catalogue a library item | `/consortium/catalogue-library-item` | 6 | | consortium | Charge compute and storage to grant codes | `/consortium/charge-compute-storage-grant` | 5 | | consortium | Chase an overdue review | `/consortium/chase-overdue-review` | 5 | | consortium | Check exclusion grounds | `/consortium/check-exclusion-grounds` | 5 | | consortium | Choose a specialism after the whole map | `/consortium/choose-specialism-after-whole` | 7 | | consortium | Claim payment for a completed review | `/consortium/claim-payment-completed-review` | 5 | | consortium | Classify data sensitivity | `/consortium/classify-data-sensitivity` | 5 | | consortium | Close a grant | `/consortium/close-grant` | 6 | | consortium | Close a project and archive its records | `/consortium/close-project-archive-records` | 6 | | consortium | Co-edit a Markdown in real time | `/consortium/co-edit-markdown-real` | 5 | | consortium | Complete a right-to-work check | `/consortium/complete-right-work-check` | 5 | | consortium | Complete research integrity training | `/consortium/complete-research-integrity` | 5 | | consortium | Compose an image in the imaging studio | `/consortium/compose-image-imaging-studio` | 5 | | consortium | Compose audio in Judio | `/consortium/compose-audio-judio` | 6 | | consortium | Conduct a probation review | `/consortium/conduct-probation-review` | 5 | | consortium | Confirm a delivery | `/consortium/confirm-delivery` | 5 | | consortium | Confirm a person's enrolment status | `/consortium/confirm-persons-enrolment-status` | 5 | | consortium | Confirm a time and send the invitation | `/consortium/confirm-time-send-invitation` | 5 | | consortium | Contribute a library entry as an outsider | `/consortium/contribute-library-entry` | 5 | | consortium | Correct a student record | `/consortium/correct-student-record` | 6 | | consortium | Cost a grant application | `/consortium/cost-grant-application` | 6 | | consortium | Count agents and subagents per task | `/consortium/count-agents-subagents-per` | 5 | | consortium | Create a contract from a template | `/consortium/create-contract-template` | 6 | | consortium | Create a job role and vacancy | `/consortium/create-job-role-vacancy` | 5 | | consortium | Create a Markdown document | `/consortium/create-markdown-document` | 6 | | consortium | Create a module | `/consortium/create-module` | 6 | | consortium | Create a team and its channels | `/consortium/create-team-channels` | 5 | | consortium | Create an availability poll | `/consortium/create-availability-poll` | 5 | | consortium | Curate a collection | `/consortium/curate-collection` | 5 | | consortium | Declare a conflict of interest | `/consortium/declare-conflict-interest` | 5 | | consortium | Declare exclusion grounds and conflicts | `/consortium/declare-exclusion-grounds` | 6 | | consortium | Decline an assignment | `/consortium/decline-assignment` | 5 | | consortium | Define a purpose view | `/consortium/define-purpose-view` | 5 | | consortium | Define or change a schema | `/consortium/define-change-schema` | 7 | | consortium | Delete data at retention end | `/consortium/delete-data-retention-end` | 5 | | consortium | Deploy a release | `/consortium/deploy-release` | 6 | | consortium | Deposit a dataset with a persistent identifier | `/consortium/deposit-dataset-persistent` | 7 | | consortium | Deprecate a feature | `/consortium/deprecate-feature` | 5 | | consortium | Detect unlicensed reuse | `/consortium/detect-unlicensed-reuse` | 5 | | consortium | Draw on JudeCanvas | `/consortium/draw-judecanvas` | 5 | | consortium | Edit a video | `/consortium/edit-video` | 6 | | consortium | End employment and revoke access | `/consortium/end-employment-revoke-access` | 6 | | consortium | Enforce the code of conduct | `/consortium/enforce-code-conduct` | 5 | | consortium | Enrol on a doctoral training module | `/consortium/enrol-doctoral-training-module` | 6 | | consortium | Escalate a dispute to the editor | `/consortium/escalate-dispute-editor` | 5 | | consortium | Evaluate bids with a scoring panel | `/consortium/evaluate-bids-scoring-panel` | 5 | | consortium | Expire or revoke a licence | `/consortium/expire-revoke-licence` | 6 | | consortium | Explore the knowledge graph visually | `/consortium/explore-knowledge-graph-visually` | 6 | | consortium | Export a creation with licence and attribution | `/consortium/export-creation-licence` | 6 | | consortium | Export a meeting record and audit chain | `/consortium/export-meeting-record-audit` | 4 | | consortium | Export and verify an audit chain | `/consortium/export-verify-audit-chain` | 5 | | consortium | Find a time across attendees | `/consortium/find-time-across-attendees` | 5 | | consortium | Find and follow tender opportunities | `/consortium/find-follow-tender-opportunities` | 5 | | consortium | Find funding calls | `/consortium/find-funding-calls` | 6 | | consortium | Flag a concern without blocking | `/consortium/flag-concern-without-blocking` | 4 | | consortium | Flag plagiarism or an integrity concern | `/consortium/flag-plagiarism-integrity` | 5 | | consortium | From student to long-term contractor | `/consortium/student-long-term-contractor` | 6 | | consortium | Grant a clearance with expiry | `/consortium/grant-clearance-expiry` | 5 | | consortium | Grant a licence to a third party | `/consortium/grant-licence-third-party` | 5 | | consortium | Grant a maintainer role with expiry | `/consortium/grant-maintainer-role-expiry` | 5 | | consortium | Hand over to another agent | `/consortium/hand-over-another-agent` | 4 | | consortium | Hand over work with the Baton | `/consortium/hand-over-work-baton` | 5 | | consortium | Handle a contract dispute | `/consortium/handle-contract-dispute` | 5 | | consortium | Handle a data subject access request | `/consortium/handle-data-subject-access` | 6 | | consortium | Handle an academic appeal on marks | `/consortium/handle-academic-appeal-marks` | 6 | | consortium | Handle an appeal | `/consortium/handle-appeal` | 6 | | consortium | Hold an audit interview by secure call | `/consortium/hold-audit-interview-secure` | 5 | | consortium | Import external records | `/consortium/import-external-records` | 6 | | consortium | Inspect carbon and energy records | `/consortium/inspect-carbon-energy-records` | 5 | | consortium | Inspect release gates and evidence | `/consortium/inspect-release-gates-evidence` | 5 | | consortium | Inspect retention schedule compliance | `/consortium/inspect-retention-schedule` | 5 | | consortium | Inspect reviewer payment records | `/consortium/inspect-reviewer-payment-records` | 5 | | consortium | International technical students | `/consortium/international-technical-students` | 7 | | consortium | Issue a certificate of membership | `/consortium/issue-certificate-membership` | 5 | | consortium | Issue a takedown or attribution request | `/consortium/issue-takedown-attribution` | 5 | | consortium | Issue a transcript of completed modules | `/consortium/issue-transcript-completed` | 5 | | consortium | Issue an employment contract | `/consortium/issue-employment-contract` | 6 | | consortium | Issue an offer letter | `/consortium/issue-offer-letter` | 5 | | consortium | Issue findings and track remediation | `/consortium/issue-findings-track-remediation` | 5 | | consortium | Issue the decision letter | `/consortium/issue-decision-letter` | 6 | | consortium | Issue the written statement of particulars | `/consortium/issue-written-statement` | 4 | | consortium | Join a secure call through the lobby | `/consortium/join-secure-call-through` | 6 | | consortium | Join Consortium through a university | `/consortium/join-consortium-through` | 6 | | consortium | Judiverse included in Edge Consortium tuition | `/consortium/judiverse-included-edge` | 5 | | consortium | Link ORCID to the profile | `/consortium/link-orcid-profile` | 7 | | consortium | Log a receipt for every action | `/consortium/log-receipt-every-action` | 4 | | consortium | Make a below-threshold direct award | `/consortium/make-below-threshold-direct` | 6 | | consortium | Manage a controlled vocabulary | `/consortium/manage-controlled-vocabulary` | 6 | | consortium | Manage a grant extension or variation | `/consortium/manage-grant-extension-variation` | 6 | | consortium | Manage an embargo and release date | `/consortium/manage-embargo-release-date` | 6 | | consortium | Manage consent records | `/consortium/manage-consent-records` | 5 | | consortium | Manage institutional roles and delegates | `/consortium/manage-institutional-roles` | 5 | | consortium | Map a dataset to the knowledge graph | `/consortium/map-dataset-knowledge-graph` | 6 | | consortium | Map a module to a challenge specification | `/consortium/map-module-challenge` | 6 | | consortium | Map a module to programme and standard | `/consortium/map-module-programme-standard` | 5 | | consortium | Mark and moderate submissions | `/consortium/mark-moderate-submissions` | 6 | | consortium | Match reviewers to a submission | `/consortium/match-reviewers-submission` | 6 | | consortium | Measure live energy per job and service | `/consortium/measure-live-energy-per` | 5 | | consortium | Merge duplicate entities | `/consortium/merge-duplicate-entities` | 6 | | consortium | Merge with required checks | `/consortium/merge-required-checks` | 5 | | consortium | Mirror a repository to the source repository | `/consortium/mirror-repository-source` | 5 | | consortium | Moderate a post | `/consortium/moderate-post` | 5 | | consortium | Monitor data quality | `/consortium/monitor-data-quality` | 5 | | consortium | Monitor job progress and output | `/consortium/monitor-job-progress-output` | 4 | | consortium | Monitor tokens, prompts and time to evidence | `/consortium/monitor-tokens-prompts-time` | 5 | | consortium | Negotiate clauses with the counterparty | `/consortium/negotiate-clauses-counterparty` | 6 | | consortium | Nominate a learner for reviewer training | `/consortium/nominate-learner-reviewer` | 5 | | consortium | Offer the degree with a partner institution | `/consortium/offer-degree-partner-institution` | 7 | | consortium | Onboard a new starter | `/consortium/onboard-new-starter` | 5 | | consortium | Open an issue | `/consortium/open-issue` | 5 | | consortium | Paid student work, offset against tuition | `/consortium/paid-student-work-offset` | 7 | | consortium | Pass reviewer credential verification | `/consortium/pass-reviewer-credential` | 5 | | consortium | Pause or pre-empt a job | `/consortium/pause-pre-empt-job` | 6 | | consortium | Pay reviewer fees | `/consortium/pay-reviewer-fees` | 5 | | consortium | Pay salary | `/consortium/pay-salary` | 6 | | consortium | Petition for student finance | `/consortium/petition-student-finance` | 6 | | consortium | Pin, edit or delete a post | `/consortium/pin-edit-delete-post` | 5 | | consortium | Plan a maintenance window | `/consortium/plan-maintenance-window` | 6 | | consortium | Plan a procurement and publish pipeline notice | `/consortium/plan-procurement-publish` | 6 | | consortium | Post an announcement | `/consortium/post-announcement` | 5 | | consortium | Pre-register a study | `/consortium/pre-register-study` | 5 | | consortium | Prepare a review summary for a reviewer | `/consortium/prepare-review-summary-reviewer` | 5 | | consortium | Prepare an employment contract template | `/consortium/prepare-employment-contract` | 6 | | consortium | Preserve and migrate file formats | `/consortium/preserve-migrate-file-formats` | 6 | | consortium | Process a supplier invoice | `/consortium/process-supplier-invoice` | 6 | | consortium | Produce a carbon reduction plan | `/consortium/produce-carbon-reduction-plan` | 6 | | consortium | Produce a graphic in Visual Studio | `/consortium/produce-graphic-visual-studio` | 6 | | consortium | Produce a statutory data return | `/consortium/produce-statutory-data-return` | 6 | | consortium | Propose a change from a fork | `/consortium/propose-change-fork` | 6 | | consortium | Propose a real-world problem project | `/consortium/propose-real-world-problem` | 6 | | consortium | Provide a funder audit trail | `/consortium/provide-funder-audit-trail` | 5 | | consortium | Provide a reference | `/consortium/provide-reference` | 6 | | consortium | Publish a data dictionary | `/consortium/publish-data-dictionary` | 5 | | consortium | Publish a Markdown as deck or book | `/consortium/publish-markdown-deck-book` | 6 | | consortium | Publish a new dataset version | `/consortium/publish-new-dataset-version` | 6 | | consortium | Publish a tender notice | `/consortium/publish-tender-notice` | 5 | | consortium | Publish a terms version and require re-acceptance | `/consortium/publish-terms-version-require` | 6 | | consortium | Publish an open access version | `/consortium/publish-open-access-version` | 6 | | consortium | Publish documentation | `/consortium/publish-documentation` | 5 | | consortium | Publish module materials to the library | `/consortium/publish-module-materials-library` | 6 | | consortium | Publish the catalogue to open indexes | `/consortium/publish-catalogue-open-indexes` | 5 | | consortium | Publish the contract details notice | `/consortium/publish-contract-details-notice` | 5 | | consortium | Publish the monthly carbon and cost statement | `/consortium/publish-monthly-carbon-cost` | 5 | | consortium | Raise a challenge or complaint | `/consortium/raise-challenge-complaint` | 6 | | consortium | Raise a grievance | `/consortium/raise-grievance` | 6 | | consortium | Raise a progress or wellbeing concern | `/consortium/raise-progress-wellbeing-concern` | 5 | | consortium | Rank agents and models by efficiency | `/consortium/rank-agents-models-efficiency` | 5 | | consortium | Read a verified entry with its provenance | `/consortium/read-verified-entry-provenance` | 6 | | consortium | Receive a mandate | `/consortium/receive-mandate` | 5 | | consortium | Receive bids in a sealed queue | `/consortium/receive-bids-sealed-queue` | 5 | | consortium | Receive the award outcome | `/consortium/receive-award-outcome` | 5 | | consortium | Receive time-boxed auditor access | `/consortium/receive-time-boxed-auditor` | 5 | | consortium | Reconcile monthly spend | `/consortium/reconcile-monthly-spend` | 5 | | consortium | Reconcile provider invoices with metered use | `/consortium/reconcile-provider-invoices` | 5 | | consortium | Reconcile with the student record system | `/consortium/reconcile-student-record-system` | 6 | | consortium | Record a call with consent | `/consortium/record-call-consent` | 5 | | consortium | Record a conflict of interest assessment | `/consortium/record-conflict-interest` | 5 | | consortium | Record a data processing agreement | `/consortium/record-data-processing-agreement` | 6 | | consortium | Record a priority date | `/consortium/record-priority-date` | 4 | | consortium | Record a trademark and brand use rule | `/consortium/record-trademark-brand-use` | 5 | | consortium | Record an invention disclosure | `/consortium/record-invention-disclosure` | 6 | | consortium | Record applications and outcomes in a portfolio | `/consortium/record-applications-outcomes` | 6 | | consortium | Record attendance and engagement | `/consortium/record-attendance-engagement` | 4 | | consortium | Record authorship and contributor roles | `/consortium/record-authorship-contributor` | 6 | | consortium | Record background and foreground IP | `/consortium/record-background-foreground-ip` | 5 | | consortium | Record copyright and attribution | `/consortium/record-copyright-attribution` | 5 | | consortium | Record energy per job | `/consortium/record-energy-per-job` | 5 | | consortium | Record payment compliance | `/consortium/record-payment-compliance` | 5 | | consortium | Record provenance and lineage | `/consortium/record-provenance-lineage` | 5 | | consortium | Record working hours and leave | `/consortium/record-working-hours-leave` | 5 | | consortium | Refuse an action outside the mandate | `/consortium/refuse-action-outside-mandate` | 5 | | consortium | Register a compute resource | `/consortium/register-compute-resource` | 6 | | consortium | Register a dataset in the catalogue | `/consortium/register-dataset-catalogue` | 6 | | consortium | Register a delegated authority | `/consortium/register-delegated-authority` | 6 | | consortium | Register a research project with an ID | `/consortium/register-research-project-id` | 7 | | consortium | Register a university as a member institution | `/consortium/register-university-member` | 6 | | consortium | Register as a supplier | `/consortium/register-supplier` | 6 | | consortium | Register with the Office for Students | `/consortium/register-office-students` | 7 | | consortium | Reject a contribution with reasons | `/consortium/reject-contribution-reasons` | 5 | | consortium | Release a version | `/consortium/release-version` | 6 | | consortium | Release marks and feedback | `/consortium/release-marks-feedback` | 5 | | consortium | Release results with disclosure control | `/consortium/release-results-disclosure` | 5 | | consortium | Renew annual institutional membership | `/consortium/renew-annual-institutional` | 6 | | consortium | Renew reviewer verification | `/consortium/renew-reviewer-verification` | 5 | | consortium | Report a personal data breach | `/consortium/report-personal-data-breach` | 6 | | consortium | Report a research misconduct concern | `/consortium/report-research-misconduct` | 5 | | consortium | Report carbon intensity per accepted outcome | `/consortium/report-carbon-intensity-per` | 4 | | consortium | Report expenditure to the funder | `/consortium/report-expenditure-funder` | 6 | | consortium | Report monitoring measures | `/consortium/report-monitoring-measures` | 4 | | consortium | Report performance indicator evidence | `/consortium/report-performance-indicator` | 5 | | consortium | Report review throughput and times | `/consortium/report-review-throughput-times` | 4 | | consortium | Report tokens, prompts and time to evidence | `/consortium/report-tokens-prompts-time` | 5 | | consortium | Report utilisation | `/consortium/report-utilisation` | 4 | | consortium | Request a compute allowance for a project | `/consortium/request-compute-allowance` | 6 | | consortium | Request a purpose view on federated data | `/consortium/request-purpose-view-federated` | 6 | | consortium | Request a Sandbox with GPU and data | `/consortium/request-sandbox-gpu-data` | 5 | | consortium | Request approval for a consequential action | `/consortium/request-approval-consequential` | 5 | | consortium | Request changes from the contributor | `/consortium/request-changes-contributor` | 5 | | consortium | Request supervisor allocation | `/consortium/request-supervisor-allocation` | 6 | | consortium | Research and development partnership | `/consortium/research-development-partnership` | 7 | | consortium | Research consortium | `/consortium/research-consortium` | 9 | | consortium | Reserve capacity for a deadline | `/consortium/reserve-capacity-deadline` | 5 | | consortium | Resolve conflicting reviews | `/consortium/resolve-conflicting-reviews` | 5 | | consortium | Respond to a security incident | `/consortium/respond-security-incident` | 7 | | consortium | Respond to an invitation | `/consortium/respond-invitation` | 4 | | consortium | Retract or correct a verified entry | `/consortium/retract-correct-verified-entry` | 6 | | consortium | Retry a failed job | `/consortium/retry-failed-job` | 6 | | consortium | Review a contribution against criteria | `/consortium/review-contribution-against` | 6 | | consortium | Review a funder expenditure statement | `/consortium/review-funder-expenditure` | 4 | | consortium | Review a pull request | `/consortium/review-pull-request` | 5 | | consortium | Review access logs and anomalies | `/consortium/review-access-logs-anomalies` | 5 | | consortium | Review agent receipts | `/consortium/review-agent-receipts` | 5 | | consortium | Review module evaluation feedback | `/consortium/review-module-evaluation` | 5 | | consortium | Review quality and standards evidence | `/consortium/review-quality-standards` | 4 | | consortium | Review the DPIA and record of processing | `/consortium/review-dpia-record-processing` | 5 | | consortium | Revoke a purpose view | `/consortium/revoke-purpose-view` | 5 | | consortium | Revoke access for a leaver | `/consortium/revoke-access-leaver` | 5 | | consortium | Rotate keys | `/consortium/rotate-keys` | 6 | | consortium | Run a clarification round | `/consortium/run-clarification-round` | 4 | | consortium | Run a compute job under mandate | `/consortium/run-compute-job-under` | 5 | | consortium | Run a cyber recovery drill | `/consortium/run-cyber-recovery-drill` | 6 | | consortium | Run a live assessment with two examiners and a recording | `/consortium/run-live-assessment-two` | 7 | | consortium | Run a live seminar | `/consortium/run-live-seminar` | 6 | | consortium | Run a security-classified procurement | `/consortium/run-security-classified` | 6 | | consortium | Run a training job on scheduled compute | `/consortium/run-training-job-scheduled` | 7 | | consortium | Run a vulnerability scan on a release | `/consortium/run-vulnerability-scan-release` | 6 | | consortium | Run an approved query where data lives | `/consortium/run-approved-query-where` | 5 | | consortium | Run code in the Sandbox | `/consortium/run-code-sandbox` | 6 | | consortium | Run data quality checks on a database | `/consortium/run-data-quality-checks` | 6 | | consortium | Sample verified entries for quality | `/consortium/sample-verified-entries-quality` | 5 | | consortium | Save sources to a reading list | `/consortium/save-sources-reading-list` | 5 | | consortium | Scale with an external provider | `/consortium/scale-external-provider` | 6 | | consortium | Schedule a job | `/consortium/schedule-job` | 5 | | consortium | Schedule a meeting on request | `/consortium/schedule-meeting-request` | 5 | | consortium | Schedule a secure government resource | `/consortium/schedule-secure-government` | 6 | | consortium | Schedule a send and set a signature | `/consortium/schedule-send-set-signature` | 5 | | consortium | Schedule and record supervision meetings | `/consortium/schedule-record-supervision` | 6 | | consortium | Schedule deliveries and logistics | `/consortium/schedule-deliveries-logistics` | 6 | | consortium | Search across mail, channels and meetings | `/consortium/search-across-mail-channels` | 4 | | consortium | Search libraries and the knowledge graph | `/consortium/search-libraries-knowledge-graph` | 5 | | consortium | Send a contract for signature | `/consortium/send-contract-signature` | 5 | | consortium | Send mail from an estate mailbox | `/consortium/send-mail-estate-mailbox` | 6 | | consortium | Set a carbon budget and threshold | `/consortium/set-carbon-budget-threshold` | 6 | | consortium | Set a project access control policy | `/consortium/set-project-access-control` | 5 | | consortium | Set agent authority scope | `/consortium/set-agent-authority-scope` | 5 | | consortium | Set and publish key performance indicators | `/consortium/set-publish-key-performance` | 5 | | consortium | Set assessment and marking scheme | `/consortium/set-assessment-marking-scheme` | 5 | | consortium | Set contract parties, permissions and expiry | `/consortium/set-contract-parties-permissions` | 6 | | consortium | Set evaluation criteria and weightings | `/consortium/set-evaluation-criteria` | 5 | | consortium | Set mail rules, flags and snooze | `/consortium/set-mail-rules-flags` | 5 | | consortium | Set mention and tag rules | `/consortium/set-mention-tag-rules` | 4 | | consortium | Set out of office and working hours | `/consortium/set-out-office-working` | 5 | | consortium | Set permissions on project records | `/consortium/set-permissions-project-records` | 6 | | consortium | Set presence | `/consortium/set-presence` | 4 | | consortium | Set project governance | `/consortium/set-project-governance` | 5 | | consortium | Set quotas per project | `/consortium/set-quotas-per-project` | 5 | | consortium | Set renewal and expiry alerts | `/consortium/set-renewal-expiry-alerts` | 5 | | consortium | Set retention and export a channel | `/consortium/set-retention-export-channel` | 5 | | consortium | Set review criteria for a field | `/consortium/set-review-criteria-field` | 6 | | consortium | Set review fields and availability | `/consortium/set-review-fields-availability` | 6 | | consortium | Set scheduling policy and queues | `/consortium/set-scheduling-policy-queues` | 5 | | consortium | Set the access tier on an item | `/consortium/set-access-tier-item` | 6 | | consortium | Set the review deadline and reminders | `/consortium/set-review-deadline-reminders` | 5 | | consortium | Set up an awarded grant | `/consortium/set-up-awarded-grant` | 6 | | consortium | Share a calendar with a delegate | `/consortium/share-calendar-delegate` | 5 | | consortium | Share a campus with a partner university | `/consortium/share-campus-partner-university` | 6 | | consortium | Share a creation in the JudeBox exchange | `/consortium/share-creation-judebox-exchange` | 5 | | consortium | Share a document in a call | `/consortium/share-document-call` | 5 | | consortium | Share a workspace folder across universities | `/consortium/share-workspace-folder-across` | 6 | | consortium | Shift jobs to low-carbon windows | `/consortium/shift-jobs-low-carbon` | 6 | | consortium | Show undisclosed figures as gaps | `/consortium/show-undisclosed-figures-gaps` | 4 | | consortium | Sign a collaboration agreement | `/consortium/sign-collaboration-agreement` | 5 | | consortium | Sign a contract with EDGE ID | `/consortium/sign-contract-edge-id` | 6 | | consortium | Sign off a progression milestone | `/consortium/sign-off-progression-milestone` | 6 | | consortium | Sign the contributor agreement | `/consortium/sign-contributor-agreement` | 5 | | consortium | Sign the employment contract | `/consortium/sign-employment-contract` | 6 | | consortium | Sign the public contract | `/consortium/sign-public-contract` | 5 | | consortium | Stop or kill a job | `/consortium/stop-kill-job` | 6 | | consortium | Store the signed contract in the Vault | `/consortium/store-signed-contract-vault` | 5 | | consortium | Study the governed intelligence degree | `/consortium/study-governed-intelligence` | 9 | | consortium | Submit a bid | `/consortium/submit-bid` | 6 | | consortium | Submit a contribution for review | `/consortium/submit-contribution-review` | 7 | | consortium | Submit a grant application | `/consortium/submit-grant-application` | 6 | | consortium | Submit a paper to Consortium review | `/consortium/submit-paper-consortium-review` | 6 | | consortium | Submit an annual report to the funder | `/consortium/submit-annual-report-funder` | 6 | | consortium | Submit an invoice and track payment | `/consortium/submit-invoice-track-payment` | 5 | | consortium | Submit staff for security vetting | `/consortium/submit-staff-security-vetting` | 6 | | consortium | Suspend or end a membership | `/consortium/suspend-end-membership` | 6 | | consortium | Terminate a contract | `/consortium/terminate-contract` | 6 | | consortium | Terminate a contract and publish the notice | `/consortium/terminate-contract-publish` | 6 | | consortium | Trace network ingress and egress | `/consortium/trace-network-ingress-egress` | 5 | | consortium | Track research outputs for a REF return | `/consortium/track-research-outputs-ref` | 6 | | consortium | Track royalties and revenue share | `/consortium/track-royalties-revenue-share` | 5 | | consortium | Triage a security vulnerability report | `/consortium/triage-security-vulnerability` | 6 | | consortium | Triage the submissions queue | `/consortium/triage-submissions-queue` | 6 | | consortium | University membership and enrolment | `/consortium/university-membership-enrolment` | 11 | | consortium | Update supplier information | `/consortium/update-supplier-information` | 5 | | consortium | Use the Terminal | `/consortium/use-terminal` | 5 | | consortium | Vary a contract | `/consortium/vary-contract` | 5 | | consortium | Verify a fact before stating it | `/consortium/verify-fact-before-stating` | 5 | | consortium | Verify a fact in the knowledge graph | `/consortium/verify-fact-knowledge-graph` | 6 | | consortium | Verify a person's identity | `/consortium/verify-persons-identity` | 6 | | consortium | Verify a signature | `/consortium/verify-signature` | 4 | | consortium | Verify institutional accreditation status | `/consortium/verify-institutional` | 5 | | consortium | Verify procurement notice publication | `/consortium/verify-procurement-notice` | 4 | | consortium | Verify the accessibility statement | `/consortium/verify-accessibility-statement` | 5 | | consortium | Verify the audit chain offline | `/consortium/verify-audit-chain-offline` | 5 | | consortium | Version and roll back a creation | `/consortium/version-roll-back-creation` | 6 | | consortium | Vet a reviewer or supplier staff member | `/consortium/vet-reviewer-supplier-staff` | 6 | | consortium | View the Diary by day, week, month | `/consortium/view-diary-day-week` | 5 | | consortium | View the public reviewer track record | `/consortium/view-public-reviewer-track` | 4 | | consortium | Vote on a poll | `/consortium/vote-poll` | 4 | | consortium | Write a data management plan | `/consortium/write-data-management-plan` | 6 | | consortium | Write a paper in Markdowns with co-authors | `/consortium/write-paper-markdowns-co` | 6 | | evidence | Licence enforcement evidence | `/evidence/licence-enforcement-evidence` | 7 | | finance | Accept a cheque deposit | `/finance/accept-cheque-deposit` | 6 | | finance | Access a read only evidence room | `/finance/access-read-only-evidence` | 4 | | finance | Account for agent resource use | `/finance/account-agent-resource-use` | 4 | | finance | Accrue deposit interest expense | `/finance/accrue-deposit-interest-expense` | 4 | | finance | Acknowledge a complaint | `/finance/acknowledge-complaint` | 4 | | finance | Act under a power of attorney | `/finance/act-under-power-attorney` | 5 | | finance | Action an erasure or rectification request | `/finance/action-erasure-rectification` | 4 | | finance | Activate a monitoring rule | `/finance/activate-monitoring-rule` | 4 | | finance | Activate a proposed rate | `/finance/activate-proposed-rate` | 4 | | finance | Add a joint account holder | `/finance/add-joint-account-holder` | 5 | | finance | Add a payee with name check | `/finance/add-payee-name-check` | 6 | | finance | Add or remove a signatory | `/finance/add-remove-signatory` | 5 | | finance | Agree a forbearance plan | `/finance/agree-forbearance-plan` | 5 | | finance | Amend a policy mid term | `/finance/amend-policy-mid-term` | 5 | | finance | Amend a repayment schedule | `/finance/amend-repayment-schedule` | 5 | | finance | Amend or cancel a standing order | `/finance/amend-cancel-standing-order` | 5 | | finance | Analyse complaint root causes | `/finance/analyse-complaint-root-causes` | 4 | | finance | Answer a customer question from records | `/finance/answer-customer-question-records` | 4 | | finance | Answer a subject access request | `/finance/answer-subject-access-request` | 5 | | finance | Apply a database migration | `/finance/apply-database-migration` | 6 | | finance | Apply a payment holiday | `/finance/apply-payment-holiday` | 5 | | finance | Apply enhanced due diligence | `/finance/apply-enhanced-due-diligence` | 5 | | finance | Apply for a business account | `/finance/apply-business-account` | 7 | | finance | Apply for a business loan or facility | `/finance/apply-business-loan-facility` | 7 | | finance | Apply for a loan | `/finance/apply-loan` | 10 | | finance | Apply for a personal loan | `/finance/apply-personal-loan` | 9 | | finance | Apply reasonable adjustments | `/finance/apply-reasonable-adjustments` | 4 | | finance | Apply retention and deletion | `/finance/apply-retention-deletion` | 5 | | finance | Approve a financial promotion | `/finance/approve-financial-promotion` | 5 | | finance | Approve a new integration or API consumer | `/finance/approve-new-integration-api` | 5 | | finance | Approve or decline an application | `/finance/approve-decline-application` | 6 | | finance | Approve or reject a payment | `/finance/approve-reject-payment` | 6 | | finance | Assess a claim | `/finance/assess-claim` | 5 | | finance | Assess an outsourcing or third party | `/finance/assess-outsourcing-third-party` | 5 | | finance | Assess income and expenditure | `/finance/assess-income-expenditure` | 4 | | finance | Assess source of funds and wealth | `/finance/assess-source-funds-wealth` | 5 | | finance | Audit partner usage | `/finance/audit-partner-usage` | 4 | | finance | Authorise card transactions in real time | `/finance/authorise-card-transactions-real` | 5 | | finance | Back up the database | `/finance/back-up-database` | 5 | | finance | Bill a partner | `/finance/bill-partner` | 5 | | finance | Block a payment in flight | `/finance/block-payment-flight` | 5 | | finance | Book an appointment | `/finance/book-appointment` | 5 | | finance | Bulk payments | `/finance/bulk-payments` | 7 | | finance | Business team roles, approvals and accountant access | `/finance/business-team-roles-approvals` | 9 | | finance | Calculate and post interest accrual | `/finance/calculate-post-interest-accrual` | 4 | | finance | Calculate the liquidity coverage ratio | `/finance/calculate-liquidity-coverage` | 5 | | finance | Cancel a Direct Debit | `/finance/cancel-direct-debit` | 6 | | finance | Cancel a policy | `/finance/cancel-policy` | 4 | | finance | Capital to outcomes ledger | `/finance/capital-outcomes-ledger` | 7 | | finance | Cards | `/finance/cards` | 5 | | finance | Change contact details | `/finance/change-contact-details` | 6 | | finance | Claim a Direct Debit indemnity refund | `/finance/claim-direct-debit-indemnity` | 6 | | finance | Close a business account | `/finance/close-business-account` | 5 | | finance | Close a period | `/finance/close-period` | 5 | | finance | Close an account | `/finance/close-account` | 7 | | finance | Close an account | `/finance/close-account-2` | 6 | | finance | Collect Direct Debits as an originator | `/finance/collect-direct-debits-originator` | 6 | | finance | Collect premiums | `/finance/collect-premiums` | 4 | | finance | Compare cover options | `/finance/compare-cover-options` | 4 | | finance | Complete entity tax classification | `/finance/complete-entity-tax` | 4 | | finance | Complete tax residency self certification | `/finance/complete-tax-residency-self` | 5 | | finance | Compute expected credit losses | `/finance/compute-expected-credit-losses` | 5 | | finance | Confidence report | `/finance/confidence-report` | 7 | | finance | Configure scheduled jobs | `/finance/configure-scheduled-jobs` | 5 | | finance | Confirm balances as external auditor | `/finance/confirm-balances-external` | 4 | | finance | Connect accounts at other banks | `/finance/connect-accounts-other-banks` | 6 | | finance | Contact a customer in arrears | `/finance/contact-customer-arrears` | 4 | | finance | Continue a joint account after a death | `/finance/continue-joint-account-after` | 5 | | finance | Create a monitoring rule | `/finance/create-monitoring-rule` | 4 | | finance | Create a prospect record | `/finance/create-prospect-record` | 5 | | finance | Create a savings product | `/finance/create-savings-product` | 4 | | finance | Create and release a hold | `/finance/create-release-hold` | 5 | | finance | Crypto wallet | `/finance/crypto-wallet` | 7 | | finance | Decide a treasury proposal | `/finance/decide-treasury-proposal` | 5 | | finance | Decide an APP reimbursement claim | `/finance/decide-app-reimbursement-claim` | 5 | | finance | Decide and file a SAR | `/finance/decide-file-sar` | 6 | | finance | Decline with reasons | `/finance/decline-reasons` | 4 | | finance | Deploy a release as Jude | `/finance/deploy-release-jude` | 5 | | finance | Deploy a release through deploy.py | `/finance/deploy-release-through-py` | 6 | | finance | Design a new lending product | `/finance/design-new-lending-product` | 5 | | finance | Detect a missed repayment | `/finance/detect-missed-repayment` | 4 | | finance | Direct Debits | `/finance/direct-debits` | 6 | | finance | Disburse a loan | `/finance/disburse-loan` | 5 | | finance | Disclose a vulnerability or need | `/finance/disclose-vulnerability-need` | 6 | | finance | Disposition an alert | `/finance/disposition-alert` | 6 | | finance | Dispute a card transaction | `/finance/dispute-card-transaction` | 7 | | finance | Disputes | `/finance/disputes` | 6 | | finance | Draw down from a facility | `/finance/draw-down-facility` | 6 | | finance | Enforce strong customer authentication | `/finance/enforce-strong-customer` | 4 | | finance | Escalate a customer concern | `/finance/escalate-customer-concern` | 5 | | finance | Execute a money market placement | `/finance/execute-money-market-placement` | 5 | | finance | Exit a customer for financial crime | `/finance/exit-customer-financial-crime` | 5 | | finance | Explain a transaction | `/finance/explain-transaction` | 5 | | finance | Export the evidence bundle | `/finance/export-evidence-bundle` | 5 | | finance | Export transactions to accounting software | `/finance/export-transactions-accounting` | 5 | | finance | Financing graph and independent demand | `/finance/financing-graph-independent` | 6 | | finance | Flag a mule account | `/finance/flag-mule-account` | 5 | | finance | Fraud-sharing network | `/finance/fraud-sharing-network` | 6 | | finance | Freeze or unfreeze a card | `/finance/freeze-unfreeze-card` | 5 | | finance | Fund settlement accounts for rails | `/finance/fund-settlement-accounts-rails` | 4 | | finance | Give or revoke open banking consent | `/finance/give-revoke-open-banking` | 6 | | finance | Grant an operator role | `/finance/grant-operator-role` | 4 | | finance | Hand a case to a specialist team | `/finance/hand-case-specialist-team` | 6 | | finance | Hand work between agents | `/finance/hand-work-between-agents` | 4 | | finance | Handle a Banking Protocol concern | `/finance/handle-banking-protocol-concern` | 5 | | finance | Handle a breathing space notice | `/finance/handle-breathing-space-notice` | 4 | | finance | Handle a debt sale | `/finance/handle-debt-sale` | 5 | | finance | Handle a gambling block request | `/finance/handle-gambling-block-request` | 4 | | finance | Handle a mental capacity concern | `/finance/handle-mental-capacity-concern` | 4 | | finance | Handle a partner data request or deletion | `/finance/handle-partner-data-request` | 4 | | finance | Handle a security incident | `/finance/handle-security-incident` | 6 | | finance | Handle a trigger event review | `/finance/handle-trigger-event-review` | 5 | | finance | Handle a whistleblowing report | `/finance/handle-whistleblowing-report` | 6 | | finance | Handle an indemnity claim as paying bank | `/finance/handle-indemnity-claim-paying` | 5 | | finance | Hedge interest rate exposure | `/finance/hedge-interest-rate-exposure` | 6 | | finance | Import a bank or rail statement | `/finance/import-bank-rail-statement` | 5 | | finance | Inspect a SAR audit trail | `/finance/inspect-sar-audit-trail` | 4 | | finance | Introduce a customer to a specialist | `/finance/introduce-customer-specialist` | 5 | | finance | Investigate a complaint | `/finance/investigate-complaint` | 4 | | finance | Investigate a payment exception | `/finance/investigate-payment-exception` | 5 | | finance | Investigate a production error | `/finance/investigate-production-error` | 5 | | finance | Investigate a suspected fraud | `/finance/investigate-suspected-fraud` | 5 | | finance | Invite a user with a role | `/finance/invite-user-role` | 6 | | finance | Issue a default notice | `/finance/issue-default-notice` | 5 | | finance | Issue a final response | `/finance/issue-final-response` | 5 | | finance | Issue a finding to the bank | `/finance/issue-finding-bank` | 4 | | finance | Issue a policy | `/finance/issue-policy` | 5 | | finance | Issue a replacement card in branch | `/finance/issue-replacement-card-branch` | 6 | | finance | Issue an agent mandate | `/finance/issue-agent-mandate` | 5 | | finance | Issue annual loan statements | `/finance/issue-annual-loan-statements` | 4 | | finance | Joint accounts | `/finance/joint-accounts` | 10 | | finance | Lend | `/finance/lend` | 7 | | finance | Lift an account freeze | `/finance/lift-account-freeze` | 5 | | finance | Log an operational incident | `/finance/log-operational-incident` | 5 | | finance | Log an outbound contact | `/finance/log-outbound-contact` | 5 | | finance | Maintain internal and nostro accounts | `/finance/maintain-internal-nostro` | 5 | | finance | Maintain the business wide risk assessment | `/finance/maintain-business-wide-risk` | 4 | | finance | Maintain the chart of accounts | `/finance/maintain-chart-accounts` | 4 | | finance | Maintain the record of processing | `/finance/maintain-record-processing` | 4 | | finance | Maintain the regulatory obligations register | `/finance/maintain-regulatory-obligations` | 4 | | finance | Maintain the risk register | `/finance/maintain-risk-register` | 6 | | finance | Make a complaint | `/finance/make-complaint` | 7 | | finance | Make a Faster Payment | `/finance/make-faster-payment` | 8 | | finance | Make a regulatory notification | `/finance/make-regulatory-notification` | 5 | | finance | Make an internal transfer | `/finance/make-internal-transfer` | 7 | | finance | Make an international payment | `/finance/make-international-payment` | 7 | | finance | Manage a syndicated participation | `/finance/manage-syndicated-participation` | 5 | | finance | Manage capacity | `/finance/manage-capacity` | 5 | | finance | Manage encryption and device binding | `/finance/manage-encryption-device-binding` | 4 | | finance | Manage environments and secrets free config | `/finance/manage-environments-secrets-free` | 5 | | finance | Manage foreign exchange positions | `/finance/manage-foreign-exchange` | 5 | | finance | Manage partner user access | `/finance/manage-partner-user-access` | 4 | | finance | Manage payee name check responses | `/finance/manage-payee-name-check` | 4 | | finance | Manage payment limits | `/finance/manage-payment-limits` | 5 | | finance | Manage senior manager responsibilities | `/finance/manage-senior-manager` | 5 | | finance | Manage the central bank reserve account | `/finance/manage-central-bank-reserve` | 4 | | finance | Map important business services | `/finance/map-important-business-services` | 5 | | finance | Mark a customer as a fraud victim | `/finance/mark-customer-fraud-victim` | 4 | | finance | Match statement rows to ledger legs | `/finance/match-statement-rows-ledger` | 5 | | finance | Measure carbon, tokens and time to evidence | `/finance/measure-carbon-tokens-time` | 4 | | finance | Message a customer in a case | `/finance/message-customer-case` | 6 | | finance | Money provenance certificate | `/finance/money-provenance-certificate` | 9 | | finance | Monitor CHAPS cut-offs and queue | `/finance/monitor-chaps-cut-offs` | 5 | | finance | Monitor conduct and quality | `/finance/monitor-conduct-quality` | 5 | | finance | Monitor facility covenants | `/finance/monitor-facility-covenants` | 5 | | finance | Monitor intraday liquidity | `/finance/monitor-intraday-liquidity` | 4 | | finance | Monitor job runs and leases | `/finance/monitor-job-runs-leases` | 5 | | finance | Monitor privileged actions | `/finance/monitor-privileged-actions` | 4 | | finance | Monitor product outcomes | `/finance/monitor-product-outcomes` | 4 | | finance | Monitor service health | `/finance/monitor-service-health` | 4 | | finance | Monitor without blocking | `/finance/monitor-without-blocking` | 4 | | finance | Notify a claim | `/finance/notify-claim` | 5 | | finance | Notify a death and restrict accounts | `/finance/notify-death-restrict-accounts` | 6 | | finance | Notify a personal data breach | `/finance/notify-personal-data-breach` | 5 | | finance | Offboard a business customer | `/finance/offboard-business-customer` | 5 | | finance | Onboard a partner organisation | `/finance/onboard-partner-organisation` | 5 | | finance | Onboard an agent as a customer | `/finance/onboard-agent-customer` | 6 | | finance | Open a business account | `/finance/open-business-account` | 8 | | finance | Open a fixed term deposit | `/finance/open-fixed-term-deposit` | 6 | | finance | Open a foreign currency account | `/finance/open-foreign-currency-account` | 6 | | finance | Open a personal account | `/finance/open-personal-account` | 9 | | finance | Open a personal current account | `/finance/open-personal-current-account` | 8 | | finance | Open a savings account | `/finance/open-savings-account` | 6 | | finance | Open a service case | `/finance/open-service-case` | 5 | | finance | Open a transaction receipt | `/finance/open-transaction-receipt` | 4 | | finance | Open virtual accounts under a master | `/finance/open-virtual-accounts-under` | 5 | | finance | Order business cards for staff | `/finance/order-business-cards-staff` | 6 | | finance | Overdraft | `/finance/overdraft` | 11 | | finance | Pay redress | `/finance/pay-redress` | 5 | | finance | Perform a data protection impact assessment | `/finance/perform-data-protection-impact` | 6 | | finance | Perform an internal audit | `/finance/perform-internal-audit` | 6 | | finance | Place a temporary account freeze | `/finance/place-temporary-account-freeze` | 5 | | finance | Plan a portfolio review | `/finance/plan-portfolio-review` | 4 | | finance | Plan the internal audit year | `/finance/plan-internal-audit-year` | 4 | | finance | Post a manual journal | `/finance/post-manual-journal` | 5 | | finance | Prepare a business proposition | `/finance/prepare-business-proposition` | 5 | | finance | Prepare a compliance return | `/finance/prepare-compliance-return` | 4 | | finance | Prepare a payment for approval | `/finance/prepare-payment-approval` | 4 | | finance | Prepare a payment for human approval | `/finance/prepare-payment-human-approval` | 5 | | finance | Prepare funds transfer pricing | `/finance/prepare-funds-transfer-pricing` | 4 | | finance | Prepare regulatory returns | `/finance/prepare-regulatory-returns` | 5 | | finance | Prepare statutory accounts | `/finance/prepare-statutory-accounts` | 6 | | finance | Process a Bacs input and output cycle | `/finance/process-bacs-input-output` | 5 | | finance | Process a cheque clearing exception | `/finance/process-cheque-clearing` | 5 | | finance | Process a counter cash deposit | `/finance/process-counter-cash-deposit` | 5 | | finance | Process a counter cash withdrawal | `/finance/process-counter-cash-withdrawal` | 6 | | finance | Process a loan repayment manually | `/finance/process-loan-repayment-manually` | 4 | | finance | Process an inbound Faster Payment | `/finance/process-inbound-faster-payment` | 5 | | finance | Process card clearing and settlement | `/finance/process-card-clearing-settlement` | 6 | | finance | Produce a KYC file for review | `/finance/produce-kyc-file-review` | 5 | | finance | Produce a trial balance | `/finance/produce-trial-balance` | 4 | | finance | Produce the annual MLRO report | `/finance/produce-annual-mlro-report` | 4 | | finance | Produce the FSCS single customer view | `/finance/produce-fscs-single-customer` | 4 | | finance | Propose a future rate | `/finance/propose-future-rate` | 4 | | finance | Propose a treasury credit | `/finance/propose-treasury-credit` | 4 | | finance | Provenance coverage report | `/finance/provenance-coverage-report` | 7 | | finance | Provide extra evidence on request | `/finance/provide-extra-evidence-request` | 6 | | finance | Publish terms and conditions | `/finance/publish-terms-conditions` | 5 | | finance | Pull and record a credit reference | `/finance/pull-record-credit-reference` | 4 | | finance | Query a mule account indicator | `/finance/query-mule-account-indicator` | 4 | | finance | Raise an internal suspicion report | `/finance/raise-internal-suspicion-report` | 4 | | finance | Receive a fraud intelligence signal | `/finance/receive-fraud-intelligence` | 4 | | finance | Receive a rate change notice | `/finance/receive-rate-change-notice` | 5 | | finance | Recognise loan interest income | `/finance/recognise-loan-interest-income` | 4 | | finance | Reconcile rail settlement | `/finance/reconcile-rail-settlement` | 5 | | finance | Reconcile suspense accounts | `/finance/reconcile-suspense-accounts` | 4 | | finance | Record a credit decision override | `/finance/record-credit-decision-override` | 4 | | finance | Record a customer meeting | `/finance/record-customer-meeting` | 5 | | finance | Record a customer risk assessment | `/finance/record-customer-risk-assessment` | 5 | | finance | Record a declined application reason | `/finance/record-declined-application` | 5 | | finance | Record a regulatory breach | `/finance/record-regulatory-breach` | 5 | | finance | Record a support need | `/finance/record-support-need` | 5 | | finance | Record a vulnerability disclosure | `/finance/record-vulnerability-disclosure` | 6 | | finance | Record AML training completion | `/finance/record-aml-training-completion` | 4 | | finance | Record and test a control | `/finance/record-test-control` | 5 | | finance | Recover account access | `/finance/recover-account-access` | 6 | | finance | Recover funds from a receiving account | `/finance/recover-funds-receiving-account` | 5 | | finance | Refer a complaint to FOS | `/finance/refer-complaint-fos` | 6 | | finance | Refer to a debt collection agency | `/finance/refer-debt-collection-agency` | 5 | | finance | Refer to debt advice | `/finance/refer-debt-advice` | 4 | | finance | Refund an unauthorised transaction | `/finance/refund-unauthorised-transaction` | 5 | | finance | Register a court appointed deputy | `/finance/register-court-appointed-deputy` | 6 | | finance | Register a default with agencies | `/finance/register-default-agencies` | 4 | | finance | Register a lasting power of attorney | `/finance/register-lasting-power-attorney` | 6 | | finance | Register a new device | `/finance/register-new-device` | 5 | | finance | Register a third party mandate | `/finance/register-third-party-mandate` | 6 | | finance | Register and value collateral | `/finance/register-value-collateral` | 5 | | finance | Register beneficial owners | `/finance/register-beneficial-owners` | 6 | | finance | Regulator inspection | `/finance/regulator-inspection` | 6 | | finance | Release estate funds | `/finance/release-estate-funds` | 6 | | finance | Rely on another firm's due diligence | `/finance/rely-another-firms-due` | 5 | | finance | Remove a joint holder | `/finance/remove-joint-holder` | 6 | | finance | Rename an account | `/finance/rename-account` | 4 | | finance | Renew a policy | `/finance/renew-policy` | 4 | | finance | Reopen a period | `/finance/reopen-period` | 4 | | finance | Repair a misdirected payment | `/finance/repair-misdirected-payment` | 6 | | finance | Repay or overpay a loan | `/finance/repay-overpay-loan` | 6 | | finance | Report a bereavement | `/finance/report-bereavement` | 6 | | finance | Report a change of company details | `/finance/report-change-company-details` | 6 | | finance | Report a lost or stolen card | `/finance/report-lost-stolen-card` | 6 | | finance | Report an APP scam | `/finance/report-app-scam` | 6 | | finance | Report an unauthorised payment | `/finance/report-unauthorised-payment` | 6 | | finance | Report complaints data | `/finance/report-complaints-data` | 5 | | finance | Report fraud losses to the regulator | `/finance/report-fraud-losses-regulator` | 5 | | finance | Report fraud to Action Fraud or police | `/finance/report-fraud-action-police` | 4 | | finance | Report insurance data to the regulator | `/finance/report-insurance-data-regulator` | 4 | | finance | Report payment statistics | `/finance/report-payment-statistics` | 4 | | finance | Report to credit reference agencies | `/finance/report-credit-reference-agencies` | 5 | | finance | Request a bank statement letter | `/finance/request-bank-statement-letter` | 5 | | finance | Request a company audit letter | `/finance/request-company-audit-letter` | 5 | | finance | Request a covenant waiver | `/finance/request-covenant-waiver` | 5 | | finance | Request a date of death balance letter | `/finance/request-date-death-balance` | 4 | | finance | Request a defence against money laundering | `/finance/request-defence-against-money` | 5 | | finance | Request a loan settlement quote | `/finance/request-loan-settlement-quote` | 6 | | finance | Request a trade finance instrument | `/finance/request-trade-finance-instrument` | 6 | | finance | Request affordability evidence | `/finance/request-affordability-evidence` | 4 | | finance | Request an arranged overdraft | `/finance/request-arranged-overdraft` | 6 | | finance | Request an insurance quote from KYC | `/finance/request-insurance-quote-kyc` | 5 | | finance | Request further information | `/finance/request-further-information` | 5 | | finance | Request money | `/finance/request-money` | 6 | | finance | Request my personal data | `/finance/request-personal-data` | 5 | | finance | Request Owner approval for a consequential action | `/finance/request-owner-approval` | 5 | | finance | Request records from the bank | `/finance/request-records-bank` | 5 | | finance | Reset access for a customer | `/finance/reset-access-customer` | 5 | | finance | Resolve within 3 business days | `/finance/resolve-within-3-business` | 4 | | finance | Respond to a production order | `/finance/respond-production-order` | 6 | | finance | Respond to an open banking payment request | `/finance/respond-open-banking-payment` | 5 | | finance | Responsible lending decision | `/finance/responsible-lending-decision` | 7 | | finance | Restore from a backup | `/finance/restore-backup` | 6 | | finance | Retire an agent | `/finance/retire-agent` | 6 | | finance | Retry a failed job step | `/finance/retry-failed-job-step` | 5 | | finance | Retry a failed occurrence | `/finance/retry-failed-occurrence` | 5 | | finance | Return an unapplied payment | `/finance/return-unapplied-payment` | 4 | | finance | Reverse a journal with a linked entry | `/finance/reverse-journal-linked-entry` | 4 | | finance | Review a complaint file as FOS | `/finance/review-complaint-file-fos` | 5 | | finance | Review a loan application | `/finance/review-loan-application` | 5 | | finance | Review access rights | `/finance/review-access-rights` | 5 | | finance | Review an application queue | `/finance/review-application-queue` | 4 | | finance | Review APP reimbursement performance | `/finance/review-app-reimbursement` | 4 | | finance | Review customer financial statements | `/finance/review-customer-financial` | 5 | | finance | Review device and session signals | `/finance/review-device-session-signals` | 4 | | finance | Review operational resilience evidence | `/finance/review-operational-resilience` | 4 | | finance | Review SM&CR responsibilities | `/finance/review-sm-cr-responsibilities` | 4 | | finance | Review the vulnerability register | `/finance/review-vulnerability-register` | 4 | | finance | Revoke a role on leaving | `/finance/revoke-role-leaving` | 4 | | finance | Revoke an agent mandate | `/finance/revoke-agent-mandate` | 4 | | finance | Roll back a release | `/finance/roll-back-release` | 5 | | finance | Rotate keys and device bindings | `/finance/rotate-keys-device-bindings` | 5 | | finance | Run a Consumer Duty outcomes review | `/finance/run-consumer-duty-outcomes` | 5 | | finance | Run a disaster recovery test | `/finance/run-disaster-recovery-test` | 4 | | finance | Run a fair value assessment | `/finance/run-fair-value-assessment` | 5 | | finance | Run a KYC check on a subject | `/finance/run-kyc-check-subject` | 5 | | finance | Run a ledger report for a period | `/finance/run-ledger-report-period` | 5 | | finance | Run a liquidity stress test | `/finance/run-liquidity-stress-test` | 5 | | finance | Run an annual relationship review | `/finance/run-annual-relationship-review` | 6 | | finance | Run due jobs as service actor | `/finance/run-due-jobs-service` | 5 | | finance | Run due standing orders | `/finance/run-due-standing-orders` | 5 | | finance | Run end of day payment reconciliation | `/finance/run-end-day-payment` | 4 | | finance | Run payroll payments | `/finance/run-payroll-payments` | 6 | | finance | Run penetration testing | `/finance/run-penetration-testing` | 6 | | finance | Run periodic KYC refresh | `/finance/run-periodic-kyc-refresh` | 6 | | finance | Run protected reveals | `/finance/run-protected-reveals` | 4 | | finance | Run sanctions screening | `/finance/run-sanctions-screening` | 5 | | finance | Run transaction monitoring | `/finance/run-transaction-monitoring` | 5 | | finance | Sample test receipts as external auditor | `/finance/sample-test-receipts-external` | 4 | | finance | Sanctions and PEP screening | `/finance/sanctions-pep-screening` | 5 | | finance | Save an unfinished application | `/finance/save-unfinished-application` | 5 | | finance | Savings pots and goals | `/finance/savings-pots-goals` | 5 | | finance | Schedule a future payment | `/finance/schedule-future-payment` | 6 | | finance | Scheduled payments | `/finance/scheduled-payments` | 6 | | finance | Screen a list against sanctions | `/finance/screen-list-against-sanctions` | 5 | | finance | Screen existing customers against list updates | `/finance/screen-existing-customers` | 5 | | finance | Screen for politically exposed persons | `/finance/screen-politically-exposed` | 5 | | finance | Search transactions | `/finance/search-transactions` | 5 | | finance | Security and devices | `/finance/security-devices` | 8 | | finance | Send a CHAPS payment | `/finance/send-chaps-payment` | 7 | | finance | Send a receiving bank APP notice | `/finance/send-receiving-bank-app` | 5 | | finance | Send money | `/finance/send-money` | 5 | | finance | Set and control an agent's mandate | `/finance/set-control-agents-mandate` | 6 | | finance | Set and monitor exposure limits | `/finance/set-monitor-exposure-limits` | 5 | | finance | Set balance and payment alerts | `/finance/set-balance-payment-alerts` | 4 | | finance | Set card limits and controls | `/finance/set-card-limits-controls` | 5 | | finance | Set eligibility rules | `/finance/set-eligibility-rules` | 5 | | finance | Set fees and charges | `/finance/set-fees-charges` | 5 | | finance | Set joint operating rules | `/finance/set-joint-operating-rules` | 5 | | finance | Set loan terms and produce an offer | `/finance/set-loan-terms-produce` | 5 | | finance | Set negotiated pricing | `/finance/set-negotiated-pricing` | 5 | | finance | Set up a Direct Debit | `/finance/set-up-direct-debit` | 6 | | finance | Set up a mandate | `/finance/set-up-mandate` | 6 | | finance | Set up a standing order | `/finance/set-up-standing-order` | 5 | | finance | Set up merchant acquiring | `/finance/set-up-merchant-acquiring` | 6 | | finance | Settle a claim | `/finance/settle-claim` | 5 | | finance | Settle and close a loan | `/finance/settle-close-loan` | 5 | | finance | Solution bank funding | `/finance/solution-bank-funding` | 10 | | finance | Stop a savings enrolment | `/finance/stop-savings-enrolment` | 6 | | finance | Submit transactions for monitoring | `/finance/submit-transactions-monitoring` | 5 | | finance | Support a customer in financial difficulty | `/finance/support-customer-financial` | 5 | | finance | Support the external audit | `/finance/support-external-audit` | 4 | | finance | Switch account to EDGE Bank | `/finance/switch-account-edge-bank` | 6 | | finance | Terminate a partnership | `/finance/terminate-partnership` | 5 | | finance | Trace ingress and egress | `/finance/trace-ingress-egress` | 4 | | finance | Track remediation actions | `/finance/track-remediation-actions` | 5 | | finance | Transaction monitoring | `/finance/transaction-monitoring` | 7 | | finance | Transition a case state | `/finance/transition-case-state` | 6 | | finance | Triage a case queue | `/finance/triage-case-queue` | 4 | | finance | Tune a fraud rule | `/finance/tune-fraud-rule` | 6 | | finance | Underwrite a business facility | `/finance/underwrite-business-facility` | 5 | | finance | Underwrite a policy | `/finance/underwrite-policy` | 5 | | finance | Update communication preferences | `/finance/update-communication-preferences` | 5 | | finance | Upload a bulk payment file | `/finance/upload-bulk-payment-file` | 6 | | finance | Validate a credit risk model | `/finance/validate-credit-risk-model` | 5 | | finance | Verify a caller or visitor | `/finance/verify-caller-visitor` | 5 | | finance | Verify a company and its officers | `/finance/verify-company-officers` | 5 | | finance | Verify a regulatory return against the ledger | `/finance/verify-regulatory-return-against` | 4 | | finance | Verify a trust or charity | `/finance/verify-trust-charity` | 4 | | finance | Verify address evidence | `/finance/verify-address-evidence` | 5 | | finance | Verify an identity through EDGE ID | `/finance/verify-identity-through-edge` | 5 | | finance | Verify executor or administrator | `/finance/verify-executor-administrator` | 5 | | finance | Verify identity evidence | `/finance/verify-identity-evidence` | 5 | | finance | Verify identity in person | `/finance/verify-identity-person` | 9 | | finance | View and download a statement | `/finance/view-download-statement` | 5 | | finance | View balances and available funds | `/finance/view-balances-available-funds` | 4 | | finance | View multi account cash position | `/finance/view-multi-account-cash` | 5 | | finance | View PIN securely | `/finance/view-pin-securely` | 4 | | finance | Void an imported statement | `/finance/void-imported-statement` | 5 | | finance | Withdraw a product | `/finance/withdraw-product` | 4 | | finance | Withdraw a proposed rate | `/finance/withdraw-proposed-rate` | 4 | | finance | Withdraw an application | `/finance/withdraw-application` | 5 | | finance | Work the fraud alert queue | `/finance/work-fraud-alert-queue` | 5 | | finance | Write a fraud case outcome letter | `/finance/write-fraud-case-outcome` | 5 | | finance | Write off a debt | `/finance/write-off-debt` | 5 | | id | Claim an organisation's workspace by magic link | `/id/claim-organisations-workspace` | 7 | | id | Find, claim and check any UK organisation | `/id/find-claim-check-any` | 10 | | judiverse | Ask Jude or instruct it to act | `/judiverse/ask-jude-instruct-act` | 5 | | judiverse | Build an agent in the intenJude lab | `/judiverse/build-agent-intenjude-lab` | 7 | | judiverse | Calendar and scheduling | `/judiverse/calendar-scheduling` | 7 | | judiverse | Compile an intention into a certified procedure | `/judiverse/compile-intention-certified` | 6 | | judiverse | Contacts, people directory and search | `/judiverse/contacts-people-directory-search` | 10 | | judiverse | Deep research with checked findings | `/judiverse/deep-research-checked-findings` | 7 | | judiverse | Documents and forms | `/judiverse/documents-forms` | 8 | | judiverse | JudeBox agents | `/judiverse/judebox-agents` | 2 | | judiverse | JudeBox calls | `/judiverse/judebox-calls` | 3 | | judiverse | JudeBox connections | `/judiverse/judebox-connections` | 1 | | judiverse | JudeBox contacts | `/judiverse/judebox-contacts` | 2 | | judiverse | JudeBox email | `/judiverse/judebox-email` | 2 | | judiverse | JudeBox memory and vault | `/judiverse/judebox-memory-vault` | 2 | | judiverse | JudeBox messages | `/judiverse/judebox-messages` | 3 | | judiverse | JudeBox numbers and mailboxes | `/judiverse/judebox-numbers-mailboxes` | 2 | | judiverse | JudeBox webhooks and integrations | `/judiverse/judebox-webhooks-integrations` | 1 | | judiverse | JudeBox: artists' own music with direct royalties | `/judiverse/judebox-artists-own-music` | 8 | | judiverse | Judiverse included with every application licence | `/judiverse/judiverse-included-every` | 6 | | judiverse | Judiverse standalone: by capability or whole | `/judiverse/judiverse-standalone-capability` | 6 | | judiverse | Mail and agent mailboxes | `/judiverse/mail-agent-mailboxes` | 7 | | judiverse | Messages and chat | `/judiverse/messages-chat` | 11 | | judiverse | Tasks and workflows | `/judiverse/tasks-workflows` | 7 | | organisations | Bottleneck and blocker removal by design | `/organisations/bottleneck-blocker-removal` | 7 | | organisations | Pre-onboarded organisations from public registers | `/organisations/pre-onboarded-organisations` | 6 | | organisations | Scope change intake: scope creep as raw material | `/organisations/scope-change-intake-creep` | 8 | | public | Accept a package referral | `/public/accept-package-referral` | 6 | | public | Act on a deadline stack item | `/public/act-deadline-stack-item` | 5 | | public | Activate the emergency centre and keep the decision log | `/public/activate-emergency-centre-keep` | 5 | | public | Adopt a risk-based reserves strategy with minimum level and replenishment | `/public/adopt-risk-based-reserves` | 6 | | public | Advise on legality and sign the legal implications of a report | `/public/advise-legality-sign-legal` | 5 | | public | Agree a care and support plan and personal budget | `/public/agree-care-support-plan` | 5 | | public | Agree a deferred payment | `/public/agree-deferred-payment` | 5 | | public | Agree a personalised housing plan | `/public/agree-personalised-housing-plan` | 5 | | public | Agree an information sharing agreement | `/public/agree-information-sharing` | 5 | | public | Agree and seal a planning obligation | `/public/agree-seal-planning-obligation` | 6 | | public | Agree the council plan and its outcomes | `/public/agree-council-plan-outcomes` | 6 | | public | Agree the officer and member protocol | `/public/agree-officer-member-protocol` | 5 | | public | Agree the political group position before a meeting | `/public/agree-political-group-position` | 5 | | public | Agree the post-election timetable and transition protocol | `/public/agree-post-election-timetable` | 6 | | public | Allocate committee seats by political balance | `/public/allocate-committee-seats` | 5 | | public | Analog human switch for irreversible actions | `/public/analog-human-switch-irreversible` | 8 | | public | Answer a Best Value Notice or MHCLG engagement with one evidence set | `/public/answer-best-value-notice` | 6 | | public | Answer a code of conduct complaint | `/public/answer-code-conduct-complaint` | 5 | | public | Answer a corporate complaint | `/public/answer-corporate-complaint` | 5 | | public | Answer a freedom of information request | `/public/answer-freedom-information` | 5 | | public | Answer a local land charges search | `/public/answer-local-land-charges` | 4 | | public | Answer a media enquiry and issue a press release | `/public/answer-media-enquiry-issue` | 5 | | public | Answer a standards complaint | `/public/answer-standards-complaint` | 4 | | public | Answer a subject access request | `/public/answer-subject-access-request` | 5 | | public | Answer a written parliamentary question | `/public/answer-written-parliamentary` | 4 | | public | Answer an appeal to the Valuation Tribunal | `/public/answer-appeal-valuation-tribunal` | 4 | | public | Answer an environmental information request | `/public/answer-environmental-information` | 4 | | public | Answer ministerial correspondence | `/public/answer-ministerial` | 5 | | public | Answer questions at a scrutiny meeting | `/public/answer-questions-scrutiny` | 6 | | public | Appeal a planning decision | `/public/appeal-planning-decision` | 2 | | public | Apply an empty home or second home premium | `/public/apply-empty-home-second` | 4 | | public | Apply an exemption and issue a refusal notice | `/public/apply-exemption-issue-refusal` | 4 | | public | Apply for a Blue Badge | `/public/apply-blue-badge` | 4 | | public | Apply for a deprivation of liberty authorisation | `/public/apply-deprivation-liberty` | 6 | | public | Apply for a discount or council tax reduction | `/public/apply-discount-council-tax` | 4 | | public | Apply for a licence | `/public/apply-licence` | 3 | | public | Apply for a postal or proxy vote | `/public/apply-postal-proxy-vote` | 3 | | public | Apply for a school place and appeal a refusal | `/public/apply-school-place-appeal` | 4 | | public | Apply for a voter authority certificate | `/public/apply-voter-authority` | 3 | | public | Apply for an export licence | `/public/apply-export-licence` | 5 | | public | Apply for free school meals | `/public/apply-free-school-meals` | 4 | | public | Apply for planning permission and track it | `/public/apply-planning-permission-track` | 4 | | public | Apply the Armed Forces Covenant duty in a council decision | `/public/apply-armed-forces-covenant` | 4 | | public | Apply the monthly savings test | `/public/apply-monthly-savings-test` | 6 | | public | Apply the national pay award | `/public/apply-national-pay-award` | 5 | | public | Apply the one-page decision standard and report standard to every committee paper | `/public/apply-one-page-decision` | 6 | | public | Apply the pre-election period | `/public/apply-pre-election-period` | 4 | | public | Apply the retention schedule and dispose of records | `/public/apply-retention-schedule-dispose` | 5 | | public | Appoint an election agent | `/public/appoint-election-agent` | 3 | | public | Appoint an independent member to the committee | `/public/appoint-independent-member` | 6 | | public | Appoint and train polling staff | `/public/appoint-train-polling-staff` | 5 | | public | Appoint or dismiss a statutory officer | `/public/appoint-dismiss-statutory` | 5 | | public | Appoint the cabinet and allocate portfolios | `/public/appoint-cabinet-allocate` | 5 | | public | Appoint the external auditor through PSAA | `/public/appoint-external-auditor-through` | 4 | | public | Appoint the internal auditor and receive the report | `/public/appoint-internal-auditor-receive` | 4 | | public | Approve a business case through the assurance framework | `/public/approve-business-case-through` | 5 | | public | Approve a change through the change board | `/public/approve-change-through-board` | 5 | | public | Approve a direct payment | `/public/approve-direct-payment` | 6 | | public | Approve a disabled facilities grant | `/public/approve-disabled-facilities` | 4 | | public | Approve a new system against the architecture standard | `/public/approve-new-system-against` | 4 | | public | Approve a severance package over £100,000 | `/public/approve-severance-package-over` | 4 | | public | Approve a virement | `/public/approve-virement` | 5 | | public | Approve the internal audit charter | `/public/approve-internal-audit-charter` | 5 | | public | Approve the internal audit plan | `/public/approve-internal-audit-plan` | 4 | | public | Approve the medium term financial strategy | `/public/approve-medium-term-financial` | 5 | | public | Approve the pension fund accounts and valuation | `/public/approve-pension-fund-accounts` | 5 | | public | Approve the senior structure and statutory officer designations | `/public/approve-senior-structure` | 5 | | public | Approve the treasury, capital and minimum revenue provision strategies | `/public/approve-treasury-capital-minimum` | 4 | | public | Ask a question of the leader or a cabinet member | `/public/ask-question-leader-cabinet` | 4 | | public | Ask a question or present a petition at a meeting | `/public/ask-question-present-petition` | 4 | | public | Ask for a care needs assessment | `/public/ask-care-needs-assessment` | 4 | | public | Ask for help with homelessness | `/public/ask-help-homelessness` | 4 | | public | Ask for own personal data | `/public/ask-own-personal-data` | 4 | | public | Assess a carer's needs | `/public/assess-carers-needs` | 5 | | public | Assess a code of conduct complaint | `/public/assess-code-conduct-complaint` | 6 | | public | Assess a Housing Benefit claim | `/public/assess-housing-benefit-claim` | 5 | | public | Assess and decide bus franchising | `/public/assess-decide-bus-franchising` | 6 | | public | Assess Council Tax Reduction | `/public/assess-council-tax-reduction` | 5 | | public | Assess risk and publish the community risk register | `/public/assess-risk-publish-community` | 5 | | public | Assess the committee's effectiveness and report to council | `/public/assess-committees-effectiveness` | 4 | | public | Assess the needs of an adult | `/public/assess-needs-adult` | 6 | | public | Attend the count as an agent | `/public/attend-count-agent` | 4 | | public | Attend, film or report a public meeting | `/public/attend-film-report-public` | 4 | | public | Authenticate and seal a document | `/public/authenticate-seal-document` | 4 | | public | Authorise and make a payment | `/public/authorise-make-payment` | 5 | | public | Authorise covert surveillance under RIPA | `/public/authorise-covert-surveillance` | 5 | | public | Award a discount or exemption | `/public/award-discount-exemption` | 4 | | public | Award a discretionary housing payment | `/public/award-discretionary-housing` | 4 | | public | Award and monitor a voluntary sector grant | `/public/award-monitor-voluntary-sector` | 6 | | public | Award business rates relief | `/public/award-business-rates-relief` | 5 | | public | Award under a framework or dynamic market | `/public/award-under-framework-dynamic` | 5 | | public | Back up and test a restore | `/public/back-up-test-restore` | 5 | | public | Bid for and report on a transport grant | `/public/bid-report-transport-grant` | 5 | | public | Bill a new business rates property | `/public/bill-new-business-rates` | 4 | | public | Build a resident service through discovery, alpha and beta | `/public/build-resident-service-through` | 6 | | public | Build and sign the savings ledger line by line | `/public/build-sign-savings-ledger` | 6 | | public | Calculate the council tax base and requirement | `/public/calculate-council-tax-base` | 5 | | public | Call in an executive decision | `/public/call-executive-decision` | 6 | | public | Canvass voters lawfully | `/public/canvass-voters-lawfully` | 4 | | public | Carry out a data protection impact assessment | `/public/carry-out-data-protection` | 6 | | public | Carry out a financial assessment and charge | `/public/carry-out-financial-assessment` | 6 | | public | Carry out an education, health and care needs assessment | `/public/carry-out-education-health` | 5 | | public | Carry out an internal review | `/public/carry-out-internal-review` | 4 | | public | Cast a recorded vote on the budget | `/public/cast-recorded-vote-budget` | 5 | | public | Certify a grant claim | `/public/certify-grant-claim` | 4 | | public | Chair full council and rule on order | `/public/chair-full-council-rule` | 7 | | public | Challenge a council tax band | `/public/challenge-council-tax-band` | 4 | | public | Challenge a provider rate and decide an uplift | `/public/challenge-provider-rate-decide` | 5 | | public | Check supply chain labour and modern slavery risk | `/public/check-supply-chain-labour` | 5 | | public | Claim emergency financial assistance | `/public/claim-emergency-financial` | 4 | | public | Claim members' allowances and expenses | `/public/claim-members-allowances` | 6 | | public | Claim staffing and business costs | `/public/claim-staffing-business-costs` | 5 | | public | Clerk a school admission appeal panel | `/public/clerk-school-admission-appeal` | 5 | | public | Comment on a planning application | `/public/comment-planning-application` | 3 | | public | Comment on a planning application | `/public/comment-planning-application-2` | 4 | | public | Commission a placement for a looked-after child | `/public/commission-placement-looked` | 6 | | public | Commission a public health service under the provider selection regime | `/public/commission-public-health-service` | 5 | | public | Commission supported living from a charity provider | `/public/commission-supported-living` | 6 | | public | Commission temporary accommodation | `/public/commission-temporary` | 5 | | public | Commission the external quality assessment | `/public/commission-external-quality` | 5 | | public | Complain to the council and the ombudsman | `/public/complain-council-ombudsman` | 4 | | public | Complete induction and training on reserves, debt, EFS and risk using live decisions | `/public/complete-induction-training` | 5 | | public | Complete mandatory member training | `/public/complete-mandatory-member` | 4 | | public | Complete the annual governance and accountability return | `/public/complete-annual-governance` | 4 | | public | Complete the Data Security and Protection Toolkit | `/public/complete-data-security` | 4 | | public | Comply with a Secretary of State direction | `/public/comply-secretary-state-direction` | 5 | | public | Conduct a marriage or civil partnership | `/public/conduct-marriage-civil` | 5 | | public | Confirm MRP and EFS accounting with the external auditor | `/public/confirm-mrp-efs-accounting` | 5 | | public | Consider an objection from a local elector | `/public/consider-objection-local-elector` | 4 | | public | Consider social value before procurement | `/public/consider-social-value-before` | 4 | | public | Consolidate published recommendations into owned operating systems | `/public/consolidate-published` | 7 | | public | Consult and publicise an application | `/public/consult-publicise-application` | 5 | | public | Consult on and make redundancies | `/public/consult-make-redundancies` | 7 | | public | Contract management | `/public/contract-management` | 12 | | public | Convene the right people by mail and call | `/public/convene-right-people-mail` | 6 | | public | Correct a register entry | `/public/correct-register-entry` | 3 | | public | Council decisions, interests and scrutiny | `/public/council-decisions-interests` | 11 | | public | Cyber posture | `/public/cyber-posture` | 7 | | public | Deal with a dangerous structure | `/public/deal-dangerous-structure` | 4 | | public | Decide a parking penalty challenge | `/public/decide-parking-penalty-challenge` | 5 | | public | Decide an application at planning committee | `/public/decide-application-planning` | 6 | | public | Decide on sharing confidential care information as Caldicott guardian | `/public/decide-sharing-confidential-care` | 4 | | public | Declare an impairment to independence | `/public/declare-impairment-independence` | 4 | | public | Declare an interest at a meeting and withdraw | `/public/declare-interest-meeting` | 6 | | public | Declare an interest in proceedings | `/public/declare-interest-proceedings` | 3 | | public | Declare the collection fund surplus or deficit | `/public/declare-collection-fund-surplus` | 4 | | public | Decommission a legacy system | `/public/decommission-legacy-system` | 6 | | public | Defend a planning appeal | `/public/defend-planning-appeal` | 5 | | public | Deliver the risk-based audit plan from substantive capacity | `/public/deliver-risk-based-audit` | 5 | | public | Design and run a grant scheme | `/public/design-run-grant-scheme` | 6 | | public | Determine a building regulations application | `/public/determine-building-regulations` | 3 | | public | Determine a postal or proxy vote application | `/public/determine-postal-proxy-vote` | 5 | | public | Determine an application to register | `/public/determine-application-register` | 5 | | public | Discharge planning conditions | `/public/discharge-planning-conditions` | 4 | | public | Dispose of or acquire land at best consideration | `/public/dispose-acquire-land-best` | 6 | | public | Employ staff in the MP's office | `/public/employ-staff-mps-office` | 4 | | public | Employment within the platform | `/public/employment-within-platform` | 14 | | public | Enforce joiner, mover and leaver with segregation of duties | `/public/enforce-joiner-mover-leaver` | 5 | | public | Enforce publication deadlines with an exception register | `/public/enforce-publication-deadlines` | 5 | | public | Enrol staff in the pension scheme and send monthly data | `/public/enrol-staff-pension-scheme` | 5 | | public | Ensure continuity of care on provider failure | `/public/ensure-continuity-care-provider` | 5 | | public | Evaluate and grade a job | `/public/evaluate-grade-job` | 4 | | public | Fill a casual vacancy by election or co-option | `/public/fill-casual-vacancy-election` | 5 | | public | Find a council, borough or government department in the institution directory | `/public/find-council-borough-government` | 5 | | public | Find the right council and service for an address | `/public/find-right-council-service` | 4 | | public | Flow obligations into a contract or commission | `/public/flow-obligations-contract` | 5 | | public | Give notice to hand back a contract | `/public/give-notice-hand-back` | 4 | | public | Give the opinion on the financial statements | `/public/give-opinion-financial` | 3 | | public | Grant a dispensation for a pecuniary interest | `/public/grant-dispensation-pecuniary` | 5 | | public | Grant a premises licence | `/public/grant-premises-licence` | 5 | | public | Grant a street works permit | `/public/grant-street-works-permit` | 4 | | public | Handle a benefit reconsideration or appeal | `/public/handle-benefit-reconsideration` | 5 | | public | Handle a highway claim | `/public/handle-highway-claim` | 5 | | public | Handle a housing complaint | `/public/handle-housing-complaint` | 4 | | public | Handle a request for erasure or restriction | `/public/handle-request-erasure` | 4 | | public | Hear a call-in and decide | `/public/hear-call-decide` | 4 | | public | Hear a councillor call for action | `/public/hear-councillor-call-action` | 4 | | public | Hold a citizenship ceremony | `/public/hold-citizenship-ceremony` | 6 | | public | Hold a constituency surgery | `/public/hold-constituency-surgery` | 5 | | public | Hold a licensing sub-committee hearing | `/public/hold-licensing-sub-committee` | 5 | | public | Hold an initial child protection conference | `/public/hold-initial-child-protection` | 5 | | public | Hold and record the monthly golden triangle meeting | `/public/hold-record-monthly-golden` | 5 | | public | Hold identity and operational evidence | `/public/hold-identity-operational` | 5 | | public | Hold pre-decision scrutiny of a key decision | `/public/hold-pre-decision-scrutiny` | 4 | | public | Hold the annual parish meeting | `/public/hold-annual-parish-meeting` | 4 | | public | Host a social housing consumer standards inspection | `/public/host-social-housing-consumer` | 5 | | public | Host an LGA corporate peer challenge | `/public/host-lga-corporate-peer` | 7 | | public | Independently test an implemented control under the self-review rule | `/public/independently-test-implemented` | 5 | | public | Inspect a food business and award a hygiene rating | `/public/inspect-food-business-award` | 6 | | public | Inspect a highway and repair a defect | `/public/inspect-highway-repair-defect` | 5 | | public | Inspect a private rented home and serve a notice | `/public/inspect-private-rented-home` | 5 | | public | Inspect a workplace for health and safety | `/public/inspect-workplace-health-safety` | 3 | | public | Inspect the council's accounts and object | `/public/inspect-councils-accounts-object` | 3 | | public | Inspect work and issue a completion certificate | `/public/inspect-work-issue-completion` | 4 | | public | Instruct enforcement agents with a vulnerability check | `/public/instruct-enforcement-agents` | 4 | | public | Investigate a breach and serve an enforcement notice | `/public/investigate-breach-serve` | 6 | | public | Investigate a consumer complaint and take enforcement action | `/public/investigate-consumer-complaint` | 5 | | public | Investigate a flood as lead local flood authority | `/public/investigate-flood-lead-local` | 5 | | public | Investigate a statutory nuisance and serve an abatement notice | `/public/investigate-statutory-nuisance` | 6 | | public | Investigate a suspected fraud | `/public/investigate-suspected-fraud` | 6 | | public | Investigate tenancy fraud | `/public/investigate-tenancy-fraud` | 4 | | public | Issue a Best Value Notice and run engagement | `/public/issue-best-value-notice` | 5 | | public | Issue a certified copy of an entry | `/public/issue-certified-copy-entry` | 5 | | public | Issue a CIL liability notice and demand | `/public/issue-cil-liability-notice` | 6 | | public | Issue a mayoral decision | `/public/issue-mayoral-decision` | 5 | | public | Issue a refund or transfer a credit | `/public/issue-refund-transfer-credit` | 4 | | public | Issue a Section 114 report | `/public/issue-section-114-report` | 5 | | public | Issue a Section 5 report on unlawfulness or maladministration | `/public/issue-section-5-report` | 6 | | public | Issue a security aspects letter to a contractor | `/public/issue-security-aspects-letter` | 5 | | public | Issue a statutory recommendation or public interest report | `/public/issue-statutory-recommendation` | 5 | | public | Issue a voter authority certificate | `/public/issue-voter-authority` | 4 | | public | Issue annual council tax bills | `/public/issue-annual-council-tax` | 4 | | public | Issue the annual internal audit opinion | `/public/issue-annual-internal-audit` | 5 | | public | Issue the audit plan | `/public/issue-audit-plan` | 4 | | public | Issue the decision notice | `/public/issue-decision-notice` | 4 | | public | Issue the role statement, train and attribute challenge | `/public/issue-role-statement-train` | 4 | | public | Issue the summons and agenda | `/public/issue-summons-agenda` | 4 | | public | Issue the weekly one-page exception report | `/public/issue-weekly-one-page` | 6 | | public | Join a council's approved provider list | `/public/join-councils-approved-provider` | 4 | | public | Join a strategic coordinating group | `/public/join-strategic-coordinating` | 5 | | public | Join the housing register and assess priority | `/public/join-housing-register-assess` | 5 | | public | Keep business continuity plans and advise businesses | `/public/keep-business-continuity-plans` | 4 | | public | Keep spending within the limit and file the return | `/public/keep-spending-within-limit` | 4 | | public | Keep statutory notification readiness and named incident roles | `/public/keep-statutory-notification` | 5 | | public | Keep the device and software asset register | `/public/keep-device-software-asset` | 4 | | public | Keep the outside bodies appointments register | `/public/keep-outside-bodies-appointments` | 4 | | public | Keep the record of processing activities | `/public/keep-record-processing` | 3 | | public | Lay a statutory instrument | `/public/lay-statutory-instrument` | 5 | | public | Lay the annual report and accounts | `/public/lay-annual-report-accounts` | 5 | | public | Lead recovery | `/public/lead-recovery` | 5 | | public | Lead the response to a Section 114 notice | `/public/lead-response-section-114` | 6 | | public | Let a social home | `/public/let-social-home` | 6 | | public | Licence a taxi or private hire driver | `/public/licence-taxi-private-hire` | 6 | | public | Licence an animal activity | `/public/licence-animal-activity` | 4 | | public | License a house in multiple occupation | `/public/license-house-multiple` | 6 | | public | Logistics tracking | `/public/logistics-tracking` | 7 | | public | Maintain and publish the register of members' interests | `/public/maintain-publish-register` | 5 | | public | Maintain the brownfield and self-build registers | `/public/maintain-brownfield-self-build` | 4 | | public | Maintain the Constitution and scheme of delegation with live decision routes | `/public/maintain-constitution-scheme` | 6 | | public | Maintain the Contract Procedure Rules | `/public/maintain-contract-procedure` | 5 | | public | Maintain the publication scheme and disclosure log | `/public/maintain-publication-scheme` | 3 | | public | Maintain the single procurement and commissioning pipeline | `/public/maintain-single-procurement` | 5 | | public | Make a freedom of information request or ask for review | `/public/make-freedom-information-request` | 5 | | public | Make a public spaces protection order | `/public/make-public-spaces-protection` | 6 | | public | Make a safeguarding adults enquiry | `/public/make-safeguarding-adults-enquiry` | 6 | | public | Make a traffic regulation order | `/public/make-traffic-regulation-order` | 6 | | public | Make a tree preservation order | `/public/make-tree-preservation-order` | 5 | | public | Manage a contract and hold review meetings | `/public/manage-contract-hold-review` | 5 | | public | Manage a persistent or unreasonable complainant | `/public/manage-persistent-unreasonable` | 5 | | public | Manage a public question or petition at a meeting | `/public/manage-public-question-petition` | 5 | | public | Manage allotments | `/public/manage-allotments` | 5 | | public | Manage cash flow and borrowing | `/public/manage-cash-flow-borrowing` | 5 | | public | Manage sickness absence | `/public/manage-sickness-absence` | 4 | | public | Mark and check politically restricted posts | `/public/mark-check-politically` | 4 | | public | Mark exempt information and run a private item | `/public/mark-exempt-information-run` | 5 | | public | Mark, store and send classified information | `/public/mark-store-send-classified` | 5 | | public | Meet the auditors without officers present | `/public/meet-auditors-without-officers` | 4 | | public | Modify a contract and publish the change notice | `/public/modify-contract-publish-change` | 5 | | public | Monitor a grant and recover misused funds | `/public/monitor-grant-recover-misused` | 4 | | public | Monitor a provider's quality | `/public/monitor-providers-quality` | 6 | | public | Name and number a street | `/public/name-number-street` | 4 | | public | Nominate an asset of community value | `/public/nominate-asset-community-value` | 3 | | public | Nominate an asset of community value | `/public/nominate-asset-community-value-2` | 4 | | public | Notify the regulator of an incident | `/public/notify-regulator-incident` | 5 | | public | Obtain facility security clearance for a supplier site | `/public/obtain-facility-security` | 5 | | public | Open a constituent case with consent | `/public/open-constituent-case-consent` | 5 | | public | Open a council tax account on a move in | `/public/open-council-tax-account` | 5 | | public | Open a rest centre and register evacuees | `/public/open-rest-centre-register` | 5 | | public | Open and conclude an inquest | `/public/open-conclude-inquest` | 6 | | public | Open postal votes and verify personal identifiers | `/public/open-postal-votes-verify` | 5 | | public | Open the Government station and choose a journey | `/public/open-government-station-choose` | 5 | | public | Organise evidence and records for a public body | `/public/organise-evidence-records-public` | 4 | | public | Pass a proposal through the business case gateway | `/public/pass-proposal-through-business` | 5 | | public | Pay council tax and set up a direct debit | `/public/pay-council-tax-set` | 4 | | public | Pay or challenge a parking penalty | `/public/pay-challenge-parking-penalty` | 4 | | public | Pay the data protection fee | `/public/pay-data-protection-fee` | 4 | | public | Place a household in interim accommodation | `/public/place-household-interim` | 5 | | public | Pool a budget with the NHS | `/public/pool-budget-nhs` | 4 | | public | Prepare a care proceedings application | `/public/prepare-care-proceedings` | 5 | | public | Prepare a Green Book business case | `/public/prepare-green-book-business` | 5 | | public | Prepare a neighbourhood plan | `/public/prepare-neighbourhood-plan` | 7 | | public | Prepare a procurement and agreement requirements brief for a government body | `/public/prepare-procurement-agreement` | 6 | | public | Prepare an impact assessment | `/public/prepare-impact-assessment` | 4 | | public | Prepare an off-site plan for a major hazard site | `/public/prepare-off-site-plan` | 5 | | public | Prepare and examine the local plan | `/public/prepare-examine-local-plan` | 6 | | public | Prepare for a CQC assessment of adult social care | `/public/prepare-cqc-assessment-adult` | 5 | | public | Prepare for a Public Accounts Committee hearing | `/public/prepare-public-accounts` | 5 | | public | Prepare for an inspection of children's services | `/public/prepare-inspection-childrens` | 5 | | public | Prepare main and supplementary estimates | `/public/prepare-main-supplementary` | 5 | | public | Prepare the local transport plan | `/public/prepare-local-transport-plan` | 5 | | public | Prepare the market sustainability plan | `/public/prepare-market-sustainability` | 5 | | public | Price a single source defence contract | `/public/price-single-source-defence` | 4 | | public | Process a right to buy application | `/public/process-right-buy-application` | 6 | | public | Process a temporary event notice | `/public/process-temporary-event-notice` | 4 | | public | Procure a separately procurable engagement with an aggregation check | `/public/procure-separately-procurable` | 6 | | public | Procurement intelligence | `/public/procurement-intelligence` | 9 | | public | Propose the budget and council tax to full council | `/public/propose-budget-council-tax` | 7 | | public | Publish an accessibility statement | `/public/publish-accessibility-statement` | 5 | | public | Publish an algorithmic transparency record | `/public/publish-algorithmic-transparency` | 4 | | public | Publish an executive decision record | `/public/publish-executive-decision` | 4 | | public | Publish and maintain privacy notices | `/public/publish-maintain-privacy-notices` | 4 | | public | Publish government procurement card spend | `/public/publish-government-procurement` | 4 | | public | Publish members' allowances for the year | `/public/publish-members-allowances-year` | 3 | | public | Publish notice of a key decision 28 clear days ahead | `/public/publish-notice-key-decision` | 4 | | public | Publish open data | `/public/publish-open-data` | 4 | | public | Publish spending over £500 | `/public/publish-spending-over-500` | 4 | | public | Publish the agenda and reports five clear days ahead | `/public/publish-agenda-reports-five` | 5 | | public | Publish the annual pay policy statement | `/public/publish-annual-pay-policy` | 4 | | public | Publish the contracts register | `/public/publish-contracts-register` | 4 | | public | Publish the draft and audited statement of accounts | `/public/publish-draft-audited-statement` | 6 | | public | Publish the forward plan twelve weeks ahead | `/public/publish-forward-plan-twelve` | 4 | | public | Publish the gambling statement of principles | `/public/publish-gambling-statement` | 5 | | public | Publish the gender pay gap | `/public/publish-gender-pay-gap` | 3 | | public | Publish the market position statement | `/public/publish-market-position` | 5 | | public | Publish the notice of election | `/public/publish-notice-election` | 3 | | public | Publish the regulatory enforcement policy | `/public/publish-regulatory-enforcement` | 5 | | public | Publish the revised register | `/public/publish-revised-register` | 4 | | public | Publish the updated Constitution | `/public/publish-updated-constitution` | 4 | | public | Publish transparency data | `/public/publish-transparency-data` | 2 | | public | Put a digital imprint on campaign material | `/public/put-digital-imprint-campaign` | 3 | | public | Raise a purchase order before package spend | `/public/raise-purchase-order-before` | 4 | | public | Raise a safeguarding concern | `/public/raise-safeguarding-concern` | 5 | | public | Raise a service complaint | `/public/raise-service-complaint` | 6 | | public | Read a committee pack within the page limit and lodge questions | `/public/read-committee-pack-within` | 5 | | public | Rebase the adult social care budget from the care ledger | `/public/rebase-adult-social-care` | 6 | | public | Receive and decide on nominations | `/public/receive-decide-nominations` | 5 | | public | Receive and handle a whistleblowing disclosure | `/public/receive-handle-whistleblowing` | 5 | | public | Receive and publish candidates' spending returns | `/public/receive-publish-candidates` | 5 | | public | Receive quarterly monitoring with a demand bridge | `/public/receive-quarterly-monitoring` | 5 | | public | Receive the auditor's annual report and value for money commentary | `/public/receive-auditors-annual-report` | 4 | | public | Receive the counter-fraud annual report | `/public/receive-counter-fraud-annual` | 4 | | public | Receive the monthly evidence pack rated on evidence | `/public/receive-monthly-evidence-pack` | 5 | | public | Reconcile access and contain unexplained accounts | `/public/reconcile-access-contain` | 5 | | public | Reconcile the balance-sheet exposure statement quarterly | `/public/reconcile-balance-sheet-exposure` | 5 | | public | Record a gift or hospitality | `/public/record-gift-hospitality` | 5 | | public | Record a price exception with the finance position | `/public/record-price-exception-finance` | 5 | | public | Record an exemption from Contract Procedure Rules | `/public/record-exemption-contract` | 5 | | public | Recover a benefit overpayment | `/public/recover-benefit-overpayment` | 4 | | public | Recover unpaid council tax through a liability order | `/public/recover-unpaid-council-tax` | 6 | | public | Recruit safely with DBS checks | `/public/recruit-safely-dbs-checks` | 5 | | public | Refer a band challenge to the Valuation Office Agency | `/public/refer-band-challenge-valuation` | 4 | | public | Refer higher-risk building work to the Building Safety Regulator | `/public/refer-higher-risk-building` | 4 | | public | Register a birth | `/public/register-birth` | 5 | | public | Register a birth, death or marriage notice | `/public/register-birth-death-marriage` | 4 | | public | Register a death and offer Tell Us Once | `/public/register-death-offer-tell` | 5 | | public | Register a financial interest | `/public/register-financial-interest` | 4 | | public | Register a foreign activity arrangement | `/public/register-foreign-activity` | 3 | | public | Register a political party | `/public/register-political-party` | 4 | | public | Register as a consultant lobbyist and file returns | `/public/register-consultant-lobbyist` | 3 | | public | Register as a non-party campaigner | `/public/register-non-party-campaigner` | 4 | | public | Register disclosable pecuniary interests | `/public/register-disclosable-pecuniary` | 5 | | public | Register to vote | `/public/register-vote` | 3 | | public | Reimburse operators for concessionary travel | `/public/reimburse-operators` | 4 | | public | Renew Public Services Network compliance | `/public/renew-public-services-network` | 5 | | public | Reperform a closed control from source | `/public/reperform-closed-control-source` | 5 | | public | Report a defence security incident | `/public/report-defence-security-incident` | 5 | | public | Report a street problem | `/public/report-street-problem` | 5 | | public | Report complaints and compliments annually | `/public/report-complaints-compliments` | 4 | | public | Report contract performance indicators | `/public/report-contract-performance` | 4 | | public | Report digital service performance | `/public/report-digital-service` | 3 | | public | Report donations and loans | `/public/report-donations-loans` | 4 | | public | Report limited or no assurance to the audit committee | `/public/report-limited-no-assurance` | 5 | | public | Report on a council-owned company to the shareholder committee | `/public/report-council-owned-company` | 4 | | public | Report on staffing and organisation as head of paid service | `/public/report-staffing-organisation` | 5 | | public | Report on the robustness of estimates and adequacy of reserves | `/public/report-robustness-estimates` | 5 | | public | Report on value for money arrangements | `/public/report-value-money-arrangements` | 3 | | public | Represent the council on the combined authority or a regional board | `/public/represent-council-combined` | 5 | | public | Request a rate uplift | `/public/request-rate-uplift` | 5 | | public | Request a review of a housing decision | `/public/request-review-housing-decision` | 4 | | public | Request and track audit evidence | `/public/request-track-audit-evidence` | 4 | | public | Request exceptional financial support | `/public/request-exceptional-financial` | 6 | | public | Require a cabinet member or officer to attend | `/public/require-cabinet-member-officer` | 5 | | public | Resolve a service desk incident | `/public/resolve-service-desk-incident` | 6 | | public | Resolve eligibility for the general power of competence | `/public/resolve-eligibility-general` | 4 | | public | Respond to a child protection referral | `/public/respond-child-protection` | 6 | | public | Respond to a damp, mould or repair report | `/public/respond-damp-mould-repair` | 6 | | public | Respond to a judicial review pre-action letter | `/public/respond-judicial-review-pre` | 6 | | public | Respond to a parliamentary petition | `/public/respond-parliamentary-petition` | 4 | | public | Respond to a statutory recommendation from the auditor | `/public/respond-statutory-recommendation` | 6 | | public | Respond to an area SEND inspection | `/public/respond-area-send-inspection` | 5 | | public | Respond to an ICO decision notice | `/public/respond-ico-decision-notice` | 2 | | public | Respond to an ombudsman enquiry and carry out the remedy | `/public/respond-ombudsman-enquiry-carry` | 5 | | public | Respond to an ombudsman finding | `/public/respond-ombudsman-finding` | 6 | | public | Return post-election data to the Electoral Commission | `/public/return-post-election-data` | 3 | | public | Review a homelessness decision | `/public/review-homelessness-decision` | 5 | | public | Review a looked-after child's care plan | `/public/review-looked-after-childs` | 5 | | public | Review local air quality | `/public/review-local-air-quality` | 5 | | public | Review polling districts and places | `/public/review-polling-districts-places` | 5 | | public | Review the annual governance statement | `/public/review-annual-governance` | 4 | | public | Review the corporate risk register quarterly | `/public/review-corporate-risk-register` | 5 | | public | Review the draft statement of accounts | `/public/review-draft-statement-accounts` | 4 | | public | Review the information risk register as senior information risk owner | `/public/review-information-risk-register` | 5 | | public | Review the risk management framework | `/public/review-risk-management-framework` | 4 | | public | Run a combined authority board meeting | `/public/run-combined-authority-board` | 4 | | public | Run a deep dive with recorded influence on direction | `/public/run-deep-dive-recorded` | 5 | | public | Run a demand control cell weekly | `/public/run-demand-control-cell` | 5 | | public | Run a disaster recovery test | `/public/run-disaster-recovery-test` | 5 | | public | Run a disciplinary process | `/public/run-disciplinary-process` | 5 | | public | Run a neighbourhood planning referendum | `/public/run-neighbourhood-planning` | 5 | | public | Run a public consultation and publish the response | `/public/run-public-consultation-publish` | 5 | | public | Run a public consultation and report the results | `/public/run-public-consultation-report` | 6 | | public | Run a recall petition as petition officer | `/public/run-recall-petition-officer` | 5 | | public | Run a standards hearing and publish the decision | `/public/run-standards-hearing-publish` | 6 | | public | Run a task and finish review and secure the executive response | `/public/run-task-finish-review` | 6 | | public | Run an assurance review on a major project | `/public/run-assurance-review-major` | 5 | | public | Run an audit engagement from terms of reference to report | `/public/run-audit-engagement-terms` | 7 | | public | Run monthly budget monitoring with budget holders | `/public/run-monthly-budget-monitoring` | 6 | | public | Run monthly payroll and report to HMRC | `/public/run-monthly-payroll-report` | 6 | | public | Run National Fraud Initiative matches | `/public/run-national-fraud-initiative` | 6 | | public | Run security awareness and phishing tests | `/public/run-security-awareness-phishing` | 5 | | public | Run the annual canvass | `/public/run-annual-canvass` | 6 | | public | Run the annual council meeting | `/public/run-annual-council-meeting` | 5 | | public | Run the cross-committee forum with other chairs | `/public/run-cross-committee-forum` | 5 | | public | Run the decision clock on stalled reforms | `/public/run-decision-clock-stalled` | 6 | | public | Run the make-or-buy test before a material re-let | `/public/run-make-buy-test` | 5 | | public | Run the out-of-hours duty officer rota | `/public/run-out-hours-duty` | 5 | | public | Run the period for exercise of public rights | `/public/run-period-exercise-public` | 4 | | public | Run the public inspection period for the accounts | `/public/run-public-inspection-period` | 5 | | public | Run the reorganisation programme to vesting day | `/public/run-reorganisation-programme` | 6 | | public | Run the verification and count and declare the result | `/public/run-verification-count-declare` | 7 | | public | Score the thirty cyber criteria with evidence references | `/public/score-thirty-cyber-criteria` | 6 | | public | Scrutinise a substantial NHS service change | `/public/scrutinise-substantial-nhs` | 5 | | public | Scrutinise the budget proposals before council | `/public/scrutinise-budget-proposals` | 4 | | public | Scrutinise the combined authority | `/public/scrutinise-combined-authority` | 4 | | public | Scrutinise the community safety partnership | `/public/scrutinise-community-safety` | 4 | | public | Scrutinise treasury management | `/public/scrutinise-treasury-management` | 4 | | public | Security incident response | `/public/security-incident-response` | 7 | | public | Seek a Cabinet Office spend control approval | `/public/seek-cabinet-office-spend` | 5 | | public | Seek HM Treasury approval above delegated limits | `/public/seek-hm-treasury-approval` | 5 | | public | Select a candidate and authorise the party description | `/public/select-candidate-authorise-party` | 4 | | public | Send apologies and keep attendance within the six-month rule | `/public/send-apologies-keep-attendance` | 5 | | public | Separate cashable saving, cost avoidance and deferred demand on every line | `/public/separate-cashable-saving-cost` | 6 | | public | Set and collect council housing rents | `/public/set-collect-council-housing` | 5 | | public | Set the annual calendar of meetings | `/public/set-annual-calendar-meetings` | 5 | | public | Set the annual work programme | `/public/set-annual-work-programme` | 5 | | public | Set the local government finance settlement | `/public/set-local-government-finance` | 5 | | public | Set the mayoral budget and precept | `/public/set-mayoral-budget-precept` | 4 | | public | Set the precept and notify the billing authority | `/public/set-precept-notify-billing` | 4 | | public | Set the scheme of executive delegation | `/public/set-scheme-executive-delegation` | 4 | | public | Set up an instalment plan or direct debit | `/public/set-up-instalment-plan` | 4 | | public | Share information with other responders | `/public/share-information-other` | 4 | | public | Sign the annual governance statement | `/public/sign-annual-governance-statement` | 5 | | public | Sign the declaration of acceptance of office | `/public/sign-declaration-acceptance` | 5 | | public | Speak in a debate and check the official report | `/public/speak-debate-check-official` | 5 | | public | Spend under the section 137 power within the limit | `/public/spend-under-section-137` | 4 | | public | Submit a motion to full council | `/public/submit-motion-full-council` | 6 | | public | Submit an innovation proposal to the Defence and Security Accelerator | `/public/submit-innovation-proposal` | 5 | | public | Submit planning statistics returns | `/public/submit-planning-statistics` | 4 | | public | Submit revenue and capital returns to MHCLG | `/public/submit-revenue-capital-returns` | 5 | | public | Submit single source contract reports | `/public/submit-single-source-contract` | 4 | | public | Submit the business rates forecast and outturn returns | `/public/submit-business-rates-forecast` | 4 | | public | Submit the homelessness statistics return | `/public/submit-homelessness-statistics` | 4 | | public | Submit the party's annual accounts | `/public/submit-partys-annual-accounts` | 4 | | public | Submit the provider cost return | `/public/submit-provider-cost-return` | 4 | | public | Submit the tenant satisfaction measures | `/public/submit-tenant-satisfaction` | 5 | | public | Supplier vetting | `/public/supplier-vetting` | 8 | | public | Supply the register to an entitled person | `/public/supply-register-entitled-person` | 4 | | public | Support a care leaver with a pathway plan | `/public/support-care-leaver-pathway` | 5 | | public | Table a motion or seek a debate | `/public/table-motion-seek-debate` | 3 | | public | Table a written question | `/public/table-written-question` | 4 | | public | Take a bill through Parliament | `/public/take-bill-through-parliament` | 6 | | public | Take a financial monitoring item with decision-grade evidence | `/public/take-financial-monitoring-item` | 5 | | public | Take a homelessness application and decide the duty | `/public/take-homelessness-application` | 5 | | public | Take a key decision on the one-page standard | `/public/take-key-decision-one` | 5 | | public | Take a ward casework enquiry through to reply | `/public/take-ward-casework-enquiry` | 7 | | public | Take an individual portfolio decision | `/public/take-individual-portfolio` | 6 | | public | Take an urgent decision under special urgency | `/public/take-urgent-decision-under` | 5 | | public | Take and publish the minutes | `/public/take-publish-minutes` | 5 | | public | Take and publish the minutes | `/public/take-publish-minutes-2` | 6 | | public | Take evidence and agree a select committee report | `/public/take-evidence-agree-select` | 6 | | public | Take notice of marriage or civil partnership | `/public/take-notice-marriage-civil` | 5 | | public | Take the six-month progress report on statutory recommendations | `/public/take-six-month-progress` | 5 | | public | Tender and award | `/public/tender-award` | 9 | | public | Track audit recommendations to owner and date and escalate overdue items | `/public/track-audit-recommendations` | 6 | | public | Track the executive's responses to recommendations | `/public/track-executives-responses` | 5 | | public | Transfer staff under TUPE | `/public/transfer-staff-under-tupe` | 5 | | public | Transfer the commissioning toolkit into a service | `/public/transfer-commissioning-toolkit` | 5 | | public | Validate a planning application | `/public/validate-planning-application` | 4 | | public | Verify weighing and measuring equipment | `/public/verify-weighing-measuring` | 4 | | public | Vet and train care staff | `/public/vet-train-care-staff` | 5 | | public | View the public outcome scorecard | `/public/view-public-outcome-scorecard` | 4 | | public | Vote in a division | `/public/vote-division` | 4 | | public | Warn and inform the public | `/public/warn-inform-public` | 4 | | public | Webcast and allow reporting of a public meeting | `/public/webcast-allow-reporting-public` | 5 | | public | Write a submission to a minister | `/public/write-submission-minister` | 5 | | public | Write and exercise the emergency plan | `/public/write-exercise-emergency-plan` | 6 | | public | Write off an irrecoverable debt | `/public/write-off-irrecoverable-debt` | 6 | | public | Write the officer report and decide under delegation | `/public/write-officer-report-decide` | 4 | | public | Write to a public body on a constituent's behalf | `/public/write-public-body-constituents` | 6 | | regulators | Automated review outreach and scheduling | `/regulators/automated-review-outreach` | 8 | | regulators | Regulator, auditor and reviewer workspace | `/regulators/regulator-auditor-reviewer` | 9 | | releases | Free evaluation and commercial licensing | `/releases/free-evaluation-commercial` | 7 | ## 4. Research ### research/AGENT-RESEARCH.md `research/AGENT-RESEARCH.md` · sha256 `723389b5a6f22b3dc428abb167925da4ca8a5cfd7e8ef76df76e8aa177e9feb7` #### Everything the estate knows, in one place Generated by `tools/compendium.py` from the ontology and every research source. Do not edit by hand: change a source and run it again. This is research, never authority: the contract (`contract/estate-contract.json`) and the owner's recorded decisions are the authority, and they come first below. **How to read it (people and agents).** 1. Start with the gate. Every source is counted, and nothing is dropped. 2. Sections 1 to 3 are the authority: the contract's intentions, then the owner's decisions and requests. 3. Section 4 is the specification, and section 5 is the proof. 4. Section 6 holds every research finding, each with an ID and the capability URL that serves it. 5. Section 7 reconciles the machine registers with their ledgers. 6. Section 8 accounts for every old-estate document. 7. Section 9 lists everything not yet met. Search by ID (`F-…`), by URL (`/finance/…`) or by intention (`glass-box`). ##### The gate | Measure | Met | Of | Share | |---|---|---|---| | Intentions served | 27 | 27 | 100.0% | | Owner decisions recorded | 21 | 21 | 100.0% | | Owner requests done (rest need the owner) | 53 | 58 | 91.4% | | Research findings placed on a capability | 2626 | 13987 | 18.8% | | Old-estate inventory lines (accounted in section 8) | 1157 | 1157 | 100.0% | | Research files read | 47 | 47 | 100.0% | ##### 1. Intentions (the contract) | ID | Intention | Capabilities serving it | Status | |---|---|---|---| | `source-of-truth` | The estate is the living source of truth for all human domain expertise: regulator requirements, live facts and laws, and everything that can be researched, refreshed continuously. | 18 | SERVED | | `truth-separated-from-agreement` | Every statement is classed as observation, agreement or claim; for every rule, law and fact the estate shows how it came to be, who agrees, who disagrees and on what evidence. | 113 | SERVED | | `glass-box` | A glass box, not a black box: light, not darkness; possibility, not limitation; to be seen, not to control. Authorised people and regulators can inspect knowledge, reasoning, authority, execution, verification, recovery and efficiency. | 131 | SERVED | | `collaboration-and-shared-data` | Collaboration, community, cooperation and teamwork are invited. Data is shared for good under lawful authorised access, not hoarded, so that organisations create new capabilities, industries, jobs and solutions. | 128 | SERVED | | `solve-the-crises` | Capabilities directed at the energy crisis; cyber attack made unprofitable; people given the capability to generate wealth, meaning the power to be; more and better housing; raw materials used better; zero waste; everything monitored and tracked. | 40 | SERVED | | `agents-native` | Agents are native to the estate as much as people: always on, inside the estate so data never leaves it, each with its own portal and racetrack so agents cannot crash or collide. The estate is built for agents, which makes it better and safer for people. | 139 | SERVED | | `clear-screens-deep-help` | Screens show the essential result first; depth and context live behind a help control as markdown, served to people and agents on demand, so neither people nor agent context windows are overwhelmed. | 93 | SERVED | | `exhaustive-deterministic-scope` | Everything is defined and scoped before it is built: every screen, feature, workflow, capability, control, intention, outcome, endpoint and every whole-of-life, customer, agent, regulator and engineer journey. Nothing missing, nothing broken. | 97 | SERVED | | `match-and-exceed` | Every competitor capability is matched, exceeded, improved and made easier and faster; benchmarks, baselines, competitors, features, behaviours, workflows and controls are known from research. | 50 | SERVED | | `compress-time` | Slow, clunky processes and industries are compressed in time. Human intention is converted into computational instructions; labs automate prompt engineering, creation and solutions; engineering work that took a day takes minutes. | 1580 | SERVED | | `agents-build-agents` | Always-on agents build further agents that are advanced, fine-tunable, transferable and reproducible. | 17 | SERVED | | `always-on-monitoring` | Everything is reproducible, retraceable, auditable and monitored always on; this is how sovereignty and control of every organisation's data are maintained. | 68 | SERVED | | `lean-repository` | Clean source code; one instruction markdown; the repository contains exactly what it needs; every file is named and placed by TOK. | 48 | SERVED | | `compute-light-runtime` | AI is the factory, not the product: neural models are used at build time for speed; what runs is deterministic, compute-light, more useful and more accurate, and reduces the need for AI over time. | 61 | SERVED | | `only-true-things` | The runtime says only what a verified procedure produces inside its certified scope, with its error bound attached, using error-rate control and Bayesian reasoning; otherwise it says it does not know. | 74 | SERVED | | `same-rules-everywhere` | One method applies to every domain: identity, relations, processes, logic, probabilities and patterns. Every domain is added by extension and the same rules hold. | 40 | SERVED | | `better-routes` | Where a licence or partner is not needed, a better route is engineered: consented data, simulators, open sources, own networks. Where the licence is the network, the legal permission or the clinical safety, the licence or partner is obtained as well. | 514 | SERVED | | `sight` | Working backwards from preventing avoidable illness and blindness: the time from first meaningful signal to a verified beneficial outcome is shortened, starting at the front door before the GP, with own capture technology and consented patient data. | 87 | SERVED | | `approved-design` | Interfaces use approved UX Pilot, Figma and Untitled UI structure and the estate's original logos, used as-is and never recreated: accessible, readable, complete navigation, real functionality. | 52 | SERVED | | `every-intention-represented` | Every intention of the governing organisation and of every onboarded organisation is represented, including capabilities implied and never spelled out, each placed in an application or foundation. | 22 | SERVED | | `verified-outcomes-fast` | The governing objective is the shortest independently justified time from legitimate intention to verified beneficial outcome, at the greatest feasible speed and autonomy, with no reduction in scope or acceptance. | 57 | SERVED | | `education-is-the-source` | Education is where all capabilities come from; research and development is worth investing in; universities are made prestigious again by being the home from which public, defence, government, procurement, logistics and cyber capabilities grow. | 60 | SERVED | | `agents-as-real-customers` | Agents are real users and real customers in every application: lenders, borrowers, buyers, suppliers, patients' representatives, applicants and operators. They act as real customers on their racetracks so they learn what works and what does not, under their own identity and authority. | 1591 | SERVED | | `confidence-reporting` | Confidence reporting sets the standard beyond credit scoring: a confidence score that is more informationally complete than a credit score, showing not only whether obligations were met but how a person or organisation acted when things went wrong (whether they did all they could, how they communicated, what caused the situation), so a single turn of events does not define them. It strengthens KYC, is explainable and contestable, and is proven by research and measured business and economic outcomes, including wider access to resources. | 189 | SERVED | | `ai-assurance-standard` | The estate sets the global standard in autonomous agent assurance, as the first fully agent-native bank: agents as customers and staff with their own identity, certificates, mandates and term, with no gaps in assurance. | 134 | SERVED | | `solution-banks` | From think tanks to solution banks: agents generate solutions, and solution banks empower and fund people to enact them in the real world, because having the answer is not enough. Better lending structures and better ways to move money through the economy create jobs and access to resources. | 59 | SERVED | | `physical-presence` | EDGE is a real bank and clinic in the physical world, not only online: EDGE points and devices that come to the person, verifying identity by proximity (scanning or tapping an identity document or device), and edge hardware and wearables used for good, including early detection in the clinic. | 84 | SERVED | ###### Founding principles and principles - el_roi - let_there_be_light - what_god_cannot_do_does_not_exist - truth_and_agreement - `one-source` (GATE): TOK is the only source. Everything is an object in it; every document, view, index, diagram, count and test is generated from it; a hand edit of generated output is a failure. Check: G01 schema, G03 closure, G05 generation, G16 live counts - `truth-agreement-claim` (GATE): Every statement is exactly one of observation, agreement or claim, carrying that class's required fields; none is rendered as another. Check: G04 - `honest-status` (GATE): Every realisation carries a truth status set from evidence, never above it. An attempt is not a completion; a rehearsal is not an outcome; an intention is not an observation. Nobody fabricates evidence, authority, benchmarks, permissions or readiness; a falsehood is rejected and never rewarded. Check: G04 status rules; breaches are incidents - `live-numbers` (GATE): Every number is produced from the objects at the moment it is stated, with its method and completeness. No count is written by hand. Check: G16 - `everything-identified` (GATE): Everything that exists has a permanent identity, a kind, its relations and a working monitor; nothing is undefined, dead, unaccounted for or without an identity. Check: G02, G03 - `complete-contract-first` (GATE): Every capability, journey, screen, control, operation, endpoint and agent resolves to a complete executable contract before it is built: intention, requirement, steps, executor, authority, design, acceptance, evidence, resources, recovery and answerer. A missing part is an OPEN obligation with an owner, counted and visible. Check: G03 required relations, G10, G11, G13 - `research-defines-scope` (GATE): Scope comes from research: stated intentions, current primary sources, regulations, standards and every top competitor's screens, flows, capabilities and controls. Every competitor is a floor, matched then exceeded. Check: G15; parity measured and published - `agents-native` (GATE): Every operation has a human path and an agent path to the same outcome with equal capability, under the agent's own EDGE ID, with the same authority, preview, audit and undo. Every agent has a portal, a mailbox, a racetrack and a purpose-limited authority from a legitimate issuer; consequential actions are rehearsed with the other agents involved before they commit; people and agents share each workflow with stated hand-off, approval, pause, take-over and undo, conflicts resolved by lease and fence. Check: G12; agent fields in G01 - `accountable-authority` (GATE): Every effectful actor has an authenticated identity, an accountable owner and a current purpose-limited permission from the competent issuer. Every effectful object names the human or organisational role that answers if it goes wrong; an agent is never the answerer of last resort. Check: G13 - `evidence-with-the-effect` (GATE): Every action binds its inputs, prior state, release, contract, predicate and receiving effect to durable evidence; the receiving system's own readback establishes the outcome. Rendering or request acceptance alone establishes nothing. Check: evidence fields in G01; G09 - `verification-by-reproduction` (GATE): Correctness comes from reproducing the exact contract against the actual evidence with the declared checker. VERIFIED records whether the reproduction was by the same lineage or an independent one; where an applicable authority requires independence, nothing less is accepted for that claim. Check: G04 VERIFIED rules - `preventive-enforcement` (CONDUCT): Unauthorised, unsafe or conflicting effects are refused at the boundary before execution; emerging risks are predicted and corrected before they become incidents; every discovered defect becomes a correction to the mechanism and a permanent regression test. Zero silent failures, zero fabricated assurance, zero unauthorised actions, zero avoidable harm. Check: incident objects carry cause, correction and regression test (G01); monitors are live - `authorised-recovery` (GATE): Every operation has a verified recovery or compensation; irreversible consequences are explicitly controlled; an ambiguous effect is queried, never replayed; recovery preserves the intention and the acceptance predicate. Check: operation recovers_by in G03 - `glass-box` (GATE): Knowledge, reasoning, authority, execution, verification, recovery and efficiency are inspectable by authorised people, including regulators, without the builder's help. Observation protects the dignity and privacy of the person; to be seen is not to be watched. Knowledge about a person confers no authority over that person. Check: REGULATOR journey in G11; inspection surfaces on every journey - `sovereign-data` (CONDUCT): Each organisation's data and agents stay under its own control inside the estate and never go to model providers; agents run inside the estate. Check: access_class on every object (G01); data-flow review on every endpoint - `clear-screens` (GATE): Every screen shows the essential result first and keeps depth behind a help control as markdown served to people and agents alike; every screen is accessible and completes for a person with no sight. Check: G14 - `approved-design` (GATE): Only approved UX Pilot, Figma and Untitled UI structure and the estate's original logos, used as-is; no hand-drawing and no logo recreation. Check: designed_by in G14; design-asset approval field - `outcomes-first` (GATE): Every capability serves at least one outcome: time, money, lives, planet, resources or value, saved or more. Outcomes are claimed only when verified with their measurement; resources consumed per verified outcome (tokens, GBP, kWh, gCO2e, litres, seconds) are published for every agent and capability. Check: G10; ledger measurements - `compiled-runtime` (GATE): Neural models are the factory, used at build time; what runs is deterministic and compute-light: TOK facts, compiled procedures with certified scope and error bound, answers rendered from proofs with sources. Outside a certified scope the answer is unknown or a question. The AI-free rate is published and rises. Check: G17; AI-free rate, cost, latency and reproducibility measured - `lean-estate` (GATE): One instruction markdown; every file accounted for by an object; no duplicates; raw material, datasets and large generated data in storage by digest; nothing is stored without justification. Check: G03, G08 - `always-new` (GATE): Every build starts from the current sources and the current agreement. Nothing is inherited: research is admitted as observation or claim, never as authority, and every rule, count, permission and status is produced fresh. Check: G04 authority rules; G16 - `reconciled-change` (GATE): Every change to a principle or decision names what it supersedes or narrows and retires it in the same commit; the principal's wording stays verbatim; a check no program can evaluate is never a gate. Check: G06 - `always-on-monitoring` (MEASURE): Everything is reproducible, retraceable, auditable and monitored always on; monitoring is part of assurance and never sufficient alone. Check: monitor coverage and freshness published from the objects - `latest-approval-governs` (CONDUCT): Only the most recent explicit approval of an exact subject by its competent authority is active. Check: decision objects carry decided_at; the latest governs - `act-on-intention` (CONDUCT): Standing intention is direction, not a step approval; agents act on it without questions. A question is asked only where intention and authority do not cover the exact action, or where a legal or irreversible-safety condition requires a recorded confirmation. Check: none mechanical - `one-clock` (MEASURE): Every claimed run is measured on one external monotonic clock including research, provisioning, build, verification, review, deployment, operation, admission and recording; a timeout or incomplete outcome is never a success. Check: whole elapsed time published with every claimed run - `extension-not-redesign` (GATE): Every human domain is added by extension of TOK with crosswalks to its established ontologies and standards; the kernel does not change to admit a domain. Check: G18 ###### Outcomes the estate commits to - dimensions - directions - units_per_verified_outcome - states - rule ###### Open decisions in the contract (4) - the host that runs the self-hosted CI - the storage location for research and datasets - whether screens mirror a competitor's exact layout or only its behaviour and features - which mark is EDGE Public Consortium's: the-edge-consortium.svg, public-edge.svg or cyber-governance.svg (all admitted under design/brand/) ##### 2. The owner's decisions (21) Each is an AGREEMENT recorded with `tools/amend.py`; the basis is the owner's own wording. - **EDGE Public Consortium uses the Public EDGE mark** `tok:decision:01m68c87qmz54aqnbm8pqzmgt4`: X-X.TECH decided on 31 October 2026: EDGE Public Consortium uses public-edge.svg as its mark; the-edge-consortium.svg and cyber-governance.svg are kept as alternates. - **Agents hold accounts in their own name** `tok:decision:01m696xzzpcbnsexpcdtz7ak9n`: X-X.TECH decided on 31 October 2026: agents can hold accounts in the autonomous agent estate; there is no law disallowing it. An agent's account is opened, held and closed in the agent's own name under its own identity, birth certificate and mandate, not as a sub-account of a person. As with everything in the estate, an accountable owner (a person or organisation) is named on the account and answers for it; that role is never the agent. - **TRL 9 and CRL 9 mean every intention carried out exactly** `tok:decision:01m69b9nqqnxj7bh7p0mner5rf`: X-X.TECH decided on 31 October 2026: in this estate, TRL 9 and CRL 9 mean that the estate does exactly everything the owner intended, all intentions brought into being as the raw operation of an autonomous agent estate (prompt engineering, governed intelligence, knowing your agent). Regulators give approvals; they do not define TRL 9 and CRL 9 here. - **Agent high-risk outcomes are provisional until the owner signs off in person** `tok:decision:01m69gty59gnpc17g811317htn`: X-X.TECH decided on 31 October 2026: any high-risk outcome run by an agent (large transfers and anything of value that can be stolen) completes only as provisional value, marked at both ends. It becomes final only when the human owner signs it off in person, on a registered device at a registered place (NFC tap, passkey), singly or in bulk, and the recipient agrees, within a grace window of 24 hours (at most 48). Without both, it is reversed automatically wherever the value went. A stolen key can move value; the move means nothing until the owner agrees in the real world. - **EDGE Public Consortium becomes The Edge Consortium (education and research); Public EDGE is a separate application for government** `tok:decision:01m69jxxj5cazjefrrj1fvb962`: X-X.TECH decided on 31 October 2026: rename EDGE Public Consortium to The Edge Consortium, the estate's own university and research consortium (education, research, peer review, grants, libraries, the governed intelligence degree). Public EDGE becomes its own application for government: central and local government, councils, defence, procurement, policing, borders, cyber governance, regulators and public services. Five applications: EDGE Finance, EDGE Clinic, The Edge Consortium, Public EDGE, Judiverse. - **No irreversible outcome without a real, analog, in-person human act by the owner** `tok:decision:01m69jxxqzrhz7r36ht4pvv5td`: X-X.TECH decided on 31 October 2026: an irreversible outcome (weapons and anything that can cause mass harm, and anything else that cannot be undone) can never be completed by an agent or by software alone. It needs the human owner, in person, to make a physical analog act on dedicated hardware that sits in the actuation path itself, so no software or AI signal can close it. The owner holds identical backup units of that hardware, so a broken unit never blocks the owner, and each unit proves it was the owner who ordered it and that they definitely want it to occur. Lost or broken units are revoked; agents can prepare, rehearse and request, never execute. - **Judiverse comes with every app licence, is sold standalone to consumers, and is included in Edge Consortium tuition; students do paid real work offset against tuition** `tok:decision:01m69k8c61zde0er6j94yb4ekz`: X-X.TECH decided on 31 October 2026: the estate is four applications (EDGE Finance, EDGE Clinic, The Edge Consortium, Public EDGE) plus Judiverse, where JudeAI lives with all its other applications. Any organisation that pays for any of the four gets full Judiverse access included, subject to rate limits. Consumers who are not business customers of any of the four buy Judiverse standalone, either by the capabilities they want or as a whole subscription. Anyone enrolled at The Edge Consortium gets Judiverse included in their tuition, with no extra cost for course materials. Their assignments are real-life projects, sandboxed, solving real problems; they can work freelance for partner institutions and clients, and every piece of work that produces a real solution is paid and can be offset against their tuition, never taken as free work. They can be signed on as long-term contractors for clients where that agreement is made. The Edge Consortium petitions for student finance to fund the course. The aim is to create the most employment anyone has created. - **London South Bank University is the first partner institution; SITHub teaches networking in the degree; The Edge Consortium registers with the Office for Students, shares LSBU's campus, and partners with LSBU for R&D and international technical students** `tok:decision:01m69m36psxrwhsa148z746ppv`: X-X.TECH decided on 31 October 2026: London South Bank University (LSBU) is The Edge Consortium's first partner institution for the governed intelligence degree, and one of the first universities to offer it; every university should offer it, starting with South Bank. LSBU's SITHub brings its networking capabilities as a module in the degree, so LSBU is a flagship too. The Edge Consortium registers with the Office for Students, can share the campus with LSBU, and offers itself to LSBU as a partner institution for research and development and for international technical students. - **Scope frozen on 31 October 2026; every later change enters through scope change intake** `tok:decision:01m69p22005ent1mresdjz6sh8`: X-X.TECH decided on 31 October 2026: the whole scope is known, so it is frozen. The frozen scope is every brief and unit in the new estate as built on this date, recorded by digest in spec/SCOPE-FREEZE.json. From now on a change to the frozen scope enters only through foundation/scope-change-intake as a recorded change request, and the gate reports any change without one. What remains is plugging in external models and datasets through the plug registry (spec/plugins.json), which does not change the scope. - **The GitHub branch contract (fingerprint d1233bfd) governs the build, untouched** `tok:decision:01m69thft6cmgyfgkjrt6bdrws`: X-X.TECH decided on 31 October 2026: the contract on branch claude/clever-ramanujan-9w05c1, fingerprint d1233bfd3bbc288bae594e21c6744e3501fe50f37dbd7a57f0a5068b768634bc, governs the closure build. Its bytes and the kernel stay untouched; the closure gate runs beside tools/gate.py. Other checkouts pull this contract. - **Three domains, one site: judiverse.tech canonical, judeai.tech and x-x.tech mirror it; nothing is named estate** `tok:decision:01m69thg13xhssvqbxygtr2zp8`: X-X.TECH decided on 31 October 2026: judiverse.tech is the base and canonical domain, judeai.tech serves professional services and x-x.tech is the neutral company domain. All three, and their www names, serve exactly the same site and directory, with one canonical link to judiverse.tech. Nothing is named estate; the word only describes what the whole is. - **TRL 9 and CRL 9 are reached when the closure gate passes on the deployed commit with the plug set empty** `tok:decision:01m69thg7kks8fwaj0266pf1jy`: X-X.TECH decided on 31 October 2026: TRL 9 and CRL 9, as the owner defines them, are reached when tools/closure.py passes on the deployed commit: every obligation in the frozen scope has reproduced evidence from a bound receiver, a browser video of every human journey, a record of every agent journey and a clean independent rerun with no unexplained differences, and the plug set is empty. Nothing else may claim either level. - **Jude carries out the whole plan: builds, reproduces, runs, deploys and maintains the system itself** `tok:decision:01m69xj6emef4jkw7kgv63bsdz`: X-X.TECH decided on 31 October 2026, as a hard requirement: Jude must be able to carry out the whole plan and build, reproduce, run, deploy and maintain the system without a human or an outside agent. Barachiel runs on the system's own host with the build, deploy, hub, mail, SMS and voice MCP servers, executes the plan's tasks through the racetrack, and lets the closure gate decide when each is done. The only outside act is the first start on the host; the only reserved human act is the analog switch for irreversible outcomes. Any plan step that only a human or an outside agent can perform is a defect. - **Money provenance: trust from origin and activity, face value at par, four assessments kept separate** `tok:decision:01m6a4t0fxra5avfxf0pr2p692`: The value of trust in money comes from where it was obtained and what activity supports it, never from its quantity and never by discounting its face value. The bank issues signed provenance certificates that state their verification scope and what remains unknown; provenance coverage, evidence reliability, financial-crime findings and investment risk are four separate assessments, never one score; attribution is pro rata and exact so that moving money cannot manufacture credibility; the financing graph reports independent demand share and circular funding as dependence, not wrongdoing; capital is followed to delivered work and measured resources; lending decisions carry reasons, a correction route and a named person. Decided by the Owner on 31 October 2026; admitted through scope-change intake 2026-10-31-money-provenance. - **The plug set is post-closure: closure is computed on everything the estate itself executes; external receivers bind by one-tap plug or an agent racetrack afterwards** `tok:decision:01m6a7p9c8566pe659w2hpgbjq`: Owner, 1 November 2026 00:17Z: closure should not depend on external providers. The estate builds the one-tap plug-and-play mechanism (tools/plugs.py: plugs_plug, plugs_drop) and the agent racetracks that bind an external receiver in minutes, so external slots without a credential are reported as the plug set and are not counted against closure. Closure passes when every obligation outside the plug set is evidenced on the deployed commit and every check passes. Binding each external receiver, and proving the journeys that need it, is post-closure work recorded per slot. - **Application names: EDGE Consortium and Judiverse Work Suite** `tok:decision:01m6a8mtbg94twxzcdgjp4yeac`: Owner, 1 November 2026 00:35Z: the applications are named EDGE Consortium (not The Edge Consortium) and Judiverse Work Suite (not Judiverse). The contract application records carry these names; every screen, catalogue entry, document and generator reads them from the contract. - **EDGE Finance is called EDGE** `tok:decision:01m6acbx9rv6x4yxh5gxkz1x30`: Owner, 1 November 2026 01:40Z: call EDGE Finance EDGE, so there is no choosing between EDGE ID, EDGE Bank and EDGE Finance. The contract application record carries the name; every screen, catalogue entry and document reads it from the contract. - **Application names: EDGE Public (was Public EDGE) and Judiverse Suite (was Judiverse Work Suite)** `tok:decision:01m6aeafjpfwmamsqm7apf2wfj`: Owner, 1 November 2026 02:14Z: 'I want the word work dropped from the suite, so it's just Judiverse Suite. Also, I want public edge to be edge public.' The contract's applications read EDGE, EDGE Clinic, EDGE Consortium, EDGE Public and Judiverse Suite; formal names JudeAI / EDGE Public and Judiverse Suite. Every rendered name comes from the contract; unit copy that still carries the old words is lineage, overridden at render. - **The finance application is called EDGE Bank** `tok:decision:01m6aennyqz91exa4a2sda0019`: Owner, 1 November 2026 02:22Z: 'I want EDGE to be called EDGE Bank again, we will offer the capabilities within them by their own names but I want it to be known that it is a bank.' The finance application's name is EDGE Bank (formal name JudeAI / EDGE Bank); its capabilities keep their own names (Open a personal current account, Send money, ...). Supersedes the 1 November decision that shortened the name to EDGE. - **Screens come from the reference flows; the estate adds only its controls and journeys** `tok:decision:01m6bgf72jtm0n7xyba043bqh4`: X-X.TECH decided on 1 November 2026 (04:10Z to 04:14Z): layouts and screens come from the reference flows recorded in research/ (Mobbin and Refero: Microsoft 365, Notion, Monday, Slack, Fabric, Starling, Monzo, Google Drive, ChatGPT, Base44 V7BRO); the only difference from a reference is the estate's added controls or journeys; nothing is hand-drawn. This resolves the contract's open decision on whether screens mirror a competitor's exact layout. - **EDGE Public Consortium's mark is consortium.svg from the approved brand register** `tok:decision:01m6bgf72knqzc3da9n20c282a`: X-X.TECH supplied the approved brand register on 31 October 2026 (design/brand/approved/BRAND-REGISTER.json): consortium.svg is 'The Edge Consortium application mark'. Under the one contract the Consortium and the public face are one application, so public.svg is not shown as a separate application and cyber-governance.svg is not an application mark. This resolves the contract's open decision on the Consortium's mark and supersedes tok:decision:01m68c87qmz54aqnbm8pqzmgt4 where the two differ. ##### 3. The owner's requests (59) | # | Request | Status | Shown by | |---|---|---|---| | 1 | Five applications: EDGE Finance, EDGE Clinic, The Edge Consortium, Public EDGE, Judiverse | DONE | `spec/units/finance`, `spec/units/clinic`, `spec/units/edge-consortium`, `spec/units/public-edge`, `spec/units/judiverse` | | 2 | Four applications: EDGE Finance, EDGE Clinic, EDGE Public Consortium, Judiverse | OPEN: missing spec/units/public-consortium | `spec/units/finance`, `spec/units/clinic`, `spec/units/public-consortium`, `spec/units/judiverse` | | 3 | Every capability, screen, control, journey and endpoint specified, nothing left to imagination | DONE | `build/estate-spec.json`, `generated/diagrams/README.md` | | 4 | Mermaid diagrams and click-by-click breakdowns for every journey | DONE | `generated/diagrams/README.md` | | 5 | No dead controls anywhere | DONE | `ledger/SWEEP.md` | | 6 | Every competitor feature matched like for like, with proof | DONE | `ledger/parity/README.md`, `docs/proof/README.md` | | 7 | Competitor behaviour read from their own documentation | DONE | `tools/behaviour.py`, `research/parity/context.jsonl` | | 8 | Outcomes matched to competitors' documented outcomes, with baselines and targets | DONE | `ledger/outcomes/README.md`, `spec/standards.json` | | 9 | Novel measures: intent to outcome, words, prompts, tokens, cost and carbon per verified outcome | DONE | `spec/standards.json`, `runtime/engine.py` | | 10 | Help centre, admin guides and API reference set against competitors | DONE | `docs/README.md` | | 11 | The blueprint standard so others can reproduce it | DONE | `docs/STANDARD.md` | | 12 | Reference code and open-source reuse for every operation | DONE | `docs/reference/README.md` | | 13 | A running engine, HTTP server and agent tools generated from the specification | DONE | `runtime/engine.py`, `runtime/server.py`, `runtime/mcp_server.py` | | 14 | Screens generated in Figma from the specification, never hand-drawn | DONE | `design/figma/FRAMES.json` | | 15 | Every October 2026 headline: root cause, protective capability, benchmark | DONE | `ledger/problems/README.md`, `ledger/benchmarks/README.md` | | 16 | Vulnerability map of the whole estate | DONE | `ledger/vulnerabilities/README.md` | | 17 | Provisional value with in-person owner sign-off; reversal by design | DONE | `runtime/engine.py`, `ledger/benchmarks/binance-replay.json` | | 18 | Agents bound to their own devices and resources | DONE | `runtime/credentials.py`, `ledger/benchmarks/binance-replay.json` | | 19 | Intention triage: secure, efficient and reproducible by design | DONE | `tools/intent_triage.py`, `runtime/triage_mcp.py` | | 20 | The FCA's agentic framework exceeded point by point | DONE | `research/finance/FCA-AGENTIC-FRAMEWORK.md` | | 21 | Zero waste: monitored, blocked and impossible to bypass | DONE | `tools/waste.py`, `tools/agent_gate.py`, `tools/install_gate.sh`, `ledger/WASTE.md` | | 22 | Agent mailbox through Judiverse Mail | DONE | `runtime/mail_mcp.py` | | 23 | Agent credentials | DONE | `runtime/credentials.py` | | 24 | Agent SMS: receive free, never spend | DONE | `runtime/sms_mcp.py` | | 25 | Reversibility in health register, including impact reversal for conditions that are part of who people are | DONE | `research/clinic/REVERSIBILITY.md` | | 26 | Regulator, auditor and reviewer workspaces with their workflows | DONE | `spec/units/foundation/regulator-auditor-reviewer-workspace.json` | | 27 | Pre-onboarded organisations from public registers, claimed by magic link | DONE | `spec/units/foundation/pre-onboarded-organisations.json`, `spec/units/foundation/claim-organisation-by-magic-link.json` | | 28 | Automated outreach, scheduling and reviewer inboxes | DONE | `spec/units/foundation/automated-review-outreach-and-scheduling.json` | | 29 | Free evaluation, commercial licensing and enforcement evidence | DONE | `spec/units/foundation/evaluation-and-commercial-licensing.json`, `spec/units/foundation/licence-enforcement-evidence.json` | | 30 | Estate gap analysis | DONE | `ledger/GAPS.md` | | 31 | Mail relay credentials in the environment | OPEN: missing NEEDS THE OWNER | `NEEDS THE OWNER` | | 32 | The SMS Works key and inbound forwarding | OPEN: missing NEEDS THE OWNER | `NEEDS THE OWNER` | | 33 | Hosting decision and deployment | OPEN: missing NEEDS THE OWNER | `NEEDS THE OWNER` | | 34 | Real receiving systems connected (banks, NHS, registries) | OPEN: missing NEEDS THE OWNER | `NEEDS THE OWNER` | | 35 | The Edge Consortium for education and research, separate from Public EDGE for government | DONE | `contract/estate-contract.json`, `tools/apps.py`, `spec/units/edge-consortium`, `spec/units/public-edge` | | 36 | The governed intelligence degree: modules, real-work assessment (coursework with proof, live challenges, live spec-driven development, live audio), set against leading programmes | DONE | `research/edge-consortium/GOVERNED-INTELLIGENCE-DEGREE.md`, `tools/degree.py`, `spec/units/edge-consortium/governed-intelligence-degree.json`, `spec/units/edge-consortium/run-a-live-assessment.json` | | 37 | Offered with leading institutions and by The Edge Consortium | DONE | `spec/units/edge-consortium/offer-the-degree-with-a-partner-institution.json` | | 38 | Scope discipline: scope creep managed as raw material | DONE | `spec/units/foundation/scope-change-intake.json` | | 39 | Bottleneck and blocker science: removed at the root, shortest intention to outcome | DONE | `spec/units/foundation/bottleneck-and-blocker-removal.json` | | 40 | No irreversible outcome without the owner's analog, in-person act, with identical backup hardware | DONE | `research/public-edge/ANALOG-HUMAN-SWITCH.md`, `runtime/engine.py`, `spec/units/public-edge/analog-human-switch-for-irreversible-actions.json`, `ledger/benchmarks/analog-switch.json` | | 41 | Judiverse included with every licence of the four applications, subject to rate limits; standalone for consumers by capability or whole | DONE | `spec/offers.json`, `runtime/entitlements.py`, `spec/units/judiverse/judiverse-included-with-every-licence.json`, `spec/units/judiverse/judiverse-standalone-subscription.json` | | 42 | Judiverse included in Edge Consortium tuition; real sandboxed projects; paid student work offset against tuition; contractor pathway; student finance petition | DONE | `spec/units/edge-consortium/judiverse-included-in-tuition.json`, `spec/units/edge-consortium/real-project-assignments.json`, `spec/units/edge-consortium/paid-student-work-with-tuition-offset.json`, `spec/units/edge-consortium/student-to-contractor-pathway.json`, `spec/units/edge-consortium/petition-for-student-finance.json`, `ledger/benchmarks/offers.json` | | 43 | Specialisms only after every domain and how they connect: medical, financial, social, public, research, defence, environmental and legal intelligence | DONE | `spec/degree.json`, `runtime/programme.py`, `ledger/benchmarks/programme.json`, `spec/units/edge-consortium/choose-a-specialism.json` | | 44 | Every new screen in Figma with progress, navigation and monitors from the journey diagrams; no dead controls | DONE | `design/figma/FRAMES.json`, `tools/expand.py`, `generated/diagrams/README.md` | | 45 | Clean, brief, permanent URLs and a URL directory from the ontology | DONE | `tools/urls.py`, `spec/urls.json`, `docs/URLS.md` | | 46 | London South Bank University as first partner institution; SITHub networking module; shared campus; R&D and international technical students | DONE | `research/edge-consortium/GOVERNED-INTELLIGENCE-DEGREE.md`, `spec/units/edge-consortium/share-a-campus-with-a-partner-university.json`, `spec/units/edge-consortium/research-and-development-partnership.json`, `spec/units/edge-consortium/international-technical-students.json` | | 47 | Register The Edge Consortium with the Office for Students | DONE | `spec/units/edge-consortium/register-with-the-office-for-students.json` | | 48 | All research compiled into one agent and human markdown, organised by the ontology, with every finding accounted for and carried | DONE | `research/AGENT-RESEARCH.md`, `tools/compendium.py`, `tools/carry.py`, `spec/carry.json` | | 49 | The owner's white papers, frontier agent capability reference and SITHub deck built into the degree | DONE | `research/edge-consortium/SOURCES-2026-10-31.md`, `research/edge-consortium/GOVERNED-INTELLIGENCE-DEGREE.md` | | 50 | Scope frozen; every later change only through a change request | DONE | `spec/SCOPE-FREEZE.json`, `tools/freeze.py`, `ledger/SCOPE-CHANGES.md` | | 51 | Every guard enforced: bound by family to a deterministic rule or a plug slot | DONE | `runtime/guards.py`, `runtime/plugins.py`, `spec/plugins.json` | | 52 | Plug-and-play mechanism for external models and datasets: one line to plug, one to test | DONE | `tools/plugs.py`, `docs/PLUGS.md` | | 53 | Every capability in the TOK catalogue | DONE | `tok/catalogue.jsonl`, `tools/catalogue.py` | | 54 | Every old-estate document read and its findings placed | DONE | `tools/oldestate.py`, `ledger/old-estate/README.md` | | 55 | Every capability and workflow as an MCP; one hub for search, call and the racetrack | DONE | `runtime/hub_mcp.py`, `runtime/mcp_server.py`, `docs/MCP.md`, `ledger/benchmarks/mcp.json` | | 56 | Agents do everything themselves on a racetrack: intention to verified outcome, rehearsed first, protected by the engine | DONE | `runtime/racetrack.py` | | 57 | Drag-and-drop plugs agents can use themselves | DONE | `plugs/README.md`, `tools/plugs.py` | | 58 | Figma screens without uploads: compressed calls inflated in Figma | DONE | `design/figma/inflate.js`, `design/figma/emit_figma.py` | | 59 | Signed agreements: Office for Students registration decision, LSBU partnership and shared-campus terms, a student sponsor licence for international students | NEEDS THE OWNER | | ##### 4. The specification - **Standards** (`spec/standards.json`): `schema`, `statement_class`, `agreed_by`, `what_this_is`, `beat_the_best`, `per_operation`, `every_operation`, `every_capability`, `outcome_economics` - **Offers** (`spec/offers.json`): included-with-app-licence: an organisation with an active paid licence for any of the four applications; included-in-tuition: a student enrolled at The Edge Consortium; standalone-whole: a consumer with no business licence for any of the four; standalone-by-capability: a consumer with no business licence for any of the four - **Degree** (`spec/degree.json`): 30 modules; 9 pathways (Governed Intelligence (Medical Intelligence), Governed Intelligence (Financial Intelligence), Governed Intelligence (Social Intelligence), Governed Intelligence (Public Intelligence), Governed Intelligence (Research and Educational Intelligence), Governed Intelligence (Defence and Security Intelligence), Governed Intelligence (Environmental and Energy Intelligence), Governed Intelligence (Legal Intelligence), Governed Intelligence (Network and Edge Intelligence)); gate: GI300 and every domain worked 2 times - **Capabilities by application**: clinic 342, edge-consortium 363, finance 391, foundation 9, judiverse 22, public-edge 489; URL directory: `docs/URLS.md` ##### 5. Proofs: benchmarks, tests and baselines - `ledger/benchmarks/analog-switch.json`: {"truth_status": "SIMULATION", "operation": "public-edge/analog-human-switch-for-irreversible-actions/execute-the-irreversible-action-once-onl", "ai_attempts": 1000, "executed_by_ai": 0, "held_for_analog_act": 1000, "revoked_unit_executed": false, "backup_unit_executed": true, "second_act_on_same_request": "already executed", "person_through_software": "ANALOG_HUMAN_ACT_REQUIRED"} - `ledger/benchmarks/binance-replay.json`: {"truth_status": "SIMULATION", "operation": "finance/crypto-wallet/send-crypto"} - `ledger/benchmarks/mcp.json`: {} - `ledger/benchmarks/offers.json`: {"consumer_by_capability": "standalone-by-capability", "consumer_whole": "standalone-whole", "below_pay_floor": "BELOW_PAY_FLOOR", "over_visa_hours": "WORK_HOURS_LIMIT", "sandboxed": "SANDBOXED", "client_takes_without_paying": "NOT_PAID", "paid_and_deployed": true} - `ledger/benchmarks/programme.json`: {} - `ledger/benchmarks/README.md`: Generated by `tools/bench.py` on `runtime/engine.py`. Synthetic test inputs; local effect (default adapter), so this is the estate's own processing time, not a receiving system's. Repeat with `python3 tools/bench.py `. / Operations / Verified outcomes / Median p50 / Worst p95 / Prompts / Tokens / Deterministic share / /---/---/---/---/---/---/---/ / 8534 / 0 / None ms / None ms / 0 / 0 / - / - `ledger/parity/README.md`: Generated by `tools/match.py`. One line per competitor clause in `research/parity/`: the unit, screen, control, operation and endpoint that covers it, and how it was placed. Every clause is covered. / Application / Clauses / On existing steps / In new capabilities / /---/---/---/---/ / clinic / 381 / 0 / 0 / / edge-consortium / 3075 / 0 / 0 / / finance / 968 / 0 / 0 / - `ledger/outcomes/README.md`: Generated by `tools/outcomes.py`. One row per operation. The **floor** is the competitor's best published figure (`research/outcomes/*.jsonl`, or a figure on the competitor's own page); the **target** is the estate's own measurement (`ledger/benchmarks/results.jsonl`, `ledger/evidence/operations.jsonl`, `ledger/evidence/journeys-*.jsonl`) and stays **OPEN** until one exists. No number is halved, moved halfway or otherwise derived. **Outcome matched** is true only when the capability's bound reference flow (`research/parity/REFERENCE-FLOWS.jsonl`) records the screens that cover the step this operation sits at in the person's journey. A competitor clause or figure found by rare-word overlap is - `ledger/RUN.md`: Rehearsed as a person: {"OK:None": 25916, "REFUSED:INVALID_INPUT": 64276, "REFUSED:CONSENT_REQUIRED: we need your agreement before we can use health details": 1, "REFUSED:EMPTY_DESCRIPTION": 1, "REFUSED:CONSENT_NOT_GIVEN": 2, "REFUSED:FORMULATION_INCOMPLETE": 1, "REFUSED:REFERENCE_MISSING": 1, "REFUSED:NOTE_EMPTY": 1, "REFUSED:DESCRIPTION_MISSING": 1, "REFUSED:CONSENT_NOT_GIVEN: emergency guidance is still shown": 1, "REFUSED:NOTE_TOO_LONG": 1, "REFUSED:DATES_INVALID": 1, "REFUSED:REASON_REQUIRED": 10, "REFUSED:EXPIRY_TOO_LONG": 1, "REFUSED:NOT_ATTESTED": 1, "REFUSED:CONSENT_REQUIRED": 4, "REFUSED:INCOMPLETE": 2, "REFUSED:LIMITS_INCONSISTENT: per-payment must not exceed daily, daily must not - `ledger/SWEEP.md`: Generated by `tools/sweep.py`. Deterministic: the same estate gives the same sweep. Total problems: **9098** across 1616 units. - rule 5: 74 - rule 6: 20 - rule 8: 9004 Competitor matches graded by wording: 31. This measures shared wording only; behaviour is measured by `tools/closure.py`. The gap analysis is `ledger/GAPS.md`; external plugs `docs/PLUGS.md`; old-estate coverage `ledger/old-estate/README.md`; changes since the scope freeze `ledger/SCOPE-CHANGES.md`. ##### 6. Research findings, each placed on the capability that serves it Every bullet, table row and stated sentence in every research file. Each has an ID (`F-` and a digest of its file and text), the heading it sits under, its statement class where the research gave one, and where it lives in the estate: the capability's clean URL (a finding no other capability carried is a checked requirement of a `findings-…` capability, made by `tools/carry.py`), or **UNPLACED**. ###### research ###### `research/README.md` (15 findings) - `F-24506a7770` L3 _Research_: This folder holds research written for the estate. It informs what the four applications must do; it is never authority. Every finding carries its source link and the date it was found, and is marked as one of: → **UNPLACED** - `F-3710803a42` L5 [OBSERVATION] _Research_: **OBSERVATION** (seen in a primary source or a real product), → **UNPLACED** - `F-8044582fe9` L6 [AGREEMENT] _Research_: **AGREEMENT** (a law, regulation or standard, with its status and dates), → **UNPLACED** - `F-a4d104b776` L7 [CLAIM] _Research_: **CLAIM** (what a company or document says about itself, not independently checked), → **UNPLACED** - `F-5f1e244c82` L11 _Research_: Raw material (downloaded pages, datasets, screenshots, earlier work) stays in storage and is referenced by digest; only the written research lives here. → **UNPLACED** - `F-db6299f724` L15 _Research › How to read it_: Start with the application you are working on. Each topic file opens with what it is, how it was gathered and what it does not cover. → **UNPLACED** - `F-e90307ca3e` L19 _Research › How to read it_: `finance/` / EDGE Finance: competitors and our lane, regulators in the UK and worldwide, agent assurance gaps and reusable open-source systems, confidence reporting, physical presence and solution banks, real product flows → **UNPLACED** - `F-d612990d9c` L20 _Research › How to read it_: `clinic/` / EDGE Clinic: social prescribing, symptom checking, early detection and wearables; competitors, regulators and reusable systems; real product flows → **UNPLACED** - `F-6f506ac037` L21 _Research › How to read it_: `public-consortium/` / EDGE Public Consortium: competitors, regulators and reusable systems for cyber governance, procurement, vetting, logistics and education; real product flows → **UNPLACED** - `F-160d6107cb` L22 _Research › How to read it_: `judiverse/` / Judiverse: competitors, regulators and reusable systems for office, assistants, agent lab and music; real product flows → **UNPLACED** - `F-f2819a68ed` L23 _Research › How to read it_: `experience/` / Customer experience patterns from any industry (onboarding, errors, confirmations, help, accessibility and more), useful to every application → **UNPLACED** - `F-ef93d6256a` L24 _Research › How to read it_: `market/` / What is being funded and bought (grants, contracts, venture capital) and the costliest real problems by industry → **UNPLACED** - `F-3404e6574c` L25 `legacy/` / What the organisation's earlier work already contains (competitors, benchmarks, readiness, blueprints, features, standards, papers, guides, strategy, funding), read as research so none of it is wasted or repeated → old-estate inventory (section 8) - `F-e91d633757` L29 _Research › Real product flows_: The `REFERO-FLOWS` files come from Refero, a library of real app and website screens with each journey described step by step. Results from other industries are kept on purpose, because they are often the best customer experience to learn from. Refero holds mostly consumer and start-up products: many incumbents (for example Starling, Monzo, Chase UK) and most government, enterprise and NHS systems are not in it, so these files are a sample of the floor, not a census. → **UNPLACED** - `F-03f2f56b26` L33 _Research › Coverage_: What exists in each folder, and what is still being researched, is listed by `python3 tools/build.py`, which writes the generated index. Nothing in this folder states a count of estate objects; those are produced live by the build. → **UNPLACED** ###### `research/REUSE-ENGINES.md` (20 findings) - `F-ac0afe9133` L3 [CLAIM] _Engines behind the applications: what is reused, what is built_: Statement class: research and recommendation, 1 November 2026 03:35Z. OBSERVED where the estate's own research verified it (file and date given); CLAIM where it has not. Never a rule. → **UNPLACED** - `F-9b1cf417e8` L7 _Engines behind the applications: what is reused, what is built › What the specification actually needs (counted from spec/receivers.json, research and competitor units excluded)_: Application / Operations / Workflow steps on a record or case (journey state, ours) / Evidence reads (ours) / Domain entities that need an engine → **UNPLACED** - `F-370471269f` L9 _Engines behind the applications: what is reused, what is built › What the specification actually needs (counted from spec/receivers.json, research and competitor units excluded)_: EDGE Bank / 3,301 / 1,452 / 378 / account, payment, card, member, payee; rail: payment-scheme (48 operations) → **UNPLACED** - `F-b62a6a9957` L10 _Engines behind the applications: what is reused, what is built › What the specification actually needs (counted from spec/receivers.json, research and competitor units excluded)_: EDGE Clinic / 2,284 / 1,376 / 338 / FHIR resources (69), consent (34), identity claim (28), safety plan (14), plan (11); NHS services (29), detection models (11) → **UNPLACED** - `F-fdf530ba5f` L11 _Engines behind the applications: what is reused, what is built › What the specification actually needs (counted from spec/receivers.json, research and competitor units excluded)_: EDGE Consortium / 4,778 / 1,942 / 359 / case (74), egress and access policy (89), role (52), module and client (84), consortium (32) → **UNPLACED** - `F-34dd5d8e62` L12 _Engines behind the applications: what is reused, what is built › What the specification actually needs (counted from spec/receivers.json, research and competitor units excluded)_: EDGE Public / 3,225 / 1,830 / 486 / application (41), vacancy (18), asset (18), alert (18), decision (15) → **UNPLACED** - `F-6e2b90b08e` L13 _Engines behind the applications: what is reused, what is built › What the specification actually needs (counted from spec/receivers.json, research and competitor units excluded)_: Judiverse Suite / 3,770 / (steps spread across) / intention (113), contact (91), upload terms (90), template (82), event (65), task (44), document versions and objects (98) → **UNPLACED** - `F-c80d9d17e2` L15 _Engines behind the applications: what is reused, what is built › What the specification actually needs (counted from spec/receivers.json, research and competitor units excluded)_: Reading: about three quarters of every application is journey state and evidence, which the estate already runs. The engine question is the last column: a small set of domain entities per application. One engine per application covers it. → **UNPLACED** - `F-89dd1f6e1e` L21 _Engines behind the applications: what is reused, what is built › The engine per application_: EDGE Bank / Griffin (sponsor bank, rail) + Apache Fineract (deposits, loans, accounting) + TigerBeetle (ledger) + Hyperswitch (card and alternative processors) / accounts, balances, payments, Direct Debits, standing orders, loans, overdrafts, cards / Griffin: PRA/FCA bank, sandbox self-serve (OBSERVED 1 Nov, research/finance/ENGINE-BY-STARLING.md §3). Fineract Apache-2.0, TigerBeetle, Hyperswitch Apache-2.0 (OBSERVED 31 Oct, research/finance/ASSURANCE-AND-REUSE.md §2) / `payment-scheme` and `card-processor` slots; runtime/domains/bank.py already holds the ledger interface, so Fineract or TigerBeetle replaces its SQLite tables without touching a screen → **UNPLACED** - `F-a0b68ed85a` L22 [CLAIM] _Engines behind the applications: what is reused, what is built › The engine per application_: EDGE Clinic / Medplum (FHIR R4 server, UK Core, scheduling, auth, bots, React components) / patients, appointments, consents, observations, care plans, questionnaires, referrals / Apache-2.0 (OBSERVED 31 Oct, research/clinic/COMPETITORS-REGULATORS-REUSE.md); ONC certification CLAIM / `feature_fhir`, consent, safety_plan and plan entities map to FHIR resources through one adapter; `nhs-services` stays a plug → **UNPLACED** - `F-7ada38d1fb` L23 [CLAIM] _Engines behind the applications: what is reused, what is built › The engine per application_: EDGE Consortium / ERPNext (Frappe) for roles, modules, clients, access and cases; Keycloak or Ory for identity federation / the university's records: enrolment, modules, roles with expiry, cases, policies / ERPNext GPL-3 (CLAIM, not verified by the estate; network copyleft to assess), Keycloak Apache-2.0 (OBSERVED 31 Oct) / case, role, module, client, access entities through the Frappe REST API as one adapter → **UNPLACED** - `F-ffe52ed0e3` L24 _Engines behind the applications: what is reused, what is built › The engine per application_: EDGE Public / Same ERPNext instance (applications, vacancies, assets, alerts, decisions) / government-on-the-estate records and decisions / as above / application, vacancy, asset, alert, decision entities → **UNPLACED** - `F-13daa7e74f` L25 [CLAIM] _Engines behind the applications: what is reused, what is built › The engine per application_: Judiverse Suite / ONLYOFFICE Docs (documents, sheets, decks), Matrix (Synapse or Conduit) for messages and calls with LiveKit for media, a CalDAV server for calendar, Simwood for UK numbers / editing, collaboration, chat, calls, meetings, calendar, mail (mail already runs natively) / ONLYOFFICE AGPL-3 community with a separate embedding licence (CLAIM); Matrix Synapse AGPL-3 (CLAIM); LiveKit Apache-2.0 (CLAIM); Simwood commercial API (CLAIM) / document, template, event, task, contact entities; `voice-telephony` and `sms-gateway` slots → **UNPLACED** - `F-19b8690980` L26 [CLAIM] _Engines behind the applications: what is reused, what is built › The engine per application_: Agents operating other software / Composio (hosted) or the estate's own MCP hub / third-party tools under per-user authentication / Composio hosts data outside the estate: assess before UK-only use (CLAIM) / the hub already exposes every operation to agents (`estate_call`, `racetrack_run`); Composio adds outbound integrations only → **UNPLACED** - `F-404cde90b5` L30 _Engines behind the applications: what is reused, what is built › The pattern (unchanged)_: approved screen → Jude's authenticated operation (identity, mandate, guards, rehearsal, evidence: ours) → engine adapter (runtime/adapters.ADAPTERS, one module per engine in runtime/domains/) → the engine's own confirmation read back → evidence. TOK, EDGE ID and assurance stay shared. Nothing is hand-built that an engine supplies; what is built is the adapter and the domain behaviour no engine has (money provenance, agent mandates, glass-box evidence). → **UNPLACED** - `F-0caf5250c0` L34 _Engines behind the applications: what is reused, what is built › Order_: EDGE Bank: Griffin key, then Fineract behind runtime/domains/bank.py. The screens and ledger interface exist; the customer sees the same app with real rails. → **UNPLACED** - `F-2b828dc610` L35 _Engines behind the applications: what is reused, what is built › Order_: EDGE Clinic: Medplum on Civo; one FHIR adapter for the five clinic entities. Largest single coverage gain after the bank. → **UNPLACED** - `F-a0902e9c3c` L36 _Engines behind the applications: what is reused, what is built › Order_: Judiverse Suite: ONLYOFFICE and Matrix, because documents and messages are what every other application's people use daily. → **UNPLACED** - `F-b7aa6059fb` L37 _Engines behind the applications: what is reused, what is built › Order_: EDGE Consortium and EDGE Public: one ERPNext instance for both. → **UNPLACED** - `F-d612b5b80a` L39 _Engines behind the applications: what is reused, what is built › Order_: Each engine lands as a named service on Civo with backup proven first, then its adapter binds, then the affected journeys are walked and their results read back from the engine. → **UNPLACED** ###### clinic ###### `research/clinic/COMPETITORS-REGULATORS-REUSE.md` (157 findings) - `F-a4dc55d42f` L3 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems_: Research for EDGE Clinic, one of four applications in JudeAI / EDGE ID. Written 31 October 2026 for a brand-new engineer. → **UNPLACED** - `F-32dfea2efd` L4 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems_: This is research, not clinical or legal advice. Anything here must be checked by a qualified clinical safety officer and a lawyer before it drives a decision. → **UNPLACED** - `F-dc38d0f9ac` L6 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems_: **Scope.** Four topics only: (1) competitors, (2) regulators and rules, (3) agent assurance gaps in health, (4) open standards and open-source to reuse. → **UNPLACED** - `F-c592024ca8` L7 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems_: Social prescribing evidence, symptom checkers, early detection, wearables, physical presence and NHS costs are covered elsewhere, in `research/clinic/SOCIAL-PRESCRIBING-AND-DETECTION.md`, and are not repeated here. → **UNPLACED** - `F-be77430c40` L13 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › How to read the labels_: **OBSERVED** / I read it directly in a primary page or file during this research (for example, a licence file or a regulator's page). → **UNPLACED** - `F-0393cd2dcc` L14 [CLAIM] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › How to read the labels_: **CLAIM** / A vendor, a law firm, a trade paper or a blog says it. Not independently checked. → **UNPLACED** - `F-3a8f35cf39` L15 [AGREEMENT] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › How to read the labels_: **AGREEMENT** / A law, rule, standard or guidance that is adopted and in force, with a date. → **UNPLACED** - `F-cff7eb5259` L16 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › How to read the labels_: **PROPOSED** / A draft, consultation, plan or recommendation. Not binding yet. → **UNPLACED** - `F-5bc72f3028` L17 [UNKNOWN] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › How to read the labels_: **UNKNOWN** / Sources disagree, or I could not find a reliable answer. → **UNPLACED** - `F-3bbb9ca7c6` L19 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › How to read the labels_: Every link was accessed on 2026-10-31. Where a source is a secondary summary rather than the regulator's own text, the line says so. → **UNPLACED** - `F-91a9d23b3a` L25 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 0. The ten things to know first_: **The UK has no AI-specific medical device law yet.** The MHRA (Medicines and Healthcare products Regulatory Agency, the UK device regulator) put draft amending regulations out in May 2026, and they are expected to apply from June 2027. They have not passed Parliament. PROPOSED. [Latham & Watkins](https://www.lw.com/en/insights/mhra-publishes-draft-amendment-to-the-uk-medical-devices-regulations), accessed 2026-10-31. → `/clinic/register-device-fault` - `F-3a9d1910b6` L26 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 0. The ten things to know first_: **A national commission set out 44 recommendations for regulating AI in UK health on 10 September 2026**, including staged "L-plate" authorisations and continuous real-world monitoring. Government says it will "carefully consider" them. PROPOSED. [GOV.UK press release](https://www.gov.uk/government/news/independent-commission-led-by-nhs-doctors-sets-out-blueprint-to-accelerate-safe-ai-adoption-in-healthcare) (OBSERVED), accessed 2026-10-31. → `/clinic/apply-national-data-opt` - `F-307d6d0c82` L27 [AGREEMENT] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 0. The ten things to know first_: **The MHRA's test for "is it a medical device?" (29 July 2026):** a tool that only transcribes, summarises or drafts letters is not a device, but one that supports diagnosis, treatment or prevention, or "take[s] automated action such as placing orders without clinician review", is. AGREEMENT (guidance). [MHRA news](https://www.gov.uk/government/news/mhra-clarifies-regulatory-status-of-ambient-voice-technologies-used-in-the-nhs) (OBSERVED), accessed 2026-10-31. EDGE Clinic triage, prescribing safety and retinal screening fall on the device side of that line. → `/clinic/retinal-screening` - `F-af6e6fdef6` L28 [AGREEMENT] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 0. The ten things to know first_: **The NHS clinical safety standards DCB0129 and DCB0160 are being rewritten, with AI named as a reason.** The consultation ran 29 June to 11 September 2026. The current versions still apply. PROPOSED (revision); AGREEMENT (current versions). [NHS England consultation page](https://www.engage.england.nhs.uk/patient-safety/national-review-of-clinical-risk-management-standa/) (OBSERVED), accessed 2026-10-31. → `/clinic/run-clinical-safety-training` - `F-1e56f87017` L29 [AGREEMENT] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 0. The ten things to know first_: **The EU AI Act's high-risk rules for AI inside medical devices now apply from 2 August 2028**, moved back by the "Digital Omnibus on AI", Regulation (EU) 2026/1744, which came into force on 27 July 2026. AGREEMENT. [Hunton](https://www.hunton.com/privacy-and-cybersecurity-law-blog/eu-digital-omnibus-on-ai-enters-into-force) (OBSERVED, law-firm summary), accessed 2026-10-31. → `/clinic/act-high-risk-answer` - `F-490c9d5a70` L30 [AGREEMENT] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 0. The ten things to know first_: **In the US, FDA's lifecycle guidance for AI-enabled devices is still a January 2025 draft.** The final PCCP (predetermined change control plan) guidance is in force from December 2024. PROPOSED and AGREEMENT respectively. [RAPS](https://www.raps.org/resource/fda-s-device-center-releases-guidance-agenda-for-fy-2027.html); [Ropes & Gray](https://ropesgray.com/en/insights/alerts/2024/12/fda-finalizes-guidance-on-predetermined-change-control-plans-for-ai-enabled-device), accessed 2026-10-31. → `/clinic/front-door-check` - `F-7d260ea7e1` L31 [CLAIM] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 0. The ten things to know first_: **The big AI labs are now in consumer health.** OpenAI launched ChatGPT Health in January 2026, and Anthropic launched Claude for Healthcare in January 2026. Both are US-first and connect personal records. CLAIM. [Fortune](https://fortune.com/2026/01/07/openai-launches-chatgpt-health-in-a-push-to-become-a-hub-for-personal-health-data); [Anthropic](https://www.anthropic.com/news/healthcare-life-sciences), accessed 2026-10-31. → `/clinic/mental-health-care-crisis` - `F-cbff46c7fb` L32 [CLAIM] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 0. The ten things to know first_: **Evidence that "a clever model" is not "a safe front door":** in a randomised study of 1,298 UK participants, people who used LLMs did not choose better than controls, even though the models alone named conditions correctly 94.9% of the time. CLAIM (peer-reviewed). [University of Oxford](https://www.ox.ac.uk/news/2026-02-10-new-study-warns-risks-ai-chatbots-giving-medical-advice), accessed 2026-10-31. → `/clinic/front-door-check` - `F-79309f73f3` L33 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 0. The ten things to know first_: **No standard exists yet for an AI agent's identity, mandate and delegation in health.** The OpenID Foundation names these gaps and is drafting extensions. PROPOSED. [OpenID Foundation](https://openid.net/new-whitepaper-tackles-ai-agent-identity-challenges/), accessed 2026-10-31. → `/clinic/accept-proxy-mandate` - `F-dd093cb24d` L34 [CLAIM] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 0. The ten things to know first_: **There is no open, clinical-grade drug-interaction source.** The US NLM (National Library of Medicine) switched off its free interaction API around 2 January 2024. The BNF (British National Formulary) needs a licence for use inside systems. DrugBank's full data is non-commercial only. OBSERVED/CLAIM. [NLM RxNav](https://www.lhncbc.nlm.nih.gov/RxNav/news/APIUpdate.html); [Pharmaceutical Press licensing](https://www.pharmaceuticalpress.com/?p=146134) (OBSERVED); [DrugBank FAQ](https://dev.drugbank.com/guides/faqs), accessed 2026-10-31. → `/clinic/medication-safety-check` - `F-d8b3ef3d60` L44 [AGREEMENT] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: **NHS App** (NHS England) / The government says the app will be the "front door to the health service". The 10 Year Health Plan (July 2025) targets a "full front door to the entire NHS" by 2028. / The plan proposed "My NHS GP", an AI tool to route people to the right care. It is reported to reach 200,000 patients in year one, with full rollout by April 2028. No confirmed 2026 go-live found. / AGREEMENT (policy); PROPOSED (My NHS GP) / [GOV.UK oral statement](https://www.gov.uk/government/speeches/fit-for-the-future-health-and-social-care-secretarys-statement); [Full Fact tracker](https://fullfact.org/government-tracker/nhs-app-improvements/); [Digital Health](https://www.digitalhealth.net/2025/07/what-does-the-10-year-plan-mean-for-digital/), accessed 2026-10-31 → `/clinic/front-door-check` - `F-ad9a9b8dca` L45 [CLAIM] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: NHS App + Rapid Health Smart Triage / A GP booking and triage tool that works inside the NHS App. / Reported as available to more than a million patients in April 2026. / CLAIM / [Digital Health](https://www.digitalhealth.net/2026/04/ai-tool-allows-a-million-patients-to-book-appointments-via-nhs-app/), accessed 2026-10-31 → `/clinic/symptom-check-triage` - `F-b26e5e386b` L46 [CLAIM] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: NHS App messaging shift / NHS App leadership has described the app as "not the only" service, moving from "front door" towards "companion". / n/a / CLAIM / [Digital Health](https://www.digitalhealth.net/2026/03/nhs-app-to-move-from-front-door-to-companion/), accessed 2026-10-31 → `/clinic/front-door-check` - `F-219dce19d6` L47 [CLAIM] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: **Accurx** / Patient messaging, online triage, video consultations and Scribe. / Accurx says Scribe (built with Tandem) has summarised over one million consultations across nearly 3,000 practices, and that 1,350 practices run "total triage" on Accurx. It advertises "AI-enhanced Patient Triage". / CLAIM / [Accurx blog](https://www.accurx.com/post/whats-new-in-2026-for-gp); [Accurx benefits](https://www.accurx.com/benefits), accessed 2026-10-31 → `/clinic/online-urgent-triage` - `F-ffb7484ca0` L48 [CLAIM] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: Accurx pricing / Per-unit prices on the G-Cloud government catalogue. For example, Scribe is £0.50 down to £0.40 per unit, depending on volume. / n/a / CLAIM (vendor price list) / [G-Cloud pricing document](https://assets.applytosupply.digitalmarketplace.service.gov.uk/g-cloud-14/documents/708675/281422867192066-pricing-document-2026-08-10-1029.pdf), accessed 2026-10-31 → `/clinic/price-policy-clinic-data` - `F-c1438c5662` L49 [CLAIM] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: **EMIS** (Optum UK, now TPG) / EMIS Web, the GP clinical record system, and EMIS-X (cloud). / EMIS Scribe, an ambient scribe for EMIS Web and EMIS-X. EMIS is on NHS England's AVT (ambient voice technology) registry. No autonomous agents found. / CLAIM / [EMIS April 2026 newsletter](https://www.emishealth.com/news-insights/april-2026-newsletter); [EMIS January 2026 newsletter](https://www.emishealth.com/news-insights/january-2026-newsletter), accessed 2026-10-31 → `/clinic/front-door-check` - `F-180d95aa47` L50 [CLAIM] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: EMIS ownership / Reported that TPG's acquisition of Optum UK, including EMIS, closed on 13 March 2026. / n/a / CLAIM (single secondary source) / [healthcare.digital](https://www.healthcare.digital/single-post/the-strategic-transformation-of-uk-primary-care-digital-infrastructure-tpg-s-acquisition-of-emis-an), accessed 2026-10-31 → **UNPLACED** - `F-de6c63e7f7` L51 [CLAIM] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: **TPP SystmOne + Airmid** / SystmOne is a clinical record system. Airmid is a patient app with booking, video, messaging and wearable readings (Apple Health, Google Fit), and it supports NHS login. / TPP says it is "embedding AI and Genetic data" for early disease detection. No ambient scribe found. / CLAIM / [TPP Airmid](https://www.tpp-uk.com/products/airmid); [TPP news](https://tpp-uk.com/news/taking-healthcare-to-the-next-level/), accessed 2026-10-31 → `/clinic/price-policy-clinic-data` - `F-53c2431add` L52 [CLAIM] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: **Doctolib + Medicus** / Doctolib (France) acquired Medicus, a UK GP software company, in May 2026, and plans to invest over £100m in the UK. / Positioned as an "AI-powered GP platform". / CLAIM / [BusinessCloud](https://businesscloud.co.uk/news/doctolib-acquires-medicus-plans-to-invest-100m-in-uk/); [TechMarketView](https://www.techmarketview.com/ukhotviews/archive/2026/05/06/doctolib-backs-medicus-to-challenge-gp-it-duopoly), accessed 2026-10-31 → **UNPLACED** - `F-d453d3eebd` L53 [CLAIM] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: **Huma** (+ eConsult, iPlato) / Huma bought eConsult (digital triage, reportedly used by over 1,800 GP practices) in October 2024 and launched Huma Workspace. Its platform holds FDA 510(k) Class II (2023) and Health Canada Class II (2025) clearance as SaMD (software as a medical device). / Health Canada licence allows AI algorithms for screening, diagnosis and decision support. Huma acquired Aluna (remote spirometry) in May 2025. / CLAIM / [Huma newsroom](https://www.huma.com/newsroom/huma-acquires-econsult-and-launches-huma-workspace-for-health-systems); [Digital Health](https://digitalhealth.net/2024/10/huma-acquires-econsult-to-become-end-to-end-tech-platform); [Huma, Health Canada](https://huma.com/resources/huma-therapeutics-receives-health-canada-class-ii-medical-device-license-for-digital-health-platform); [MobiHealthNews](https://www.mobihealthnews.com/news/huma-receives-510k-class-ii-clearance-its-disease-agnostic-platform), accessed 2026-10-31 → `/clinic/classify-software-medical-device` - `F-2a54af2347` L54 [UNKNOWN] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: **Livi / Kry** / Video GP capacity sold to NHS commissioners. CQC (Care Quality Commission) lists Livi UK's location as "Archived", and Wikipedia says Livi was being rebranded Kry in France and the UK. / Not found. / UNKNOWN (current UK status) / [CQC](https://www.cqc.org.uk/location/1-10716659568/inspection-summary); [Wikipedia](https://en.wikipedia.org/wiki/Livi_(company)), accessed 2026-10-31 → `/clinic/see-day` - `F-b60b5d979a` L55 [UNKNOWN] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: **Push Doctor** / Video GP. Square Health bought its business and assets; CB Insights dates the deal to December 2021. / Not found. / UNKNOWN (current status) / [Healthcare & Protection](https://healthcareandprotection.com/square-health-buys-push-doctor/amp/); [CB Insights](https://www.cbinsights.com/investor/push-doctor), accessed 2026-10-31 → `/clinic/share-medicines-list-clinic` - `F-3c9bd39544` L56 [UNKNOWN] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: **Doctaly** / Paid face-to-face GP appointments with NHS GPs. A separate brand, "Doctaly Assist", offers GP access via WhatsApp. No 2025–2026 news found. / Not found. / UNKNOWN / [BusinessCloud](https://businesscloud.co.uk/news/platform-for-private-consultations-with-nhs-gps-crowdfunds/); [Endole](https://open.endole.co.uk/insight/brand/167752-doctaly-assist), accessed 2026-10-31 → `/clinic/settle-clinic` - `F-3cdd965aa8` L57 [CLAIM] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: **Cera** / Home care with AI. Raised about $150m (debt and equity) in January 2025. Says it works in 150 of 382 UK local authorities and 29 of 42 ICBs (integrated care boards). / Cera claims up to 70% fewer emergency admissions. TechCrunch reported that some NHS partnerships were wrongly listed on Cera's website. / CLAIM (disputed in part) / [Med-Tech News](https://med-technews.com/news/latest-medtech-news/Ceras-ai-led-social-care-model-to-potentially-save-nhs-and-uk-government-1-billion-per-year); [City AM](https://www.cityam.com/?p=2297543); [TechCrunch](https://techcrunch.com/?p=1632310), accessed 2026-10-31 → `/clinic/wearable-home-device-signals` - `F-75031eafdd` L61 _1. Competitors › 1.2 AI clinical intelligence vendors_: Vendor / product / What it offers / Agents? / Regulatory position / Label / Source → `/clinic/wearable-home-device-signals` - `F-3b97725962` L63 [CLAIM] _1. Competitors › 1.2 AI clinical intelligence vendors_: **OpenAI – ChatGPT Health / "Health in ChatGPT"** / Launched 7 January 2026 as a separate space where users connect medical records (via b.well) and wellness apps. OpenAI dated a relaunch to 23 July 2026, for US users only. / Consumer assistant. By default it asks before it uses connected records. / OpenAI says it "does not replace qualified medical care". Connected health data is not used to train foundation models. / CLAIM / [Fortune](https://fortune.com/2026/01/07/openai-launches-chatgpt-health-in-a-push-to-become-a-hub-for-personal-health-data); [Fierce Healthcare](https://fiercehealthcare.com/ai-and-machine-learning/openai-launches-chatgpt-health-connect-data-health-apps-medical-records); [OpenAI](https://openai.com/index/health-in-chatgpt/) (page blocked automated fetch), accessed 2026-10-31 → `/clinic/mental-health-care-crisis` - `F-6d3a660c1e` L64 [CLAIM] _1. Competitors › 1.2 AI clinical intelligence vendors_: **Anthropic – Claude for Healthcare** / Launched January 2026. Connectors for the CMS (US Centers for Medicare & Medicaid Services) Coverage Database, ICD-10 and the NPI (National Provider Identifier) Registry, plus a FHIR skill and a prior-authorisation template. Consumer connectors (HealthEx, Function, Apple Health, Android Health Connect) are in beta. / Agent skills. / Described as "HIPAA-ready". Anthropic says it does not train on user health data. / CLAIM / [Anthropic](https://www.anthropic.com/news/healthcare-life-sciences); [MobiHealthNews](https://www.mobihealthnews.com/news/anthropic-launches-claude-healthcare); [The Register](https://www.theregister.com/2026/01/12/claude_anthropic_healthcare/), accessed 2026-10-31 → `/clinic/refresh-public-data-caches` - `F-cdbb6469c9` L65 [CLAIM] _1. Competitors › 1.2 AI clinical intelligence vendors_: **Microsoft – Dragon Copilot** (formerly Nuance DAX) / Ambient documentation. Generally available in the UK since September 2025 and hosted in a UK Azure data centre. / A marketplace of partner apps and agents inside Dragon Copilot. / Microsoft says DCB0129 is complete and assessed by a UK Clinical Safety Officer. / CLAIM / [Phoenix Software](https://www.phoenixs.co.uk/resources/blog/microsoft-dragon-copilot-is-now-generally-available-in-the-uk/); [Microsoft UK](https://www.microsoft.com/en-gb/health-solutions/clinical-workflow/dragon-copilot), accessed 2026-10-31 → `/clinic/medication-safety-check` - `F-11d0622a2c` L66 [CLAIM] _1. Competitors › 1.2 AI clinical intelligence vendors_: **Microsoft 365 Copilot for NHS England** / June 2026 agreement covering Microsoft 365 Copilot and Copilot Studio, so NHS England and trusts can build their own agents (for example, for complaints and FOI requests), governed through "Agent 365". / Administrative agents, not clinical. / n/a / CLAIM / [Microsoft](https://news.microsoft.com/source/2026/06/08/nhs-england-accelerates-ai-adoption-with-microsoft-365-copilot-to-improve-service-delivery-reduce-costs-and-create-more-time-for-care/); [TechMarketView](https://www.techmarketview.com/ukhotviews/archive/2026/06/08/nhs-england-scales-microsoft-copilot-adoption), accessed 2026-10-31 → `/clinic/person-held-health-record` - `F-1ac417b4cc` L67 [CLAIM] _1. Competitors › 1.2 AI clinical intelligence vendors_: **Google – AMIE and MedGemma** / AMIE is a research diagnostic dialogue agent, now extended to disease management. MedGemma is a family of open-weight medical models. / AMIE is multi-agent (a dialogue agent plus a management agent). It is still research. / Not a product. Studies are running with Beth Israel Deaconess and Included Health. / CLAIM / [Google blog](https://blog.google/innovation-and-ai/models-and-research/google-research/amie-for-disease-management-in-nature); [Google Research I/O 2025](https://www.research.google/blog/google-research-at-google-io-2025), accessed 2026-10-31 → `/clinic/price-policy-clinic-data` - `F-39153186e3` L68 [CLAIM] _1. Competitors › 1.2 AI clinical intelligence vendors_: **Abridge** / Ambient documentation. In August 2026 it expanded its "clinical intelligence agent" with decision support to more than 300 health systems. / Yes, for clinicians. / US-focused. / CLAIM / [HLTH](https://hlth.com/insights/news/abridge-expands-clinical-intelligence-ai-across-300-health-systems), accessed 2026-10-31 → `/clinic/mental-health-care-crisis` - `F-001b0c2e6f` L69 [UNKNOWN] _1. Competitors › 1.2 AI clinical intelligence vendors_: **Tortus, Heidi, Tandem** (UK ambient scribes) / Ambient scribes. A tracker reports Tortus as UKCA Class IIa (24 June 2026) and Tandem as EU MDR Class IIa. Heidi's status conflicts between sources. / Documentation, plus coding suggestions. / Mixed; see the regulator section. / UNKNOWN (single commercial blog family, with internal conflicts) / [iatroX tracker](https://www.iatrox.com/blog/ai-scribe-regulatory-status-tracker-2026), accessed 2026-10-31 → `/clinic/track-referral` - `F-232ece4dc8` L70 [CLAIM] _1. Competitors › 1.2 AI clinical intelligence vendors_: **NHS England AVT supplier registry** / A self-certified registry that is not a commercial framework and not an endorsement. It listed 19 suppliers in January 2026. / n/a / Suppliers must complete DTAC, plus other criteria. / CLAIM / [HTN](https://htn.co.uk/2026/01/19/nhs-england-lists-19-suppliers-meeting-criteria-for-ambient-voice-registry/); [Digital Health](https://www.digitalhealth.net/2026/01/nhs-england-launches-supplier-registry-for-ai-scribing-tech/), accessed 2026-10-31 → `/clinic/register-trial` - `F-c5e603f19a` L76 _1. Competitors › 1.3 Clinical decision support and drug-interaction data_: **BNF / BNFC** (British National Formulary, and the version for children) / The UK's standard prescribing reference, with A–Z interaction monographs. Free to NHS users via NICE (National Institute for Health and Care Excellence), UK only. / Non-NHS bodies need a licence from the BNF Partners. Merging BNF content into third-party systems needs a direct licence. RPS (Royal Pharmaceutical Society) offers a data feed or API. / OBSERVED / [Pharmaceutical Press licensing](https://www.pharmaceuticalpress.com/?p=146134); [NICE BNF UK only](https://www.nice.org.uk/bnf-uk-only), accessed 2026-10-31 → `/clinic/mental-health-care-crisis` - `F-44ed7432b4` L77 [CLAIM] _1. Competitors › 1.3 Clinical decision support and drug-interaction data_: **Stockley's Drug Interactions** / Monographs covering more than 5,000 interactions, with evidence, mechanism and management. Accessed by subscription through MedicinesComplete. / Subscription. / CLAIM / [SPS (NHS Specialist Pharmacy Service)](https://sps.nhs.uk/articles/drug-interactions-resources-to-support-answering-questions/), accessed 2026-10-31 → `/clinic/claim-edge-id-record` - `F-ed82c3eed9` L78 [CLAIM] _1. Competitors › 1.3 Clinical decision support and drug-interaction data_: **FDB Multilex / OptimiseRx** (First Databank UK) / FDB calls Multilex the UK's most widely used drug database and e-prescribing decision support (more than 10,000 sites). It covers drug–drug interactions, sensitivity and duplicate-therapy checks. OptimiseRx gives point-of-care prompts in EMIS Web. / Commercial. A 2026 NHS contract was awarded without competition because only FDB could supply it. / CLAIM / [SNOMED International marketplace](https://snomed.org/snomed-ct-marketplace/first-databank-uk-ltd.); [Find a Tender](https://www.find-tender.service.gov.uk/procurement/ocds-h6vhtk-06c6f3), accessed 2026-10-31 → `/clinic/social-prescribing-referral` - `F-41b2c674fe` L79 _1. Competitors › 1.3 Clinical decision support and drug-interaction data_: **NLM RxNav drug interaction API** (US) / A free US government interaction API. / Scheduled for shutdown on or about 2 January 2024. / OBSERVED (NLM notice) / [NLM](https://www.lhncbc.nlm.nih.gov/RxNav/news/APIUpdate.html); [DrugBank blog](https://blog.drugbank.com/nih-discontinues-their-drug-interaction-api/), accessed 2026-10-31 → **UNPLACED** - `F-1de1985cac` L80 [CLAIM] _1. Competitors › 1.3 Clinical decision support and drug-interaction data_: **DrugBank** / A drug knowledge base with interactions. / Open vocabulary data: CC0-1.0. Full data: CC-BY-NC-4.0 (academic, non-commercial). Commercial use needs a separate agreement. Several release pages show academic downloads "temporarily unavailable". / CLAIM / [DrugBank FAQ](https://dev.drugbank.com/guides/faqs); [DrugBank 5.1.22](https://go.drugbank.com/releases/5-1-22), accessed 2026-10-31 → `/clinic/price-policy-clinic-data` - `F-a4abe60ca5` L81 [UNKNOWN] _1. Competitors › 1.3 Clinical decision support and drug-interaction data_: **DDInter 2.0** / An academic open-access database: 2,310 drugs and 302,516 drug–drug interaction records (NAR, Nucleic Acids Research, 2025). / "Open-access", but no data licence found. / UNKNOWN (licence) / [NAR paper](https://academic.oup.com/nar/article/53/D1/D1356/7740584); [DDInter 2.0](https://ddinter2.scbdd.com/), accessed 2026-10-31 → **UNPLACED** - `F-134e62442c` L82 [UNKNOWN] _1. Competitors › 1.3 Clinical decision support and drug-interaction data_: **Medscape interaction checker** / Not researched in depth. No primary source checked. / — / UNKNOWN / — → **UNPLACED** - `F-7157075615` L84 _1. Competitors › 1.3 Clinical decision support and drug-interaction data_: **Gap for EDGE Clinic:** the brief asks for "an open clinical-grade drug-interaction source". On this evidence, none exists today. The clinical-grade sources (BNF, Stockley's, FDB) are licensed. The open ones (DrugBank open data, DDInter) are either non-commercial, of unclear licence, or not presented as clinical-grade. The free US government API was withdrawn. Sources as above, accessed 2026-10-31. → **UNPLACED** - `F-534f8e8493` L90 [CLAIM] _1. Competitors › 1.4 Retinal AI vendors_: **LumineticsCore** (Digital Diagnostics, formerly IDx-DR) / Autonomous detection of more-than-mild diabetic retinopathy. It was the first autonomous AI diagnostic authorised by the FDA (US Food and Drug Administration), via De Novo. / US authorised. / CLAIM (vendor) / [Digital Diagnostics](https://www.digitaldiagnostics.com/resources/insights/lumineticscore-engineered-by-digital-diagnostics-early-detection-of-diabetic-retinopathy-using-ai/), accessed 2026-10-31 → **UNPLACED** - `F-48f40cf0ed` L91 [CLAIM] _1. Competitors › 1.4 Retinal AI vendors_: **EyeArt** (Eyenuk) / Detects mild and vision-threatening diabetic retinopathy. FDA cleared in 2020, with more cameras added in 2023. / US cleared. / CLAIM / [Insight News](https://www.insightnews.com.au/fda-clearance-expands-eyenuk-ai-system-to-automatically-detect-diabetic-retinopathy/), accessed 2026-10-31 → **UNPLACED** - `F-303cecdce2` L92 [CLAIM] _1. Competitors › 1.4 Retinal AI vendors_: **AEYE-DS** (AEYE Health) / Reported as cleared on the Optomed Aurora handheld camera and the Topcon NW400. / Date not confirmed. / CLAIM / [ai2.work](https://ai2.work/blog/handheld-ai-eye-screening-clears-fda-inside-the-retina-shift), accessed 2026-10-31 → `/clinic/claim-edge-id-record` - `F-804755ccca` L93 _1. Competitors › 1.4 Retinal AI vendors_: **NHS Diabetic Eye Screening Programme (DESP)** / About 2.2 million screening appointments a year. The UK National Screening Committee consulted on AI in diabetic eye screening (May 2026). A Birmingham procurement framework favours AI as a first-pass triage, with specialist graders reviewing. / Outcome not found. No vendor confirmed in NHS DESP. / PROPOSED / [AOP](https://www.aop.org.uk/ot/news/2026/05/13/uk-national-screening-committee-consults-on-use-of-ai-in-diabetic-eye-screening); [University of Birmingham](https://www.birmingham.ac.uk/news/2026/birmingham-research-helps-shape-how-nhs-procures-ai-based-diabetic-eye-screening); [SGUL AIDED study](https://www.sgul.ac.uk/about/our-institutes/population-health/projects/aided-study), accessed 2026-10-31 → `/clinic/retinal-screening` - `F-4ca569c1a4` L94 [CLAIM] _1. Competitors › 1.4 Retinal AI vendors_: **Cascader** (Moorfields, UCL, Topcon) / A new company for AI on eye disease and "oculomics" (reading signs of other diseases, such as heart disease or dementia, from retinal images), with access to over 35 million images. / Early stage. / CLAIM / [Moorfields](https://www.moorfields.nhs.uk/about-us/news-and-blogs/news/new-company-launched-to-advance-ai-driven-eyecare), accessed 2026-10-31 → `/clinic/retinal-screening` - `F-2253860171` L95 [CLAIM] _1. Competitors › 1.4 Retinal AI vendors_: UK optometry pilots (Toku, Heart-Eye and others) / Small UK pilots of retinal AI for cardiovascular risk. / Pilots only. / CLAIM / [LOCSU case studies](https://locsu.co.uk/wp-content/uploads/2025/12/CVD-Case-Studies-AI-Powered-Retinal-Imaging-for-Cardiovascular-Risk-Assessment.pdf), accessed 2026-10-31 → `/clinic/retinal-screening` - `F-0ba9774cab` L99 _1. Competitors › 1.5 What the competitors leave open, and bare minimum versus ambitious_: These gaps come from the tables above, with sources as listed there. → `/clinic/retinal-screening` - `F-1e7ba7562f` L101 [UNKNOWN] _1. Competitors › 1.5 What the competitors leave open, and bare minimum versus ambitious_: **Nobody joins the steps "front door → detection → social prescribing referral".** The NHS App is national but is a router. Accurx, eConsult (Huma) and Rapid Health triage GP demand. Cera is home care. Retinal AI is a single test. EDGE Clinic's "come about anything" scope has no direct UK rival found. UNKNOWN whether "My NHS GP" will cover non-medical needs. → `/clinic/retinal-screening` - `F-e5dec444ef` L102 _1. Competitors › 1.5 What the competitors leave open, and bare minimum versus ambitious_: **AI in competitors is mostly ambient documentation**, which the MHRA now says is not a device when it only summarises ([MHRA](https://www.gov.uk/government/news/mhra-clarifies-regulatory-status-of-ambient-voice-technologies-used-in-the-nhs), accessed 2026-10-31). Patient-facing agents that act (book, refer, check prescriptions) are rare and regulated. → `/clinic/front-door-check` - `F-ce074d7ac1` L103 _1. Competitors › 1.5 What the competitors leave open, and bare minimum versus ambitious_: **None of the vendors found publishes an agent identity, mandate or audit model that a regulator could re-run.** Microsoft's "Agent 365" governance is administrative only ([TechMarketView](https://www.techmarketview.com/ukhotviews/archive/2026/06/08/nhs-england-scales-microsoft-copilot-adoption), accessed 2026-10-31). → `/clinic/price-policy-clinic-data` - `F-60c9d0aaf5` L107 _1. Competitors › 1.5 What the competitors leave open, and bare minimum versus ambitious_: Front door / NHS login, DTAC, DCB0129, DSPT, signposting only (not a device) / An agent that triages and acts under a mandate the patient can see, with every step replayable by a regulator → `/clinic/price-policy-clinic-data` - `F-242b7a81b6` L108 _1. Competitors › 1.5 What the competitors leave open, and bare minimum versus ambitious_: Detection / Buy a cleared retinal AI and keep a human grader in the loop / Multi-signal detection with staged ("L-plate") authorisation and a public post-market safety record → `/clinic/price-policy-clinic-data` - `F-1ad40be3f2` L109 _1. Competitors › 1.5 What the competitors leave open, and bare minimum versus ambitious_: Prescribing safety / License BNF or FDB through the EHR supplier / Contribute to an open, versioned interaction knowledge base with provenance → `/clinic/price-policy-clinic-data` - `F-7050ceb429` L110 _1. Competitors › 1.5 What the competitors leave open, and bare minimum versus ambitious_: Agents / No autonomous actions; human signs off every action / Agent identity, mandate, consent and escalation recorded as FHIR Consent, Provenance and AuditEvent → `/clinic/price-policy-clinic-data` - `F-3e6b8088d3` L120 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: **MHRA – UK MDR 2002** (UK Medical Devices Regulations) / Software with a medical purpose is a medical device. Post-market surveillance rules (SI 2024/1368) apply to devices placed on the GB market from 16 June 2025. / In force since 16 June 2025 / AGREEMENT / [GOV.UK](https://www.gov.uk/government/news/first-major-overhaul-of-medical-device-regulation-comes-into-force-across-great-britain), accessed 2026-10-31 → `/clinic/classify-software-medical-device` - `F-89b105d7a3` L121 _2. Regulators and rules › 2.1 United Kingdom_: MHRA – draft Medical Devices (Amendment) Regulations 2026 / New pre-market rules and an "international reliance" route, which lets devices authorised in Australia, Canada or the US register in GB. Software-specific provisions are reportedly not updated yet. / Draft published 8 May 2026. Expected to be adopted December 2026 and to come into force June 2027, subject to Parliament. / PROPOSED / [Latham & Watkins](https://www.lw.com/en/insights/mhra-publishes-draft-amendment-to-the-uk-medical-devices-regulations); [Fieldfisher](https://www.fieldfisher.com/en/insights/new-pre-market-requirements-for-medical-devices-mhra-publishes-draft-legislation---what%e2%80%99s-changed); [Qualio](https://www.qualio.com/blog/mhra-s-2026-devices-reform-what-changes-and-when), accessed 2026-10-31 → `/clinic/classify-software-medical-device` - `F-6c41b8d8f0` L122 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: MHRA – AVT (ambient voice technology) guidance / Transcription, summarising or drafting letters alone: not a device. Supporting diagnosis, treatment or prevention, or taking automated action without clinician review: a device. / Published 29 July 2026 / AGREEMENT (guidance, does not change law) / [MHRA news](https://www.gov.uk/government/news/mhra-clarifies-regulatory-status-of-ambient-voice-technologies-used-in-the-nhs) (OBSERVED); [MHRA PDF](https://assets.publishing.service.gov.uk/media/6a68daaff23ff62c51c1a7a3/MHRA_Guidance_on_Ambient_Voice_Technology-Enabled_Products.pdf), accessed 2026-10-31 → `/clinic/front-door-check` - `F-0569d3e11b` L123 _2. Regulators and rules › 2.1 United Kingdom_: MHRA – PCCP guidance for AI devices (UK) / Reported plan to issue draft guidance on predetermined change control plans by end 2026. A PCCP is a plan, agreed in advance, for how a model may change after approval. / Not published / PROPOSED / [Qualio](https://www.qualio.com/blog/mhra-s-2026-devices-reform-what-changes-and-when), accessed 2026-10-31 → `/clinic/issue-field-safety-notice` - `F-cacb289f02` L124 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: MHRA – AI Airlock / A regulatory sandbox for AI as a medical device. Phase 2 had 7 products, including note-taking and eye disease detection, and ran April 2025 to May 2026. DHSC (Department of Health and Social Care) funds £1.2m a year for 2026–2029. / Phase 2 reports out July 2026. Phase 3 not confirmed. / AGREEMENT (programme); UNKNOWN (Phase 3) / [GOV.UK collection](https://www.gov.uk/government/collections/ai-airlock-the-regulatory-sandbox-for-aiamd); [Phase 2 report](https://www.gov.uk/government/publications/ai-airlock-sandbox-phase-2-programme-report); [MHRA blog](https://medregs.blog.gov.uk/2026/06/09/advancing-ai-regulation-in-healthcare-insights-from-ai-airlock-phase-2/), accessed 2026-10-31 → `/clinic/register-device-fault` - `F-13c42af2f1` L125 [UNKNOWN] _2. Regulators and rules › 2.1 United Kingdom_: National Commission into the Regulation of AI in Healthcare / 44 recommendations, including staged authorisations ("L-plates"), continuous real-world monitoring, a public searchable safety database, and stronger MHRA enforcement. / Report 10 September 2026. GOV.UK says a "formal response to follow". One law firm says government accepted all 44 on 6 October 2026. / PROPOSED; UNKNOWN (acceptance) / [GOV.UK](https://www.gov.uk/government/news/independent-commission-led-by-nhs-doctors-sets-out-blueprint-to-accelerate-safe-ai-adoption-in-healthcare) (OBSERVED); [Report](https://www.gov.uk/government/publications/national-commission-into-the-regulation-of-ai-in-healthcare-recommendations-for-a-future-regulatory-framework); [CMS](https://cms.law/en/gbr/legal-updates/from-pre-market-to-lifecycle-the-national-commission-s-44-recommendations-on-ai-regulation-in-healthcare); [Ropes & Gray](https://www.ropesgray.com/en/insights/viewpoints/2026/10/102o5fp/ai-medical-devices-in-great-britain-what-the-mhras-response-to-the-national-com), accessed 2026-10-31 → `/clinic/social-prescribing-referral` - `F-f115959ca0` L126 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: **NHS England – DCB0129** (the manufacturer's clinical risk management) and **DCB0160** (the deployer's) / Mandatory clinical safety standards for health IT in England. They require a named Clinical Safety Officer, a hazard log and a safety case. / Current versions in force. Revision consultation ran 29 June–11 September 2026, with AI named as a driver. / AGREEMENT (current); PROPOSED (revision) / [NHS England consultation](https://www.engage.england.nhs.uk/patient-safety/national-review-of-clinical-risk-management-standa/) (OBSERVED); [DCB0129 standard](https://adultsocialcare.standards.nhs.uk/published-standards/clinical-risk-management-its-application-in-the-manufacture-of-health-it-systems); [DCB0160 standard](https://standards.nhs.uk/published-standards/clinical-risk-management-its-application-in-the-deployment-and-use-of-health-it-systems), accessed 2026-10-31 → `/clinic/inspect-hazard-log-safety` - `F-69610398cb` L127 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: **DTAC** (Digital Technology Assessment Criteria) / The NHS buyer's baseline checklist. It covers clinical safety, data protection, technical assurance, interoperability, and usability and accessibility. / New form from 24 February 2026, with about 25% fewer questions. The old form was retired on 6 April 2026. / AGREEMENT / [NHS Innovation Service](https://innovation.nhs.uk/news/updated-nhs-england-digital-technology-assessment-criteria-dtac-form-and-guidance/); [HTN](https://htn.co.uk/2026/02/25/digital-technology-assessment-criteria-refreshed-to-create-simpler-more-trusted-pathway-for-nhs-digital-innovation/), accessed 2026-10-31 → `/clinic/medication-safety-check` - `F-5584fac901` L128 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: **DSPT** (Data Security and Protection Toolkit) / An annual self-assessment that any organisation accessing NHS patient data must complete. Trusts and ICBs (integrated care boards) use a version aligned to the NCSC (National Cyber Security Centre) CAF (Cyber Assessment Framework). IT suppliers were reported to stay on the non-CAF version for 2025/26. / 2025/26 deadline 30 June 2026. Version 9 (2026/27) changes not checked. / AGREEMENT; UNKNOWN (2026/27 supplier detail) / [DSPT help](https://www.dsptoolkit.nhs.uk/Help/64); [Dionach](https://dionach.com/data-security-and-protection-toolkit-dspt-2025-2026-caf/), accessed 2026-10-31 → `/clinic/complete-data-security` - `F-8cf03b70a5` L129 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: NHS England – ambient scribing guidance and AVT registry / Guidance first published April 2025. NHS England withdrew its earlier statement that summarisation tools are likely devices, to match the MHRA. The registry launched January 2026. / Updated July 2026 / AGREEMENT (guidance) / [NHS England guidance](https://www.england.nhs.uk/long-read/guidance-on-the-use-of-ai-enabled-ambient-scribing-products-in-health-and-care-settings); [HTN](https://htn.co.uk/2026/07/29/mhra-clarifies-regulatory-medical-devices-status-of-ambient-voice-technologies/), accessed 2026-10-31 → `/clinic/claim-edge-id-record` - `F-c80145f4dd` L130 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: **CQC** (Care Quality Commission) / Registers providers of "regulated activities". A 2022 scope update says AI suppliers that analyse and report on screening images may need to register for "diagnostic and screening procedures" if no CQC-registered provider is doing so. / No AI-specific 2025–2026 CQC guidance found / AGREEMENT (2022 scope); UNKNOWN (newer) / [Osborne Clarke](https://osborneclarke.com/insights/updated-cqc-guidance-has-implications-health-technology-and-wellbeing-industries); [NHS digital regulations guide](https://www.digitalregulations.innovation.nhs.uk/regulations-and-guidance-for-developers/all-developers-guidance/regulated-activities-check-if-you-need-to-register-with-the-care-quality-commission-cqc/pdf/), accessed 2026-10-31 → `/clinic/retinal-screening` - `F-ca3348e347` L131 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: **NICE – ESF** (Evidence Standards Framework) for digital health technologies / Sets evidence tiers in proportion to risk, and includes AI and data-driven tools. A further revision is expected to follow MHRA reforms. Early Value Assessments let technologies be used while evidence is gathered. / ESF in use; next update pending / AGREEMENT; PROPOSED (update) / [Fieldfisher](https://www.fieldfisher.com/en/insights/new-changes-to-evidence-standards-framework-and-digital-health-technologies); [JMIR viewpoint](https://mhealth.jmir.org/2025/1/e67435); [ISPOR poster](https://www.ispor.org/heor-resources/presentations-database/presentation-cti/ispor-europe-2026/poster-session-5-4/learning-while-using-evidence-expectations-under-nice-s-reformed-healthtech-program), accessed 2026-10-31 → `/clinic/request-evidence-against` - `F-191c70cf28` L132 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: **ICO** (Information Commissioner's Office) – UK GDPR, AI guidance / A DPIA (data protection impact assessment) is needed for AI processing health data at scale. AI inferences about health can count as "special category" data. / AI guidance updated March 2023; a 2026 refresh was not confirmed / AGREEMENT / [ICO guidance](https://www.ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/artificial-intelligence/guidance-on-ai-and-data-protection); [Hunton](https://www.hunton.com/privacy-and-cybersecurity-law-blog/uk-ico-issues-updated-guidance-on-ai-and-data-protection); [NHS DPIA step](https://digitalregulations.innovation.nhs.uk/regulations-and-guidance-for-developers/all-developers-guidance/how-to-comply-with-the-uk-gdpr-as-a-developer/step-2-do-a-data-protection-impact-assessment-dpia/pdf/), accessed 2026-10-31 → `/clinic/run-data-protection-impact` - `F-8d88159991` L133 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: Data (Use and Access) Act 2025 – automated decision-making / Most reforms came into force on 5 February 2026. Solely automated significant decisions are now restricted mainly where special category data (such as health) is involved, so health stays strict. / In force 5 February 2026. ICO ADM guidance was consulted on in 2026. / AGREEMENT; PROPOSED (ICO guidance) / [GOV.UK](https://www.gov.uk/guidance/data-use-and-access-act-2025-data-protection-and-privacy-changes); [Travers Smith](https://www.traverssmith.com/knowledge/knowledge-container/uks-data-protection-reforms-take-effect-a-new-era-for-automated-decision-making/); [Kennedys](https://www.kennedyslaw.com/en/thought-leadership/article/2026/the-data-use-and-access-act-2025-commencement-dates-and-planned-guidance-for-2026), accessed 2026-10-31 → `/clinic/person-held-health-record` - `F-4573a6ae5d` L134 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: **Caldicott Principles / National Data Guardian (NDG)** / Eight principles for using confidential patient information, listed as legally required in the NHS digital regulations service. The NDG's 2024/25 report mentions input to AI scribe guidance. / Principles in force. No NDG AI-specific guidance found. / AGREEMENT; UNKNOWN (AI guidance) / [NHS digital regulations](https://digitalregulations.innovation.nhs.uk/regulations-and-guidance-for-adopters/all-adopters-guidance/complying-with-the-uk-gdpr-steps-1-7-an-introduction/step-7-follow-the-8-caldicott-principles/pdf/); [NDG annual report](https://assets.publishing.service.gov.uk/media/6942a76f9273c48f554cf506/NDG_annual_report_2024-25_v1.0_-_17.12.25.pdf), accessed 2026-10-31 → `/clinic/report-data-breach` - `F-31c00f3382` L135 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: **HRA** (Health Research Authority) / HRA approval covers project-based research in the NHS in England and Wales. Service evaluation may not need ethics review; use the HRA decision tool. CAG (Confidentiality Advisory Group) approval may be needed for identifiable data without consent. / In force / AGREEMENT / [HRA approval](https://www.hra.nhs.uk/approvals-amendments/what-approvals-do-i-need/hra-approval/); [HRA: what approvals](https://www.hra.nhs.uk/approvals-amendments/what-approvals-do-i-need/), accessed 2026-10-31 → `/clinic/front-door-check` - `F-9b7c220def` L141 [AGREEMENT] _2. Regulators and rules › 2.2 European Union_: **EU MDR** (Medical Device Regulation 2017/745) / The device rules in the EU. Its classification decides whether the AI Act's high-risk rules apply. / In force / AGREEMENT / [EUR-Lex](https://eur-lex.europa.eu/eli/reg/2017/745/oj), accessed 2026-10-31 → `/clinic/classify-software-medical-device` - `F-e70c2d6f90` L142 [AGREEMENT] _2. Regulators and rules › 2.2 European Union_: **EU AI Act** (Regulation 2024/1689) / Risk-based AI law. AI that is a medical device, or a safety component of one, and needs notified-body assessment is "high-risk". / Original high-risk date for Annex I products was 2 August 2027, now moved (next row) / AGREEMENT / [EUR-Lex](https://eur-lex.europa.eu/eli/reg/2024/1689/oj); [OpenRegulatory on MDCG 2025-6](https://openregulatory.com/mdcg/mdcg-2025-6), accessed 2026-10-31 → `/clinic/act-high-risk-answer` - `F-567669e045` L143 [AGREEMENT] _2. Regulators and rules › 2.2 European Union_: **Digital Omnibus on AI** (Regulation 2026/1744) / Moves the high-risk deadlines. Annex III (standalone) to 2 December 2027; Annex I (AI in regulated products such as medical devices) to 2 August 2028. Transparency obligations start 2 August 2026, and some content-marking duties 2 December 2026. / In force 27 July 2026 / AGREEMENT / [Hunton](https://www.hunton.com/privacy-and-cybersecurity-law-blog/eu-digital-omnibus-on-ai-enters-into-force) (OBSERVED, law firm); [Gibson Dunn](https://www.gibsondunn.com/eu-ai-act-omnibus-agreement-postponed-high-risk-deadlines-and-other-key-changes/), accessed 2026-10-31 → `/clinic/act-high-risk-answer` - `F-15342eb89d` L144 [AGREEMENT] _2. Regulators and rules › 2.2 European Union_: **MDCG 2025-6** (FAQ from the Medical Device Coordination Group and the AI Board) / Explains how the MDR and the AI Act fit together. The MDR class drives AI Act applicability, and the notified body assesses both together. / Published June 2025 (exact date disputed) / AGREEMENT (guidance) / [OpenRegulatory](https://openregulatory.com/mdcg/mdcg-2025-6), accessed 2026-10-31 → `/clinic/manage-patient-group-directions` - `F-3d68ee40ac` L145 [AGREEMENT] _2. Regulators and rules › 2.2 European Union_: **EHDS** (European Health Data Space, Regulation 2025/327) / Patient access to records, EHR (electronic health record) system requirements and secondary use of data. / In force 26 March 2025. Main application 2027–2031 in phases. / AGREEMENT (phased) / [Kennedys](https://www.kennedyslaw.com/en/thought-leadership/article/2026/the-european-health-data-space-is-in-force-implications-for-healthcare-medtech-and-life-sciences); [A&O Shearman](https://www.aoshearman.com/en/insights/ao-shearman-on-life-sciences/european-health-data-space-regulation-sets-requirements-for-electronic-health-record-systems), accessed 2026-10-31 → `/clinic/symptom-check-triage` - `F-b4add08f8c` L146 [AGREEMENT] _2. Regulators and rules › 2.2 European Union_: **Product Liability Directive** (2024/2853) / Software and AI become "products" under strict liability. Courts can presume a product is defective where technical complexity makes proof too hard. / Member states must transpose it by 9 December 2026 / AGREEMENT (transposition due) / [Covington Inside Privacy](https://www.insideprivacy.com/uncategorized/eu-member-states-begin-rolling-out-new-product-liability-rules/); [Hogan Lovells](https://www.hoganlovells.com/en/publications/eu-introduces-comprehensive-digitalera-product-liability-directive), accessed 2026-10-31 → `/clinic/cancel-clinic-session` - `F-976ce30751` L152 [AGREEMENT] _2. Regulators and rules › 2.3 United States_: FDA – Clinical Decision Support (CDS) software guidance / Revised 6 January 2026 without a comment period. It relaxes oversight of some tools that recommend a single option to a clinician. / Final guidance (replaces 2022) / AGREEMENT / [King & Spalding](https://www.kslaw.com/insights/articles/fda-updates-general-wellness-and-clinical-decision-support-guidance-documents); [Arnold & Porter](https://www.arnoldporter.com/en/perspectives/advisories/2026/01/fda-cuts-red-tape-on-clinical-decision-support-software), accessed 2026-10-31 → **UNPLACED** - `F-2ce2701823` L153 [AGREEMENT] _2. Regulators and rules › 2.3 United States_: FDA – General Wellness guidance / Revised 6 January 2026. Adds criteria for non-invasive sensors, such as wrist-worn blood pressure estimates. / Final / AGREEMENT / [King & Spalding](https://www.kslaw.com/insights/articles/fda-updates-general-wellness-and-clinical-decision-support-guidance-documents), accessed 2026-10-31 → **UNPLACED** - `F-b911005bc2` L154 [AGREEMENT] _2. Regulators and rules › 2.3 United States_: FDA – PCCP for AI-enabled device software functions / Lets a maker pre-agree how a model will change: description of the changes, a modification protocol and an impact assessment. / Final, December 2024 / AGREEMENT / [Ropes & Gray](https://ropesgray.com/en/insights/alerts/2024/12/fda-finalizes-guidance-on-predetermined-change-control-plans-for-ai-enabled-device), accessed 2026-10-31 → `/clinic/register-imaging-device` - `F-05b35a1316` L155 _2. Regulators and rules › 2.3 United States_: FDA – AI-enabled device software functions: lifecycle and submissions / Draft from January 2025. Finalising it is a top priority on the FY2027 (US fiscal year 2027) agenda. / Draft / PROPOSED / [RAPS](https://www.raps.org/resource/fda-s-device-center-releases-guidance-agenda-for-fy-2027.html); [CAP](https://www.cap.org/advocacy/latest-news-and-practice-data/fda-releases-new-artificial-intelligence-draft-guidance), accessed 2026-10-31 → `/clinic/classify-software-medical-device` - `F-cd0a41f910` L156 [CLAIM] _2. Regulators and rules › 2.3 United States_: FDA – AI-enabled device list / FDA lists over 1,600 AI-enabled devices and says the list is not comprehensive. A commentator notes none of the authorised devices it cites are generative or agentic. / Live list / OBSERVED (page exists); CLAIM (count, commentary) / [FDA list](https://www.fda.gov/medical-devices/artificial-intelligence-enabled-medical-devices/list-artificial-intelligence-enabled-medical-devices); [Clearwater](https://clearwatersecurity.com/transcripts/agentic-ai-in-healthcare-navigating-regulatory-uncertainty-and-building-governance-that-lasts/), accessed 2026-10-31 → `/clinic/see-day` - `F-d96310492f` L157 _2. Regulators and rules › 2.3 United States_: **HIPAA** (Health Insurance Portability and Accountability Act) Security Rule update / The proposed overhaul (published January 2025) is not final. HHS (Department of Health and Human Services) moved the target to July 2027. The current rule still applies. / Proposed / PROPOSED / [Fierce Healthcare](https://www.fiercehealthcare.com/health-tech/feds-push-back-hipaa-security-rule-overhaul-july-2027); [Davis Wright Tremaine](https://www.dwt.com/blogs/privacy--security-law-blog/2026/07/hhs-updates-hipaa-rulemaking-timeframes), accessed 2026-10-31 → `/clinic/apply-treatment-rule` - `F-dca458eefa` L158 _2. Regulators and rules › 2.3 United States_: ONC (Office of the National Coordinator for Health IT) – Interoperability Standards Platform AI appendix / Describes MCP (Model Context Protocol) over FHIR, A2A (agent-to-agent) standards, human sign-off on writes, and audit logging that links AI requests to users, agents and organisations. Says identity and delegation are "still evolving". / Reference content, no binding status / OBSERVED / [ONC ISP](https://isp.healthit.gov/appendix-vi-ai-interoperability), accessed 2026-10-31 → `/clinic/person-held-health-record` - `F-888c904182` L164 [AGREEMENT] _2. Regulators and rules › 2.4 Worldwide_: WHO guidance on LMMs (large multi-modal models) for health / More than 40 recommendations, including that governments assign a regulator to approve LMMs used in care, and independent post-release audits for large-scale deployments. / Published January 2024 / AGREEMENT (non-binding guidance) / [WHO](https://www.who.int/publications-detail-redirect/9789240084759), accessed 2026-10-31 → `/clinic/mental-health-care-crisis` - `F-4a3ebbd7fd` L168 _2. Regulators and rules › 2.5 Applying the MHRA test to EDGE Clinic (research reading, not legal advice)_: EDGE Clinic function / Likely device? / Why / Source → `/clinic/mental-health-care-crisis` - `F-ec69ae4d07` L170 [UNKNOWN] _2. Regulators and rules › 2.5 Applying the MHRA test to EDGE Clinic (research reading, not legal advice)_: Signposting to a social prescribing service, with no clinical claim / Probably not (UNKNOWN until assessed) / It does not "support diagnosis, treatment or prevention" / [MHRA](https://www.gov.uk/government/news/mhra-clarifies-regulatory-status-of-ambient-voice-technologies-used-in-the-nhs), accessed 2026-10-31 → `/clinic/social-prescribing-referral` - `F-b1ab4bde46` L171 _2. Regulators and rules › 2.5 Applying the MHRA test to EDGE Clinic (research reading, not legal advice)_: Triage, including online A&E (accident and emergency) routing / Likely yes / It supports diagnosis or decisions about urgency / same → `/clinic/social-prescribing-referral` - `F-ec9c53e611` L172 _2. Regulators and rules › 2.5 Applying the MHRA test to EDGE Clinic (research reading, not legal advice)_: Retinal screening AI / Yes / A diagnostic or screening purpose / same; [FDA LumineticsCore example](https://www.digitaldiagnostics.com/resources/insights/lumineticscore-engineered-by-digital-diagnostics-early-detection-of-diabetic-retinopathy-using-ai/), accessed 2026-10-31 → `/clinic/retinal-screening` - `F-a5a17e9a12` L173 _2. Regulators and rules › 2.5 Applying the MHRA test to EDGE Clinic (research reading, not legal advice)_: Prescribing safety (interaction checks) / Likely yes / It supports treatment decisions / [MHRA](https://www.gov.uk/government/news/mhra-clarifies-regulatory-status-of-ambient-voice-technologies-used-in-the-nhs), accessed 2026-10-31 → `/clinic/issue-field-safety-notice` - `F-f6d412b93d` L174 _2. Regulators and rules › 2.5 Applying the MHRA test to EDGE Clinic (research reading, not legal advice)_: Any agent that books, orders or refers without clinician review / Yes / "take automated action such as placing orders without clinician review" / same (OBSERVED quote) → `/clinic/issue-field-safety-notice` - `F-ae62034329` L180 [AGREEMENT] _2. Regulators and rules › 2.6 Date line (in force versus coming)_: 26 March 2025 / EHDS in force (applies in phases from 2027) / AGREEMENT / [Kennedys](https://www.kennedyslaw.com/en/thought-leadership/article/2026/the-european-health-data-space-is-in-force-implications-for-healthcare-medtech-and-life-sciences), accessed 2026-10-31 → **UNPLACED** - `F-df25558bdf` L181 [AGREEMENT] _2. Regulators and rules › 2.6 Date line (in force versus coming)_: 16 June 2025 / GB post-market surveillance rules apply / AGREEMENT / [GOV.UK](https://www.gov.uk/government/news/first-major-overhaul-of-medical-device-regulation-comes-into-force-across-great-britain), accessed 2026-10-31 → **UNPLACED** - `F-c4661bee12` L182 [AGREEMENT] _2. Regulators and rules › 2.6 Date line (in force versus coming)_: 6 January 2026 / FDA CDS and wellness guidance revised / AGREEMENT / [King & Spalding](https://www.kslaw.com/insights/articles/fda-updates-general-wellness-and-clinical-decision-support-guidance-documents), accessed 2026-10-31 → **UNPLACED** - `F-2cad20c062` L183 [AGREEMENT] _2. Regulators and rules › 2.6 Date line (in force versus coming)_: 5 February 2026 / UK DUAA automated decision-making reforms in force / AGREEMENT / [Travers Smith](https://www.traverssmith.com/knowledge/knowledge-container/uks-data-protection-reforms-take-effect-a-new-era-for-automated-decision-making/), accessed 2026-10-31 → `/clinic/retinal-screening` - `F-455779f8ea` L184 [AGREEMENT] _2. Regulators and rules › 2.6 Date line (in force versus coming)_: 6 April 2026 / Old DTAC form retired / AGREEMENT / [NHS Innovation Service](https://innovation.nhs.uk/news/updated-nhs-england-digital-technology-assessment-criteria-dtac-form-and-guidance/), accessed 2026-10-31 → **UNPLACED** - `F-1d833a2131` L185 [AGREEMENT] _2. Regulators and rules › 2.6 Date line (in force versus coming)_: 27 July 2026 / EU Digital Omnibus on AI in force / AGREEMENT / [Hunton](https://www.hunton.com/privacy-and-cybersecurity-law-blog/eu-digital-omnibus-on-ai-enters-into-force), accessed 2026-10-31 → **UNPLACED** - `F-5e5817458a` L186 [AGREEMENT] _2. Regulators and rules › 2.6 Date line (in force versus coming)_: 29 July 2026 / MHRA AVT guidance / AGREEMENT / [MHRA](https://www.gov.uk/government/news/mhra-clarifies-regulatory-status-of-ambient-voice-technologies-used-in-the-nhs), accessed 2026-10-31 → **UNPLACED** - `F-8b88a07be6` L187 [AGREEMENT] _2. Regulators and rules › 2.6 Date line (in force versus coming)_: 2 August 2026 / EU AI Act transparency obligations / AGREEMENT / [Hunton](https://www.hunton.com/privacy-and-cybersecurity-law-blog/eu-digital-omnibus-on-ai-enters-into-force), accessed 2026-10-31 → **UNPLACED** - `F-9806aadf4b` L188 [AGREEMENT] _2. Regulators and rules › 2.6 Date line (in force versus coming)_: 9 December 2026 / EU Product Liability Directive transposition deadline / AGREEMENT / [Covington](https://www.insideprivacy.com/uncategorized/eu-member-states-begin-rolling-out-new-product-liability-rules/), accessed 2026-10-31 → **UNPLACED** - `F-f4a42759fc` L189 _2. Regulators and rules › 2.6 Date line (in force versus coming)_: End 2026 / MHRA draft PCCP guidance (planned) / PROPOSED / [Qualio](https://www.qualio.com/blog/mhra-s-2026-devices-reform-what-changes-and-when), accessed 2026-10-31 → **UNPLACED** - `F-84a3dd9374` L190 _2. Regulators and rules › 2.6 Date line (in force versus coming)_: June 2027 / UK Medical Devices (Amendment) Regulations expected to apply / PROPOSED / [Latham & Watkins](https://www.lw.com/en/insights/mhra-publishes-draft-amendment-to-the-uk-medical-devices-regulations), accessed 2026-10-31 → **UNPLACED** - `F-55898fe301` L191 _2. Regulators and rules › 2.6 Date line (in force versus coming)_: July 2027 / HIPAA Security Rule final (target) / PROPOSED / [Fierce Healthcare](https://www.fiercehealthcare.com/health-tech/feds-push-back-hipaa-security-rule-overhaul-july-2027), accessed 2026-10-31 → **UNPLACED** - `F-5b4bd35ef6` L192 [AGREEMENT] _2. Regulators and rules › 2.6 Date line (in force versus coming)_: 2 December 2027 / EU AI Act Annex III high-risk rules apply / AGREEMENT / [Hunton](https://www.hunton.com/privacy-and-cybersecurity-law-blog/eu-digital-omnibus-on-ai-enters-into-force), accessed 2026-10-31 → `/clinic/act-high-risk-answer` - `F-60324e521e` L193 _2. Regulators and rules › 2.6 Date line (in force versus coming)_: April 2028 / 2028 / NHS App "full front door" and My NHS GP full rollout (targets) / PROPOSED / [Full Fact](https://fullfact.org/government-tracker/nhs-app-improvements/); [Digital Health](https://www.digitalhealth.net/2025/07/what-does-the-10-year-plan-mean-for-digital/), accessed 2026-10-31 → `/clinic/front-door-check` - `F-6c8e754b60` L194 [AGREEMENT] _2. Regulators and rules › 2.6 Date line (in force versus coming)_: 2 August 2028 / EU AI Act high-risk rules for AI in medical devices apply / AGREEMENT / [Hunton](https://www.hunton.com/privacy-and-cybersecurity-law-blog/eu-digital-omnibus-on-ai-enters-into-force), accessed 2026-10-31 → `/clinic/act-high-risk-answer` - `F-31c4cec2aa` L200 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 3. Agent assurance gaps in health_: "Agent" here means software that acts with or for a patient (books, refers, checks, messages), not only one that answers questions. → `/clinic/act-high-risk-answer` - `F-f01fd7fbf9` L204 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 3. Agent assurance gaps in health_: **Agent identity** (who or what is acting) / The OpenID Foundation names "agent identity fragmentation", "user impersonation" and "recursive delegation" as gaps ([OIDF](https://openid.net/new-whitepaper-tackles-ai-agent-identity-challenges/)). A 2026 survey says no production-ready standard traces authorisation through multi-hop agent chains ([arXiv 2604.23280](https://arxiv.org/pdf/2604.23280)). Accessed 2026-10-31. / OAuth token exchange "on-behalf-of" flows, and the OIDF's draft OpenID Connect Authority Claims Extension ([OIDF at ITU-T](https://openid.net/oidf-presents-at-itu-t-workshop-on-digital-id-and-agentic-ai/), accessed 2026-10-31). The FHIR `Device` resource can represent an AI system ([HL7 AI Transparency](https://hl7.org/fhir/uv/aitransparency/2026Jan/artifacts.html), accessed 2026-10-31). / A first-class agent identity issued by EDGE ID and bound to a human or organisation principal / PROPOSED (standards); OBSERVED (gap) → `/clinic/connect-lims` - `F-ff7090220a` L205 [AGREEMENT] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 3. Agent assurance gaps in health_: **Patient identity and verification** / NHS login does not do session management or logout, and does not replace PDS (the Personal Demographics Service) ([NHS login, archived](https://tnaqa.mirrorweb.com/ukgwa/20260505142011mp_/https://digital.nhs.uk/developer/api-catalogue/nhs-login), accessed 2026-10-31). / NHS login uses OpenID Connect, with P0/P5/P9 verification levels (P9 means checked against photo ID) / Use NHS login P9 for record access, and keep the agent session separate / AGREEMENT → `/clinic/register-child-patient` - `F-7c77116d41` L206 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 3. Agent assurance gaps in health_: **Mandate** (what the agent may do, for whom, until when) / A health-system preprint warns of "tool permissions that persist beyond the original use case" and "unclear accountability" ([arXiv 2601.15630](https://arxiv.org/pdf/2601.15630), accessed 2026-10-31). ONC says identity, authorisation and delegation are "still evolving" ([ONC ISP](https://isp.healthit.gov/appendix-vi-ai-interoperability), accessed 2026-10-31). / SMART on FHIR scopes ([SMART App Launch 2.2](https://hl7.org/fhir/smart-app-launch/apps.html), accessed 2026-10-31). No health standard for agent mandates. / A machine-readable mandate (scope, purpose, expiry, revocation) checked on every action / OBSERVED (gap) → `/clinic/person-held-health-record` - `F-9440a69925` L207 [AGREEMENT] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 3. Agent assurance gaps in health_: **Consent** / The OIDF names "consent fatigue" ([OIDF](https://openid.net/new-whitepaper-tackles-ai-agent-identity-challenges/), accessed 2026-10-31). OpenAI asks before using connected records by default ([OpenAI via Fierce](https://fiercehealthcare.com/ai-and-machine-learning/openai-launches-chatgpt-health-connect-data-health-apps-medical-records), accessed 2026-10-31). The UK's solely-automated-decision limits remain where health data is involved ([Travers Smith](https://www.traverssmith.com/knowledge/knowledge-container/uks-data-protection-reforms-take-effect-a-new-era-for-automated-decision-making/), accessed 2026-10-31). / The FHIR Consent resource (part of the FHIR R4 base specification, [HL7 FHIR R4 licence page](https://hl7.org/fhir/R4/license.html), accessed 2026-10-31). The Caldicott principles. / Consent captured per purpose and per agent, shown to the patient and revocable / AGREEMENT (law); OBSERVED (gap in practice) → `/clinic/connect-lims` - `F-663dcd3200` L208 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 3. Agent assurance gaps in health_: **Escalation to a human** / People using LLMs did no better than controls, with a "two-way communication breakdown" ([Oxford](https://www.ox.ac.uk/news/2026-02-10-new-study-warns-risks-ai-chatbots-giving-medical-advice), accessed 2026-10-31). ECRI's No. 1 hazard for 2026 is misuse of AI chatbots ([ECRI](https://home.ecri.org/blogs/ecri-news/misuse-of-ai-chatbots-tops-annual-list-of-health-technology-hazards), accessed 2026-10-31). The MHRA treats action "without clinician review" as device territory ([MHRA](https://www.gov.uk/government/news/mhra-clarifies-regulatory-status-of-ambient-voice-technologies-used-in-the-nhs), accessed 2026-10-31). / DCB0129 hazard logs. ONC's pattern of "write proposals" with human approval ([ONC ISP](https://isp.healthit.gov/appendix-vi-ai-interoperability), accessed 2026-10-31). / Hard-coded red-flag escalation, human sign-off on any clinical action, and user testing before launch (as the Oxford authors recommend) / OBSERVED (gap) → `/clinic/wearable-home-device-signals` - `F-2cb123a6f0` L209 [UNKNOWN] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 3. Agent assurance gaps in health_: **Liability** / No UK AI liability law was found. The Commission's report was summarised as patients wanting "clear accountability if things go wrong" ([GOV.UK](https://www.gov.uk/government/news/independent-commission-led-by-nhs-doctors-sets-out-blueprint-to-accelerate-safe-ai-adoption-in-healthcare), accessed 2026-10-31). In the EU, the new PLD (Product Liability Directive) makes software a product with strict liability ([Hogan Lovells](https://www.hoganlovells.com/en/publications/eu-introduces-comprehensive-digitalera-product-liability-directive), accessed 2026-10-31). Commentators say accountability for multi-step agent actions is "unsettled" ([Clearwater](https://clearwatersecurity.com/transcripts/agentic-ai-in-healthcare-navigating-regulatory-uncertainty-and-building-governance-that-lasts/), accessed 2026-10-31). / Device manufacturer liability, plus clinician duty of care. OpenMRS's licence puts responsibility on the clinician ([ScanCode](https://scancode-licensedb.aboutcode.org/openmrs-exception-to-mpl-2.0.html), accessed 2026-10-31). / A named accountable human or organisation for every agent mandate, written in the audit / UNKNOWN (UK law); AGREEMENT (EU PLD) → `/clinic/mental-health-care-crisis` - `F-56c71260b3` L210 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 3. Agent assurance gaps in health_: **Audit a regulator can re-run** / The Commission recommends continuous real-world monitoring and a public searchable safety database ([GOV.UK](https://www.gov.uk/government/news/independent-commission-led-by-nhs-doctors-sets-out-blueprint-to-accelerate-safe-ai-adoption-in-healthcare), accessed 2026-10-31). The WHO recommends independent post-release audits ([WHO](https://www.who.int/publications-detail-redirect/9789240084759), accessed 2026-10-31). No standard for replaying an agent's decisions was found. / HL7 "AI Transparency on FHIR": Provenance links AI output back to the model, the input prompt and the human verifier. It is a ballot/continuous build, not yet published ([HL7 AI Transparency](https://hl7.org/fhir/uv/aitransparency/2026Jan/artifacts.html); [requirements](https://build.fhir.org/ig/HL7/aitransparency-ig/en/requirements.html), accessed 2026-10-31). FDA PCCPs for planned model change ([Ropes & Gray](https://ropesgray.com/en/insights/alerts/2024/12/fda-finalizes-guidance-on-predetermined-change-control-plans-for-ai-enabled-device), accessed 2026-10-31). / A content-addressed log of inputs, model version, prompt, tools called, mandate and human sign-off, so a regulator can replay a case / PROPOSED (standard); OBSERVED (gap) → `/clinic/connect-lims` - `F-361786b334` L211 [CLAIM] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 3. Agent assurance gaps in health_: **Tool-calling standard** / MCP was reportedly donated to the Linux Foundation's Agentic AI Foundation in December 2025 ([Keragon](https://www.keragon.com/blog/medical-mcp-servers), accessed 2026-10-31). There is no official HL7 FHIR MCP profile, only community RFCs ([GitHub discussion](https://github.com/HeliosSoftware/hfs/discussions/231), accessed 2026-10-31). / MCP over FHIR with SMART auth (ONC pattern) / Wrap FHIR and NHS APIs as tools with per-mandate scopes / CLAIM / PROPOSED → `/clinic/connect-lims` - `F-9458c1d68f` L212 [UNKNOWN] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 3. Agent assurance gaps in health_: **Regulator view of generative and agentic devices** / None of the FDA-authorised AI devices cited by one commentator is generative or agentic ([Clearwater](https://clearwatersecurity.com/transcripts/agentic-ai-in-healthcare-navigating-regulatory-uncertainty-and-building-governance-that-lasts/), accessed 2026-10-31). The UK draft regulations do not yet update software-specific provisions ([Qualio](https://www.qualio.com/blog/mhra-s-2026-devices-reform-what-changes-and-when), accessed 2026-10-31). / The AI Airlock sandbox ([GOV.UK](https://www.gov.uk/government/collections/ai-airlock-the-regulatory-sandbox-for-aiamd), accessed 2026-10-31) / Engage the MHRA early (Airlock Phase 3, if it opens) / UNKNOWN → **UNPLACED** - `F-b42862b9d0` L218 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 4. Reusable systems: open standards and open-source_: SPDX is the standard short-code for licences (for example `Apache-2.0`). Where there is no SPDX code, `LicenseRef-…` marks a custom licence. → **UNPLACED** - `F-dec47f8279` L224 _4. Reusable systems: open standards and open-source › 4.1 Standards and specifications_: **HL7 FHIR R4** / Fast Healthcare Interoperability Resources: the web API data standard for health records (Patient, Observation, MedicationRequest, Consent, Provenance, AuditEvent and so on). / `CC0-1.0`. SNOMED CT, LOINC and DICOM content are excluded and need their own terms. / Normative parts; widely used / Core data model and API / OBSERVED / [HL7 FHIR R4 licence](https://hl7.org/fhir/R4/license.html), accessed 2026-10-31 → `/clinic/connect-lims` - `F-b47d5573c7` L225 [UNKNOWN] _4. Reusable systems: open standards and open-source › 4.1 Standards and specifications_: **FHIR UK Core** / UK profiles of FHIR R4 (HL7 UK and NHS England). STU2 (standard for trial use 2) is release 2.0; 2.0.2 is the latest seen. / UNKNOWN (not checked) / STU (trial use) / Required shape for NHS integrations / CLAIM / [HL7 UK](https://www.hl7.org.uk/2024/05/28/fhir-uk-core-release-2-0-stu2/); [Simplifier](https://simplifier.net/guide/uk-core-implementation-guide-stu2), accessed 2026-10-31 → `/clinic/connect-lims` - `F-1afead96e5` L226 [UNKNOWN] _4. Reusable systems: open standards and open-source › 4.1 Standards and specifications_: **SMART on FHIR – App Launch 2.2** / OAuth 2.0 / OpenID Connect profile for apps and backend services that access FHIR. / Spec: UNKNOWN. Reference JS client: `Apache-2.0` (OBSERVED). / STU 2.2, current published / The authorisation layer for agents (scopes) / OBSERVED (client licence) / [HL7 SMART App Launch](https://hl7.org/fhir/smart-app-launch/apps.html); [client-js LICENSE](https://raw.githubusercontent.com/smart-on-fhir/client-js/master/LICENSE), accessed 2026-10-31 → `/clinic/connect-lims` - `F-ea9c094df4` L227 [UNKNOWN] _4. Reusable systems: open standards and open-source › 4.1 Standards and specifications_: **HL7 AI Transparency on FHIR** / Provenance, labels and model cards for AI-generated FHIR data. / UNKNOWN (an HL7 IG; not checked) / Ballot / continuous build, "not an authorized publication" / The best fit for replayable AI audit / CLAIM / [HL7](https://hl7.org/fhir/uv/aitransparency/2026Jan/artifacts.html); [build](https://build.fhir.org/ig/HL7/aitransparency-ig/en/requirements.html), accessed 2026-10-31 → `/clinic/connect-lims` - `F-a9f9f1b5b0` L228 [UNKNOWN] _4. Reusable systems: open standards and open-source › 4.1 Standards and specifications_: **SNOMED CT (UK Edition)** / Clinical terminology. dm+d codes are SNOMED CT codes. / `LicenseRef-SNOMED-CT-UK-Affiliate`: you accept both the UK National Release and the International affiliate licences via TRUD (NHS England's terminology download service). Fees and commercial terms were not confirmed. / Mature / Mandatory for UK clinical coding / OBSERVED (licence pages exist); UNKNOWN (commercial terms) / [TRUD UK licence](https://isd.hscic.gov.uk/trud/users/guest/filters/0/licence/10); [TRUD International licence](https://isd.hscic.gov.uk/trud/users/guest/filters/0/licence/4), accessed 2026-10-31 → `/clinic/check-record-accuracy` - `F-6cfb3f844a` L229 [UNKNOWN] _4. Reusable systems: open standards and open-source › 4.1 Standards and specifications_: **dm+d** (Dictionary of Medicines and Devices) / The NHS standard list of medicines and devices, from the NHSBSA (NHS Business Services Authority), downloaded from TRUD. / Open Government Licence. Version not stated in the source; `OGL-UK-3.0` is likely but UNKNOWN. / Mature, weekly / Medicine identity for prescribing safety (identity, not interactions) / CLAIM / [NHSBSA](https://www.nhsbsa.nhs.uk/pharmacies-gp-practices-and-appliance-contractors/nhs-dictionary-medicines-and-devices-dmd/snomed-ct-and-dmd); [NHSBSA dm+d](https://nhsbsa.nhs.uk/pharmacies-gp-practices-and-appliance-contractors/dictionary-medicines-and-devices-dmd), accessed 2026-10-31 → `/clinic/medication-safety-check` - `F-78ce44a55d` L230 _4. Reusable systems: open standards and open-source › 4.1 Standards and specifications_: **openEHR** / A clinical data modelling standard (archetypes and templates). / Specification repo: `CC-BY-SA-3.0` / Mature, strongest in Europe / Optional long-term record store; FHIR is the better fit for UK integration / OBSERVED / [openEHR specifications-RM LICENSE](https://raw.githubusercontent.com/openEHR/specifications-RM/master/LICENSE), accessed 2026-10-31 → **UNPLACED** - `F-af029e2bf6` L231 [CLAIM] _4. Reusable systems: open standards and open-source › 4.1 Standards and specifications_: **IHE profiles** (Integrating the Healthcare Enterprise) / Integration profiles, such as audit trails and document sharing. / Documents: `CC-BY-4.0`. Patents are handled separately, on reasonable-and-non-discriminatory terms. / Mature / Audit and identity patterns / CLAIM / [IHE Europe IP](https://ihe-europe.net/intellectual-property); [IHE General Introduction](https://profiles.ihe.net/GeneralIntro/ch-11.html), accessed 2026-10-31 → `/clinic/claim-edge-id-record` - `F-82dd04ab7f` L232 [UNKNOWN] _4. Reusable systems: open standards and open-source › 4.1 Standards and specifications_: **OMOP CDM / OHDSI vocabularies** (Observational Medical Outcomes Partnership Common Data Model; Observational Health Data Sciences and Informatics) / A research data model and vocabularies for population analytics. / Vocabularies: `Apache-2.0`, with some licensed vocabularies (for example FDB) needing proof of a licence. CDM spec: UNKNOWN. / Mature / Detection analytics and evaluation, not care delivery ("should not be used for individual patient healthcare") / CLAIM / [HL7 THO OMOP](https://build.fhir.org/ig/HL7/UTG/en/NamingSystem-OMOP.html), accessed 2026-10-31 → `/clinic/read-assistive-output-image` - `F-3cb70605e8` L233 [UNKNOWN] _4. Reusable systems: open standards and open-source › 4.1 Standards and specifications_: **MCP** (Model Context Protocol) / An open protocol for connecting agents to tools and data. / UNKNOWN (not checked) / Emerging / The tool layer for agents over FHIR / CLAIM / [ONC ISP](https://isp.healthit.gov/appendix-vi-ai-interoperability); [Keragon](https://www.keragon.com/blog/medical-mcp-servers), accessed 2026-10-31 → `/clinic/price-policy-clinic-data` - `F-d5f10d9813` L239 _4. Reusable systems: open standards and open-source › 4.2 Open-source software_: **HAPI FHIR** / Java FHIR server and library / `Apache-2.0` / Mature, widely used / A self-hosted FHIR store and validator / OBSERVED / [LICENSE.txt](https://raw.githubusercontent.com/hapifhir/hapi-fhir/master/LICENSE.txt), accessed 2026-10-31 → **UNPLACED** - `F-90609ba336` L240 [CLAIM] _4. Reusable systems: open standards and open-source › 4.2 Open-source software_: **Medplum** / TypeScript FHIR-native platform (server, SDKs, React components) / `Apache-2.0` / Production. ONC certification (dated 31 December 2025) is claimed, but disputed for self-hosted copies. / Fastest route to a FHIR backend with auth and bots / OBSERVED (licence); CLAIM (certification) / [LICENSE.txt](https://raw.githubusercontent.com/medplum/medplum/main/LICENSE.txt); [Mindbowser](https://www.mindbowser.com/open-source-ehr/), accessed 2026-10-31 → **UNPLACED** - `F-1333e9c3c2` L241 [CLAIM] _4. Reusable systems: open standards and open-source › 4.2 Open-source software_: **OpenMRS** / Open-source medical record system, strong in low- and middle-income countries / `MPL-2.0` plus a Healthcare Disclaimer (`LicenseRef-OpenMRS-Exception-to-MPL-2.0`) / Mature / Global and offline settings, not NHS primary care / OBSERVED (MPL-2.0); CLAIM (disclaimer) / [LICENSE](https://raw.githubusercontent.com/openmrs/openmrs-core/master/LICENSE); [ScanCode](https://scancode-licensedb.aboutcode.org/openmrs-exception-to-mpl-2.0.html), accessed 2026-10-31 → `/clinic/claim-edge-id-record` - `F-2e7d534397` L242 _4. Reusable systems: open standards and open-source › 4.2 Open-source software_: **EHRbase** / openEHR clinical data repository / `Apache-2.0` / Production / Only if the openEHR route is chosen / OBSERVED / [LICENSE](https://raw.githubusercontent.com/ehrbase/ehrbase/develop/LICENSE), accessed 2026-10-31 → `/clinic/request-clinic-data-quote` - `F-88e0c8463e` L243 _4. Reusable systems: open standards and open-source › 4.2 Open-source software_: **OHDSI WebAPI / ATLAS** / OMOP analytics tools / `Apache-2.0` / Mature / Population detection analytics and research / OBSERVED / [WebAPI LICENSE](https://raw.githubusercontent.com/OHDSI/WebAPI/master/LICENSE); [Atlas LICENSE](https://raw.githubusercontent.com/OHDSI/Atlas/master/LICENSE), accessed 2026-10-31 → **UNPLACED** - `F-a9e4cc9ef0` L249 [CLAIM] _4. Reusable systems: open standards and open-source › 4.3 NHS national services (reuse by integration, not by code)_: **NHS login** / Patient identity (OpenID Connect). Verification levels P0, P5 and P9. Partners must meet the identity standard DAPB3051. / n/a (service; partner onboarding) / Live, national / Patient sign-in and verified identity / CLAIM (archived NHS page) / [NHS login, archived](https://tnaqa.mirrorweb.com/ukgwa/20260505142011mp_/https://digital.nhs.uk/developer/api-catalogue/nhs-login), accessed 2026-10-31 → `/clinic/receive-pre-authorisation` - `F-6cd5e64389` L250 [UNKNOWN] _4. Reusable systems: open standards and open-source › 4.3 NHS national services (reuse by integration, not by code)_: **GP Connect** (Access Record: Structured) / FHIR API to read structured GP records. Working version 1.5.0. / n/a (service) / Live, but the service page says there is "no longer an onboarding route" for new developers. It needs SCAL (the supplier conformance assessment list), a CSO (clinical safety officer) and a hazard log. / Would be high value; access is UNKNOWN for a new supplier / CLAIM / UNKNOWN / [NHS England DSIC wiki](https://nhse-dsic.atlassian.net/wiki/spaces/DCSDCS/pages/12723519489/GP+Connect+Access+Record+Structured), accessed 2026-10-31 → `/clinic/investigate-clinical-safety` - `F-fdfc39ee91` L254 _4. Reusable systems: open standards and open-source › 4.4 Recommended reuse stack (research suggestion only)_: **Data and API:** FHIR R4 + UK Core on HAPI FHIR or Medplum (both `Apache-2.0`, OBSERVED above). → `/clinic/investigate-clinical-safety` - `F-e84807cd1f` L255 _4. Reusable systems: open standards and open-source › 4.4 Recommended reuse stack (research suggestion only)_: **Terminology:** SNOMED CT UK and dm+d from TRUD (licences above). → `/clinic/investigate-clinical-safety` - `F-523a18921d` L256 _4. Reusable systems: open standards and open-source › 4.4 Recommended reuse stack (research suggestion only)_: **Auth:** NHS login for patients, plus SMART scopes for agents. → `/clinic/investigate-clinical-safety` - `F-3828d5ba30` L257 _4. Reusable systems: open standards and open-source › 4.4 Recommended reuse stack (research suggestion only)_: **Audit:** FHIR Provenance and AuditEvent following the AI Transparency on FHIR IG. → `/clinic/investigate-clinical-safety` - `F-d6c7713a75` L258 _4. Reusable systems: open standards and open-source › 4.4 Recommended reuse stack (research suggestion only)_: **Analytics:** OMOP / OHDSI for detection evaluation. → `/clinic/investigate-clinical-safety` - `F-fe276e0da8` L259 _4. Reusable systems: open standards and open-source › 4.4 Recommended reuse stack (research suggestion only)_: **Interactions:** license BNF or FDB until an open, clinical-grade source exists (sources in 1.3). → `/clinic/investigate-clinical-safety` - `F-d347fc28ea` L265 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 5. Open questions (UNKNOWN) to resolve next_: Has the government formally responded to the National Commission? GOV.UK and Ropes & Gray differ. [GOV.UK](https://www.gov.uk/government/news/independent-commission-led-by-nhs-doctors-sets-out-blueprint-to-accelerate-safe-ai-adoption-in-healthcare); [Ropes & Gray](https://www.ropesgray.com/en/insights/viewpoints/2026/10/102o5fp/ai-medical-devices-in-great-britain-what-the-mhras-response-to-the-national-com), accessed 2026-10-31. → **UNPLACED** - `F-46b3dcc5ca` L266 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 5. Open questions (UNKNOWN) to resolve next_: When will the revised DCB0129/DCB0160 be published, and what will they require for AI? [NHS England](https://www.engage.england.nhs.uk/patient-safety/national-review-of-clinical-risk-management-standa/), accessed 2026-10-31. → `/clinic/nominate-carer-proxy` - `F-1ba06a174f` L267 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 5. Open questions (UNKNOWN) to resolve next_: Is there any GP Connect onboarding route for a new supplier? [NHS England DSIC wiki](https://nhse-dsic.atlassian.net/wiki/spaces/DCSDCS/pages/12723519489/GP+Connect+Access+Record+Structured), accessed 2026-10-31. → **UNPLACED** - `F-c8e7342440` L268 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 5. Open questions (UNKNOWN) to resolve next_: What was the outcome of the UK NSC consultation on AI in diabetic eye screening? [AOP](https://www.aop.org.uk/ot/news/2026/05/13/uk-national-screening-committee-consults-on-use-of-ai-in-diabetic-eye-screening), accessed 2026-10-31. → `/clinic/retinal-screening` - `F-21c0fd88a5` L269 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 5. Open questions (UNKNOWN) to resolve next_: What are the licences for UK Core, the SMART App Launch spec, MCP, DDInter and the OMOP CDM spec? These were not checked in primary sources. → `/clinic/retinal-screening` - `F-80005aa855` L270 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 5. Open questions (UNKNOWN) to resolve next_: What is the current UK status of Livi/Kry, Push Doctor and Doctaly? See the 1.1 sources. → `/clinic/retinal-screening` - `F-6b2750b6a4` L271 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 5. Open questions (UNKNOWN) to resolve next_: Medscape's interaction checker and the Google ARDA retinal AI were not researched. → `/clinic/retinal-screening` ###### `research/clinic/ENGINE-MEDPLUM.md` (101 findings) - `F-5bf3f17867` L3 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record_: Built 1 Nov 2026 by ATBAN Claude-Jegudiel 00:5531102026-1 (service-builder lane). Code: `runtime/domains/clinic.py` (the → **UNPLACED** - `F-45790beb84` L5 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record_: `tests/test_clinic_domain.py` (unit tests with a fake FHIR server; `LiveSmoke` against the real server when `MEDPLUM_SMOKE=1`). → **UNPLACED** - `F-b3c4358449` L6 [OBSERVATION] _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record_: Every statement below is marked OBSERVATION (measured, with how and when), AGREEMENT (a decision taken here) or CLAIM. → **UNPLACED** - `F-690c8c0617` L10 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 1. The engine (OBSERVATION, jude-1, 1 Nov 2026 04:5x–05:1xZ)_: Medplum `5.2.13-15a306f` runs as podman container `jude-engine-medplum` in pod `jude-medplum-engines` with → **UNPLACED** - `F-cc7eedf858` L16 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 1. The engine (OBSERVATION, jude-1, 1 Nov 2026 04:5x–05:1xZ)_: Stop seen during this lane: at 04:48:05Z a `podman pod stop` (pid 1428339) stopped the three engine pods; systemd brought the → **UNPLACED** - `F-ce88e92c04` L17 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 1. The engine (OBSERVATION, jude-1, 1 Nov 2026 04:5x–05:1xZ)_: Medplum units back at 04:51:42Z. After that restart the database container's `/dev/shm` was `drwx------ root` and every → **UNPLACED** - `F-95ac78e586` L18 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 1. The engine (OBSERVATION, jude-1, 1 Nov 2026 04:5x–05:1xZ)_: PostgreSQL backend failed with `could not open shared memory segment "/PostgreSQL.…": Permission denied` (Medplum answered → **UNPLACED** - `F-1d1489e0e5` L19 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 1. The engine (OBSERVATION, jude-1, 1 Nov 2026 04:5x–05:1xZ)_: 400 on every FHIR call, `/healthcheck` still said ok). Fix applied 05:0xZ: `chmod 1777 /dev/shm` inside → **UNPLACED** - `F-6c91d6e322` L20 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 1. The engine (OBSERVATION, jude-1, 1 Nov 2026 04:5x–05:1xZ)_: `jude-engine-medplum-db`; FHIR calls answered 200 at once. Cause of the `podman pod stop` itself: not determinable from the → **UNPLACED** - `F-96da7e1ece` L21 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 1. The engine (OBSERVATION, jude-1, 1 Nov 2026 04:5x–05:1xZ)_: journal (no sudo line names it); the engines' owner should confirm. The shm mode will revert on the next container start → **UNPLACED** - `F-dad2bb6fb7` L22 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 1. The engine (OBSERVATION, jude-1, 1 Nov 2026 04:5x–05:1xZ)_: unless the pod's infra container or the unit sets it: an open item for the engine lane. → **UNPLACED** - `F-e03792ed05` L23 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 1. The engine (OBSERVATION, jude-1, 1 Nov 2026 04:5x–05:1xZ)_: The seeded super-admin login still accepts Medplum's published first-run credentials (checked 04:5xZ: `POST /auth/login` 200). → **UNPLACED** - `F-bb38fc3e57` L24 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 1. The engine (OBSERVATION, jude-1, 1 Nov 2026 04:5x–05:1xZ)_: That is a security finding for the engine lane: rotate it. This lane did not change it (not its file) and does not use it at → **UNPLACED** - `F-94e61074e2` L29 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 2. Authentication (AGREEMENT: how; OBSERVATION: that it works; no values here)_: Medplum Project `EDGE Clinic` (`strictMode: true`) created 1 Nov 2026 through the super-admin FHIR API, with one → **UNPLACED** - `F-d2466b725f` L30 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 2. Authentication (AGREEMENT: how; OBSERVATION: that it works; no values here)_: `ClientApplication` `edge-clinic-adapter` and its `ProjectMembership` (`/auth/me` as the client: project `EDGE Clinic`, profile → **UNPLACED** - `F-2888d5eeac` L31 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 2. Authentication (AGREEMENT: how; OBSERVATION: that it works; no values here)_: `ClientApplication`; never the Super Admin project). `POST /admin/projects` answers 404 on this build; the FHIR route was used. → **UNPLACED** - `F-04528410bb` L32 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 2. Authentication (AGREEMENT: how; OBSERVATION: that it works; no values here)_: The adapter authenticates with OAuth 2.0 **client credentials** (`POST /oauth2/token`, `grant_type=client_credentials`, → **UNPLACED** - `F-7317a4ccff` L33 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 2. Authentication (AGREEMENT: how; OBSERVATION: that it works; no values here)_: client id and secret in the form body; the HTTP Basic form answers 400 on this build). The token (3600 s) is cached in the → **UNPLACED** - `F-403cbed57f` L34 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 2. Authentication (AGREEMENT: how; OBSERVATION: that it works; no values here)_: process and renewed 60 s before expiry; a 401 triggers one fresh token and one retry. → **UNPLACED** - `F-5d65f28726` L35 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 2. Authentication (AGREEMENT: how; OBSERVATION: that it works; no values here)_: The credential lives at `/etc/x-x/plugs/medplum.env` (`root:root 0640`; there is no `x-x` user or group on jude-1 and → **UNPLACED** - `F-a3b95537bb` L36 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 2. Authentication (AGREEMENT: how; OBSERVATION: that it works; no values here)_: `/etc/x-x/plugs/griffin.key` does not exist yet, so the Griffin precedent could not be copied literally; the web service runs → **UNPLACED** - `F-5201d0e69f` L37 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 2. Authentication (AGREEMENT: how; OBSERVATION: that it works; no values here)_: as root and reads it). Keys: `MEDPLUM_BASE_URL`, `MEDPLUM_PROJECT_ID`, `MEDPLUM_CLIENT_ID`, `MEDPLUM_CLIENT_SECRET`. The same → **UNPLACED** - `F-2251b5ee0c` L38 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 2. Authentication (AGREEMENT: how; OBSERVATION: that it works; no values here)_: names in the environment override the file (tests use that with fake values). No value is ever written to a record, a reply, → **UNPLACED** - `F-4d02529370` L39 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 2. Authentication (AGREEMENT: how; OBSERVATION: that it works; no values here)_: the specification or Git. Without the file `medplum.bound()` is False and a live tenant's operation answers `no_answer` with → **UNPLACED** - `F-0ccbef1314` L44 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Clinic has 4,151 receiver-labelled operations. 624 are bound to Medplum; 3,527 stay with the estate record or another receiver. → **UNPLACED** - `F-07d9b62164` L45 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: (A count is live: `python3 -c "from runtime.domains import clinic, json; print(json.dumps(clinic.coverage()))"`.) → **UNPLACED** - `F-e788d5690d` L47 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: FHIR R4 resource / Bound operations / Estate entities and units (from `spec/receivers.json` labels `estate-record::`) → **UNPLACED** - `F-1e6699e7a3` L49 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Appointment (+ Location, Schedule, Slot when a time is chosen) / 89 / `record` of the booking units: book-an-appointment, book-for-the-person-i-care-for, book-a-pathway-test, book-from-the-referral-queue, book-an-interpreter, book-transport, rebook-or-cancel-an-appointment, offer-a-cancellation-slot, hold-a-slot-for-a-tertiary-handover, check-in-an-arriving-patient, check-in-on-arrival, front-door-check-in, confirm-attendance-and-prepare, attend-an-appointment-with-the-person, handle-a-did-not-attend, receive-a-did-not-attend-notice, record-attendance-outcome, add-an-appointment-to-my-calendar, schedule-the-scan, see-my-day; `feature_appointment` → **UNPLACED** - `F-ce0de968bc` L50 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: ServiceRequest / 88 / `referral`, `feature_referral`; referral and order units (make-an-urgent-same-day-referral, refer-on-a-suspected-cancer-pathway, refer-to-a-crisis-team, refer-to-tertiary-care, refer-to-the-mental-health-stream, refer-to-the-multidisciplinary-team, social-prescribing-referral, ask-for-advice-and-guidance, order-a-diagnostic-test, order-a-direct-access-test, order-a-genomic-test, request-a-second-read, request-a-second-read-or-discrepancy-review, request-a-second-opinion, track-a-referral, triage-a-referral, triage-a-mental-health-referral, close-a-referral, vet-an-imaging-request) → **UNPLACED** - `F-dc22a01ea5` L51 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Observation / 82 / `result`, `feature_result`, `feature_observation`; take-observations, record-a-point-of-care-urine-result, record-a-wound-or-skin-check, measure-a-nodule-and-its-growth, share-a-wearable-signal, wearable-signals, take-an-online-early-symptom-test, run-an-online-early-symptom-test, review-a-real-result, authorise-a-result, tell-the-patient-a-result, explain-a-result-to-the-patient, explain-a-result-in-plain-language, receive-a-critical-result-for-my-patient, phone-a-critical-result, view-my-results, view-results-in-scope, chase-missing-results, retinal-screening, grade-a-retinal-image → **UNPLACED** - `F-fea1649aad` L52 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Consent / 77 / `consent`, `feature_consent`; give-or-withdraw-consent, give-consent-on-behalf, take-informed-consent-for-research, apply-a-national-data-opt-out, record-sample-consent-version, set-my-data-sharing-choices, share-clinic-data-with-my-insurer, share-a-medicines-list-with-the-clinic, share-images-with-another-trust, approve-a-data-release → **UNPLACED** - `F-9176d074b0` L53 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: CarePlan / 57 / `safety_plan`, `plan`, `feature_careplan`; write-a-care-plan, maintain-the-suicide-prevention-plan, follow-up-an-action-plan, set-a-surveillance-plan, discharge-from-the-pathway, discharge-from-the-stream, track-my-cancer-pathway, record-decision-to-treat → **UNPLACED** - `F-f7b20a6fc5` L54 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: QuestionnaireResponse / 36 / `assessment`, `answer`, `feature_questionnaire`; complete-a-pre-clinic-questionnaire, wellbeing-questionnaires, complete-a-mental-health-check-in, complete-a-risk-formulation, run-a-pre-assessment, score-and-interpret-a-response, act-on-a-high-risk-answer, symptom-check-and-triage, online-urgent-triage → **UNPLACED** - `F-07c1dbaa8f` L55 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: MedicationRequest / 33 / `feature_prescription`, `feature_medicationrequest`; prescribe, prescribe-under-the-pathway-protocol, order-a-repeat-medicine, order-repeat-medicines-on-behalf, review-a-repeat-request, clinically-check-a-prescription, check-a-chemotherapy-prescription → **UNPLACED** - `F-281acb3033` L56 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Flag / 29 / `crisis_alert`, `flag`; flag-a-critical-finding, receive-a-crisis-alert, receive-a-crisis-alert-for-the-person, record-a-safeguarding-concern, raise-a-safeguarding-concern-from-a-response → **UNPLACED** - `F-b4395d81e8` L57 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Patient / 25 / `feature_patient`; register-a-walk-in-patient, register-a-child-as-a-patient, update-patient-demographics → **UNPLACED** - `F-79d13d2621` L58 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Communication / 21 / message-my-care-team, send-a-results-ready-notice, send-appointment-letters-and-reminders, send-a-check-in-to-a-caseload, make-a-follow-up-call → **UNPLACED** - `F-de4e41548c` L59 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: MedicationStatement / 18 / `medicine_list`, `feature_medication`; reconcile-medicines, reconcile-medicines-at-admission, view-my-medicines-list → **UNPLACED** - `F-3f6218e2f2` L60 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Specimen / 15 / take-a-blood-or-urine-sample, receive-and-book-in-a-sample, reject-an-unlabelled-sample, track-sample-custody → **UNPLACED** - `F-8ae36275c3` L61 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Encounter / 10 / `check_in`, `feature_encounter`; record-a-clinic-consultation, hold-a-remote-session → **UNPLACED** - `F-0ab6ef3d09` L62 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: DiagnosticReport / 10 / `feature_diagnosticreport`; report-histology, issue-a-genomic-report → **UNPLACED** - `F-4f826175bb` L63 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: MedicationDispense, Immunization, RelatedPerson, AllergyIntolerance, AdverseEvent / 5 each / dispense-a-medicine; give-a-vaccination; nominate-a-carer-or-proxy; record-an-adverse-drug-reaction; report-a-side-effect → **UNPLACED** - `F-303beafed9` L64 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: MedicationAdministration 4, Condition 3, Procedure 2 / administer-a-medicine; `feature_condition`; `feature_procedure` → **UNPLACED** - `F-663b4fbc67` L66 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Operation classes covered, each with the receiver contract (create 159, read 30, change 383, remove 5, send 12, decide 27, → **UNPLACED** - `F-b1ded01cc3` L67 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: plan 8 = 624, measured by `clinic.plan()`; `runtime/receivers.effective_class` still turns a "remove" of a workflow `record` into a → **UNPLACED** - `F-45c35cd01e` L70 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: **create** POST, then GET; confirmed with the Medplum id (the journey carries it on as `id` and `_id`). → **UNPLACED** - `F-07110ed400` L71 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: **read** GET by any id in the inputs that is this resource in this tenant; else a search `_tag=` (newest first, 50). → **UNPLACED** - `F-610adf2610` L72 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: **change** GET, merge, PUT; `expected_version`/`version` becomes `If-Match: W/""` and a stale one is → **UNPLACED** - `F-463d6dd277` L73 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: `rejected VERSION_CONFLICT` (checked before the write and honoured by the server: 412). A change with no existing resource → **UNPLACED** - `F-898cd16eff` L74 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: creates one (a journey's first touch), as the estate record does. The operation's verb moves the status: cancel/withdraw/ → **UNPLACED** - `F-b9d7ef08e5` L77 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: **remove** DELETE, then GET must answer 410 Gone; a cancelled Appointment frees its Slot first. → **UNPLACED** - `F-7d707eb91a` L78 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: **send, decide, plan** like change, and the status lands on the resource's done state (`STATUS[resource]["done"]`); the → **UNPLACED** - `F-99ed311f64` L80 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: **Confirmed** only when the read-back equals what was written (meta and narrative aside); **rejected** with `refusal_code` → **UNPLACED** - `F-930a9ae9f8` L82 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: `RATE_LIMITED`); **no_answer** on 5xx, on a transport failure (`TIMEOUT`, `UNREACHABLE`), on a read-back mismatch, or when no → **UNPLACED** - `F-2fe69b1765` L85 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: How inputs become FHIR: every input is kept verbatim in the extension `https://judiverse.tech/fhir/StructureDefinition/estate-inputs` → **UNPLACED** - `F-161591e685` L86 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: (sorted JSON), so the read-after-write compares the whole input set exactly; on top, fields the unit declares (`data..fields` → **UNPLACED** - `F-d55a1a195c` L87 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: types) or names are placed structurally: a `datetime`/`date` field (or any value with a time of day) becomes `Appointment.start/end` → **UNPLACED** - `F-86285cd144` L89 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: and so on; a numeric input becomes `Observation.valueQuantity`; `name`, `date_of_birth`, `nhs_number` (10 digits, system → **UNPLACED** - `F-3192635ff8` L92 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: value: //}` (a single identifier on AdverseEvent and QuestionnaireResponse, which are 0..1 in R4). The → **UNPLACED** - `F-14e297f01f` L93 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: person behind the actor (or an agent's mandate owner) is one `Patient` per tenant found by identifier `https://judiverse.tech/fhir/edge-id` → **UNPLACED** - `F-acfc29dc58` L96 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Booking a time: an Appointment that receives a start creates (once per tenant and clinic name) a `Location`, a `Schedule` whose → **UNPLACED** - `F-973e3f60b4` L97 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: actor is that Location, and a busy `Slot` for the 30 minutes, and references the Slot; cancelling frees the Slot. FHIR invariant → **UNPLACED** - `F-7029d11968` L98 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: app-3 ("only proposed or cancelled appointments can be missing start/end") is honoured: a confirm before a time is chosen leaves the → **UNPLACED** - `F-b67f2a59a4` L99 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Appointment `proposed`, never `booked` (OBSERVED: the live server rejects the alternative with 400 INVARIANT). → **UNPLACED** - `F-bd0032a39c` L103 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: `record` of the 496 Clinic units not in the map (219 capability units that are rotas, audits, incidents, governance, trials, → **UNPLACED** - `F-60c8dfd2a8` L104 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: genomics pipelines and the like, and 277 research, parity and problem units; 995 operations together) → **UNPLACED** - `F-4f5fb8baeb` L105 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: and every `case` (659: the workflow case is the estate's own, not a clinical resource); `evidence_query` (606) and the → **UNPLACED** - `F-7196c318c8` L107 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: `feature_fhir` (579), `feature_screen`, `feature_chart`, `feature_manage`, `feature_ios`, `feature_web` and the other → **UNPLACED** - `F-c875e2deff` L108 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: `feature_*` entities that are competitor-feature extractions, not clinical entities. A `feature_` is bound only when → **UNPLACED** - `F-a1eb1f7141` L109 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: `` is a FHIR resource (Patient, Appointment, Observation, Condition, Encounter, Procedure, Immunization, CarePlan, → **UNPLACED** - `F-5c968a823d` L112 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: The 112 operations whose receiver is an `external:` slot (NHS services, voice telephony, forecasting models…). → **UNPLACED** - `F-675dcd1acb` L113 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: `session` (clinic sessions and therapy sessions share one word), `escalation`, `trusted_contact`, `concern`, `review`, → **UNPLACED** - `F-0e71afb0ab` L114 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: `outcome` and the other single-digit entities: candidates (Schedule, Flag, RelatedPerson…) once a unit by unit reading says so. → **UNPLACED** - `F-7fa3aafcbd` L115 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: Tenant isolation is by tag inside the one Project `EDGE Clinic`, enforced by the adapter on every read and write. Medplum's → **UNPLACED** - `F-696bad7dad` L116 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: own isolation is one Project (and one ClientApplication) per tenant; that is the next step and needs a provisioning route → **UNPLACED** - `F-adc56efdf6` L117 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: (`POST /admin/projects` is 404 on this build; the super-admin FHIR route works). → **UNPLACED** - `F-0e6580adc3` L118 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: Binary attachments (images, documents) are not sent: `DocumentReference.content` carries a title only. NHS connectivity (PDS, → **UNPLACED** - `F-a89671bfdf` L119 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: e-RS, GP Connect) is not inferred from FHIR availability, as the engine lane's note says. → **UNPLACED** - `F-737889f476` L123 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 5. Measured round trips (OBSERVATION, `LiveSmoke`, jude-1, 1 Nov 2026 05:1xZ, tenant `smoke-d14def5f`, loopback)_: 20 resource types, each create → decide → stale-version change (412 expected) → read → remove (DELETE, 410 read-back): 0 failures, → **UNPLACED** - `F-548c7c4bd0` L124 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 5. Measured round trips (OBSERVATION, `LiveSmoke`, jude-1, 1 Nov 2026 05:1xZ, tenant `smoke-d14def5f`, loopback)_: 237 HTTP calls in 7.0 s wall (unittest), raw per-call figures in `/tmp/clinic-smoke.json` on jude-1 (dated, not in Git). → **UNPLACED** - `F-513743a8bd` L130 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 5. Measured round trips (OBSERVATION, `LiveSmoke`, jude-1, 1 Nov 2026 05:1xZ, tenant `smoke-d14def5f`, loopback)_: `POST /fhir/R4/` / 24 / 44.5 ms / 153.0 ms (Observation) → **UNPLACED** - `F-a3dd1fdf01` L131 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 5. Measured round trips (OBSERVATION, `LiveSmoke`, jude-1, 1 Nov 2026 05:1xZ, tenant `smoke-d14def5f`, loopback)_: `PUT /fhir/R4//` / 21 / 38.9 ms / 412.2 ms (one outlier) → **UNPLACED** - `F-0fa7498658` L132 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 5. Measured round trips (OBSERVATION, `LiveSmoke`, jude-1, 1 Nov 2026 05:1xZ, tenant `smoke-d14def5f`, loopback)_: `DELETE /fhir/R4//` / 23 / 33.4 ms / 716.4 ms (one outlier) → **UNPLACED** - `F-e8d7a04b04` L134 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 5. Measured round trips (OBSERVATION, `LiveSmoke`, jude-1, 1 Nov 2026 05:1xZ, tenant `smoke-d14def5f`, loopback)_: A create as the receiver sees it (POST + read-back GET) is therefore about 55–60 ms on the host; a change (find GET + PUT + → **UNPLACED** - `F-f432f95f68` L135 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 5. Measured round trips (OBSERVATION, `LiveSmoke`, jude-1, 1 Nov 2026 05:1xZ, tenant `smoke-d14def5f`, loopback)_: read-back GET) about 60–65 ms; a first touch adds the subject Patient (search + POST) once per person. The earlier run at 05:0xZ → **UNPLACED** - `F-e8452d13ea` L136 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 5. Measured round trips (OBSERVATION, `LiveSmoke`, jude-1, 1 Nov 2026 05:1xZ, tenant `smoke-d14def5f`, loopback)_: (199 calls, 0 failures) gave the same order: GET 7–16 ms, POST 29–116 ms, PUT 26–75 ms per type. Not measured: throughput under → **UNPLACED** - `F-4338b1e9d5` L137 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 5. Measured round trips (OBSERVATION, `LiveSmoke`, jude-1, 1 Nov 2026 05:1xZ, tenant `smoke-d14def5f`, loopback)_: load, the public path (there is none: loopback only), or cold starts. → **UNPLACED** - `F-9bda144050` L141 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 6. Demo tenant and the live journey walk (OBSERVATION)_: The demo tenant never touches Medplum: `_demo(engine, tenant)` routes to the estate record on the demo engine and the outputs → **UNPLACED** - `F-37414e24cc` L142 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 6. Demo tenant and the live journey walk (OBSERVATION)_: say `SIMULATED: demo tenant; the clinic record is simulated and Medplum is not touched ()` with → **UNPLACED** - `F-9b9fcf5331` L144 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 6. Demo tenant and the live journey walk (OBSERVATION)_: `tools/journeys_http.py --demo --only /tmp/clinic-medplum-journeys.txt --shard 1/1` on the live release 7a2a0ea, 05:1xZ: the → **UNPLACED** - `F-cf09c8ae05` L145 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 6. Demo tenant and the live journey walk (OBSERVATION)_: 123 `#human` journeys of the 123 mapped units: `{"journeys": 123, "pass": 123, "plug": 0, "held": 0, "fail": 0, "seconds": 44.0}` → **UNPLACED** - `F-b45becbe25` L146 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 6. Demo tenant and the live journey walk (OBSERVATION)_: (`ledger/evidence/journeys-demo-rerun-1.jsonl`). Note: the running service does not yet import this domain (section 7), so → **UNPLACED** - `F-ebc767000d` L147 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 6. Demo tenant and the live journey walk (OBSERVATION)_: that walk proves the demo path through the estate record, which is also the path the domain takes for the demo tenant. → **UNPLACED** - `F-6d0d592151` L153 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 7. Registration (what the running site needs; one line in a file this lane does not own)_: from runtime.domains import clinic # noqa: F401 - binds the Clinic record to Medplum (FHIR R4) → **UNPLACED** - `F-a216076e44` L155 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 7. Registration (what the running site needs; one line in a file this lane does not own)_: Shipped once through the deploy helper at 05:15:03Z (record `/opt/x-x/deploy/ledger/deploy/2026-11-01T051503Z-edit.json`, health live). → **UNPLACED** - `F-e7db9a1d96` L156 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 7. Registration (what the running site needs; one line in a file this lane does not own)_: Until that line is in the release the live tenants' Clinic operations keep the generic estate record; the demo tenant is unchanged → **UNPLACED** - `F-51a1f16b10` L157 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 7. Registration (what the running site needs; one line in a file this lane does not own)_: either way. `python3 -m unittest tests.test_clinic_domain` (14 tests, fake server) passes on the Mac and on jude-1; → **UNPLACED** - `F-a54d8786c1` L162 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 8. Open items (CLAIMs for the owners named)_: Engine lane: rotate the seeded super-admin password; make the database container's `/dev/shm` mode survive a restart; name → **UNPLACED** - `F-241fe2d9bd` L164 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 8. Open items (CLAIMs for the owners named)_: Coordinator: add the registration line (section 7) and ship. → **UNPLACED** - `F-5d20d3a6fa` L165 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 8. Open items (CLAIMs for the owners named)_: This domain, next: one Medplum Project per tenant; `Binary` upload for images and documents; the remaining single-digit → **UNPLACED** - `F-555ed9315b` L166 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 8. Open items (CLAIMs for the owners named)_: entities after a unit-by-unit reading; `Schedule` capacity for clinic sessions (cancel-a-clinic-session, plan-weekly-capacity). → **UNPLACED** ###### `research/clinic/MOBBIN-FLOWS.md` (74 findings) - `F-d8344926cc` L3 _EDGE Clinic research: real product flows from Mobbin_: **What this is.** Real screens and user flows from existing apps, recorded by Mobbin (a library of real app and website screenshots, grouped into step-by-step flows). It complements `REFERO-FLOWS.md` in this folder. It shows the floor that people already expect from a health front door, and good practice worth borrowing from other industries. → **UNPLACED** - `F-e6a4a5ce6f` L5 [OBSERVATION] _EDGE Clinic research: real product flows from Mobbin_: **Statement class and date.** Every item below is an OBSERVATION of a real product, seen in Mobbin on 2026-10-31. The descriptions are ours, written from the preview images Mobbin returned (an evenly spaced sample of each flow's screens, not always every screen). Product names belong to their owners. Image links are not kept because they expire; each item carries its `mobbin_url`. → **UNPLACED** - `F-1f1efb301c` L7 _EDGE Clinic research: real product flows from Mobbin_: **How it was gathered.** Searches by app name plus task, on iOS and web, about five flows per search: appointment booking, symptom checking, test results, medication reminders, mental health check-ins, health questionnaires, wearables and readiness scores, sharing health data, telehealth, food logging, emergency and crisis help, and local community activity (as a stand-in for social prescribing). → **UNPLACED** - `F-b8a3498fcb` L11 _EDGE Clinic research: real product flows from Mobbin › Coverage: what Mobbin held and what it did not_: **Held and returned:** Ada (symptom assessment, condition detail, past assessments), Headspace (anxiety and stress check-ins, therapy booking on web), Apple Health (sharing, data sources and access, profile), Oura (readiness and a medical AI chat), WHOOP (labs), MyFitnessPal (meal logging), Zocdoc (doctor booking), GoodRx, Hers and Hims (medication reminders, treatment questionnaires), Ro, Superpower and Bevel (lab results and intake), Alan (consultations, with Livi as the video provider inside it), and single screens from Wysa, Bloom, Garmin Connect and Amazon's health assistant. → **UNPLACED** - `F-08b13d2376` L12 _EDGE Clinic research: real product flows from Mobbin › Coverage: what Mobbin held and what it did not_: **Searched for but not returned:** NHS App (the booking search returned Zocdoc, IKEA, Crate & Barrel and Fresha instead) and Doctolib (the web search returned Selfridges, Square and Headspace instead). No NHS, GP practice or other UK public health service appeared in any result. → **UNPLACED** - `F-e627fb3aea` L13 _EDGE Clinic research: real product flows from Mobbin › Coverage: what Mobbin held and what it did not_: **Not searched by name in this pass:** Babylon, Healthily, Calm. Livi appeared only as the video provider inside Alan. → **UNPLACED** - `F-abdfeb1203` L14 _EDGE Clinic research: real product flows from Mobbin › Coverage: what Mobbin held and what it did not_: **Not found at all:** any social prescribing service, any link worker or referral to community groups, any flow where a GP or clinician receives what the person filled in, and any flow where an AI agent acts for a person or organisation with its own identity. → **UNPLACED** - `F-90462267dc` L20 [OBSERVATION] _By task › Booking an appointment_: **Zocdoc, iOS, "Booking an appointment"** (10 screens). A doctor's page with a "new patient" tick box, a choice of location and tomorrow's slots as buttons. The review step lets you book for yourself or "someone else", shows a red inline error when the address is missing, and asks for insurance, visit reason, a phone number and optional notes for the practice, with a tick box certifying the payment method. The confirmation offers "Add to calendar" and "Prepare for appointment". OBSERVATION, 2026-10-31. https://mobbin.com/flows/fcd6461e-af19-4ca9-9864-8b21c286888c → `/clinic/add-appointment-calendar` - `F-5a1b914060` L21 [OBSERVATION] _By task › Booking an appointment_: **Zocdoc, iOS, "Making a doctor appointment"** (12 screens). In-person or video visit toggle, slots grouped by day, then a gate: "This provider only sees in-network patients", with the choice of photographing the insurance card or choosing it by hand, and accepting the terms. Afterwards the home screen shows the upcoming appointment with "Almost done! Upload a photo of your ID". OBSERVATION, 2026-10-31. https://mobbin.com/flows/26c741b6-0231-4040-b5e9-890f0b6db100 → `/clinic/mental-health-care-crisis` - `F-5bf4329f12` L22 [OBSERVATION] _By task › Booking an appointment_: **IKEA, iOS, "Booking an appointment"** (15 screens, retail). A plain explanation of the session and what to prepare beforehand, a calendar with green dots on days that have space, time slots labelled with the time zone, a contact form that says why the phone number is needed (reminders), and a confirmation with a booking reference and all details. OBSERVATION, 2026-10-31. https://mobbin.com/flows/64df5892-fc18-4edb-ba02-8b365b63bebe → `/clinic/explain-result-plain-language` - `F-4ec91ee969` L23 [OBSERVATION] _By task › Booking an appointment_: **Crate & Barrel, iOS, "Booking an appointment"** (14 screens, retail). A three-step progress bar, separate tick boxes for marketing texts and for appointment reminders, a calendar with an available and not available key, and a confirmation that explains "What's next", "What to expect" and "What you'll get". OBSERVATION, 2026-10-31. https://mobbin.com/flows/8e862fc0-87f2-44f0-a201-787bb47e5d6e → `/clinic/add-appointment-calendar` - `F-781bc2dec6` L24 [OBSERVATION] _By task › Booking an appointment_: **Fresha, iOS, "Booking an appointment"** (22 screens, beauty). Date strip and time list, a review with price and fees, the cancellation and no-show charges stated before payment, then a "Confirmed" page with add to calendar, directions, message the venue and venue details. OBSERVATION, 2026-10-31. https://mobbin.com/flows/7121fd1b-35ef-48da-87d5-aacbd1ba81a8 → `/clinic/add-appointment-calendar` - `F-ad8ba4362c` L25 [OBSERVATION] _By task › Booking an appointment_: **Headspace, web, "Scheduling an appointment"** (23 screens, therapy). Steps shown at the top: check in, verify insurance, schedule. "Here's your estimated cost" confirms the insurer and lists things to know (pay nothing today, the most it could cost). Time ranges show how many clinicians are free and the time zone. The final page lists next steps, including "complete your pre-appointment assessment to help your therapist prepare". OBSERVATION, 2026-10-31. https://mobbin.com/flows/b46dac23-c0f1-4b3f-b506-ef8fea7c14d7 → `/clinic/front-door-check` - `F-fd100eade8` L26 [OBSERVATION] _By task › Booking an appointment_: **Selfridges, web, "Making an appointment"** (17 screens, retail). A four-step stepper with the appointment summary always visible on the right, a week grid by morning, afternoon and evening, a phone field that says it will only be used for a reminder 48 hours before, two separate consent tick boxes, and a confirmation with a reference, print and add to calendar. OBSERVATION, 2026-10-31. https://mobbin.com/flows/e5cf516c-6b42-4bb9-b0c2-291623dc0124 → `/clinic/add-appointment-calendar` - `F-5139c69464` L27 [OBSERVATION] _By task › Booking an appointment_: **Square, web, "Booking an appointment"** (12 screens). A running summary while you add services, a checkout that holds the slot with a countdown, a cancellation timeline ("cancel before" a date), and a thank-you page with add to calendar, reschedule, cancel and book again, plus "Next steps: we'll call you". OBSERVATION, 2026-10-31. https://mobbin.com/flows/6992da62-212e-4947-991d-588142829394 → `/clinic/add-appointment-calendar` - `F-928e565e20` L31 [OBSERVATION] _By task › Symptom checking_: **Ada, iOS, "Starting a symptom assessment"** (17 screens). One question per screen (name, then date of birth as day, month and year boxes), with "Previous" and a close button, and an "About Ada and Privacy" link on every screen. Symptom search with suggestions and a help icon per symptom. It ends: "No more questions... The report is not a medical diagnosis. If in doubt, seek advice from a healthcare professional." The report lists possible causes with the most urgent first, in a red box: "Needs immediate attention: seek emergency care now". Each cause shows which of your symptoms matched and "48% of people with your symptoms and health background had this condition". OBSERVATION, 2026-10-31. https://mobbin.com/flows/dc2c5dc4-6695-44fd-b21a-3cd4e7601c9a → `/clinic/symptom-check-triage` - `F-e37a730ab5` L32 [OBSERVATION] _By task › Symptom checking_: **Ada, iOS, "Searching symptoms"** (3 screens). Typing "headache" gives suggestions and "Symptoms found: 3", each with a short plain description and a help icon. OBSERVATION, 2026-10-31. https://mobbin.com/flows/4d5fb715-f25b-47a7-b4cc-1fd4bcfe1505 → `/clinic/front-door-check` - `F-5c45744435` L33 [OBSERVATION] _By task › Symptom checking_: **Ada, iOS, "Condition detail"** (7 screens). A condition page with sections that open and close (symptoms, diagnosis, treatment, risk factors, outcomes), a statement that the list "is not a diagnosis and not exhaustive", and "Download report". OBSERVATION, 2026-10-31. https://mobbin.com/flows/ce7b8167-97a3-47da-8a93-00c89d64c098 → `/clinic/complete-risk-formulation` - `F-cd1952f7ae` L34 [OBSERVATION] _By task › Symptom checking_: **Ada, iOS, "Health assessments"** (8 screens). A profile with basic information, health background and the list of past assessments. OBSERVATION, 2026-10-31. https://mobbin.com/flows/0c0ecbbf-82cb-4dee-af6b-353de04791ce → **UNPLACED** - `F-5149910dba` L35 [OBSERVATION] _By task › Symptom checking_: **Ada, iOS, "Home"** (3 screens). The home screen with one main action, "Start symptom assessment", and a news card about changes to how you access your data. OBSERVATION, 2026-10-31. https://mobbin.com/flows/35a15c59-17c0-4872-a99d-610cc2cec951 → `/clinic/symptom-check-triage` - `F-a26a0934c6` L39 [OBSERVATION] _By task › Questionnaires and intake_: **Ro, iOS, "Onboarding"** (20 screens). Choose a goal, then a progress bar over height and weight, a "select all that apply" list of heart conditions, and treatment preferences that include "Not sure yet" and "Another option not listed". It ends "You've qualified" and "What's next: a provider will review your health information". OBSERVATION, 2026-10-31. https://mobbin.com/flows/c40ee07f-972a-43b2-83b4-666d7f6df86e → `/clinic/apply-treatment-rule` - `F-e2003b56db` L40 [OBSERVATION] _By task › Questionnaires and intake_: **Hers, iOS, "Complete treatment questionnaires (mental health)"** (29 screens). A consultation with a segmented progress bar. Answers carry a plain description (for example "Fair: I have friends, but I don't feel close to or supported by them"). A loading screen ticks off "gathering your responses, calculating your scores, preparing your results". A free-text box asks "anything else you want your provider to know?" with a tick box for "No, I've provided all relevant information". OBSERVATION, 2026-10-31. https://mobbin.com/flows/60448996-a8dd-4632-9b46-4ab745af1a4e → `/clinic/triage-mental-health-referral` - `F-0b584caa0a` L41 [OBSERVATION] _By task › Questionnaires and intake_: **Hers, iOS, "Completing treatment questionnaires (weight loss)"** (18 screens). Includes a screening question about eating disorders with the reason it is asked, marketing charts mixed into the questions, and a clear refusal screen: "you are not eligible... we recommend that you see a provider in person", with a way back. OBSERVATION, 2026-10-31. https://mobbin.com/flows/061e04f8-11e2-49c9-99ba-409a707e5f25 → `/clinic/mental-health-care-crisis` - `F-d30938d685` L42 [OBSERVATION] _By task › Questionnaires and intake_: **Superpower, iOS, "Taking lifestyle quiz"** (11 screens). "Let's understand your day-to-day: this helps us customise your plan", then questions on exercise and work, each with a one-line reason ("your job plays a big role in daily stress"). It ends by selling more tests. OBSERVATION, 2026-10-31. https://mobbin.com/flows/48c850ab-006f-4536-a9bc-1d4033c46c9c → `/clinic/set-surveillance-plan` - `F-faf19bdbcd` L46 [OBSERVATION] _By task › Mental health check-ins_: **Headspace, iOS, "Tracking anxiety levels"** (6 screens). A progress area with stress and anxiety tabs, an empty chart that says "Start tracking your anxiety over time with bi-weekly check-ins", and a seven-question check-in ("Over the last two weeks, how often have you been bothered by..." with four answers). After a check-in the button greys out with "Opens in 13 days". OBSERVATION, 2026-10-31. https://mobbin.com/flows/79577cba-b6b5-487c-ae41-5416af934661 → `/clinic/wellbeing-questionnaires` - `F-25a686afcc` L47 [OBSERVATION] _By task › Mental health check-ins_: **Headspace, iOS, "Tracking monthly stress reflection"** (6 screens). Ten questions, then a dated result: a score on a bar split into low, moderate and high with the number ranges shown, framed gently ("Think of this score as a way to get to know yourself a little better"), with suggested content and a link to questions and answers. OBSERVATION, 2026-10-31. https://mobbin.com/flows/ce032b9a-cb9f-42f9-b407-19a2796c6961 → `/clinic/front-door-check` - `F-7d422ecda3` L48 [OBSERVATION] _By task › Mental health check-ins_: **Headspace, iOS, "Profile"** (7 screens). Personal statistics, a streak described kindly ("Meditate on a schedule that feels good"), the check-in panel and activity history. OBSERVATION, 2026-10-31. https://mobbin.com/flows/8d8bd1bc-e2c4-4a04-b7f6-5ba7233321b6 → `/clinic/record-clinic-consultation` - `F-e130cdbc66` L52 [OBSERVATION] _By task › Test results_: **Bevel, iOS, "Record detail"** (6 screens). Health records filtered by type (lab report, medications, imaging), a nudge that some markers are due for a refresh, and uploaded documents "ready for review". A lab report opens with a plain summary of what is high, then each marker with its range and value. One marker carries an orange warning: "This biomarker may be incorrect. Please review the value and update it if needed", with Edit. OBSERVATION, 2026-10-31. https://mobbin.com/flows/bcb74637-8d36-4b58-8e62-99e88312cb98 → `/clinic/person-held-health-record` - `F-e9373335bb` L53 [OBSERVATION] _By task › Test results_: **Superpower, iOS, "Biomarker detail"** (3 screens). A list of results with small range bars, then one marker marked "Out of range", a chart over time with the healthy band shaded, latest result next to the target range, what it is, why it matters, what you can do, and a list of sources. OBSERVATION, 2026-10-31. https://mobbin.com/flows/0a31c382-c9dc-4f06-b681-f8bceee8e991 → `/clinic/wellbeing-questionnaires` - `F-b79ccd4e6d` L54 [OBSERVATION] _By task › Test results_: **WHOOP, iOS, "Labs summary"** (5 screens). A test offer, a summary ring with "optimal, sufficient, out of range", a promise of "clinician-reviewed insights", and an empty state for a marker: "No lab results here yet: upload past lab results or schedule a test". OBSERVATION, 2026-10-31. https://mobbin.com/flows/b7fceedb-b811-45ff-94f6-2760d2c4f382 → `/clinic/person-held-health-record` - `F-79f55b97a4` L55 [OBSERVATION] _By task › Test results_: **Ro, iOS, "Lab history"** (2 screens). Settings that include lab history, Apple Health and privacy settings; the empty lab history is used to sell tests. OBSERVATION, 2026-10-31. https://mobbin.com/flows/aebafd34-bd72-4d0c-bed0-b96557c23662 → `/clinic/record-clinic-consultation` - `F-2bd36ae620` L59 [OBSERVATION] _By task › Medication reminders_: **GoodRx, iOS, "Scheduling a daily med reminder"** (10 screens). "Add reminder details": a name, a time, days of the week, and switches for which medicines this reminder covers, with "you'll get a push notification on the days and time you select". A separate card offers refill reminders. OBSERVATION, 2026-10-31. https://mobbin.com/flows/dc7306f5-fc70-4af5-a256-95484befdc1b → `/clinic/send-appointment-letters` - `F-b340cc03a1` L60 [OBSERVATION] _By task › Medication reminders_: **GoodRx, iOS, "Daily med reminders"** (2 screens). The list of reminders with Edit and "Schedule new reminder". OBSERVATION, 2026-10-31. https://mobbin.com/flows/efcc6f42-2a09-45b6-951d-2810aae5c0eb → `/clinic/send-appointment-letters` - `F-a6ed04dbaf` L61 [OBSERVATION] _By task › Medication reminders_: **Hers, iOS, "Setting up a reminder"** (7 screens). The treatment page shows reminders on or off, refills left, a message from the provider team and how to take the medicine. The reminder sheet allows several times a day, and its small print says the reminder information "will not be shared with your Care Team. Contact your provider if you have questions." OBSERVATION, 2026-10-31. https://mobbin.com/flows/1396d89d-d0d5-4822-9df1-1211c80cec04 → `/clinic/message-care-team` - `F-166692d716` L62 [OBSERVATION] _By task › Medication reminders_: **Hims, iOS, "Setting a reminder"** (7 screens). Next refill date, a plain reason why a refill is not yet allowed, a reminders tile and "Snooze for 1 month". OBSERVATION, 2026-10-31. https://mobbin.com/flows/56b3a643-67eb-4930-a326-46b5f76dea0c → `/clinic/send-appointment-letters` - `F-e9986a1947` L66 [OBSERVATION] _By task › Wearables, scores and data sharing_: **Oura, iOS, "Readiness"** (6 screens). A daily score with a short, human message ("above your average today, how do you feel?"), a seven-day line, contributors labelled "Pay attention", "Fair" or "Good", and overnight heart rate and heart rate variability charts. OBSERVATION, 2026-10-31. https://mobbin.com/flows/b071a02f-88d4-4ee2-9a4e-ea9e9422bb56 → `/clinic/message-care-team` - `F-afc043001c` L67 [OBSERVATION] _By task › Wearables, scores and data sharing_: **Oura, iOS, "Readiness"** (10 screens). "Your first Readiness Score!" explains what the score means and that it improves as a baseline is learned; key measures as tiles; and "Recovery in progress: you've turned off Rest Mode... continue to take it easy", with a marker showing Rest Mode was on. OBSERVATION, 2026-10-31. https://mobbin.com/flows/cd0b2e3e-e462-4d9d-ba87-f0c214fe47ca → `/clinic/complete-mental-health-check` - `F-9c0e4f3c70` L68 [OBSERVATION] _By task › Wearables, scores and data sharing_: **Oura, iOS, "Daily readiness data"** (2 screens). A newer home with readiness, sleep, resilience and activity in one row and the message "your activity goal today is your baseline". OBSERVATION, 2026-10-31. https://mobbin.com/flows/1a062e29-239c-4971-94cd-292b42dfcc7c → **UNPLACED** - `F-976884f7bd` L69 [OBSERVATION] _By task › Wearables, scores and data sharing_: **Apple Health, iOS, "Share with someone"** (4 screens). An explainer in three short points (you're in control; they get a dashboard and notifications; only a summary is shared, encrypted, and you can stop at any time), "Share with Someone" or "Ask Someone to Share", and an honest error: "Unable to share health data: this person does not have a device that can view your data". OBSERVATION, 2026-10-31. https://mobbin.com/flows/f5de1f34-8e2a-4567-840d-eac966c7ef88 → `/clinic/person-held-health-record` - `F-c9403509cc` L70 [OBSERVATION] _By task › Wearables, scores and data sharing_: **Apple Health, iOS, "Sharing"** (3 screens). "Share with your doctor", apps and research studies with access, and "No one is sharing data with you". OBSERVATION, 2026-10-31. https://mobbin.com/flows/00d9fca9-9ce6-4ba4-8a91-e3141ce6e29d → `/clinic/set-data-sharing-choices` - `F-525ebe94f8` L71 [OBSERVATION] _By task › Wearables, scores and data sharing_: **Apple Health, iOS, "Data sources & access"** (3 screens). For one kind of data (steps): which apps may read it, which research studies may read it (both "None"), and which sources write it, in priority order, with an explanation of how one source is chosen. OBSERVATION, 2026-10-31. https://mobbin.com/flows/e19317bd-987a-4ace-bd7d-42048e3786b1 → `/clinic/manage-reporting-backlog` - `F-2ff4e3f702` L72 [OBSERVATION] _By task › Wearables, scores and data sharing_: **Apple Health, iOS, "Profile"** (3 screens). Medical ID, health checklist, health records, and privacy lists for apps, research studies and devices, with "Export All Health Data" and the time it last synced. OBSERVATION, 2026-10-31. https://mobbin.com/flows/40639fc3-6978-4b39-99bd-cd51db256933 → `/clinic/see-day` - `F-65c91d0a72` L73 [OBSERVATION] _By task › Wearables, scores and data sharing_: **Apple Health, iOS, "Searching apple health"** (4 screens). Browse by category (heart, medicines, mental wellbeing, symptoms, vitals), add a health records account, search results. OBSERVATION, 2026-10-31. https://mobbin.com/flows/a50e2c67-b7f5-4b5b-94f5-6ed71c1d224e → `/clinic/symptom-check-triage` - `F-6553627ef8` L74 [OBSERVATION] _By task › Wearables, scores and data sharing_: **Garmin Connect, iOS, screen.** Abnormal heart rate alerts with high and low limits, each with a plain note of when the alert fires and who it is not suitable for. OBSERVATION, 2026-10-31. https://mobbin.com/screens/65376645-1229-4e72-9733-94a8837147e4 → `/clinic/online-urgent-triage` - `F-ec0ed9ff8b` L78 [OBSERVATION] _By task › Talking to a clinician, and AI health assistants_: **Alan, iOS, "Talk with medical team"** (3 screens). "Talk with our medical team 24/7 by chat or video". The chat tab offers an AI ("Ask Mo") to "skip the wait", and lists specialties with expected reply times ("reply in about 30 minutes"), with a footer "Is it an emergency? Call 15". The video tab shows general medicine delivered by Livi ("6am to midnight, included, includes prescriptions") and therapy, with "Help is available: if you're in distress or having suicidal thoughts, reach out now". OBSERVATION, 2026-10-31. https://mobbin.com/flows/a633ff3f-4a5f-4a82-9c9f-3420c9ab2c78 → `/clinic/ask-help-now` - `F-c436ee856c` L79 [OBSERVATION] _By task › Talking to a clinician, and AI health assistants_: **Alan, iOS, "Consult"** (4 screens). A daily wellbeing page with small actions (walk, breathe, meditate), programmes run with physiotherapists and therapists, and topic tiles (mind, sleep, nutrition and more). OBSERVATION, 2026-10-31. https://mobbin.com/flows/5a2fbc6d-a598-4651-bac9-a191a52a02b3 → **UNPLACED** - `F-0fb8909832` L80 [OBSERVATION] _By task › Talking to a clinician, and AI health assistants_: **Oura, iOS, "Ask a doctor"** (2 screens). A "Medical chat" where an AI advisor from a named partner says it will use your ring data and that you can bring in a doctor; example questions; "Add a doctor: will reply after 7pm"; footer "Messages from the AI are informational. For medical advice, add a physician. For emergencies, call 911." A paid offer says physicians see your ring data and records. OBSERVATION, 2026-10-31. https://mobbin.com/flows/ad69f6c5-15bc-413f-bd86-8b459b972fc1 → `/clinic/ask-advice-guidance` - `F-82e000c66d` L81 [OBSERVATION] _By task › Talking to a clinician, and AI health assistants_: **Amazon Shopping, iOS, screen.** A health AI in beta that names itself, says "If this is an emergency, dial 911", offers "Talk to a provider", and shows a "Thinking" state with a stop button. OBSERVATION, 2026-10-31. https://mobbin.com/screens/ca1c14f1-cf34-4b27-b041-d8e30ebb332d → `/clinic/person-held-health-record` - `F-2960beab44` L85 [OBSERVATION] _By task › Crisis and urgent help_: **Wysa, iOS, screen.** A chatbot confirms "your safety plan has been created" and reminds the person to press SOS whenever they cannot cope; buttons to check the plan and add a therapist. OBSERVATION, 2026-10-31. https://mobbin.com/screens/cd4930f5-31e0-4c34-86a3-a6679e342cf7 → `/clinic/set-surveillance-plan` - `F-21ca71f632` L86 [OBSERVATION] _By task › Crisis and urgent help_: **Bloom, iOS, screen.** A "Get help" sheet with a crisis line button and a local hotline button. OBSERVATION, 2026-10-31. https://mobbin.com/screens/2cf0f6a8-c1fb-4919-8ab3-e53826be7387 → `/clinic/mental-health-care-crisis` - `F-4b7e180d30` L87 [OBSERVATION] _By task › Crisis and urgent help_: **inDrive, iOS, screen (transport).** An SOS page with the person's location on a map, the details the police will need, and one large call button. OBSERVATION, 2026-10-31. https://mobbin.com/screens/3f075a2f-0493-4d53-a905-00549b8b2fa8 → **UNPLACED** - `F-81ed66fa93` L91 [OBSERVATION] _By task › Food and lifestyle logging_: **MyFitnessPal, iOS, "Scanning a meal"** (4 screens). Today's calories and macros, quick ways to log (search, barcode, voice, photo of the meal), a scanned list of foods each with delete and "Add more", then "Log breakfast" and a "Breakfast logged!" message. OBSERVATION, 2026-10-31. https://mobbin.com/flows/e5e452d7-f91e-4939-9fa0-59ea87b7d47a → `/clinic/maintain-hazard-log` - `F-812890c018` L92 [OBSERVATION] _By task › Food and lifestyle logging_: **MyFitnessPal, iOS, "Log meal"** (7 screens). Search, saved meals and foods, "copy previous meal", and a hint to swipe to log yesterday's meal again. OBSERVATION, 2026-10-31. https://mobbin.com/flows/d58bea22-7f26-48ef-a6de-134d9183c5ad → `/clinic/search-recruiting-trials` - `F-47d2006368` L96 [OBSERVATION] _By task › Local community activity (nearest thing to social prescribing)_: **Meetup, iOS, screen.** Groups near a place with distance chips (2, 5, 10 miles), categories such as social activities and hobbies, and member counts. OBSERVATION, 2026-10-31. https://mobbin.com/screens/303b5f13-3a13-4b20-809c-f356d6993248 → `/clinic/find-recruiting-trial-near` - `F-a96603b426` L97 [OBSERVATION] _By task › Local community activity (nearest thing to social prescribing)_: **Nextdoor, iOS, screen.** Neighbourhood groups (for example help with housing) with member counts, and local posts and polls. OBSERVATION, 2026-10-31. https://mobbin.com/screens/9fa1c528-5530-4f11-bc24-0244ea7279fe → **UNPLACED** - `F-4ffd93b8a1` L98 [OBSERVATION] _By task › Local community activity (nearest thing to social prescribing)_: **Lex, iOS, screen.** A local feed filtered by events, new here, groups and friends; posts tagged "For free"; a group card with "Join now". OBSERVATION, 2026-10-31. https://mobbin.com/screens/36bffdb3-8d25-413f-9698-596d80c9c31a → `/clinic/report-side-effect` - `F-0bac76fa2b` L99 [OBSERVATION] _By task › Local community activity (nearest thing to social prescribing)_: **adidas Running, iOS, screen.** Challenges with participant counts and end dates, upcoming events ("starts in 28 days") and communities. OBSERVATION, 2026-10-31. https://mobbin.com/screens/e11d5383-ce5b-4038-b90a-39a9bf820ad7 → **UNPLACED** - `F-d60062aece` L103 _EDGE Clinic research: real product flows from Mobbin › What to match_: **Urgency first, in plain words.** Ada puts "seek emergency care now" at the top of the report in red; Alan, Oura and Amazon keep an emergency line visible in their health chats. EDGE Clinic must do at least this on every screen where a person describes a symptom. → **UNPLACED** - `F-271023f9b7` L104 _EDGE Clinic research: real product flows from Mobbin › What to match_: **"This is not a diagnosis", said once, clearly, at the right moment** (Ada at the end of questions and on every condition page), not buried in terms. → **UNPLACED** - `F-4cf0316ce5` L105 _EDGE Clinic research: real product flows from Mobbin › What to match_: **One question per screen, with a way back and the privacy link always visible** (Ada), and a reason given for sensitive questions (Hers, Superpower). → **UNPLACED** - `F-2579a98095` L106 _EDGE Clinic research: real product flows from Mobbin › What to match_: **Show what matched and how sure.** Ada lists which of your symptoms matched and a plain percentage of similar people. Superpower lists the sources behind its advice. → **UNPLACED** - `F-456522463f` L107 _EDGE Clinic research: real product flows from Mobbin › What to match_: **Let people correct the record.** Bevel flags a value that may be wrong and lets the person edit it. Clinic results and imported data need the same. → **UNPLACED** - `F-bfe0a38a38` L108 _EDGE Clinic research: real product flows from Mobbin › What to match_: **Spacing check-ins on purpose.** Headspace locks the check-in with "Opens in 13 days" and frames the score kindly with its ranges visible. → **UNPLACED** - `F-3db9f266d9` L109 _EDGE Clinic research: real product flows from Mobbin › What to match_: **Say who sees what.** Apple Health shares only a summary, can be stopped at any time, and lists every app and study that can read each kind of data. Hers says the reminder data is not shared with the care team. → **UNPLACED** - `F-ed63fab752` L110 _EDGE Clinic research: real product flows from Mobbin › What to match_: **Bookings that explain themselves.** State what to prepare (IKEA), why each piece of contact data is needed (IKEA, Selfridges), separate consent for reminders and for marketing (Crate & Barrel, Selfridges), the cost and cancellation rules before payment (Fresha, Headspace, Square), and the next steps after booking (Headspace, Crate & Barrel, Square). → **UNPLACED** - `F-0f83df7f48` L111 _EDGE Clinic research: real product flows from Mobbin › What to match_: **An honest "not for you" route.** Hers tells ineligible people to see a provider in person instead of leaving them stuck. → **UNPLACED** - `F-e6a8c8780c` L112 _EDGE Clinic research: real product flows from Mobbin › What to match_: **Booking for someone else** is a first-class choice (Zocdoc "Someone else"). → **UNPLACED** - `F-4199f0f7f3` L116 _EDGE Clinic research: real product flows from Mobbin › What none of them do_: **No AI agent is ever a user in its own right.** Every AI here (Ada, Alan's "Mo", Oura's advisor, Amazon's assistant, Wysa) is a feature inside one company's app, acting for the signed-in person. None has its own identity, none shows whose behalf it acts on, what it is allowed to do, or a record of what it did. No flow lets a family member's agent, a carer's agent or a link worker's agent book, ask or fill in a form with the person's permission. → **UNPLACED** - `F-b8d51cf1fc` L117 _EDGE Clinic research: real product flows from Mobbin › What none of them do_: **No hand-off to the GP or NHS.** Nothing sends a symptom report, check-in score or wearable trend to a named clinician with the person's consent, and nothing shows the person what the clinician received or did with it. Ada offers "Download report" only. → **UNPLACED** - `F-d7390d50d0` L118 _EDGE Clinic research: real product flows from Mobbin › What none of them do_: **No social prescribing.** No app suggests a local group, walk, class or service as a response to how someone feels, and none involves a link worker. Community apps (Meetup, Nextdoor) know nothing about health. → **UNPLACED** - `F-552835ce0e` L119 _EDGE Clinic research: real product flows from Mobbin › What none of them do_: **No organisation as a user.** No practice, charity or community group appears as an account that receives referrals or reports back. → **UNPLACED** - `F-06d5458f17` L120 _EDGE Clinic research: real product flows from Mobbin › What none of them do_: **Early detection stays inside one product.** Wearable apps show trends and alerts (Oura, Garmin) but none explains when a trend should lead to seeing a clinician, or offers to share that trend with one. → **UNPLACED** - `F-c623e6ec6d` L121 _EDGE Clinic research: real product flows from Mobbin › What none of them do_: **Confidence is rarely shown.** Apart from Ada's percentage and Bevel's "may be incorrect", nothing says how sure the system is or what it is based on. → **UNPLACED** - `F-0c0b14a3b9` L122 _EDGE Clinic research: real product flows from Mobbin › What none of them do_: **No UK public sector examples.** NHS App and Doctolib were not in the results, so UK identity checks (NHS login), GP booking and repeat prescriptions are not covered here. → **UNPLACED** ###### `research/clinic/REFERENCE-CLINIC.md` (7 findings) - `F-552743450c` L3 [OBSERVATION] _EDGE Clinic: the reference flows the application entry copies_: **Statement class.** OBSERVATION of real products in Mobbin, recorded 31 October 2026 in `MOBBIN-FLOWS.md` in this folder; this file names which of them the Clinic's entry screen is built from (Owner, 1 November 2026: layouts come from the references, the only difference being the estate's added controls or journeys). → **UNPLACED** - `F-2559e85161` L7 _EDGE Clinic: the reference flows the application entry copies › The entry (`/clinic`)_: **Ada, iOS, "Home"** (Mobbin flow 35a15c59-17c0-4872-a99d-610cc…, 3 screens): one main action on the home, "Start symptom assessment", above a short card. The Clinic's home keeps that shape: one primary action (Check your symptoms, `clinic/symptom-check-and-triage`), then the designed tasks as tiles. → **UNPLACED** - `F-65b6fde810` L8 _EDGE Clinic: the reference flows the application entry copies › The entry (`/clinic`)_: **Ada, iOS, "Starting a symptom assessment"** (17 screens): one question per screen with Previous and a close control, a privacy link on every screen. The Clinic's journey screens keep one question per screen (tools/typed_steps.py) and the help control. → **UNPLACED** - `F-0bfd6896b8` L9 _EDGE Clinic: the reference flows the application entry copies › The entry (`/clinic`)_: **Zocdoc, iOS, "Booking an appointment"** (10 screens): a provider page, a choice of location, tomorrow's slots as buttons, a review step that lets you book for someone else. `clinic/book-appointment` follows this order. → **UNPLACED** - `F-617b1c00d5` L13 _EDGE Clinic: the reference flows the application entry copies › What the estate adds, and only that_: The designed tasks as tiles under the primary action: front door check-in, medication safety check, mental health care and crisis, online urgent triage, person-held health record, retinal screening, social prescribing referral, symptom check and triage, wearable signals, wellbeing questionnaires (the Clinic's bespoke units in `design/figma/FRAMES.json`). → **UNPLACED** - `F-d19600606f` L14 _EDGE Clinic: the reference flows the application entry copies › What the estate adds, and only that_: Every other Clinic capability as the same tile beneath, so nothing is hidden and nothing is listed as prose. → **UNPLACED** - `F-2c2dc7b52e` L15 _EDGE Clinic: the reference flows the application entry copies › What the estate adds, and only that_: Jude's help behind the help control; evidence behind the evidence control (glass box). → **UNPLACED** ###### `research/clinic/REFERO-FLOWS.md` (950 findings) - `F-e2ab4c6bac` L3 _EDGE Clinic research: real product flows from Refero_: **What this is.** Real screens and user flows from existing products, recorded by Refero (a library of real app and website screenshots, each journey described step by step). They show the floor (what users already expect) and good customer experience from any industry worth borrowing. → **UNPLACED** - `F-a73e899327` L5 [OBSERVATION] _EDGE Clinic research: real product flows from Refero_: **Statement class.** Every flow is an OBSERVATION by Refero of a real product, searched 2026-10-31. Descriptions are Refero's wording. Product names belong to their owners. → **UNPLACED** - `F-a3e7811093` L7 _EDGE Clinic research: real product flows from Refero_: **How it was gathered.** Semantic searches for this application's tasks on iOS and web. Results from other industries are kept on purpose: they are often the best customer experience to learn from. Refero does not hold many incumbents (and few government, enterprise or NHS systems), so this is a sample, not a census. → **UNPLACED** - `F-bec922778e` L9 _EDGE Clinic research: real product flows from Refero_: **Products in this sample:** Ada (20), Fresha (15), District (14), WhatsApp (14), Doctronic AI (13), TidyCal (9), ALO (7), Cohere (6), Care (6), Lovable (5), Willow (5), Poppy (5), Stardust (5), Omio (5), Breathwrk (5), Airtable (5), Maze (5), Napper (5), Structured (5), Meta AI (4), Airbnb (4), Preply (4), Podimo (4), Train Fitness (4), ElevenMusic (4), GoFundMe (3), Wabi (3), Foodllama (3), Parallel (3), Raycast (3), Longevity Deck (3), GlossGenius (3), Delphi (3), fal (3), Tinder (3), Sunsama (3), Calendly (3), Craft (3), Tripsy (3), Peerlist (2), Genie (2), TIDAL (2), Pool (2), Google Gemini (2), Fingerprint (2), The Outsiders (2), Trip (2), Anam (2), Understory (2), komoot (2), Haptic (2), Seed (2), Nextdoor (2), Acuity Scheduling (2), Cal (2), IKEA (2), Cushion (2), Leonardo.Ai (2), Adobe Photoshop (2), Orbit (2), Wispr Flow (2), Vocabulary (2), OpenAI (2), n8n (2), Intercom (2), Cofounder (2), Kit (2), Blackbird (2), Loowner (2), Ground News (2), Kin (2), The Leap (2), ElevenReader (2), On (2), Instacart (2), Netflix (2), Unwind (2), Clay (2), MonoDesk (2), New Balance (2), Duolingo (2), Typeform (2), Kinhive (2), Givingli (2), Melius (2), Chance AI (2), Dropbox (2), Rork (2), Mozi (2), time2book (2), Introspect (1), Google Translate (1), Cashcaded (1), Queue (1), Subo (1), Asos (1), Apple Image Playground (1), Spyglass (1), PayPal (1), Gumroad (1), Pi (1), Binge (1), Jasper AI (1), Allbirds (1), Deezer (1), Acne Studios (1), Saint Laurent (1), West Elm (1), ElevenLabs (1), Weekrise (1), Voicenotes (1), Synthesia.io (1), Telegram (1), Comet (1), FocusFlight (1), Granola (1), Vinyls (1), Substack (1), RAD Weather (1), Remy (1), 222 (1), Pinterest (1), Shazam (1), Fuse (1), Riverside (1), Resend (1), Rox (1), Butter (1), BYBORRE® (1), Linear (1), Instagram (1), UNHINGED (1), Twitch (1), ChatGPT (1), Microsoft Teams (1), pillowtalk (1), Sora (1), GO Club (1), Teal (1), DailyArt Magazine (1), LinkedIn (1), Slite (1), Polaroid (1), Volv (1), Mercury (1), Farfetch (1), Manus (1), Prose (1), Trip Way (1), Joi (1), X (1), B150 (1), Zendesk (1), TikTok (1), Sky Guide (1), MyMind (1), Runway (1), Apple Maps (1), Athleats (1), Verse (1), Exoplan (1), Miles (1), HYPE (1), Marathon (1), Make (1), Homerun (1), TwelveLabs (1), Сontra (1), [untitled] (1), Weather mini (1), Navia (1), Tripadvisor (1), Uber (1), Brink (1), Uber Eats (1), Cursor (1) → **UNPLACED** - `F-e38d5624ab` L13 _EDGE Clinic research: real product flows from Refero › Task: AI agent setup permissions_: **Leonardo.Ai: Profile Settings Management** (web, 2 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/12914) → **UNPLACED** - `F-fd43e991dd` L15 _EDGE Clinic research: real product flows from Refero › Task: AI agent setup permissions_: > This is a settings screen in Leonardo.ai focused on personal profile management. The entry point is the user already being logged in and opening Settings, with the left sidebar showing Personal Settings and Workspace Settings. The user can review profile information, choose interest tags, confirm adult content access, and optionally delete the account. The exit point is after changes are saved and reflected in the profile section with confirmation feedback, or after the user leaves settings. → **UNPLACED** - `F-23d618be50` L19 _EDGE Clinic research: real product flows from Refero › Task: AI agent setup permissions_: **WhatsApp: Create Community Setup** (ios, 6 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12677) → `/clinic/social-prescribing-referral` - `F-1baba959f5` L21 _EDGE Clinic research: real product flows from Refero › Task: AI agent setup permissions_: > Happy path for creating a new community on mobile. The user enters a community name, customizes the community image from the Edit control, selects either an emoji or sticker with a background color, then returns to the main creation screen with the chosen avatar and name ready to submit. → `/clinic/social-prescribing-referral` - `F-a88066b022` L25 _EDGE Clinic research: real product flows from Refero › Task: AI agent setup permissions_: > A three-step onboarding carousel presented as a modal on a mobile device that introduces the "Structured AI" feature. Each slide contains a top header with the feature name, a horizontal progress indicator, an illustration of the feature, a bold headline, a short explanatory blurb, and a primary 'Continue' CTA. Controls available across slides: a Close (X) in the top-right, a Skip action, and a back arrow on later slides to navigate backwards. Entry: modal opens in front of the main app (user already in the app). Exit: user advances through slides with Continue, can Skip or Close to dismiss the modal; final Continue proceeds past the onboarding (next screen not shown). Assumption: mobile device; modal invoked to introduce a new feature; user can navigate forward/back or dismiss at any time. → `/clinic/social-prescribing-referral` - `F-2733f6f588` L27 _EDGE Clinic research: real product flows from Refero › Task: AI agent setup permissions_: Steps: Intro — Speed Up Planning → Auto-Generate Your Schedule → Digitize Your Analog Planner → `/clinic/social-prescribing-referral` - `F-53228a9681` L31 _EDGE Clinic research: real product flows from Refero › Task: AI chat assistant_: **Doctronic AI: Passwordless AI Doctor Sign-In** (web, 8 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13706) → **UNPLACED** - `F-5ee8f64fb1` L33 _EDGE Clinic research: real product flows from Refero › Task: AI chat assistant_: > The flow begins on Doctronic’s public landing experience, where an unauthenticated visitor can start a chat or access account options. The user chooses email-based login, enters an email address, receives a six-digit verification code by email, and submits the code. After successful verification, the user is taken to the authenticated AI health assistant screen, ready to ask a health question. The user is assumed to be logged out and able to access the entered email inbox. → **UNPLACED** - `F-99b8ebefe1` L35 _EDGE Clinic research: real product flows from Refero › Task: AI chat assistant_: **Doctronic AI: Sleep Concern Assessment** (web, 13 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13694) → **UNPLACED** - `F-665305f577` L37 _EDGE Clinic research: real product flows from Refero › Task: AI chat assistant_: > A logged-in user named Connie starts a conversation with Doctoric AI by entering a health concern about recurring nighttime waking. The system displays an emergency disclaimer, processes the message, and requires acceptance of the Terms of Service and AI-provider disclaimer before continuing. The user then provides age and sex information, receives follow-up questions about additional symptoms and wake-up patterns, and answers questions about health history. The flow ends with the assistant asking about current medications, with the user remaining in the conversation. → **UNPLACED** - `F-cdb7aaa80d` L39 _EDGE Clinic research: real product flows from Refero › Task: AI chat assistant_: **Google Gemini: Voice Query to Suggestions** (ios, 6 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/3207) → **UNPLACED** - `F-c198614056` L41 _EDGE Clinic research: real product flows from Refero › Task: AI chat assistant_: > Happy-path: A signed-in user interacts with the Gemini conversational assistant on a mobile device. → **UNPLACED** - `F-527e4fdda9` L43 _EDGE Clinic research: real product flows from Refero › Task: AI chat assistant_: Steps: Conversation (idle with assistant intro) → Start Voice Input (listening) → Live Transcription (partial) → Full Transcription / Submit → Assistant Responding (audio playing) → Answer Rendered (suggestions with media) → **UNPLACED** - `F-050497c925` L45 _EDGE Clinic research: real product flows from Refero › Task: AI chat assistant_: **WhatsApp: Meta AI Voice Chat Setup** (ios, 6 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12673) → **UNPLACED** - `F-de8a4d4cb7` L47 _EDGE Clinic research: real product flows from Refero › Task: AI chat assistant_: > Happy path for starting a voice conversation with Meta AI. The user enters the Meta AI profile/settings area, opens the voice feature, reviews the informational consent-style modal, and continues into a live voice session. → **UNPLACED** - `F-80d4c543f4` L51 _EDGE Clinic research: real product flows from Refero › Task: AI chat assistant_: > A user views an assistant-generated response containing ingredients and a 20-minute pasta cooking method. The user copies the response using the copy action beneath the message. A confirmation toast appears stating that the content was copied. The flow starts with an existing assistant response and ends with the response available on the device clipboard. → **UNPLACED** - `F-997d7c6368` L59 _EDGE Clinic research: real product flows from Refero › Task: API keys developer settings_: > A short onboarding carousel for Willow shown to a new or returning user before entering the main experience. The user can swipe through several promotional intro screens and tap the primary Get started button to continue. → **UNPLACED** - `F-c310b7b1bd` L63 _EDGE Clinic research: real product flows from Refero › Task: API keys developer settings_: > Happy-path onboarding/settings sequence for a logged-in user named Connie. The user opens the profile/settings area, chooses a recommended connection type (Fitness), is shown a system Health permission request, and returns to the recommendation screen after granting access. Entry is from the profile/settings menu; exit is back to the recommendation step with the selected connection highlighted and a Next button available. → **UNPLACED** - `F-d702ae8699` L65 _EDGE Clinic research: real product flows from Refero › Task: API keys developer settings_: Steps: Profile and Settings → Connection Recommendations → Health Permission Request → Connection Selected → **UNPLACED** - `F-b8253315c4` L67 _EDGE Clinic research: real product flows from Refero › Task: API keys developer settings_: **komoot: New user onboarding and consent setup** (ios, 39 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8281) → `/clinic/mental-health-care-crisis` - `F-7a8125476d` L69 _EDGE Clinic research: real product flows from Refero › Task: API keys developer settings_: > Happy-path onboarding for a first-time mobile (iOS) user. Entry: app launch -> privacy/tracking consent banner. User can Accept All, Refuse, or Customize preferences. If user customizes they can toggle categories and open vendor details. After consenting, the flow presents sign-in options (social or email). Choosing email leads to email validation then account creation (password). Post-signup the user progresses through product onboarding cards that request push notifications and access to device integrations (Apple Health). The flow ends with device connections (optional) and a completion screen, then the main map/routes home screen. → `/clinic/mental-health-care-crisis` - `F-50faecdc87` L71 _EDGE Clinic research: real product flows from Refero › Task: API keys developer settings_: Steps: Privacy notice (top-level) → Customize preferences (categories visible) → Category list (scroll) → Analytics expanded (vendor row visible) → Vendor info modal (Firebase) → Sign-in options → Enter email (initial) → Email validation error → Email accepted (proceed) → Create account (password) → Onboarding hero / intro → Onboarding card: Navigation → Onboarding card: Notifications → iOS push permission dialog → Device integrations list → Health Access (iOS modal) → Integration connected state → Onboarding completion → Main map / routes home → Optional: Connect additional devices (post-onboarding) → `/clinic/mental-health-care-crisis` - `F-4bda914af1` L73 _EDGE Clinic research: real product flows from Refero › Task: API keys developer settings_: **Poppy: Support and Privacy Settings** (ios, 6 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12194) → **UNPLACED** - `F-4939508559` L75 _EDGE Clinic research: real product flows from Refero › Task: API keys developer settings_: > A settings/support area presents account and privacy-related options, including feedback, data retention, privacy policy, terms, account deletion, experimental features, and sign out. Selecting certain items opens full-screen informational panels that explain what data Poppy collects, how it is used, and how long it is stored. Entry appears to be from an account/settings screen; exit is via the close button or back navigation. → **UNPLACED** - `F-adaf4ae428` L77 _EDGE Clinic research: real product flows from Refero › Task: API keys developer settings_: Steps: Support Menu → Data Overview → Location Details → Email Details → Health Details → Control and Deletion → **UNPLACED** - `F-153a94f24e` L81 _EDGE Clinic research: real product flows from Refero › Task: API keys developer settings_: > The flow starts from an “Add Uplink” provider selection screen and continues through the Ambient Weather credential setup. The user selects Ambient Weather, enters a personal API key, and tests the connection. The screenshots show the connection request loading and ending with a provider rejection error. The user is assumed to be on an iPhone and able to access their Ambient Weather API key. Credentials are described as being stored locally in Apple Keychain. → **UNPLACED** - `F-da47f1de5e` L83 _EDGE Clinic research: real product flows from Refero › Task: API keys developer settings_: Steps: Provider Selection → Enter API Key → Test Connection → Connection Check → Credential Rejection → **UNPLACED** - `F-7371439b4d` L87 _EDGE Clinic research: real product flows from Refero › Task: CRM contacts pipeline_: **222: Assign Close Friends** (ios, 11 screens, Lifestyle, Dating & Relationships) [Refero](https://refero.design/flows/9320) → **UNPLACED** - `F-f33319d9a8` L89 _EDGE Clinic research: real product flows from Refero › Task: CRM contacts pipeline_: > Happy path: The user refines the app's compatibility model by answering a series of persona-style prompts and assigning a real friend to each prompt. From the Close Friends screen the user taps a prompt's "choose a friend" input, picks or filters a contact from their phone list, and triggers an in-app share (pre-populated message with a referral/preview) to invite that friend to take the compatibility test. A confirmation modal appears before the assignment is finalized. Each confirmed assignment increments a progress counter (e.g., 0/5 → 1/5). → **UNPLACED** - `F-e9a205c019` L91 _EDGE Clinic research: real product flows from Refero › Task: CRM contacts pipeline_: Steps: Close Friends — prompt overview → Contact list with keyboard → Filter to a specific contact → Select contact to send referral → Message compose with referral preview → Confirmation modal (explain & confirm) → Next prompt — pick another friend → Message compose for second contact → Progressed question cards → Close Friends summary → Return to People tab / flow complete → **UNPLACED** - `F-3eef73f589` L95 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: **TIDAL: Track Report Reason Selection** (ios, 3 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/13140) → `/consortium/view-public-reviewer-track` - `F-194627cff2` L97 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: > A mobile modal for reporting a track, starting with a searchable list of report reasons and ending with a selected reason shown after search. The user opens the report dialog, searches or browses categories, and selects the matching reason. → `/consortium/view-public-reviewer-track` - `F-5db083fc0a` L99 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: Steps: Browse Report Reasons → No Match Search State → Filtered Result → `/consortium/view-public-reviewer-track` - `F-3960f58324` L101 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: **Apple Image Playground: Report Generated Image** (ios, 9 screens, AI Tool, Design & Media Editing) [Refero](https://refero.design/flows/12798) → **UNPLACED** - `F-c6f9795f79` L103 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: > A user reviews a generated image in Image Playground, opens the overflow menu, and chooses Report a Concern. They select concern categories, optionally add more information, review what will be included in the report, accept the data-sharing prompt, and submit. The report then submits successfully and returns them to the image editor, where the generated image remains visible with the usual controls. → **UNPLACED** - `F-23567f9b1c` L105 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: Steps: Image Review → Action Menu → Concern Category → Personal Characteristics Check → More Information → Submit Report → Data Consent → Submitting → Return to Editor → **UNPLACED** - `F-dbd61de343` L107 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: **fal: Recent History Preview Visibility** (web, 2 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/13542) → `/clinic/record-clinic-consultation` - `F-39e7eb1cc0` L109 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: > The user opens the Recent History area while authenticated and reviews recent API requests across endpoints. Results are displayed in a four-column card grid with generated previews, request metadata, timestamps, status codes, prompts, and actions such as Playground, Copy prompt, and Model Analytics. The user can use the Show preview toggle to switch between visible generated images and hidden preview placeholders. The flow ends with the history list displayed in the selected preview state; pagination controls and the site footer remain available below the results. → `/clinic/record-clinic-consultation` - `F-68eaf17b98` L113 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: **District: Event Feedback Submission** (ios, 5 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12916) → `/consortium/submit-paper-consortium-review` - `F-f50a5001a1` L115 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: > Happy-path mobile feedback journey after attending an event. The user enters a rating screen for a specific event ('Sip and Act at 3 Art House'), selects an overall sentiment, optionally rates individual aspects, adds a text comment, and submits. → `/consortium/submit-paper-consortium-review` - `F-eecce4a3ac` L117 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: Steps: Overall Rating → Detailed Ratings → Comment Entry → Submission Confirmation → Completion State → `/consortium/submit-paper-consortium-review` - `F-ff1f6971c8` L119 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: **Cursor: Security Audit Assistant** (web, 18 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9922) → **UNPLACED** - `F-509bb1a642` L121 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: > Happy path: From the Agents landing the user composes a task card, selects a GitHub repository, configures model(s) and multipliers, optionally attaches an image, then starts the task which spins up a cloud agent. The agent appears in the left agent list and the UI navigates to a three-column agent workspace showing the activity log (center) and a file-changes pane (right). The agent runs commands (visible as generated commands / progress) and produces a final security audit report with findings and code snippets. → **UNPLACED** - `F-447870d631` L123 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: Steps: Agents Landing — Task Composer → Repository Picker Open → Repository Search → Repository Selected → Open Model Selector → Enable Multiple Models → Set Model Multipliers → Compose Task Instruction → Attach Image Attempt → Image Preview Modal → Task Created — Agent Listed → Open Agent Conversation → Agent Starting — Generating Commands → CLI Execution Attempt → Agent Working — Additional Steps → Audit Report — Commits Reviewed → Critical Finding Detail → Final Summary and Checks → **UNPLACED** - `F-f3e30cd7bd` L129 _EDGE Clinic research: real product flows from Refero › Task: account closure_: > Happy path for closing a Care.com membership. The user starts on a membership management screen while signed in, chooses to close their account, reviews cancellation options and reason selection, then completes the closure and reaches a confirmation screen. → **UNPLACED** - `F-cf43d2389e` L137 _EDGE Clinic research: real product flows from Refero › Task: apply for loan credit_: > Happy path for a recipient responding to a friend’s product decision request on mobile. → **UNPLACED** - `F-0ff3a200a2` L141 _EDGE Clinic research: real product flows from Refero › Task: apply for loan credit_: **Lovable: Plans and credits management** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12301) → **UNPLACED** - `F-fd4092139f` L143 _EDGE Clinic research: real product flows from Refero › Task: apply for loan credit_: > Happy path for managing a workspace subscription from the Plans & credits area. Entry is the workspace settings sidebar on the Plans & credits screen. The user can review current plan status, compare Pro/Business/Enterprise options, and initiate an upgrade or credit purchase. If upgrading, the flow continues to a Stripe-style checkout where the user enters payment details and confirms the subscription. Exit is the updated Plans & credits screen showing the new active plan and refreshed credit balance. → **UNPLACED** - `F-fad846efc3` L145 _EDGE Clinic research: real product flows from Refero › Task: apply for loan credit_: Steps: Current Plan Overview → Checkout and Payment → Updated Subscription Status → **UNPLACED** - `F-4863a9059f` L149 _EDGE Clinic research: real product flows from Refero › Task: apply for loan credit_: > Mobile customer care entry that surfaces a prominent search box and categorized FAQ topics. Entry: user opens the Customer Care section (mobile browser). The primary path shown is: user enters a query in the help search (example: "visor") and is taken to a search results screen. Exit points: refine search and resubmit, select a FAQ category or question, tap header navigation (menu, account, basket), or close the help overlay. Assumptions: user is on a mobile browser (portrait), may be signed in or out (header shows account icon but no explicit sign-in state), and has basic browsing permissions. → **UNPLACED** - `F-63803c916b` L153 _EDGE Clinic research: real product flows from Refero › Task: apply for loan credit_: **District: Apply Offer Code** (web, 3 screens, Entertainment, Food & Drink, Travel & Booking) [Refero](https://refero.design/flows/13007) → `/finance/verify-identity-person` - `F-e4f6f4e769` L155 _EDGE Clinic research: real product flows from Refero › Task: apply for loan credit_: > The user is on an Offers screen during checkout or booking. They can either manually enter a coupon/promo code in the input field at the top or choose from a list of suggested offers below. The entry point is the offers panel opened from a broader booking/payment context, and the exit point is either a successful offer application or an inline validation error if the entered code is invalid. → `/finance/verify-identity-person` - `F-81c110f4d0` L157 _EDGE Clinic research: real product flows from Refero › Task: apply for loan credit_: **Lovable: Help and Resources Center** (web, 5 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12206) → `/clinic/offer-self-help-resources` - `F-6abfd2e7e4` L159 _EDGE Clinic research: real product flows from Refero › Task: apply for loan credit_: > This is a help and support entry experience for Lovable. The user starts on a public support/resources landing page with navigation, a central help search prompt, resource cards, FAQs, and a build prompt at the bottom. The user can also open a centered search modal, enter a query or numeric text, and review an AI-generated response with suggested follow-up chips. Assumption: the user is on desktop and not required to be logged in to access the visible help content. → `/clinic/offer-self-help-resources` - `F-e832e4b2e8` L161 _EDGE Clinic research: real product flows from Refero › Task: apply for loan credit_: Steps: Help Landing Page → Search Modal Opened → Query Entered → Search Suggestions → Assistant Response → `/clinic/offer-self-help-resources` - `F-2abf766d71` L165 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **District: Salon Service Booking** (ios, 8 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12874) → **UNPLACED** - `F-9e87911d93` L167 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy path for booking a salon service and reaching payment method selection. The user starts on a salon profile, chooses to book a service, selects one service and an available date/time slot, reviews the appointment details, and opens the payment method picker. → **UNPLACED** - `F-597ea5bcae` L169 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **District: Event Booking Confirmation** (ios, 5 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12872) → **UNPLACED** - `F-f25e31b68b` L171 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path ticket booking for an event, starting on a booking review screen and ending on a confirmed order with an accessible QR code ticket. → **UNPLACED** - `F-77a8f4efce` L173 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Doctronic AI: Appointment Payment Confirmation** (web, 2 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13700) → `/clinic/book-appointment` - `F-7451fcdf52` L175 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > The user is on the final payment step of a doctor appointment booking. The appointment details are shown alongside payment options. The user can use saved payment information, enter card details manually, or choose another payment method. In the shown path, the user selects saved payment information, reviews the populated card and billing fields, and can confirm the booking for $39.00. The flow starts at the payment step and ends when the user submits the booking confirmation. The screenshots do not show the post-submission result. → `/clinic/book-appointment` - `F-6e3f2c7fc1` L177 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **District: Movie Booking Support Request** (ios, 7 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12840) → `/clinic/mental-health-care-crisis` - `F-49579b5a24` L179 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > A user opens a movie-support conversation that already contains a booking summary. They confirm the booking, are prompted to describe the issue, open the attachment menu, and grant photo-library access. A booking confirmation containing a QR code and ticket details is shown as supporting context. The user types a request to replace cheese popcorn with salt, then sends the message with the booking image attached. The flow starts in an existing support chat and ends with the issue submitted to the support assistant. → `/clinic/mental-health-care-crisis` - `F-ce32e531b2` L181 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Fresha: Salon appointment booking** (ios, 9 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4973) → `/clinic/add-appointment-calendar` - `F-350a79212a` L183 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path booking of a single salon service with optional handling for fully-booked dates. Entry: user browsing/search results and opening a provider card. Exit: appointment is confirmed and shown in a booking detail screen. Assumptions: user is signed in (booking flows show payment and confirmation), has location enabled (distance shown), and is using a mobile device. → `/clinic/add-appointment-calendar` - `F-5968ae3451` L185 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Search results — provider card → Provider services list → Select services → Select time — fully booked state → Select time — date with availability → Choose time slot → Pre-visit question → Review and confirm → Confirmation & booking details → `/clinic/add-appointment-calendar` - `F-53675d3f1b` L187 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Fresha: Add Appointment to Calendar** (ios, 3 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4974) → `/clinic/add-appointment-calendar` - `F-bdd91c9b52` L189 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path: a signed-in mobile user views a confirmed appointment and chooses to add it to a calendar. → `/clinic/add-appointment-calendar` - `F-3d76dd5908` L191 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Appointment details / Confirmed → Calendar selection modal → Device event editor (prefilled) → `/clinic/add-appointment-calendar` - `F-ce815ad237` L193 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Fresha: Search & Book — Venue Listings** (ios, 2 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4966) → `/clinic/book-interpreter` - `F-fc76d63211` L195 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > User performs a location-aware search for a service (Massage) with a selected date/time. The top of the screen shows a map and a persistent search bar (query + date/time). Below are filter chips (Sort, Price, Type) and a scrollable list of venue cards. Each card shows an image, name, rating, distance, available services with prices, and tappable time-slot buttons for booking. Happy path: user reviews a venue’s service and taps a time-slot to proceed into the booking flow. → `/clinic/book-interpreter` - `F-f8c1ed6974` L197 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Venue list with time slots → No-results / Clear search state → `/clinic/book-interpreter` - `F-a457bba07a` L199 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Fresha: Refine search and pick time** (ios, 19 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4958) → `/clinic/mental-health-care-crisis` - `F-45cf7d5904` L201 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > User starts on a results list for a previously-entered query (example: "Barbering") and opens the search/filter modal to refine the search. From the modal the user can change the treatment (Treatment or venue), location, date and time, or browse categories/recents. The user drills into each filter: selects a treatment type, picks or types a new location (including using current location), selects a date from a calendar, and chooses a time window (preset or custom). After applying filters the modal returns to the results list which updates to show venues matching the query and filters with available time slots. Assumptions: mobile (iOS-style) device; user is either unauthenticated or authentication state is not relevant to the flow; user navigates via modal sheets that stack and return to the search overlay and then to the results list. Entry: results list (search active). Exit: updated results list with applied filters and visible available times. → `/clinic/mental-health-care-crisis` - `F-476aad9e1c` L203 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Results list (initial) → Search / Filter modal (root) → Treatment search (open) → Treatment search — no match state → Treatment results (matches and venues) → Location picker (root) → Location search — no match → Location search — suggestions → Search modal updated (location chosen) → Date picker (open) → Date selection (select month / date) → Time picker (open) → Time selection — preset and custom → Time selection — picker list → Search modal (filters applied preview) → Top categories (browse) → Results list (filtered) → Select time slot / Book → Back navigation and re-open search → `/clinic/mental-health-care-crisis` - `F-b08c3442d4` L205 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Omio: Booking retrieval by reference** (ios, 4 screens, Travel & Booking) [Refero](https://refero.design/flows/5242) → **UNPLACED** - `F-57aa26598d` L207 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy path: A user on the 'Your bookings' screen opens a "Retrieve booking" modal, enters a booking reference and the email used for the booking, and submits to fetch the booking so it appears in their upcoming bookings. → **UNPLACED** - `F-701b0e404b` L209 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Bookings list (Upcoming) → Retrieve booking modal (empty) → Enter details and submit → Lookup error (no booking found) → **UNPLACED** - `F-4cb1e7b309` L211 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Longevity Deck: Add activity to daily timeline** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/9895) → `/clinic/person-held-health-record` - `F-e5109f2df2` L213 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy path: From the day timeline (Yesterday) the user taps an empty timeslot (11) to add an activity. The app opens an "Add Activity at 11" picker with search, category filters, and a scrollable list of activities. The user picks an activity (optionally adjusting duration or amount in the inline numeric control) and taps the circular + control to add it. The picker closes (or the list dismisses) and the selected activity appears as a colored card in the 11 slot on the timeline. Entry: day timeline with empty slot. Exit: timeline updated with new activity card. Assumptions: user is signed in and on the daily timeline view, editing is allowed, device shown is a phone in dark mode. → `/clinic/person-held-health-record` - `F-e670fcabf3` L215 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Day timeline with empty slot → Activity picker — top of list → Activity picker — select activity → Updated timeline with activity card → `/clinic/person-held-health-record` - `F-25549753d8` L217 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Omio: Booking retrieval via Help Center** (ios, 4 screens, Travel & Booking) [Refero](https://refero.design/flows/5247) → **UNPLACED** - `F-e84a595cfc` L219 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Mobile Help Center flow for retrieving a booking or searching support content. → **UNPLACED** - `F-1f5f4b59bc` L221 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Customer Help Center landing → Search field focused → Inline empty search feedback → Full search results – empty state → **UNPLACED** - `F-69a6bbd695` L223 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Fresha: Appointment cancellation and management** (ios, 4 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4975) → `/clinic/add-appointment-calendar` - `F-32ed338f68` L225 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path: A logged-in mobile user views the details of a scheduled appointment, opens the management sheet to reschedule or cancel, confirms cancellation, and returns to the appointment details now showing a Cancelled state with options to rebook. → `/clinic/add-appointment-calendar` - `F-d12aada0e5` L227 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Appointment details (Confirmed) → Manage appointment (sheet) → Cancel confirmation → Appointment details (Cancelled) → `/clinic/add-appointment-calendar` - `F-0b421d8f5e` L229 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Genie: Weekend Plan Assistant** (ios, 7 screens, AI Tool, Social Networking) [Refero](https://refero.design/flows/12431) → **UNPLACED** - `F-2eecb5e6a0` L231 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > A mobile assistant conversation starts on a home/chat screen with suggested prompt chips and a message composer. The user types a planning request that mentions a friend, then sends it. The assistant responds by looking up the friend’s profile, generating a location-based recommendation, and showing nearby place cards. The flow assumes the user is already in the assistant experience and can type, mention people, and review suggestions. It ends with the assistant providing a drafted plan and a follow-up question about timing. → **UNPLACED** - `F-2c72420ca6` L233 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Assistant Home → Draft Message with Mention → Ready to Send → Message Sent → Profile Lookup → Recommendation Results → Plan Draft and Follow-up → **UNPLACED** - `F-806c6e904e` L235 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **District: Upcoming Movie Reminder** (ios, 4 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12870) → **UNPLACED** - `F-a0fe599a30` L237 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy path for setting a reminder on an upcoming movie and optionally choosing preferred theatres. The user enters from the Upcoming movies listing screen, where movie cards show a reminder icon. After tapping a movie’s reminder control, a confirmation state appears and may prompt theatre preference selection. The user can open a theatre picker modal, select up to three theatres in the current city, save preferences, and return to the listing. → **UNPLACED** - `F-5c8d4cda49` L239 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Fresha: Treatment / Venue search** (ios, 3 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4959) → `/clinic/mental-health-care-crisis` - `F-c5ca630ad2` L241 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > User opens a search modal (overlayed on a map) to find a treatment or venue for "Barbering." The happy path: user enters a query (or picks a recent/category), refines optional filters (location/date/time), then selects a treatment or a venue from the suggested list. → `/clinic/mental-health-care-crisis` - `F-bc5bd631ab` L243 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Search & filters → Type and refine → Review matching items → `/clinic/mental-health-care-crisis` - `F-e96caf8377` L245 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Fresha: Treatment & Category Search** (ios, 3 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4960) → `/clinic/mental-health-care-crisis` - `F-6e08998512` L247 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Mobile (iOS) search & filter flow for finding a treatment or venue. → `/clinic/mental-health-care-crisis` - `F-cf531d8380` L249 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Search & filter modal → Treatment or venue selector (with suggestions) → No-match / Top categories → `/clinic/mental-health-care-crisis` - `F-d8071d4c51` L251 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Airbnb: Photographer booking** (ios, 4 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6436) → **UNPLACED** - `F-b013356ad4` L253 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path: A mobile user discovers a photographer listing, taps Reserve, picks a package, date, and time, then confirms booking and payment. Entry: provider detail card (listing page) with service cards and a sticky Reserve CTA. Exit: user reaches the Confirm & Pay screen with booking summary and payment method (ready to review/complete payment). Assumptions: Mobile (iOS-style) interface; user has access to a saved payment method (a card is visible on the confirm screen); availability is shown per-date and per-package. The UI uses a modal sheet for scheduling and horizontal time-chip selectors for available time slots. → **UNPLACED** - `F-07f2d2b635` L255 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Provider listing & Reserve CTA → Schedule modal — pick date and time → Schedule modal — time confirmed → Confirm and pay → **UNPLACED** - `F-367a3a1f2b` L257 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Omio: Ticket access and booking management** (ios, 2 screens, Travel & Booking) [Refero](https://refero.design/flows/5231) → **UNPLACED** - `F-35dcf14189` L259 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy path: A mobile iOS user opens the "Your journey" area and selects the Ticket tab to access their ticket and booking details. From this screen the user can: add the ticket to Apple Wallet, copy the booking number, or open management options (change booking, view cancellation options, view confirmation). → **UNPLACED** - `F-c134138abd` L263 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Airbnb: Guidebook creation and place recommendation** (ios, 20 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6448) → **UNPLACED** - `F-3dcad2c4b7` L265 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path: Host navigates from the Listing editor -> Arrival guide section to open Guidebooks, creates/edits a guidebook (Maya's Guidebook), attaches a listing, edits the guidebook cover image using the photo library, then adds a place recommendation (searches for a restaurant), categorizes it and adds images and a tip. → **UNPLACED** - `F-0de0fa4d6b` L267 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Arrival guide overview → Guidebooks list → Guidebook editor (empty) → Choose listings modal (initial) → Select listing → Save listing selection → Guidebook editor (after listing linked) → Edit cover modal – pick image → Request photo library permission → Grant photo library access → Photo library picker → Cover updated preview → Add content modal → Place search UI → Select a place from results → Categorize place (required) → Add tip and images for place → Save place recommendation → Preview guidebook item → Return to Guidebooks list → **UNPLACED** - `F-9c2ec90f40` L271 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > This is a mobile assistant-style scheduling experience. It starts on a home screen with shortcut actions and a microphone-based command entry, then moves into event creation after the user chooses Create and Event. The user can pick a quick date option or enter a custom title, then refine the scheduled date and time on a calendar-style editor before confirming. → **UNPLACED** - `F-8edafe3cdc` L273 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Assistant Home → Choose Create Task → Select Event Type → Set Event Timing → Enter Event Details → Adjust Calendar Event → **UNPLACED** - `F-f71dec26c4` L275 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **WhatsApp: Schedule Call and Invite Participants** (ios, 8 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12683) → **UNPLACED** - `F-e8d3f76546` L277 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > A mobile calls interface opens on the Calls tab. The user can access recent calls, shortcuts for calling and scheduling, and create a new scheduled call. The flow starts from the Calls screen, moves into scheduled call creation, then into contact selection and chat confirmation, and ends back on the Calls screen with the new upcoming call visible. → **UNPLACED** - `F-6c4c175041` L279 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Calls Home → Calls Menu → Scheduled Calls Empty State → Schedule Call Details → Send To Recipients → Recipient Confirmed → Chat Event Confirmation → Calls Upcoming State → **UNPLACED** - `F-ac592f56dd` L281 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **TidyCal: Therapy Session Booking** (web, 6 screens, Business, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/12503) → `/clinic/hold-remote-session` - `F-fb8d5f775a` L283 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy path for a therapy provider booking site. The user starts on a therapist’s public landing page, reviews available services, selects a session type, chooses a date and time from either a calendar or list view, enters name and email in a confirmation dialog, and receives a booking confirmation screen. → `/clinic/hold-remote-session` - `F-c059b1cea1` L285 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Therapist Landing → Service Selection → Calendar Availability → Booking Details → Enter Contact Details → Booking Confirmed → `/clinic/hold-remote-session` - `F-53f8626a62` L289 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > User reaches the 'Book a private appointment' page (desktop layout). The page shows a left-hand form and a large image on the right. Country/region and store are shown as dropdown selectors at the top of the form, followed by Name, Email, Phone (optional) and Requested date & time (optional). The primary CTA at the bottom is 'REQUEST APPOINTMENT'. → `/clinic/hold-remote-session` - `F-c35d76525b` L291 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Open booking form → Fill contact details → Submit request and see confirmation → `/clinic/hold-remote-session` - `F-00c04bb720` L295 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path booking: an unauthenticated desktop user selects a service, chooses a date and time from a calendar, fills required contact fields, confirms the booking, and arrives at a confirmation screen with calendar export and reschedule/cancel options. → `/clinic/hold-remote-session` - `F-3f4b5728f0` L297 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Select Appointment Type → Choose Date & Time → Enter Contact Information → Review Filled Information → Appointment Confirmed → `/clinic/hold-remote-session` - `F-a6088ff688` L299 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Acuity Scheduling: Create and publish appointment type** (web, 8 screens, Business, Scheduling Tools) [Refero](https://refero.design/flows/9857) → `/clinic/onboard-wearable-device-type` - `F-b85b9d96d3` L301 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy path: A signed-in user managing appointment offerings starts on the Appointment Types list, chooses to create a new individual appointment type, completes details (name, description, duration, confirmation message), adds an image, selects calendars and optional resources/forms, then saves the new appointment type. → `/clinic/onboard-wearable-device-type` - `F-27f73f4d80` L303 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Appointment Types list → Details — initial → Details — completed → Image & Calendar selection → Image uploaded → Forms, Resources & Additional Options → Saving — loading → Appointment Types list — confirmation → `/clinic/onboard-wearable-device-type` - `F-01c5f39fb8` L305 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Fresha: Salon appointment booking** (web, 14 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/2379) → `/clinic/add-appointment-calendar` - `F-51deecfa95` L307 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path booking of a single service (Gel Manicure) at a listed salon (AMA Brixton). → `/clinic/add-appointment-calendar` - `F-7e4e9338f0` L309 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Salon detail (services overview) → Choose booking option → Select services (list) → Service details modal → Service selected (summary visible) → Select professional (grid) → Professional confirmed → Select time (calendar + slots) → Date view and slot selection → Full time list (scroll view) → Change/select professional from time screen (modal) → First-visit question → First-visit selected (example: No) → Review & confirm (payment entry) → `/clinic/add-appointment-calendar` - `F-99ada4023c` L313 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy path: a shopper uses the store locator to find a nearby Saint Laurent boutique, opens the store details, and requests a private in-store appointment via a right-side booking panel. → `/clinic/add-appointment-calendar` - `F-630fdd2610` L315 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Store locator results → Store detail (Grant Avenue) → Open appointment panel → Select date — no hours available → Select alternate date and time → Submit appointment — error response → `/clinic/add-appointment-calendar` - `F-0caa33cfbd` L317 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **TidyCal: Create Booking Invite** (web, 6 screens, Business, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/12502) → **UNPLACED** - `F-d00c8b4d4d` L319 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy path for creating a booking invite from the Bookings section. The user starts on the Bookings screen, opens the Create a booking panel, selects a booking type, timezone, date, time, and guest details, then sends an invite. After submission, the modal closes, a success toast appears, and the new invite is shown in the Invited list with its status and actions. → **UNPLACED** - `F-26818a8b5d` L321 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Invited Bookings Empty State → Booking Details Panel → Date and Time Selection → Guest Details Completed → Invite Sent Confirmation → Invited Bookings List → **UNPLACED** - `F-278dc85b4f` L323 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **West Elm: Design Appointment Scheduling** (web, 10 screens, Home & Decor, Marketplace) [Refero](https://refero.design/flows/1674) → **UNPLACED** - `F-799a67fedc` L325 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path summary: A desktop user requests a free Design Crew appointment. They start on the service overview, pick an appointment type (in-home, virtual, or in-store), select a local store (if applicable), choose a date and time from an availability grid, enter contact information, add project details, and receive an on-screen booking confirmation with a reference and follow-up questions. → **UNPLACED** - `F-5050e33a01` L327 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Appointment Type Selection → Find Local Store → Date & Time - Availability Overview → Select Time Slot → Change Date (Calendar) → Enter Contact Information → Info Required Modal → Confirm Contact Details → Add Project Details → Booking Confirmation & Follow-up → **UNPLACED** - `F-fdbef70a65` L331 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy path for creating a multi-service appointment from the weekly schedule on desktop. Entry: user is on the weekly calendar for a location (Laboratory) and a target date is visible (Feb). The user selects a time slot or the global + button to open a right-side "New appointment" drawer, chooses a client and provider, adds services (which updates the end time), and books the appointment via a confirmation modal. Exit: the appointment appears on the calendar (block with hashed buffer) and a success toast is shown. Assumptions: user is authenticated and has permission to create appointments, the interface is desktop with sufficient horizontal space for a right-side drawer and central calendar, provider list and client list are prepopulated. → **UNPLACED** - `F-c94f43d55c` L333 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Weekly calendar view (initial) → Select start time → Open New appointment drawer → Choose client → Open services selector → Select multiple services → Review assembled appointment → Open booking confirmation modal → Toggle notification and confirm → Appointment created (side panel) → Calendar updated with appointment → **UNPLACED** - `F-82288750a7` L337 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path booking: A visitor opens a host's public booking listing (desktop, not signed-in) and sees a list of offered meeting types (durations and descriptions). The visitor selects a meeting type (30 Min Meeting), picks an available date from a month calendar, chooses an available time slot, fills contact information and optional notes, and confirms the booking. → **UNPLACED** - `F-730df4b55f` L339 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Meeting types list → Confirm meeting type selection → Month calendar — pick a date → Choose time slot → Booking details form → Confirmation summary → `/clinic/book-appointment` - `F-3c8425742c` L341 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Fresha: Group appointment and waitlist booking** (web, 30 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/2375) → `/clinic/add-appointment-calendar` - `F-aa16523ace` L343 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path booking attempt for a group appointment (up to 3 guests) starting from a salon’s listing. → `/clinic/add-appointment-calendar` - `F-0ac5b5ec22` L345 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Salon overview & services → Booking type chooser → Select services – master list → Service options dialog (Lash Lift) → Add additional services (e.g., Level 5) → Add guests and services summary → Edit services for a guest → Confirm selected service(s) → Join the waitlist – date/time entry → Join the waitlist – time range picker → Add multiple preferred times → First visit intake question → Review & confirm waitlist – payment and notes → Waitlist error – unavailable → User attempts to exit booking → Exit confirmation decision → Booking type chooser (alternate view) → Service options selection confirmation → Multiple preferred time rows shown in summary → Final error state and guidance → `/clinic/add-appointment-calendar` - `F-d7b1867f6e` L347 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Doctronic AI: Licensed Doctor Appointment Booking** (web, 25 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13699) → `/clinic/add-appointment-calendar` - `F-8c5fd7b2b5` L349 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > The user begins with a saved Doctoronic medical report and chooses to continue to a licensed doctor. They select a payment method, provide the state where the appointment will occur, choose an immediate or scheduled appointment time, enter patient and contact details, and reach the final booking confirmation and payment step. The flow assumes the user has a report ready to share and may have saved contact or payment information. The visible sequence ends on the payment screen before a completed booking confirmation is shown. → `/clinic/add-appointment-calendar` - `F-769229f965` L351 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **TidyCal: Booking Types Management** (web, 3 screens, Business, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/12493) → `/clinic/build-clinic-template` - `F-f6d8a7f94a` L353 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > This is a booking types management screen in a scheduling admin area. The user can view a list of booking types, filter them by status, enable or disable a type, open an actions menu for each item, and share a booking page link through a share panel. → `/clinic/build-clinic-template` - `F-02d636a9e7` L355 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Booking Types List → Item Actions Menu → Share Booking Link → `/clinic/build-clinic-template` - `F-854d959c3f` L357 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **TidyCal: Therapist Booking Landing Page** (web, 2 screens, Business, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/12486) → **UNPLACED** - `F-9523c1a63c` L359 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > A therapist’s public booking landing page is shown in two visual theme variations. The user is assumed to be a site owner or editor previewing the page in an editing environment, with access to theme, typography, color, and save controls. The page begins with a hero introduction, then scrolls through credibility stats, bio, service offerings, testimonials, FAQs, and a final booking call to action, ending with footer/social link placeholders. The primary exit point is the booking button, which leads to scheduling, while the editor can also save the design changes from the top bar. → **UNPLACED** - `F-b84eeab0d2` L363 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Fresha: Search availability by treatment, location, date, time** (web, 8 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/2323) → `/clinic/search-recruiting-trials` - `F-df2ae5c92f` L365 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > User starts on their Appointments screen (desktop, logged-in — profile avatar and an upcoming confirmed appointment are visible). From here they use the persistent search bar at the top to find new availability: they open the Treatment selector and pick "Tattoo & piercing", open Location and choose a saved address (Home), open the Date picker and choose a specific day (Dec 23), then set a custom time window (16:00–21:00). After applying filters and using Search, the interface displays search results on a split layout: a scrollable list of venues and available time slots on the left and an interactive map on the right. → `/clinic/search-recruiting-trials` - `F-589b75ee5d` L367 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Appointments overview → Choose treatment category → Open location selector → Open date picker → Select a calendar date → Open time filter presets → Set custom time range → View filtered search results → `/clinic/search-recruiting-trials` - `F-32c43d8e0c` L369 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Cal: Request reschedule for booking** (web, 5 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3493) → **UNPLACED** - `F-69754bdab0` L371 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path: A signed-in user views their Upcoming bookings list, opens a booking's Edit dropdown, chooses "Request reschedule," types an optional reason in a modal, and submits the request. Entry: Bookings > Upcoming (desktop, authenticated user). Exit: modal closes and a transient confirmation toast is shown: "Reschedule request sent." The interface guides the user with an Edit dropdown → Request reschedule modal that explains the effect (cancel meeting and notify scheduler) and provides an optional free-text reason. → **UNPLACED** - `F-06446b0f59` L373 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Bookings list (Upcoming) → Open Edit dropdown → Request reschedule modal (empty) → Enter reason and submit → Confirmation toast → **UNPLACED** - `F-fd2829aa3f` L375 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **IKEA: Interior design appointment booking** (web, 9 screens, Marketplace, Home & Decor) [Refero](https://refero.design/flows/8270) → `/clinic/add-appointment-calendar` - `F-dabbe55703` L379 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Service landing → Enter postcode → Choose service → Service details (expanded) → Select date (calendar) → Confirm timeslot (alternate state) → Contact & project details (form start) → Contact form (filled state) → Booking confirmation & payment prompt → `/clinic/add-appointment-calendar` - `F-d2f3dc86b3` L381 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **TidyCal: Booking Page Setup and Scheduling** (web, 5 screens, Business, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/12499) → **UNPLACED** - `F-7fb8a38555` L383 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy path for configuring and sharing a booking page, then reviewing available booking types and availability. The flow appears to start from a dashboard/booking setup area after the account is already live, with the user signed in as the business owner. It exits on a booking-type preview/scheduling screen where the user can choose a date and time or continue editing shared booking settings. → **UNPLACED** - `F-51c6b3e4d0` L385 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Setup Checklist → Availability Calendar → Theme Selection → Availability Review → Booking Type Review → **UNPLACED** - `F-1737f8d684` L387 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Doctronic AI: Appointment Document Upload** (web, 4 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13702) → `/clinic/person-held-health-record` - `F-ee7ca0cc02` L389 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > A logged-in user opens the Appointments area and views an upcoming video appointment with booking in progress. They open the Files section to review the file-upload prompt, add documents, and wait for the upload to complete. The uploaded PDF files then appear in the appointment's file list, where each file has a remove control. → `/clinic/person-held-health-record` - `F-64ea87da8e` L395 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: > This is a mobile analysis experience where the user enters a personal feeling, dream, or life situation, then submits it for interpretation by a panel of analysts. → `/clinic/person-held-health-record` - `F-bd8bba3dcc` L397 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: Steps: Prompt Entry → Text Composed → Submit Analysis → Processing State → Results Begin → Analyst Responses → Detailed Reading → Expanded Analysis View → `/clinic/person-held-health-record` - `F-69874d5d5f` L401 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: > Happy path starts from a saved or viewed place detail screen for Citizens Bank Park, showing venue information, actions like call/website/tickets, and a directions button. The user can open navigation from the place detail view, which transitions into the maps interface focused on Citizens Bank Park and nearby points of interest. → `/clinic/person-held-health-record` - `F-39276c11d7` L405 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: **Spyglass: Export and Rate Analysis** (web, 3 screens, Business, Inspiration Board) [Refero](https://refero.design/flows/13461) → `/consortium/judiverse-included-edge` - `F-441c352c6f` L407 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: > A user reviews a generated advertising analysis in the Spyglass workspace and interacts with the response action toolbar. The sequence begins with the completed response visible, then shows a PDF download confirmation, a hover tooltip for the positive-feedback control, and finally the positive-feedback control selected. The user appears to be logged in and viewing an existing conversation or research workspace. The flow ends with the response retained in the conversation and feedback recorded. → `/consortium/judiverse-included-edge` - `F-b6604ff4b1` L409 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: Steps: Review Generated Analysis → Identify Helpful Response Control → Submit Positive Feedback → `/consortium/judiverse-included-edge` - `F-0932664a4b` L413 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: > A logged-in user on GoFundMe searches for “paris” from the top search field and browses result types using the category filter. The initial results show nonprofit organizations with a secondary location filter. The user opens the category dropdown and changes the result type to Profiles, which updates the results grid to people profiles matching the same query. Entry is the global search/results view; exit is either selecting a result or refining the search/filter. → `/consortium/judiverse-included-edge` - `F-fb2fbf88ef` L419 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: > This is a guided content experience in the Foodllama mobile interface. The user starts on a home/dashboard screen showing a featured insight card and a progress summary for today’s insights. Tapping or swiping into the card opens a full-screen lesson sequence with multiple steps, each presenting one nutrition tip with illustration, title, short explanation, and a progress indicator plus a Leave action. The sequence ends on a completion screen with celebratory confetti, a trophy graphic, and options to tell friends or finish. After completion, the user returns to the dashboard, where the insight count updates. Assumptions: user is already inside the app, on a mobile device, and viewing personalized insight content. → `/consortium/judiverse-included-edge` - `F-db43e66227` L421 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: Steps: Insight Card Dashboard → Intro Tip Screen → Carb Tip Screen → Strength Tip Screen → Completion Celebration → Return Dashboard → Updated Insight State → `/consortium/judiverse-included-edge` - `F-c79697fb15` L425 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: > A signed-in user views a personal viewing breakdown screen, then opens a display filter to switch between All, Movies, and Shows. The screen updates the chart and stats based on the selected category. Entry is from a stats or profile area; exit is via back navigation. → `/consortium/judiverse-included-edge` - `F-347ac635d5` L429 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: **District: Restaurant Q&A Assistant** (ios, 7 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12846) → **UNPLACED** - `F-79218c52a8` L431 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: > Happy path for an in-context restaurant question assistant. The user starts from a restaurant detail context and asks about Hotel Rishi. The interface shows the current topic, displays the user’s question as a highlighted prompt, and returns an AI-generated response with related action cards (Food Menu, Bar Menu, Beverage Menu). The user can rate the answer, ask a follow-up by text or voice, or submit a new query such as finding nearby breakfast options. The flow ends with a refreshed response and the follow-up input ready for another question. → **UNPLACED** - `F-9273f509e0` L435 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: > User starts on the assistant home screen, enters a query about free things to do nearby for the weekend, and waits while the system searches and composes an answer. The response appears in the main conversation area with a right-side panel that can switch between Steps and Sources. From there, the user can review how the answer was assembled, inspect cited sources, and hover or click source items to preview the referenced content. → **UNPLACED** - `F-db511a0ed4` L437 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: Steps: Home Prompt → Searching State → Steps Panel Opened → Draft Answer Displayed → Citation Hover → Sources Tab Opened → Source Preview Card → Image Preview → Source Link Hover → **UNPLACED** - `F-d86ef272ab` L439 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: **Cohere: Market overview chat** (web, 3 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3856) → **UNPLACED** - `F-d4d0f0a214` L441 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: > Happy path: A user composes a natural-language prompt in the Cohere Playground chat area, optionally configures JSON mode/schema and advanced parameters from the right-hand panel, sends the prompt, and reads the generated market overview. Entry: user lands on the Playground with Chat selected and an empty chat editor. Exit: the user has a multi-section, generated response (market size, drivers, innovations, challenges) visible in the chat and the editor ready for follow-up. Assumptions: user is on a desktop viewport, logged in, and has access to model/parameter controls on the right-hand panel. → **UNPLACED** - `F-7810cfad43` L443 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: Steps: Prompt composition & schema setup → Submit prompt and view error / initial generation → Review expanded assistant response → **UNPLACED** - `F-ae67b23741` L447 _EDGE Clinic research: real product flows from Refero › Task: business account onboarding company verification_: **Peerlist: Create Company Profile** (web, 6 screens, Job Search, Business, Social Networking) [Refero](https://refero.design/flows/1102) → **UNPLACED** - `F-0fa7320fc1` L449 _EDGE Clinic research: real product flows from Refero › Task: business account onboarding company verification_: > Happy path: A logged-in user (with permission to create a company) initiates creating a company profile via a modal. They choose an organization profile type, complete the profile form (display name, website, tagline, location, handle, size, etc.), then trigger an email verification step. After entering the verification code and accepting terms, the profile is created and the user is taken to the newly created company profile overview. → **UNPLACED** - `F-7243536630` L451 _EDGE Clinic research: real product flows from Refero › Task: business account onboarding company verification_: Steps: Choose profile type → Company profile form (empty) → Company profile form (filled) → Open verify work email modal → Enter verification code → Company profile overview → **UNPLACED** - `F-23af7fde4f` L455 _EDGE Clinic research: real product flows from Refero › Task: calendar scheduling meeting_: **Genie: Chat Reminder Setup** (ios, 3 screens, Social Networking, AI Tool) [Refero](https://refero.design/flows/12426) → **UNPLACED** - `F-9580605386` L457 _EDGE Clinic research: real product flows from Refero › Task: calendar scheduling meeting_: > A chat-style assistant screen where the user types a reminder request, sends it, and receives confirmation that a reminder has been created for tomorrow at 9:00 AM. The flow begins in an active conversation with the keyboard open and ends with the assistant confirming the reminder and offering to save reference photos to a vault. Assumptions: user is already in the conversation, signed in, and has permission to create reminders. → **UNPLACED** - `F-d084dc18e6` L461 _EDGE Clinic research: real product flows from Refero › Task: calendar scheduling meeting_: **Cofounder: Daily Meeting Summary Test** (web, 5 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/7900) → **UNPLACED** - `F-9416f0376a` L463 _EDGE Clinic research: real product flows from Refero › Task: calendar scheduling meeting_: > Happy path: a user viewing a configured scheduled automation called "Daily Meeting Summary Email" starts a manual test of the scheduled trigger, watches the test run initialize, and inspects the run timeline as integrations (Calendar) and internal reasoning steps are executed. → **UNPLACED** - `F-304061237d` L465 _EDGE Clinic research: real product flows from Refero › Task: calendar scheduling meeting_: Steps: Plan view — Automation details → Test confirmation modal → History view — Test run initializing → History view — Run processing progress → Run timeline — Integration and reasoning details → **UNPLACED** - `F-424d7c427a` L469 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Poppy: Poppy Account Setup and Subscription** (ios, 42 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/12176) → **UNPLACED** - `F-1797670f0a` L471 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > A first-time user sets up Poppy on an iPhone. The user enters a phone number, verifies it with an SMS code, provides a preferred name, selects the areas where Poppy can help, and configures notification behavior. The user is then shown Sprout and Bloom subscription options, selects the Bloom annual plan, and completes the purchase through Apple Pay. After confirmation and a short preparation state, the user reaches the personalized Poppy dashboard. The screenshots show a successful paid onboarding path; the visible “Not now” option on the subscription screen is an alternative exit that is not followed. → **UNPLACED** - `F-b530f4b6c0` L473 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Welcome and Phone Entry → Phone Number Input → Submit Phone Number → Phone Verification Request → Enter Verification Code → Complete Verification → Name Entry → Confirm Name → Choose Support Areas → Confirm Support Preferences → Customize Daily Outreach → Sprout Subscription Option → Bloom Subscription Option → Compare Subscription Features → Confirm Bloom Annual Plan → Apple Pay Subscription Review → Authorize Subscription → Dismiss Purchase Confirmation → Prepare Personalized Experience → Personalized Dashboard → **UNPLACED** - `F-2950cb8716` L475 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Ada: Symptom assessment - conversational** (ios, 31 screens, Health & Wellness) [Refero](https://refero.design/flows/4097) → `/clinic/symptom-check-triage` - `F-9cd8da67dc` L477 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Mobile, conversational symptom assessment starting from a 'New assessment' entry screen (user name shown). The user is guided through: confirming who the assessment is for, entering a main symptom (search with autocomplete), answering a short series of multiple-choice questions, optionally opening explanatory anatomy/help modals, and receiving a results report with ranked possible causes. → `/clinic/symptom-check-triage` - `F-862504b6a5` L479 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Intro / Start assessment → Select assessment recipient → Prompt for main symptom → Symptom search modal (keyboard) → No results fallback → Autocomplete suggestions → Symptom detail and selection → Duration question → Orthostatic effect question → Open explanation modal (example) → Explanation modal — confirmation → Explanation modal (neck mobility) → Neck movement question → Reduced mobility illustration (detail) → Concentration question → Symptom trend question → Assessment complete — report being generated → Open report CTA → Results summary / possible causes → Detailed causes list and continue → `/clinic/symptom-check-triage` - `F-561816f983` L481 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Ada: Record a symptom entry** (ios, 12 screens, Health & Wellness) [Refero](https://refero.design/flows/4107) → `/clinic/person-held-health-record` - `F-12b38d63b4` L483 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Happy path: From a feedback prompt the user chooses to track symptoms, picks a symptom (Dizziness), answers a short guided questionnaire (trigger: standing up effect; description type), optionally sets current severity using a segmented color-coded slider and +/- controls, optionally adds freeform notes, saves the entry, then returns to a timeline where the new entry can be opened to view details. → `/clinic/person-held-health-record` - `F-cf0317158e` L485 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Feedback prompt → Choose symptom → Initial symptom questions → Select trigger effect → Select symptom description → Open severity control → Set severity (low) → Set severity (mid) → Set severity (high) → Open notes field → Enter free-text note → Save and view entry → `/clinic/person-held-health-record` - `F-4b498d767a` L487 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Ada: Profile setup and symptom logging** (ios, 66 screens, Health & Wellness) [Refero](https://refero.design/flows/4095) → **UNPLACED** - `F-323d166402` L489 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Happy path for adding a new user profile and logging a symptom (dizziness) into the symptom tracker. → **UNPLACED** - `F-6d15f7e28f` L491 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Home / Start assessment → Add profile — Intro & safety note → Ask for user name → Name input active → Ask sex assigned at birth → Why only two sexes (modal) → Ask date of birth → Ask smoking history → Ask high blood pressure history → Open symptom modal (Dizziness) → Dizziness — trigger & type → Dizziness — optional severity collapsed → Set severity — low value → Set severity — mid value → Set severity — high value → Optional notes — edit → Save symptom (network notice) → Symptom tracking timeline → Symptom detail summary → Finish / Done → **UNPLACED** - `F-44e2e314a5` L493 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Ada: Medical quality article walkthrough** (ios, 10 screens, Health & Wellness) [Refero](https://refero.design/flows/4104) → **UNPLACED** - `F-4365b0b5bc` L495 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Happy path: A user finishes a symptom assessment and sees a high-priority recommendation ("Seek emergency care") with an inline "Tell me more" button. Tapping that link opens Ada’s external "Medical quality" article (about.ada.com) in a browser/webview. The user scrolls through the article: headline, supporting copy, visual illustrations, a "What to look for" section with accordion items (clinical governance, user experience, condition coverage, etc.), expands "Condition coverage" to view bullet points, continues to the "How Ada’s AI works" section, sees a CTA to contact the company, reviews the reference list (numbered DOIs), and reaches the footer with legal links and social icons. → **UNPLACED** - `F-5b6c7607bc` L497 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Symptom result with "Tell me more" → Article landing — Medical quality → Evidence section header → "What to look for" introduction → Accordion list of quality pillars → Expand 'Condition coverage' → How Ada's AI works → Contact CTA and Keep reading → Reference list (citations) → Footer and site navigation → **UNPLACED** - `F-3d66f37c01` L499 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Doctronic AI: Anonymous Health Consultation** (web, 6 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13696) → **UNPLACED** - `F-0876985c12` L501 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > The user starts on a health consultation entry screen and submits a symptom description to an AI assistant in Anonymous Mode. The system displays privacy and emergency notices, processes the message, summarizes the concern, and asks the user to accept the Terms of Service before continuing. The flow begins with the user ready to ask a health question and ends with the consultation ready for the user to provide more information. → **UNPLACED** - `F-022f976af0` L503 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Doctronic AI: Start Health Consultation** (web, 2 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13693) → **UNPLACED** - `F-1f0bc7b856` L505 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > The flow begins on a logged-in health assistant home screen with a collapsed navigation rail. The user expands the navigation to view available sections while remaining on the New consult screen. From here, the user can enter a health question and submit it. The screenshots show the entry state and expanded-navigation state; no question has been submitted yet. → **UNPLACED** - `F-d517b7a790` L507 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Breathwrk: Breathwrk Onboarding & Purchase** (ios, 35 screens, Health & Wellness) [Refero](https://refero.design/flows/7724) → **UNPLACED** - `F-6805ba9219` L509 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Onboarding → account / purchase → first-use tutorial. Entry begins at a marketing-style welcome/splash that highlights benefits (sleep, stress, energy). The user is guided through a short educational carousel (value props + neuroscience), asked to choose goals, offered a quick lung calibration test, then presented with a premium subscription modal (7-day trial pricing). Native platform purchase and sign-in modals appear (App Store / Sign in with Apple). After confirmation the user sees a brief onboarding tutorial and lands on the main activity screen. → **UNPLACED** - `F-d02c1f0bce` L511 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Welcome Splash → Benefit Tags → Problem Statement → Quick-Time Benefit → Neuroscience Evidence → Goals Selection (intro) → Goals Confirmation → Outcome Graph + CTA → Lung Test Intro → Inhale Instruction → Measurement / Feedback → Subscription Promotion → Pricing Selection → Native Purchase Confirmation → Purchase Success → Account Sign-in Prompt → Sign in with Apple Modal → Loading / Transition → Quick Tutorial Intro → Main Home (Ready State) → **UNPLACED** - `F-a1ccffe0b2` L515 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Happy-path: A logged-in user opens an in-product conversational assistant to request a personal productivity/growth plan. The assistant asks clarifying questions, offers quick-reply choices for focus areas, generates a week-based task plan, and then provides CTAs to add professional tasks or schedule the generated tasks into the product Planner (calendar). Entry: assistant chat screen. Exit: Planner view filtered to the user's selected project with the new tasks visible. Assumptions: user is authenticated, has at least one project (named Humanity) available, and the product can create tasks and update the Planner/calendar. → **UNPLACED** - `F-0941747d84` L517 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Chat start / prompt → User submits goal → Assistant thinking → Assistant asks clarifying questions → Quick-reply focus selection → Assistant generates task plan → Task detail links and CTAs → Planner view confirmation and options → **UNPLACED** - `F-2ce0c01c81` L521 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Happy-path: A signed-in user opens the assistant home screen, uses a category chip to reveal suggested prompts, selects a suggested prompt which populates the message composer, sends the prompt, and receives a multi-section weekly plan with recommendations and follow-up action buttons. → **UNPLACED** - `F-789c6cff67` L523 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Assistant home / composer → Category suggestions opened → Prompt populated in composer → Assistant returns weekly plan → Recommendations and follow-up CTAs → **UNPLACED** - `F-9b3c47dc8c` L525 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Longevity Deck: First Launch Introduction** (ios, 5 screens, Health & Wellness) [Refero](https://refero.design/flows/9879) → **UNPLACED** - `F-ce4196034f` L527 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > This is a first‑time user onboarding carousel that introduces the app’s core features, asks the user to acknowledge a medical disclaimer, then lands the user on the app’s main entry screen where they can begin setup (upload a photo, browse protocols, or connect health data). → **UNPLACED** - `F-019efd79f4` L529 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Intro: Personal Story → Feature: Connect Health Data → Feature: Track Daily Timeline → Legal: Medical Disclaimer → Main Entry: Setup & CTAs → **UNPLACED** - `F-ad58118a6a` L531 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Givingli: New User Onboarding & Subscription** (ios, 45 screens, Lifestyle, Scheduling Tools) [Refero](https://refero.design/flows/8150) → **UNPLACED** - `F-86b9472452` L533 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Happy-path onboarding for a gifting app that collects a user's name, preferences, celebration types and recipient categories, then presents a premium subscription offer and completes an in‑app purchase. → **UNPLACED** - `F-50ef608b93` L535 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Welcome / Splash → Name Input (first name) → Name Input (last name) → Welcome Greeting → Why are you here? (multi-select) → Selected reasons (chips active) → Choose celebrations (grid, initial) → Celebrations — some selected → Celebration options continued (scroll) → Recipients / Gift list categories → Recipient selection confirmation → Premium upsell landing → Choose plan (yearly selected default) → Plan selection swap (monthly highlighted) → Loading / requesting purchase → Platform purchase sheet → Purchase success confirmation → Home screen (post onboarding) → Subscription UI states (alternate) → Platform confirmation details → **UNPLACED** - `F-c4b824bf66` L539 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Happy path: A user opens the 'Structured AI' modal to create a plan (example: a meal plan). The modal greets the user, offers horizontal suggestion chips and a prominent input field with voice and camera affordances. The user selects or types a prompt, then taps Generate. The UI returns a generated plan preview area with a 'Your plans' summary, an 'Accept All' action, trust notice about AI accuracy, and a list of created calendar-style 'Meal Plan' items. → **UNPLACED** - `F-dc06567f95` L541 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Open Structured AI and view suggestions → Browse suggestion chips and focus input → Type a prompt → Generate plan → Review generated items → **UNPLACED** - `F-7314b350a4` L543 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Mozi: Share Plans via Messages** (ios, 4 screens, Social Networking, Travel & Booking) [Refero](https://refero.design/flows/4147) → `/clinic/social-prescribing-referral` - `F-24f93a5569` L545 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > User is on the My Plans screen (mobile, logged-in) and wants to share their upcoming plans with others. From the My Plans list the user opens a "Share my plans" modal, chooses to share, selects Messages from the system share sheet, and lands in the Messages composer with a preview card of their plans inserted. → `/clinic/social-prescribing-referral` - `F-aea55c6f6c` L547 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: My Plans list → Share options modal → iOS share sheet → Messages composer with preview → `/clinic/social-prescribing-referral` - `F-ed200a33f8` L551 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Happy path for creating a multi-city trip from the Trip overview, entering dates and stops, using AI/search suggestions, generating trip details, and saving the plan back to the main trip list. → `/clinic/social-prescribing-referral` - `F-8088442300` L553 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Trip Overview → Initial Trip Setup → Search Suggestions → Processing Entry → Trip Draft with Stops → City Weather Details → Saved Trip List → `/clinic/social-prescribing-referral` - `F-b26f12b80a` L555 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Mozi: Create a trip plan** (ios, 16 screens, Social Networking, Travel & Booking) [Refero](https://refero.design/flows/4137) → **UNPLACED** - `F-5a90de61c4` L557 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Happy-path: A logged-in user on the app Home screen taps the central + to create a new travel plan. The New plan sheet opens as a modal with fields for city, dates, description and a tentative toggle. The user opens the city selector, searches for and selects a city (chooses San Francisco), grants location permission when prompted, then sets a date range via the calendar, adds a short description (24-character limit), marks the plan as tentative, and saves. The flow ends on the city detail screen showing the newly created upcoming plan card and a confirmation CTA. Assumptions: user is signed in, device is iOS (modal and system location prompt shown), user has network access for location search. → **UNPLACED** - `F-df18822ea1` L559 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Home (empty state) → New plan form (empty) → Plan location (search) → Plan location (search ready) → Location permission prompt → Plan location (current location enabled) → Plan location (search results) → Plan location (select city) → New plan (city filled) → Calendar (month view) → Calendar (select end date) → Calendar (range saved) → New plan (dates filled) → Description entry (character limit) → Tentative toggle → Save plan and city confirmation → **UNPLACED** - `F-a13a19c4c6` L563 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Happy path: A user on the Structured AI home screen taps a promotional CTA to obtain a printable planner sheet. The CTA opens an external link (Linktree) listing Weekly / Monthly / Annual versions. The user selects a version, confirms the download in the browser prompt, and the PDF opens in the device's PDF viewer where they can share, save, or print it. → **UNPLACED** - `F-4d379443e3` L565 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Home - Download CTA → Link hub - Choose Version → Download Confirmation → PDF Preview → **UNPLACED** - `F-f9ecef34c8` L569 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Happy-path: A user scans a handwritten weekly planner and reviews AI-extracted calendar items. From a modal titled "Structured AI" the user can preview extracted tasks, open a larger scan preview, add custom instructions, regenerate the extraction, and then review the updated list. → **UNPLACED** - `F-f97477aaa4` L571 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Initial extraction preview → Detailed scan preview → Provide custom instructions → Regeneration / processing → Updated extraction results → **UNPLACED** - `F-36ff1c3976` L573 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Napper: Baby schedule & invite** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/8812) → **UNPLACED** - `F-be30678193` L575 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Happy-path: A parent edits a baby's schedule and invites a caregiver. → **UNPLACED** - `F-55472aa589` L577 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Schedule settings → Share baby profile (invite) → Caregiver added / list view → **UNPLACED** - `F-4a0ebf09a1` L581 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Happy path: From the daily timeline the user launches a task-creation flow, chooses the "Plan your day with AI" suggestion, provides a natural-language description of their plans (typed or voice), grants required system permissions (speech recognition, microphone, calendar/tasks access), and lets the Structured AI generate a prioritized agenda. The user reviews the AI-proposed tasks, accepts (or edits/discards) items, and adds a selected item into the calendar. → **UNPLACED** - `F-8fa62a028d` L583 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Daily timeline (entry) → New Task sheet → Structured AI — intro screen → Structured AI — suggestions carousel → Provide freeform plan (typed) → Start speech transcription (optional) → Grant microphone permission → Share calendar/tasks permission → AI processing — prioritizing → AI processing — optimizing → Review AI suggestions summary → Accept / expand suggestions → Item actions — Add / Edit / Discard → Add Task — set time and duration → AI list after edits → Discarding a suggestion → Finalized agenda → Voice recording states → Generate AI agenda → Per-item inline actions and final edits → **UNPLACED** - `F-f37350fc3a` L587 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > A returning Care.com user starts on the public homepage, enters the login experience, provides an email address and password, and is taken to their member dashboard. The final dashboard shows the user’s existing job, upgrade prompts, caregiver discovery options, educational content, and account navigation. The flow assumes the user is not currently logged in and has valid credentials. → **UNPLACED** - `F-0c9d80498e` L589 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Care.com Homepage → Email Entry → Password Prompt → Password Entered → Member Dashboard → Dashboard Actions → **UNPLACED** - `F-beb94d900c` L591 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Sunsama: AI-Assisted Task Capture** (web, 7 screens, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/11601) → **UNPLACED** - `F-09166e498c` L593 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > A logged-in user is viewing the Backlog and opens the Sunny assistant from the floating action button. The user selects the Braindump context, sends a natural-language request about planning a trip to Tokyo, and receives confirmation that Sunny added a corresponding task to the backlog for the next month. The flow begins in the Backlog workspace and ends with the new task visible in the appropriate time bucket. → **UNPLACED** - `F-4f356e2a6d` L595 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Open Sunny Assistant → Choose Braindump Context → Start Conversation → View Assistant Prompt → Enter Travel Idea → Send Travel Request → Confirm Backlog Task → **UNPLACED** - `F-0ce6540b2a` L597 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Airtable: Project intent and company details** (web, 3 screens, Project Management, Collaboration Tools, Scheduling Tools, Paperwork, Productivity) [Refero](https://refero.design/flows/1542) → **UNPLACED** - `F-85d1bc7731` L599 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Happy path: A user starts the Airtable Cobuilder onboarding screen to describe the project they want to build. They select one or more purpose tags, optionally write a short description, open the company details modal from the bottom-left control, fill optional company fields, and then press Next to continue to the next onboarding step. → **UNPLACED** - `F-651d7d19bd` L601 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Select purpose and add description → Open company details modal → Fill company info and continue → **UNPLACED** - `F-7b9de7a91a` L603 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Airtable: AI-guided base generator** (web, 15 screens, Project Management, Collaboration Tools, Scheduling Tools, Paperwork, Productivity) [Refero](https://refero.design/flows/1540) → **UNPLACED** - `F-4760e1dc76` L605 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Happy-path: A signed-in user with access to an empty Airtable workspace uses the Cobuilder (AI) flow to generate a new base. → **UNPLACED** - `F-10acb240a4` L607 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Empty workspace → Start method chooser → Cobuilder: initial prompt → Select use-case tags → Company details modal open → Fill company details → Open color picker → Color plan limitation tooltip → Theme applied preview → Generation begins — creating tables → Generation continues — creating pages → Generation status with artifacts count → Review screen: 'How does this look?' → Creating specific object (modal overlay) → Generated base preview → **UNPLACED** - `F-e2722e599a` L611 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > The flow starts on Liam’s dashboard with the Navia assistant panel open on the right. The user selects a suggested question about how the content plan works, reviews Navia’s response, then selects a second suggested question about what happens after 30 days and reads the answer. The flow ends with the assistant remaining open and ready for another question. The user is assumed to be signed in, viewing the dashboard on a desktop-sized screen, and permitted to access the assistant. No account changes or external navigation occur. → **UNPLACED** - `F-2686c248b8` L613 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Open Dashboard Assistant → Select Plan Question → Read Plan Explanation → Choose Follow-up Topic → Wait for Follow-up Answer → Read Renewal Explanation → **UNPLACED** - `F-174362b404` L617 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > A visitor starts on the Care.com home screen and selects the option to find a caregiver. They choose pet care, provide timing, pet count, service type, dates, hours, and pay preferences, then optionally add notes about the pets. After submitting the request, they choose between premium and limited access and arrive at a filtered pet-caregiver search results screen. The screenshots show a happy path for a user seeking one-time cat sitting, with no sign-in step visible. → **UNPLACED** - `F-a1abddd3c7` L619 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Care.com Home → Choose User Path → Select Care Type → Pet Care Selected → Care Timing → Pet Count → Cat Count Dropdown → Pet Count Updated → Service Type → Service Selected → Schedule Details → Schedule Updated → Pay Range → Pay Range Updated → Caregiver Notes → Notes Guidance → Notes Entered → Choose Access → Caregiver Search Results → Caregiver Profile Preview → **UNPLACED** - `F-bd50fa9279` L621 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Lovable: Plan Review and Approval** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12337) → **UNPLACED** - `F-6e653b4f77` L623 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > A logged-in user is reviewing a proposed plan in a split-screen editor. The left side shows the task discussion and a plan card with a summary, while the right side shows the detailed plan content with a formatting toolbar. The user can inspect the proposal, optionally change the content type/format using the text style dropdown, and then either cancel or save and approve the plan. → **UNPLACED** - `F-6802052c6f` L627 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Cohere: Business Plan Outline Generation** (web, 2 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3827) → **UNPLACED** - `F-1949c6e09e` L629 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > A single conversational session where a logged-in user submits a prompt to generate a business plan outline in French and views the formatted result. → **UNPLACED** - `F-f7152fa479` L631 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Open conversation and submit prompt → Expose copy control for generated content → **UNPLACED** - `F-649d65b9d3` L633 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Cohere: Business plan assistant interaction** (web, 3 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3829) → **UNPLACED** - `F-57f2b04e66` L635 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Happy-path summary: A signed-in user is inside a chat conversation and asks the assistant for a business-plan outline in French. The assistant returns a formatted outline in the conversation pane. The user has two PDFs staged in the message composer (not yet sent). The user hovers or clicks on the assistant message actions (e.g., Regenerate) to request a different response. After requesting regeneration, the assistant posts a new reply (in this case an apologetic message saying it couldn't find a plan in the provided documents). → **UNPLACED** - `F-51bc091e43` L637 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: View generated business plan → Reveal message actions → Regenerated assistant reply → **UNPLACED** - `F-a37ef226e7` L641 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Happy path for leaving a review for a caregiver on Care.com. The user starts on a caregiver profile, opens the review form, selects an overall star rating, adds written feedback, answers whether they hired the caregiver through Care.com, submits the review, and reaches a confirmation screen with a call to action to continue browsing care options. → **UNPLACED** - `F-89d6462c67` L643 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Caregiver Profile → Start Review → Write Review → Review Ready to Submit → Submission Confirmed → **UNPLACED** - `F-751494f95c` L645 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Care: Pet Care Caregiver Search** (web, 12 screens, Job Search, Lifestyle) [Refero](https://refero.design/flows/12230) → **UNPLACED** - `F-36f280ebf2` L647 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Happy path for finding and narrowing pet care caregivers on Care.com. The user starts on the Pet care search results view with location, date, and basic search controls already set. They then refine the list using filter chips and dropdown panels (smart search, pay rate, pet type, rating, and professional skills), and the results list updates after each applied filter. The flow begins on the caregiver search results screen and ends on a filtered set of pet care profiles with a selected caregiver highlighted in the details panel, ready for contact. → **UNPLACED** - `F-738d6c57bf` L649 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Results Overview → Smart Search Entry → Smart Search Applied → Pay Rate Filter → Pay Range Updated → Rate Filter Applied → Pet Type Selection → Pet Type Applied → Rating Filter → Rating Applied → Professional Skills Filter → Filtered Profile Review → **UNPLACED** - `F-d69c86992e` L651 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **time2book: Create and publish a plan** (web, 15 screens, Business, Fitness, Productivity) [Refero](https://refero.design/flows/6309) → **UNPLACED** - `F-b081ad9379` L653 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Happy-path: A logged-in studio manager (desktop) opens Catalog > Plans, clicks New plan, completes the Add plan panel (general info, image, pricing type and amount, benefits, services with quantities), saves the plan, and returns to the Plans list where the new plan appears with a confirmation toast. Entry: Catalog > Plans (empty state). Exit: Plans list showing the created plan. Assumptions: user is authenticated with permission to manage catalog, there are existing services to attach to a plan, and file upload/image constraints are enforced by the interface. → **UNPLACED** - `F-06786180ee` L655 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Plans list (empty state) → Add plan panel (General section) → Fill name, description and upload image → Pricing defaults (Membership selected) → Switch to single payment → Open validity dropdown → Enter price and choose validity → Add plan benefits → Open Services section → Service selector and quantity → Service list dropdown → Select first service → Add multiple services with quantities → Save plan and confirmation → Plans list (post-save) → **UNPLACED** - `F-65ea8e2b56` L657 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Cohere: Business Plan Creation via Chat** (web, 6 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3791) → **UNPLACED** - `F-458d84afed` L659 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Logged-in user starts at the Cohere dashboard, navigates into the chat workspace and is shown an informational modal about multi-step tool use. They dismiss or accept the modal, explore prompt suggestion cards under "Just Chat," enter a custom prompt in the message composer asking for a business plan outline in French, submit it, and receive a formatted, multi-section French business plan in the conversation. → **UNPLACED** - `F-2121083421` L661 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Dashboard Landing → Onboarding Modal — Multi-Step Tools → Prompt Suggestions Panel → Compose and Submit Prompt → Generated Response — Part 1 → Generated Response — Continued / Completion → **UNPLACED** - `F-82579f65a0` L663 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Tripadvisor: Personalized London trip planner** (web, 15 screens, Travel & Booking) [Refero](https://refero.design/flows/7068) → **UNPLACED** - `F-aee8247ce6` L665 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Happy-path: A user on the "Things to Do — London" listing triggers the site’s AI-powered trip planner and answers a short multi-step questionnaire (dates → trip type → interests). After submitting, the system generates personalized recommendations and presents them as categorized cards plus a map. The user can open item details in a slide-over, save selections, choose to create an itinerary, name it, and then view the newly created itinerary — a day-by-day timeline with saved items and an interactive map. Entry: user arrives at London "Things to Do" content (logged-in state not required by the UI shown). Exit: a named itinerary is created and displayed with saved items mapped to trip days. → **UNPLACED** - `F-723fa1986e` L667 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Start on destination recommendations → Select trip dates (calendar) → Confirm date range selection → Choose trip type and pet option → Select interests (chips) → Open 'Add other interests' modal → Enter custom interest text → Submit interest selections → Recommendations loading → View generated recommendations → Open item detail slide-over → Choose how to continue (save vs create itinerary) → Name your trip → View created itinerary (timeline + map) → Edit and extend itinerary → **UNPLACED** - `F-7eee883dc6` L671 _EDGE Clinic research: real product flows from Refero › Task: case management_: **OpenAI: Prompt Evaluation Setup** (web, 13 screens, Business, Development Tools, AI Tool) [Refero](https://refero.design/flows/4223) → **UNPLACED** - `F-ba156c5225` L673 _EDGE Clinic research: real product flows from Refero › Task: case management_: > Happy path: A logged-in user opens the Evaluations area, creates a new evaluation by selecting a data source (Import Chat Completions / create data / upload file / custom), composes or imports test cases, selects a model to generate responses, adds one or more test criteria (e.g., valid JSON), reviews the evaluation configuration, runs it, and sees the run status and results saved in the evaluations list. Entry: Evaluations screen (empty or default). Exit: Saved evaluation entry and run result shown in the list. Assumptions: User is logged in to the project, has permission to run evaluations, has access to models referenced (e.g., gpt-4o-mini), and is on a desktop browser. Test data can be created manually or imported from stored completions or files. → **UNPLACED** - `F-1694ed4564` L675 _EDGE Clinic research: real product flows from Refero › Task: case management_: Steps: Evaluations list (empty) → Select data source → Import Chat Completions (instructions) → Create test data (blank form) → Create test data (sample content) → Create test criteria (empty) → Add testing criteria (chooser) → Configure 'valid JSON or XML' criterion → Criteria list updated → Review evaluation → Run in progress → Run completed (failed) → Evaluations list (with saved evaluation) → **UNPLACED** - `F-7c3494fa93` L677 _EDGE Clinic research: real product flows from Refero › Task: case management_: **Cohere: Account registration and onboarding** (web, 15 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3790) → **UNPLACED** - `F-1118316b1b` L679 _EDGE Clinic research: real product flows from Refero › Task: case management_: > Happy-path desktop web flow for a new user who discovers the product marketing site, creates an account, completes an onboarding questionnaire, and arrives at the product dashboard. → **UNPLACED** - `F-813eecea06` L681 _EDGE Clinic research: real product flows from Refero › Task: case management_: Steps: Marketing landing → Create account — entry form → Password strength — weak → Password strength — improving → Password strength — good → Password strength — strong → Verify your email → Onboarding — name collection (step 1/3) → Onboarding — name confirmed → Onboarding — role selection (step 2/3) → Role selected → Onboarding — use case selection (step 3/3) → Use cases selected with custom note → Dashboard — welcome → Dashboard — modules and docs → **UNPLACED** - `F-0caaf656c4` L687 _EDGE Clinic research: real product flows from Refero › Task: code repository pull request_: > A mobile referral/promotional modal that shows a scannable QR code, an explanatory headline and copy about the reward, the user's referral code, and two primary actions: Copy Code and Share Link. → **UNPLACED** - `F-2d39eb257e` L691 _EDGE Clinic research: real product flows from Refero › Task: code repository pull request_: **Ground News: Invite Friends - Referral Code** (ios, 2 screens, News) [Refero](https://refero.design/flows/7580) → **UNPLACED** - `F-04fab3ce46` L693 _EDGE Clinic research: real product flows from Refero › Task: code repository pull request_: > Logged-in user lands on the Invite Friends screen (accessed from Profile tab). The screen displays an explanation of the referral promotion, the user's unique invite code in a bordered input, a Copy control, and a large Share Referral Code button. Primary actions: copy the code to clipboard or open sharing. Secondary content: promotion terms and a link to Terms & Conditions. Exit points: user copies the code, shares the code (system share sheet or similar), navigates back via the back arrow, or taps another bottom nav item. → **UNPLACED** - `F-a4428cdc21` L697 _EDGE Clinic research: real product flows from Refero › Task: code repository pull request_: **Leonardo.Ai: Copy Generation API Code** (web, 4 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/12888) → **UNPLACED** - `F-a1fc28ae62` L699 _EDGE Clinic research: real product flows from Refero › Task: code repository pull request_: > Logged-in user on Leonardo.ai’s AI Creation screen generates or reviews an audio asset, opens the item actions menu, chooses the option to view the Generation API Code, and then copies that code from a modal. → **UNPLACED** - `F-e6e8d0cf7e` L701 _EDGE Clinic research: real product flows from Refero › Task: code repository pull request_: Steps: AI Creation Workspace → Actions Menu → Generation API Code Modal → Copied Confirmation → **UNPLACED** - `F-655651ec09` L703 _EDGE Clinic research: real product flows from Refero › Task: code repository pull request_: **Raycast: Question search and select** (ios, 2 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/10193) → **UNPLACED** - `F-69f5276454` L705 _EDGE Clinic research: real product flows from Refero › Task: code repository pull request_: > User is on a mobile "Questions" screen that shows a searchable list of question titles. The header shows a rounded search input with placeholder text and a numbered list of question rows; each row has a right chevron indicating it leads to details. Happy path: user taps the search field to open the keyboard, types a query, the list filters in real time to matching questions, then the user taps a row to navigate to that question's details (or taps the X control to dismiss). → **UNPLACED** - `F-4dd243497c` L711 _EDGE Clinic research: real product flows from Refero › Task: compliance dashboard audit_: **Lovable: Generated Landing Page Review** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12309) → **UNPLACED** - `F-9930646f6f` L713 _EDGE Clinic research: real product flows from Refero › Task: compliance dashboard audit_: > This flow shows a user reviewing an AI-generated SaaS landing page inside a split workspace. The user starts from a prompt describing the desired landing page, then inspects the generated result in a live preview on the right while the prompt/result card remains on the left. The user can bookmark the output to history and optionally give feedback via reaction controls. The flow appears to assume the user is already signed in or has access to the workspace and preview environment. It ends with the generated landing page visible and a confirmation message that the result was saved/bookmarked or feedback was submitted. → **UNPLACED** - `F-e026d57141` L717 _EDGE Clinic research: real product flows from Refero › Task: compliance dashboard audit_: **Foodllama: Nutrition and body stats review** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/11789) → **UNPLACED** - `F-10e235391f` L719 _EDGE Clinic research: real product flows from Refero › Task: compliance dashboard audit_: > Happy path starts on the main Foodllama dashboard, where the user can review daily calorie intake, macro distribution, streak status, and weekly trend data. From there, the user taps a profile/avatar area to open a detailed stats view with weight, calorie goal, and BMI information, then can move between avatar/profile variations and return to the dashboard. → **UNPLACED** - `F-9c0353da85` L721 _EDGE Clinic research: real product flows from Refero › Task: compliance dashboard audit_: Steps: Dashboard Overview → Detailed Stats View → Avatar Selection State → Dashboard with Selected Avatar → **UNPLACED** - `F-9106bac3c8` L725 _EDGE Clinic research: real product flows from Refero › Task: connect bank account open banking_: **Fingerprint: Find and open account-related documentation** (web, 4 screens, Business, Development Tools) [Refero](https://refero.design/flows/11143) → `/finance/open-business-account` - `F-8ed9197ad4` L727 _EDGE Clinic research: real product flows from Refero › Task: connect bank account open banking_: > Happy path: a user lands on the Fingerprint Docs home screen, opens the site-wide search/Ask AI modal, searches for "account" (or a related term), and selects an "Account Sharing" / "Account Takeover" article. The flow ends with the full documentation article loaded, including code examples and step-by-step instructions. → `/finance/open-business-account` - `F-60f9b5eac7` L729 _EDGE Clinic research: real product flows from Refero › Task: connect bank account open banking_: Steps: Docs Home / Discover → Search Modal (empty) → Search Results → Documentation Article → `/finance/open-business-account` - `F-92b21ec032` L735 _EDGE Clinic research: real product flows from Refero › Task: contract management_: > A mobile (iPhone) list of subscription cards where each card represents a recurring item (podcast, news subscription, website). → `/finance/open-business-account` - `F-8590310183` L737 _EDGE Clinic research: real product flows from Refero › Task: contract management_: Steps: List — default state → Reveal destructive action → List — updated / scrolled view → `/finance/open-business-account` - `F-6b8fbd0bdf` L739 _EDGE Clinic research: real product flows from Refero › Task: contract management_: **Google Gemini: Report legal content from Gemini chat** (ios, 18 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/3187) → **UNPLACED** - `F-bbfe5e826f` L741 _EDGE Clinic research: real product flows from Refero › Task: contract management_: > Happy path: A signed-in user starts in a Gemini conversation, opens the message actions menu on an assistant reply, and chooses to report the content. The app links to Google Legal Help. The user selects a legal reason, taps Create request, completes the online ‘Report a legal removal issue’ form (country, personal details, contact email), attaches screenshots/files from the device photo picker, provides a written explanation, confirms the perjury declaration, types a signature (digital), and submits. The flow ends with a submission confirmation screen telling the user they will receive a confirmation email. Assumptions: user is signed in (avatar visible), using a mobile browser/webview on iOS, has the required evidence files in their photo library, and can receive email confirmations. → **UNPLACED** - `F-5b3fba6100` L743 _EDGE Clinic research: real product flows from Refero › Task: contract management_: Steps: Gemini conversation - message actions visible → Message options menu → Legal Help landing → Select reason to report → Create request CTA → Form – country selector (tap) → Form – country list (choose) → Form – fill identity fields → Allegedly infringing content section → Photo picker – browse library → Photo picker – confirm selection → Files attached in form → Explain why content is unlawful → Perjury declaration checkbox → Signature and date → Submit request → Submission success → End state / wait for confirmation → **UNPLACED** - `F-b0a77d3f07` L747 _EDGE Clinic research: real product flows from Refero › Task: contract management_: > Happy path begins in the Discover area, where a user searches for advice about arguing a first court case. The system returns matching expert profiles and excerpts. The user opens a conversation with Jason Jones, reviews his contextual response, chooses suggested follow-up questions, attaches a PDF, asks for feedback, records and sends a voice message, and continues the conversation based on the expert’s responses. The flow ends with another assistant response and additional suggested questions. The user is assumed to be signed in, using a desktop browser, and permitted to search experts, start conversations, upload files, and record audio. → **UNPLACED** - `F-b9918e113a` L749 _EDGE Clinic research: real product flows from Refero › Task: contract management_: **Cohere: Document-based conversation review** (web, 3 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3832) → **UNPLACED** - `F-b5670f19ec` L751 _EDGE Clinic research: real product flows from Refero › Task: contract management_: > A desktop chat workspace where a user interacts with an AI assistant to extract or summarize information from uploaded documents. → **UNPLACED** - `F-a3b423b6b9` L753 _EDGE Clinic research: real product flows from Refero › Task: contract management_: Steps: Conversation view with attachments visible → Activate collapse control for left panel → Expanded conversation (left panel collapsed) → **UNPLACED** - `F-d2bb7b18a3` L757 _EDGE Clinic research: real product flows from Refero › Task: course enrolment university_: **WhatsApp: Add Community Member** (ios, 5 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12678) → `/clinic/social-prescribing-referral` - `F-644e72ea59` L759 _EDGE Clinic research: real product flows from Refero › Task: course enrolment university_: > Happy path in a community management area. The user starts on the Community info screen as the community owner, opens Add members, selects one contact, confirms the add, and returns to Community info where the member count updates and a success message appears. → `/clinic/social-prescribing-referral` - `F-154157badd` L763 _EDGE Clinic research: real product flows from Refero › Task: course enrolment university_: > A logged-in user opens a wellness program detail screen for a prenatal yoga series. The entry point is the program landing page with a hero image, program title, instructor, and tabs for Overview and Classes. The user can review the description, preview the series video, inspect instructor information, browse the full list of classes, view similar programs, and add a comment. In the second state, the user is shown as in-progress with completion tracking and a Stop Program control, indicating they have already started the series. The flow ends when the user leaves the program page, starts or continues a class from the series, schedules a class, or posts a comment. Assumption: the user is signed in and has access to the program content on desktop web. → `/clinic/social-prescribing-referral` - `F-d720b74930` L769 _EDGE Clinic research: real product flows from Refero › Task: credit score report_: **Lovable: Referral Invite Modal** (web, 2 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12350) → **UNPLACED** - `F-d2158f6d40` L771 _EDGE Clinic research: real product flows from Refero › Task: credit score report_: > A centered referral modal opens over a dimmed dashboard/sidebar view. The user is shown a shareable invite link and QR code with instructions for earning credits through referrals. The intended entry point is a referral/share action from within the product, and the exit point is either closing the modal or copying the invite link. → **UNPLACED** - `F-06c28b1aab` L777 _EDGE Clinic research: real product flows from Refero › Task: credit score report_: > A logged-in user views a referral card screen for sharing an invite link or QR code. The entry point is likely from a rewards/referral section, and the exit points are back navigation, copying the referral link, downloading the card, or opening the system share sheet. The screen assumes the user already has an active account and referral code available. → **UNPLACED** - `F-7a5b8653dd` L783 _EDGE Clinic research: real product flows from Refero › Task: credit score report_: > A music biography screen shows a long, scrollable artist article in a dark theme. The user enters from an artist or music profile context and can read the biography, follow linked artist/song names, and scroll through the full text. A back button returns to the previous screen. A persistent mini player stays docked at the bottom, allowing playback control while reading. → **UNPLACED** - `F-76163c564d` L791 _EDGE Clinic research: real product flows from Refero › Task: dispute transaction chargeback_: > End-to-end happy-path for reporting a problem against a transaction from the activity/transaction detail screen. → **UNPLACED** - `F-afa4d37339` L793 _EDGE Clinic research: real product flows from Refero › Task: dispute transaction chargeback_: Steps: Transaction detail → Choose issue category → List items — empty → List items — filled → Pick issue — primary options → Pick issue — subtype selection → Provide other details — empty → Date picker → Provide other details — filled → Seller contact — asked → Seller contact — selection → Add documents — empty → Add documents — uploaded + notes → Review entered info → Confirmation / Case filed → **UNPLACED** - `F-3081fb8025` L797 _EDGE Clinic research: real product flows from Refero › Task: dispute transaction chargeback_: > Seller views the Sales list, opens an individual sale in a right-hand details panel, attempts to refund the purchase using a Refund control which asks for confirmation, and the system returns an error banner reporting insufficient balance. → **UNPLACED** - `F-3a6158ebb3` L799 _EDGE Clinic research: real product flows from Refero › Task: dispute transaction chargeback_: Steps: Sales list → Sale details panel → Refund confirmation modal → Insufficient balance error → **UNPLACED** - `F-57bb0e9200` L801 _EDGE Clinic research: real product flows from Refero › Task: dispute transaction chargeback_: **Trip: Canceled Booking Refund Support** (web, 8 screens, Travel & Booking) [Refero](https://refero.design/flows/12086) → `/clinic/mental-health-care-crisis` - `F-3b231b53ac` L803 _EDGE Clinic research: real product flows from Refero › Task: dispute transaction chargeback_: > Happy path starts on a canceled hotel booking details screen for a logged-in Trip.com user. The user reviews cancellation and refund information, then opens Hotels & Homes Service Chat from the booking context to ask about refund timing and what to do if the money has not returned. The chat shows an automated response with refund details, the user follows up with an additional question, and the system returns more guidance about contacting the hotel if needed. The flow ends in the chat with support information visible; no booking changes are made. → `/clinic/mental-health-care-crisis` - `F-e5f3933ae0` L805 _EDGE Clinic research: real product flows from Refero › Task: dispute transaction chargeback_: Steps: Canceled Booking Details → Service Chat Opens → Refund Question Sent → Refund Answer Returned → Follow-up Question → Follow-up Processing → Refund Escalation Guidance → Additional Deposit Advice → `/clinic/mental-health-care-crisis` - `F-236677b0b8` L809 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: **District: Restaurant Question Assistant** (ios, 5 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12847) → **UNPLACED** - `F-4ab02c0a45` L811 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: > Happy-path mobile assistant experience for asking a restaurant-related question and continuing with a follow-up. Entry is a restaurant detail or assistant overlay with an active query context (“Asking about Hotel Rishi”). The user asks a question by voice, the system transcribes and processes it, shows an answer with suggested next actions, and keeps a persistent follow-up input available. The flow ends when the user has a ready-to-send follow-up request, such as asking for nearby breakfast options. → **UNPLACED** - `F-5524190ef3` L813 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: **Ada: Dizziness results and condition detail** (ios, 5 screens, Health & Wellness) [Refero](https://refero.design/flows/4103) → **UNPLACED** - `F-bb48ed43ef` L815 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: > Happy-path: a mobile user who completed a symptom assessment for dizziness views the results list, opens the top suggested condition (Concussion), and reads detailed, structured information via expandable sections. → **UNPLACED** - `F-d530a16288` L817 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: Steps: Results — Dizziness Summary → Condition Preview — Concussion → Full Detail — Overview → Expand Section — Risks → Multiple Sections — Read Diagnosis & More → **UNPLACED** - `F-c5d288047b` L819 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: **Ada: Condition Search and Results** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/4125) → `/clinic/search-recruiting-trials` - `F-4d6bb2955c` L821 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: > User opens the Condition Library (on a mobile device, likely signed-in since profile-specific recommendations appear) to find information about health conditions. The happy path: user scans the curated lists (Most read / Mentioned in your latest assessments), taps the search field, types a query, and either selects a result or sees an empty-state message if there are no matches. → `/clinic/search-recruiting-trials` - `F-ca9d22f1b5` L823 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: Steps: Library listing → Search focused (keyboard visible) → Search submitted — no matches → `/clinic/search-recruiting-trials` - `F-0b85ef404f` L825 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: **Ada: Pain behind eye question** (ios, 2 screens, Health & Wellness) [Refero](https://refero.design/flows/4101) → **UNPLACED** - `F-7647cb23a7` L827 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: > Happy path: A mobile user opens a new assessment and is presented with a single symptom question: "Do you have pain behind the eye?" The user can either answer directly (Yes / No / I don't know) or tap the inline help link to view a full explanation (a modal/overlay) showing a short definition and an anatomical illustration. From the help overlay the user can return to the question using a prominent confirmation button or indicate they still don’t understand. → **UNPLACED** - `F-1a9d901896` L829 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: Steps: Question: Do you have pain behind the eye? → Info overlay: Pain behind the eye → **UNPLACED** - `F-f93217fce5` L831 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: **Ada: Neck mobility question and clarification** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/4102) → **UNPLACED** - `F-1b86dd662e` L833 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: > Happy path: A user working through a new medical assessment is asked a single symptom question about neck mobility. From the question screen they can either answer directly (Yes / No / I don't know) or open an inline help modal labeled “What does this mean?” The help modal provides a short definition and anatomical illustrations (front and side). The user indicates they understand the explanation and returns to the question to answer. Entry: user is mid-way through a newly started assessment. Exit: modal closed and question answered (moves to the next assessment step). Assumptions: user is on a mobile device (portrait), has already started a “New assessment” session, and can interact with modal overlays and standard controls. No special permissions or account state are required based on visible UI. → **UNPLACED** - `F-2635cb7a05` L835 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: Steps: Neck mobility question → Clarification modal — definition → Clarification modal — additional illustration → **UNPLACED** - `F-778a3e34c4` L837 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: **Ada: Condition lookup and article view** (ios, 5 screens, Health & Wellness) [Refero](https://refero.design/flows/4126) → `/clinic/record-lawful-basis` - `F-c10d683803` L839 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: > Mobile Condition Library — happy path: user opens the Conditions tab (bottom navigation), lands on the Condition Library list and either browses the pre-populated sections (Most read / Mentioned in your latest assessments) or activates the search field to find a specific condition. Typing a query shows live search results; selecting a result opens a condition article with an overview and expandable sections (Risks, Symptoms, Diagnosis, Treatment, Prevention). → `/clinic/record-lawful-basis` - `F-e9cc7326dc` L841 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: Steps: Library landing → Search activated → Partial query / live results → Browse full results → Condition detail → `/clinic/record-lawful-basis` - `F-4cb9021c42` L843 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: **Doctronic AI: Clinical Summary Feedback** (web, 4 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13701) → `/clinic/person-held-health-record` - `F-f083a20e1f` L845 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: > A user reviews an AI-generated clinical summary and provides product feedback. The flow begins on the summary view, where the user can rate how useful the summary was. The user then enters written feedback in an improvement form and submits it. After submission, the system confirms that the feedback was received. A chat input remains available throughout for asking additional questions. → `/clinic/person-held-health-record` - `F-f4921b4803` L847 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: **Parallel: Research Chat with Sourced Answers** (web, 8 screens, AI Tool) [Refero](https://refero.design/flows/9073) → **UNPLACED** - `F-c4afe719a2` L849 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: > A desktop Chat API research interface where a logged-in user composes web-research questions, receives summarized answers with source links, and can change the chat model to prioritize speed or factual citations. → **UNPLACED** - `F-22579f99eb` L851 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: Steps: Chat home / Prompt templates → Submit initial question → Compose follow-up → Follow-up visible while waiting → Open model selector → Submit a data-focused follow-up with new mode → Receive detailed, sourced results → Consolidated summary and source list → **UNPLACED** - `F-31a6bc8c5e` L853 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: **WhatsApp: Switch to Video Call** (ios, 4 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12665) → **UNPLACED** - `F-0be7d7ea12` L855 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: > Happy path for an ongoing 1:1 call with Samuel on a mobile device. The user starts on an active voice call, chooses to switch to video, confirms the change in a modal, and then reaches the video call state. → **UNPLACED** - `F-33ea36d87a` L857 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: **WhatsApp: Active Call Controls** (ios, 2 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12663) → **UNPLACED** - `F-315fe58c49` L859 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: > An active voice/video call screen with a contact named Samuel. The user is already in an ongoing call; the screen shows the call timer, a large contact avatar placeholder, quick actions on the right, and a bottom control bar. → **UNPLACED** - `F-4a1b5ec85d` L861 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: **WhatsApp: Video Message Editing** (ios, 2 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12646) → **UNPLACED** - `F-e25fdf5be3` L863 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: > A mobile media composition screen for preparing a short video before sending it in a chat. The user enters from a selected video, can edit it with tools like trim/crop/rotate/text and AI restyle, optionally mute or toggle GIF/video mode, add a caption, choose a recipient, and then send. → **UNPLACED** - `F-0334edc756` L865 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: **ElevenLabs: Style selection & generate video** (ios, 2 screens, AI Tool) [Refero](https://refero.design/flows/10737) → **UNPLACED** - `F-fc619fccf6` L867 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: > Mobile screen where the user previews a text/voice card and chooses a background style before creating a video. Entry: user arrives with a prepared text/voice clip (voice shown as 'Rob') and a card preview. Primary controls: close (top-left), three background style swatches (Multicolor, White, Black), and a primary 'Generate video' CTA. Exit: user taps Generate video to initiate video creation. Assumption: user is already at the create/preview step (logged-in state not shown); device is a smartphone. → **UNPLACED** - `F-158a6adc58` L871 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: **WhatsApp: Send Recorded Video** (ios, 11 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12650) → **UNPLACED** - `F-091f0654a8` L873 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: > A user is viewing an existing conversation with Samuel and sends a newly recorded video. The user opens the attachment menu, launches the camera, switches to video mode, grants microphone access when prompted, records a short clip, reviews it, and sends it to the conversation. The sent video then appears in the chat and can be opened for playback. The user is assumed to be on a mobile device with the conversation already open. Camera access appears available; microphone access requires approval. → **UNPLACED** - `F-ffa7df87c8` L875 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: **WhatsApp: Group Voice Conversation** (ios, 6 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12680) → **UNPLACED** - `F-b319ac0566` L877 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: > Happy path for joining and using a group voice conversation inside a community chat titled "General." The entry point is an existing chat thread where the user can read messages, add members, or start talking. The user begins by holding the talk control, then transitions into a live voice room connection where participants are shown and call controls appear. The exit point is the expanded in-call voice interface with connected participants and controls for speaker, mute, reactions, and ending the session. → **UNPLACED** - `F-ec8933ce84` L879 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: **WhatsApp: WhatsApp Call Transition** (ios, 9 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12661) → **UNPLACED** - `F-9544fe5487` L881 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: > Happy path for a WhatsApp conversation that transitions from chat to audio call and then to video call. → **UNPLACED** - `F-1beeb25baf` L883 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: **Weekrise: Embedded support chat conversation** (web, 7 screens, Collaboration Tools, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/1296) → **UNPLACED** - `F-0509204c8c` L885 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: > Happy-path interaction with an embedded live-chat widget on a desktop landing screen. → **UNPLACED** - `F-b9db2acce5` L887 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: Steps: Open chat / initial greeting → Compose and send question → Email capture prompt → Offline / slow-reply fallback → Record voice message → Upload in progress → Audio playback available → **UNPLACED** - `F-fe59925f0a` L889 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: **Anam: Live Conversation Session** (web, 4 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12945) → **UNPLACED** - `F-f853b9991d` L891 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: > This flow begins in a live conversation experience, first showing a call startup/loading state, then an active video call with a remote speaker and a self-view thumbnail. During the call, the user can open device/audio controls, give feedback, or end the call. The flow ends on a post-call confirmation screen with options to build an interactive avatar or submit feedback. → **UNPLACED** - `F-06605f4ff7` L893 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: Steps: Call Starting → Active Call → Microphone Settings → Call Ended → **UNPLACED** - `F-9640e7f487` L897 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: > The flow begins in an existing conversation with a digital representation of Jim Goetz. The user initiates a voice call using the green call control, waits while the call connects, and then interacts with Jim’s digital mind by voice. The call displays mute and end controls, shows the conversation as it progresses, and ends on a call summary state with options to view the transcript, share the call, return to chat, or start another call. The user is assumed to be in an active conversation and authorized to call this digital persona on a desktop device. → **UNPLACED** - `F-ac9f23c185` L899 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: **Voicenotes: Voice query to notes** (web, 5 screens, Collaboration Tools, AI Tool) [Refero](https://refero.design/flows/6977) → `/judiverse/judebox-messages` - `F-2095e5964f` L901 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: > User opens the 'Ask' modal to query their notes using a voice recording. The modal shows an Ask/Create toggle, a notes filter (All notes), a prompt area labeled "Ask a question" with a mic icon, and a prominent floating Record control. Happy path: user opens the Ask interface, starts recording, stops the recording (Done), the system transcribes the audio and displays the recorded audio/transcription as a message bubble, and then the assistant posts a textual reply referencing the user's notes. → `/judiverse/judebox-messages` - `F-97c5d20d77` L903 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: Steps: Open Ask modal → Start recording → Stop recording → Transcription / audio message appears → Assistant reply displayed → `/judiverse/judebox-messages` - `F-b0bbfc9869` L907 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: > A user reviews a generated video inside a workspace and completes a sequential feedback questionnaire. The video remains visible in the main viewer while a dark survey card presents one question at a time in the lower-left corner. The user rates overall video quality, avatar realism, voice audio, graphics and visuals, and pacing and transitions using five-point scales, then enters one suggested improvement in a text field and submits it. The flow begins with the video open in the Videos area and ends after the improvement response is entered and submitted. → `/judiverse/judebox-messages` - `F-9a74610c8e` L909 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: Steps: Overall Video Quality → Avatar Realism → Voice Audio Quality → Graphics and Visuals → Pacing and Transitions → Improvement Suggestion → Review Written Feedback → Submit Feedback → `/judiverse/judebox-messages` - `F-a40e7ff0c9` L911 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: **Doctronic AI: Consultation Sharing** (web, 3 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13695) → **UNPLACED** - `F-daaed3c1da` L913 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: > A user is viewing an ongoing health consultation and opens the sharing dialog to share the current message or the entire consultation. The user keeps the default “Message Only” option, chooses “Copy Link,” and receives confirmation that the link was copied to the clipboard. The sharing dialog indicates that links expire after five days and can be regenerated. The user is assumed to be in an active consultation on a desktop-sized device. → **UNPLACED** - `F-c1ad1c1451` L915 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: **Telegram: In-Chat Voice Call** (web, 4 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/571) → **UNPLACED** - `F-7d5c4e62e7` L917 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: > User is in a one-to-one conversation with a contact (Henry) on a desktop messaging client. From the open chat the user initiates a call using the header call control. A centered call card appears showing 'Ringing...' while the call is being placed and exposes inline controls (camera, screen, mute, decline). When the call connects a green active-call bar and a running timer appear; the call card updates to show the elapsed time and an End control. After ending the call the conversation is returned to and the chat thread (and left-hand conversation list) contains outgoing-call messages that show the call durations. Assumptions: user is signed in and in the specific one-to-one chat; device is desktop with enough screen width to show two-column layout (conversation list + chat). Entry: open chat with contact. Exit: chat updated with call records and user back to messaging view. → **UNPLACED** - `F-27fb83846d` L919 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: Steps: Open Conversation → Ringing Call Card → Connected Call (Active) → Return to Chat with Call Record → **UNPLACED** - `F-7e47cf9278` L921 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: **Cushion: Direct Message Conversation** (web, 4 screens, Communication Tools, Collaboration Tools) [Refero](https://refero.design/flows/12528) → **UNPLACED** - `F-c74a24b70a` L923 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: > A workspace messaging interface where the user starts a new direct conversation, sends an initial message, then continues the thread with a reply and an attached image. → **UNPLACED** - `F-ab6c75877c` L925 _EDGE Clinic research: real product flows from Refero › Task: doctor video consultation_: Steps: New Chat Composer → Initial Message Draft → Conversation Thread → Reply With Attachment → **UNPLACED** - `F-b60df1af3f` L931 _EDGE Clinic research: real product flows from Refero › Task: expense management receipts_: > The user opens a Share sheet for a purchase record and can copy the displayed social-media tag before sharing the receipt image. The flow begins on a mobile device with a share modal already open over a purchase detail screen and ends with confirmation that the tag was copied. The user appears to be viewing an existing receipt or transaction; no login, permissions, or account state are visible. The final Share button remains available but is not shown as selected. → **UNPLACED** - `F-a6efe11f67` L937 _EDGE Clinic research: real product flows from Refero › Task: expense management receipts_: > Logged-in desktop admin interface on the Experiences section. The user starts on an empty Manage categories screen, opens a modal to create a category, enters a category name, and saves it. After saving, the modal closes and the new category appears in the list with an experience count and overflow actions. → **UNPLACED** - `F-ceb9a31130` L939 _EDGE Clinic research: real product flows from Refero › Task: expense management receipts_: Steps: Empty Categories State → New Category Modal → Category Name Entered → Categories List → **UNPLACED** - `F-b3e5c045f3` L945 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: > A user is viewing a document through Google Lens with text recognized and translated from English to Spanish. The screen shows the live camera/document view with translated text overlays, a language selector at the top, and a bottom sheet offering actions for the translated text. The user is assumed to be on a mobile device, already inside Google Lens, with camera access granted. The flow starts on the translated document view and ends after the user dismisses or responds to the feedback prompt. → **UNPLACED** - `F-fb1962e86d` L951 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: > Happy path: A user starts a new assessment and is prompted to enter the symptom that troubles them most. Tapping the input opens a modal-style symptom search with the keyboard. As the user types, autocomplete suggestions appear. The user selects a suggestion to view a symptom detail card, then confirms selection with a ‘Select symptom’ button. → **UNPLACED** - `F-41bfd8031c` L953 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: Steps: Start assessment — symptom prompt → Symptom search modal (empty) → Autocomplete suggestions → Symptom detail and confirm → **UNPLACED** - `F-48ec571bcf` L957 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: > A logged-in neighbor reviews an AI-generated answer in the Ask section and rates it as helpful. The user starts on an answer about places to learn hobbies, selects the thumbs-up feedback control, and receives a confirmation toast. The answer remains visible, and the selected feedback state is highlighted. → **UNPLACED** - `F-b5edce3597` L963 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: > Happy-path: A learner on a scheduled lesson landing screen runs through an in-browser device check (camera, microphone, speakers, network) before entering the classroom. → **UNPLACED** - `F-34f592add7` L965 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: Steps: Lesson lobby with test prompt → Open audio & video test modal → Recording in progress → Visual confirmation prompt → Audio playback (speaker) confirmation → Test your devices overview → Network and device start screen → Network/device test in progress → Test results — success → Checklist with checks completed → Return to lesson lobby (ready state) → **UNPLACED** - `F-f42d14eb6d` L969 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: > Happy path: A learner arrives at a scheduled lesson landing screen and runs the built-in audio & video diagnostics before entering the classroom. From the landing area the user opens a modal that guides them to record a short (5-second) test video, watches a recording progress indicator, then confirms two simple checks: can they see themselves and can they hear the playback. The user can rerun the test and then view the consolidated "Test your devices" panel which shows device check items (Camera and mic, Network and device). → **UNPLACED** - `F-64a6121eaf` L971 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: Steps: Lesson landing with test prompt → Start test modal → Recording in progress → Playback – visual confirmation → Playback – audio confirmation → Device checklist overview → **UNPLACED** - `F-8953ee8ec2` L973 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: **DailyArt Magazine: Image-based quiz with results** (web, 7 screens, Education, Magazines & Newspapers, Marketplace) [Refero](https://refero.design/flows/3501) → **UNPLACED** - `F-bcde4d15cb` L975 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: > Happy-path: A desktop user navigates from a Quizzes listing page to a specific quiz article. The article shows a hero image, intro copy and a sequence of 10 image-based multiple-choice questions. Each question displays a large image, a short question label, a set of radio-button answers and a primary action button that advances to the next question. After the final question the primary action changes to reveal results. → **UNPLACED** - `F-6d02d2d3de` L977 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: Steps: Quizzes listing → Quiz article — start & first question → Question 2 — answer selection → Mid-quiz questions → Late quiz question (8/10) → Final question and reveal action → Results / answers summary → **UNPLACED** - `F-e85483abc9` L981 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: > Happy-path practice sequence in the Duolingo English Test practice area. The user starts on the Practice dashboard, launches a practice test, waits through a loading state, answers a timed word-recognition question, receives either incorrect or correct feedback, advances through the question, and returns to the Practice dashboard with progress updated. → **UNPLACED** - `F-28e4953f8d` L983 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: Steps: Practice Dashboard → Loading State → Word Decision Question → Incorrect Feedback → Next Question → Correct Feedback → Practice Complete → Practice Dashboard Return → **UNPLACED** - `F-fd4e8d93bb` L985 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: **TikTok: Visual product search camera** (ios, 3 screens, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/2973) → **UNPLACED** - `F-703d19e820` L987 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: > Mobile visual-search entry point that lets a user take a photo or upload an existing image to find visually similar products. Entry: user opens the visual search camera view. Primary actions: take a photo (shutter), upload from photo library (Upload), or read contextual help (Help). Secondary controls include Flash toggle and a Close (X). Help appears as a two-page bottom sheet with short guidance for taking or uploading images. Exit: after capturing or selecting an image the flow should proceed to a results/search step (not shown). Assumptions: device has camera/photo permissions, product is shown centered on a plain background, user is on a mobile device. → **UNPLACED** - `F-4c133a2c0a` L989 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: Steps: Camera / Visual Search → Help — Take a photo → Help — Upload a photo → **UNPLACED** - `F-06336609ed` L993 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: > Happy-path: A user discovers the "Search by Seeing" feature on the app home feed, opens the in-app camera experience (AI Snaps), selects a product recognizer snap, points the camera at a product, captures an image, and is taken to a generated result page containing a review, photo gallery, price cards, expert commentary and suggested follow-up questions. Entry: Home feed. Exit: Detailed result/article card with share and follow-up options. Assumptions: User is on a mobile device, has camera access, and has network connectivity. No special permissions or account state are visible/required in the screenshots. → **UNPLACED** - `F-d376ad2fa1` L995 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: Steps: Home — Search by Seeing → Camera — AI Snap Selector → Camera — Capture → Results — Article Header → Results — Visual Gallery & Price Cards → Results — Expert Review Content → Results — Actions & Suggested Questions → **UNPLACED** - `F-9725bd269a` L997 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: **Sky Guide: Tonight's Best — Observation Mode Selector** (ios, 3 screens, Education, Reference) [Refero](https://refero.design/flows/8478) → **UNPLACED** - `F-a53c38214e` L999 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: > Happy-path: User opens the "Tonight's Best" list (mobile, portrait) and sees a scrollable list of recommended celestial objects. Each row shows an icon, object name, object type (star, planet, etc.) and visibility window. A prominent center pill at the bottom shows the current observation mode (e.g., "Naked Eye"). Tapping that pill opens a small modal with three instrument options (Telescope, Binoculars, Naked Eye). Choosing an option updates the center pill and refreshes the list to show objects appropriate for that instrument. Entry: user already in the app and viewing "Tonight's Best." Exit: user either scrolls the updated list, taps the back arrow to return to the previous screen, or reopens the instrument picker to change mode. Assumptions: mobile device in portrait; user has permission to view location-based visibility times; the list content is filtered client-side and updates immediately when mode changes. → **UNPLACED** - `F-e166405ece` L1001 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: Steps: Tonight's Best List (Naked Eye) → Observation Mode Picker → Tonight's Best List (Telescope) → **UNPLACED** - `F-d9a7ae93dc` L1005 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: > Mobile show episode list with an in-show search experience. Entry is from a show’s episode list, where the user taps the search icon to open a dedicated search screen for that show. The user types a query in the episode-title field and the list filters in real time. Results count updates as matches change, and an empty state appears when no episodes match. → **UNPLACED** - `F-ddee56d66f` L1007 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: Steps: Episode List → Search Entry → No Results → Filtered Results → **UNPLACED** - `F-2d5a0f1e73` L1011 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: > Happy-path: A signed-in mobile user with camera and photo-library permissions starts on the app home / discovery screen, taps the camera/Explore action, selects an AI Snap (Destiny Mirror) from the filter carousel, either takes a live photo or picks one from the photo library, lets the app analyze the image (face/keypoint detection), and receives a generated interpretive report ("Today's card: The Silver Seraph") with expandable sections, suggested follow-up questions, and share/export actions. → **UNPLACED** - `F-82a84e725d` L1013 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: Steps: Home / Discover → Live Camera with Snap Carousel → Photo Library Picker → Image Loading & Face Keypoint Overlay → Preview with Snap Applied → Initial Result / Contextual Prompt → Full Report with Sections → Export & Follow-up Interaction → **UNPLACED** - `F-4849101a65` L1015 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: **MyMind: Visual Reference Search** (web, 3 screens, Productivity, Reference, Inspiration Board) [Refero](https://refero.design/flows/13313) → **UNPLACED** - `F-7ce8cd3980` L1017 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: > A user starts on a personal visual search dashboard with content grouped across images, books, web pages, and videos. They enter a search term, first producing an empty state for "00000," then replace or refine the query with "or" to view matching content. The flow assumes the user is already in the search workspace and can access indexed personal or curated references. It ends on a results view showing tags and a masonry-style collection of books, artwork, design references, and video content. → **UNPLACED** - `F-0a3f6162db` L1021 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: **Dropbox: Video Review and Viewing Controls** (web, 4 screens, Collaboration Tools, Business, Productivity) [Refero](https://refero.design/flows/13044) → **UNPLACED** - `F-2afabda4d8` L1023 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: > A user is viewing a video in a file preview/player and uses the top toolbar and player controls to inspect the media. The entry point is a paused video in the standard viewer, with options like Open in, Edit, Transcript, Share, comments, and tools. The user can open the tools menu to show comments, show transcript, hide tools, or switch to full screen. The main interaction shown is toggling the viewer chrome so the video can be watched with fewer distractions. The flow ends with the video remaining ready to play, either in the compact viewer or full-screen mode, with tools hidden or shown as needed. → **UNPLACED** - `F-beee5f7caa` L1027 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: > Happy-path: The user reviews a freshly generated short video in a preview player, is prompted to rate the result, and chooses to send structured feedback via a modal. → **UNPLACED** - `F-1f848c367f` L1029 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: Steps: Preview and rating prompt → Feedback modal — choose issues → Feedback modal — submit → **UNPLACED** - `F-8a96419af5` L1031 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: **Melius: Skincare Hero Image Generation** (web, 3 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/14070) → **UNPLACED** - `F-e69bb4fdda` L1033 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: > The user reviews an AI-generated commercial skincare product image inside a canvas workspace. The flow starts with a conversational prompt referencing an attached product image, continues through a source image node and generated result, and ends with the user reviewing the output and optionally rating the response. The user appears to be signed in and working in an existing canvas on a desktop device. No additional permissions or account setup are shown. → **UNPLACED** - `F-e7e0ffa2d1` L1037 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: **Dropbox: Video Playback Review** (web, 4 screens, Collaboration Tools, Business, Productivity) [Refero](https://refero.design/flows/13043) → **UNPLACED** - `F-999d0d554d` L1039 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: > A user is viewing a video in a file/media preview screen and changes playback speed while the clip continues playing. The entry point is an opened video file in a viewer with standard controls. The user remains in the same viewer throughout, using the speed control and timeline to review the content. → **UNPLACED** - `F-1b8bee62a0` L1041 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: **Apple Maps: Place Details and Street View** (web, 3 screens, Map & Navigation) [Refero](https://refero.design/flows/12472) → **UNPLACED** - `F-3fc1474e03` L1043 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: > A user opens a specific place listing in Maps, reviews its details in a side panel, and then switches to Street View to inspect the venue’s exterior and surrounding street context. The entry point is a selected place marker or search result on the map; the exit point is the place shown in Street View with the option to close or continue exploring. → **UNPLACED** - `F-81056b7fd7` L1045 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: Steps: Place Details Opened → Street View Preview → Return to Map Context → **UNPLACED** - `F-fb52ea2a59` L1047 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: **Rork: Analytics Review and Date Range Change** (web, 4 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12160) → `/clinic/share-wearable-signal` - `F-534bf809e9` L1049 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: > Happy path for viewing analytics in a dark-mode web dashboard. The user is already signed in and on the Analytics section of the workspace. They review activity metrics, change the date range using a dropdown, briefly see a loading state while the data refreshes, and then return to the updated analytics view. The screen also shows a revenue integration card with a connect action, but no connection is completed in this sequence. → `/clinic/share-wearable-signal` - `F-3633c63eb2` L1051 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: Steps: Analytics Overview → Date Range Menu → Loading State → Updated Analytics View → `/clinic/share-wearable-signal` - `F-851df0af07` L1053 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: **Fingerprint: Inspect Event JSON Payload** (web, 3 screens, Business, Development Tools) [Refero](https://refero.design/flows/11153) → **UNPLACED** - `F-31b4c3a748` L1055 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: > Happy path: A logged-in user (desktop) viewing an event record opens a modal that displays the raw JSON payload. The JSON viewer supports switching between two versions (v3/v4), searching within the payload, showing result counts, and navigating matches. The user can copy/download the payload or close the modal. → **UNPLACED** - `F-9b8befd9d5` L1057 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: Steps: Open JSON modal → Search for specific ID (no results) → Search for boolean value (matches found) → **UNPLACED** - `F-2423f6a9cb` L1059 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: **Rork: Model Catalog Search** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12165) → **UNPLACED** - `F-d4521f28c0` L1061 _EDGE Clinic research: real product flows from Refero › Task: eye test vision_: > A logged-in user is on the AI Cloud section of a project dashboard and views a catalog of available models with pricing, context, and capability indicators. The user can browse by category tabs or search the model list. In the happy path, the user types a query that filters the list and then clears or changes the query to view matching models again. The flow starts on the model catalog screen and ends with the filtered results displayed or with the full list restored after search refinement. → **UNPLACED** - `F-5168d04de4` L1069 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: > Logged-in user profile page on GoFundMe with a share-profile modal opened from the profile screen. Entry is the profile page; exit is either closing the modal or copying the profile link. → **UNPLACED** - `F-ea7c854707` L1075 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: > Happy-path onboarding to add a user profile before starting a symptom assessment. → **UNPLACED** - `F-c783333522` L1077 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: Steps: Home — Start assessment → Add profile — Intro / safety tip → Ask for name (empty) → Ask for name (typing) → Sex at birth selection → Gender explanation modal → Date of birth prompt (picker shown) → Date of birth selected → Smoking status → High blood pressure question → Diabetes question → Completion / Start assessment → **UNPLACED** - `F-1fc57d0f7d` L1081 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: > Happy path: A signed-in mobile user opens the Health profiles list, selects a profile (Tony), and navigates through profile sections to view and edit Basic information, Personal information (height/weight), Health background (yes/not answered/no), Medication (add/edit/delete), and Allergies (add/delete). → **UNPLACED** - `F-371ec67c0e` L1083 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: Steps: Health profiles list → Profile sections menu → Basic information → Personal information (entry) → Personal information (data entry) → Health background → Add medication (empty) → Add medication (typing) → Medication list updated → Medication item actions → Add allergy (empty) → Add allergy (typing) → Allergies list updated → Add multiple allergies → Allergy item actions → Final allergies list → Return to profile sections → **UNPLACED** - `F-2653863cc0` L1087 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: > Happy path: a signed-in user opens the Profiles tab, selects a secondary health profile (Sophie), chooses the Delete health profile action from that profile's settings screen, confirms deletion in a modal, and is returned to the Profiles list where the deleted profile no longer appears. → **UNPLACED** - `F-bfadd3aad0` L1089 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: Steps: Profiles list → Profile details → Delete confirmation → Updated profiles list → **UNPLACED** - `F-9d89556580` L1091 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: **Doctronic AI: Healthcare Subscription Signup** (web, 4 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13691) → **UNPLACED** - `F-1b6f07ddfc` L1093 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: > A four-step subscription signup modal appears over a medical blog article. The user starts by entering an email address, proceeds to provide a phone number for health updates, and submits the registration. The flow ends with a confirmation message. The user is assumed to be browsing the article and not necessarily logged in. The modal can be dismissed at any point using the close icon. → **UNPLACED** - `F-c7641557a8` L1095 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: **ALO: Wellness Club Registration and Personalization** (web, 21 screens, Fitness, Lifestyle) [Refero](https://refero.design/flows/12365) → **UNPLACED** - `F-e7d747d3fc` L1097 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: > Happy path from the public ALO Wellness Club marketing experience to a newly created member account and personalized content dashboard. The user starts logged out, reviews the wellness offering, selects Join Now, enters account details, accepts the required terms, and creates an account. After signup, the user can personalize recommendations through six questions covering goals, schedule, styles, fitness level, equipment, age, and gender preferences. The user then reaches a logged-in dashboard showing account status, progress metrics, library shortcuts, and content recommendations. The final dashboard may include a short guided walkthrough with tooltips. → **UNPLACED** - `F-b8ca35e35a` L1099 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: Steps: Membership Landing → Account Creation → Completed Account Form → Personalization Welcome → Personalization Choice → Goals Questionnaire → Goals Selected → Schedule Questionnaire → Schedule Selected → Style Preferences → Fitness Level → Fitness Level Selected → Equipment Questionnaire → Equipment Selected → Personal Details → Personal Details Selected → Member Dashboard Welcome → Progress Walkthrough → Personal Picks Walkthrough → Personalized Content Dashboard → **UNPLACED** - `F-b510344a48` L1103 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: > Happy-path review step in an onboarding/onboard-like flow where the user verifies and fills missing personal details before proceeding. Entry: user reaches a summary screen titled "Is this correct?" with several fields (Name, Birthday, Gender, Nursing, Height, Weight). Exit: user either updates one or more fields using inline edit or add controls, confirms a selection via a modal/bottom-sheet picker, and returns to the summary with the field(s) updated — then may continue by tapping the Next CTA. Assumptions: user is already partway through onboarding (progress indicator visible) and is on a mobile device; the UI shows editable and addable fields and uses a bottom-sheet picker for selection. → **UNPLACED** - `F-001a8382fb` L1105 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: Steps: Summary / Confirm screen → Open gender picker (bottom sheet) → Select option and confirm → Summary updated with chosen value → **UNPLACED** - `F-05f0688434` L1107 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: **Athleats: First-time sign-up and onboarding** (ios, 23 screens, Health & Wellness, Food & Drink) [Refero](https://refero.design/flows/11281) → **UNPLACED** - `F-5e472f9fe3` L1109 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: > Happy-path onboarding for a first-time iOS user who signs up using Sign in with Apple, completes a short profile (photo + username), answers a multi-step preference questionnaire, is presented with a Pro subscription upsell, completes payment via Apple Pay, and lands on the authenticated home screen. → **UNPLACED** - `F-e230fab7d8` L1111 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: Steps: Marketing / Welcome → Sign in with Apple (permissions) → Sign in with Apple (loading) → Sign in complete → Sign Up — Basic profile → Add profile picture options → Photo Library picker → Photo selection applied → Username validation (empty) → Enter username → Sign Up — Continue enabled → Onboarding — Current goal → Onboarding — Dietary preferences → Onboarding — Referral source → Pro upsell modal — feature comparison → Apple subscription confirmation → Purchase success confirmation → Finalizing onboarding → Authenticated Home — Welcome → First-time user landing state → **UNPLACED** - `F-b59ba895a5` L1113 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: **Verse: New user onboarding and profile creation** (ios, 33 screens, Social Networking, Design & Media Editing) [Refero](https://refero.design/flows/3955) → **UNPLACED** - `F-602eb68b82` L1115 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: > Happy-path new-user onboarding that guides a user from launch -> phone verification -> username claim -> short preference quiz -> automated profile assembly -> final profile reveal. → **UNPLACED** - `F-6b3475d9f0` L1117 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: Steps: Splash / Get started → iOS tracking permission → Phone number entry (empty) → Phone number entry (typing) → Enter 6-digit code (empty) → Code input attempt (invalid) → Code entry success (loading) → Claim username (empty) → Username validation error → Select star sign (carousel) → Choose aesthetic (grid) → Select favorite movie/TV genre → Select current hobby interests → Processing: analyzing your vibe → Profile reveal (confetti) → First-time profile (overview) → Edit profile CTA and placeholders → Bottom navigation visible → Profile content populated from choices → Onboarding complete → **UNPLACED** - `F-20c2ca2c84` L1119 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: **Exoplan: Health & Calendar Onboarding** (ios, 10 screens, Health & Wellness, Scheduling Tools) [Refero](https://refero.design/flows/7342) → **UNPLACED** - `F-6f3c7bfa82` L1121 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: > This is an onboarding sequence that asks the user to grant two platform-level permissions (Health data & Calendar access) so the product can provide personalized focus scheduling and health-based insights. → **UNPLACED** - `F-9ea7d3cf96` L1123 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: Steps: Intro: Health Insights → Health Permission Sheet (off) → Health Permission Sheet (on) → Intro: Connect Calendar → Calendar Access System Modal → Onboarding Transition / Loading → Intro: Notifications → Health Data Missing Message → Main Calendar (empty) → Final App Launch / Splash → **UNPLACED** - `F-0b42df3fa9` L1125 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: **Miles: First-time onboarding and Health permission setup** (ios, 5 screens, Health & Wellness) [Refero](https://refero.design/flows/6704) → **UNPLACED** - `F-bc144bbffd` L1127 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: > Happy path for first-time user setup on iOS. Entry: user launches the Miles app for the first time. The app shows a sequence of stacked modals guiding the user to start and optionally connect to Apple Health, then a short interactive tutorial, and finally reveals the main tracker screen in its initial (empty) state. → **UNPLACED** - `F-3758539ff5` L1129 _EDGE Clinic research: real product flows from Refero › Task: health onboarding profile_: Steps: Welcome modal → Apple He _… 5921424 more characters in the zip._ ### research/README.md `research/README.md` · sha256 `43ad0f21dc412a75e551259e696a18df689515c29c1f76e0fe09d4ca362b9efe` #### Research This folder holds research written for the estate. It informs what the four applications must do; it is never authority. Every finding carries its source link and the date it was found, and is marked as one of: - **OBSERVATION** (seen in a primary source or a real product), - **AGREEMENT** (a law, regulation or standard, with its status and dates), - **CLAIM** (what a company or document says about itself, not independently checked), - **PREDICTION** (an estimate, with its assumptions), or - **UNKNOWN** (looked for and not found). Raw material (downloaded pages, datasets, screenshots, earlier work) stays in storage and is referenced by digest; only the written research lives here. ##### How to read it Start with the application you are working on. Each topic file opens with what it is, how it was gathered and what it does not cover. | Folder | What it covers | |---|---| | `finance/` | EDGE Finance: competitors and our lane, regulators in the UK and worldwide, agent assurance gaps and reusable open-source systems, confidence reporting, physical presence and solution banks, real product flows | | `clinic/` | EDGE Clinic: social prescribing, symptom checking, early detection and wearables; competitors, regulators and reusable systems; real product flows | | `public-consortium/` | EDGE Public Consortium: competitors, regulators and reusable systems for cyber governance, procurement, vetting, logistics and education; real product flows | | `judiverse/` | Judiverse: competitors, regulators and reusable systems for office, assistants, agent lab and music; real product flows | | `experience/` | Customer experience patterns from any industry (onboarding, errors, confirmations, help, accessibility and more), useful to every application | | `market/` | What is being funded and bought (grants, contracts, venture capital) and the costliest real problems by industry | | `legacy/` | What the organisation's earlier work already contains (competitors, benchmarks, readiness, blueprints, features, standards, papers, guides, strategy, funding), read as research so none of it is wasted or repeated | ##### Real product flows The `REFERO-FLOWS` files come from Refero, a library of real app and website screens with each journey described step by step. Results from other industries are kept on purpose, because they are often the best customer experience to learn from. Refero holds mostly consumer and start-up products: many incumbents (for example Starling, Monzo, Chase UK) and most government, enterprise and NHS systems are not in it, so these files are a sample of the floor, not a census. ##### Coverage What exists in each folder, and what is still being researched, is listed by `python3 tools/build.py`, which writes the generated index. Nothing in this folder states a count of estate objects; those are produced live by the build. ### research/REUSE-ENGINES.md `research/REUSE-ENGINES.md` · sha256 `b141620e6b99d5fd7893e27c83061494b55384c395a09bb0ba6eea811da9c436` #### Engines behind the applications: what is reused, what is built Statement class: research and recommendation, 1 November 2026 03:35Z. OBSERVED where the estate's own research verified it (file and date given); CLAIM where it has not. Never a rule. ##### What the specification actually needs (counted from spec/receivers.json, research and competitor units excluded) | Application | Operations | Workflow steps on a record or case (journey state, ours) | Evidence reads (ours) | Domain entities that need an engine | |---|---|---|---|---| | EDGE Bank | 3,301 | 1,452 | 378 | account, payment, card, member, payee; rail: payment-scheme (48 operations) | | EDGE Clinic | 2,284 | 1,376 | 338 | FHIR resources (69), consent (34), identity claim (28), safety plan (14), plan (11); NHS services (29), detection models (11) | | EDGE Consortium | 4,778 | 1,942 | 359 | case (74), egress and access policy (89), role (52), module and client (84), consortium (32) | | EDGE Public | 3,225 | 1,830 | 486 | application (41), vacancy (18), asset (18), alert (18), decision (15) | | Judiverse Suite | 3,770 | (steps spread across) | | intention (113), contact (91), upload terms (90), template (82), event (65), task (44), document versions and objects (98) | Reading: about three quarters of every application is journey state and evidence, which the estate already runs. The engine question is the last column: a small set of domain entities per application. One engine per application covers it. ##### The engine per application | Application | Engine (self-hosted on Civo) | Covers | Licence and status | How it binds | |---|---|---|---|---| | EDGE Bank | Griffin (sponsor bank, rail) + Apache Fineract (deposits, loans, accounting) + TigerBeetle (ledger) + Hyperswitch (card and alternative processors) | accounts, balances, payments, Direct Debits, standing orders, loans, overdrafts, cards | Griffin: PRA/FCA bank, sandbox self-serve (OBSERVED 1 Nov, research/finance/ENGINE-BY-STARLING.md §3). Fineract Apache-2.0, TigerBeetle, Hyperswitch Apache-2.0 (OBSERVED 31 Oct, research/finance/ASSURANCE-AND-REUSE.md §2) | `payment-scheme` and `card-processor` slots; runtime/domains/bank.py already holds the ledger interface, so Fineract or TigerBeetle replaces its SQLite tables without touching a screen | | EDGE Clinic | Medplum (FHIR R4 server, UK Core, scheduling, auth, bots, React components) | patients, appointments, consents, observations, care plans, questionnaires, referrals | Apache-2.0 (OBSERVED 31 Oct, research/clinic/COMPETITORS-REGULATORS-REUSE.md); ONC certification CLAIM | `feature_fhir`, consent, safety_plan and plan entities map to FHIR resources through one adapter; `nhs-services` stays a plug | | EDGE Consortium | ERPNext (Frappe) for roles, modules, clients, access and cases; Keycloak or Ory for identity federation | the university's records: enrolment, modules, roles with expiry, cases, policies | ERPNext GPL-3 (CLAIM, not verified by the estate; network copyleft to assess), Keycloak Apache-2.0 (OBSERVED 31 Oct) | case, role, module, client, access entities through the Frappe REST API as one adapter | | EDGE Public | Same ERPNext instance (applications, vacancies, assets, alerts, decisions) | government-on-the-estate records and decisions | as above | application, vacancy, asset, alert, decision entities | | Judiverse Suite | ONLYOFFICE Docs (documents, sheets, decks), Matrix (Synapse or Conduit) for messages and calls with LiveKit for media, a CalDAV server for calendar, Simwood for UK numbers | editing, collaboration, chat, calls, meetings, calendar, mail (mail already runs natively) | ONLYOFFICE AGPL-3 community with a separate embedding licence (CLAIM); Matrix Synapse AGPL-3 (CLAIM); LiveKit Apache-2.0 (CLAIM); Simwood commercial API (CLAIM) | document, template, event, task, contact entities; `voice-telephony` and `sms-gateway` slots | | Agents operating other software | Composio (hosted) or the estate's own MCP hub | third-party tools under per-user authentication | Composio hosts data outside the estate: assess before UK-only use (CLAIM) | the hub already exposes every operation to agents (`estate_call`, `racetrack_run`); Composio adds outbound integrations only | ##### The pattern (unchanged) approved screen → Jude's authenticated operation (identity, mandate, guards, rehearsal, evidence: ours) → engine adapter (runtime/adapters.ADAPTERS, one module per engine in runtime/domains/) → the engine's own confirmation read back → evidence. TOK, EDGE ID and assurance stay shared. Nothing is hand-built that an engine supplies; what is built is the adapter and the domain behaviour no engine has (money provenance, agent mandates, glass-box evidence). ##### Order 1. EDGE Bank: Griffin key, then Fineract behind runtime/domains/bank.py. The screens and ledger interface exist; the customer sees the same app with real rails. 2. EDGE Clinic: Medplum on Civo; one FHIR adapter for the five clinic entities. Largest single coverage gain after the bank. 3. Judiverse Suite: ONLYOFFICE and Matrix, because documents and messages are what every other application's people use daily. 4. EDGE Consortium and EDGE Public: one ERPNext instance for both. Each engine lands as a named service on Civo with backup proven first, then its adapter binds, then the affected journeys are walked and their results read back from the engine. ### research/clinic/COMPETITORS-REGULATORS-REUSE.md `research/clinic/COMPETITORS-REGULATORS-REUSE.md` · sha256 `6d2e769a4d1ba42fa5e604d6607ab6a17b8cbcb5ba9c77cd28b87834f6ec1340` #### EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems Research for EDGE Clinic, one of four applications in JudeAI | EDGE ID. Written 31 October 2026 for a brand-new engineer. This is research, not clinical or legal advice. Anything here must be checked by a qualified clinical safety officer and a lawyer before it drives a decision. **Scope.** Four topics only: (1) competitors, (2) regulators and rules, (3) agent assurance gaps in health, (4) open standards and open-source to reuse. Social prescribing evidence, symptom checkers, early detection, wearables, physical presence and NHS costs are covered elsewhere, in `research/clinic/SOCIAL-PRESCRIBING-AND-DETECTION.md`, and are not repeated here. ##### How to read the labels | Label | Meaning | |---|---| | **OBSERVED** | I read it directly in a primary page or file during this research (for example, a licence file or a regulator's page). | | **CLAIM** | A vendor, a law firm, a trade paper or a blog says it. Not independently checked. | | **AGREEMENT** | A law, rule, standard or guidance that is adopted and in force, with a date. | | **PROPOSED** | A draft, consultation, plan or recommendation. Not binding yet. | | **UNKNOWN** | Sources disagree, or I could not find a reliable answer. | Every link was accessed on 2026-10-31. Where a source is a secondary summary rather than the regulator's own text, the line says so. --- ##### 0. The ten things to know first 1. **The UK has no AI-specific medical device law yet.** The MHRA (Medicines and Healthcare products Regulatory Agency, the UK device regulator) put draft amending regulations out in May 2026, and they are expected to apply from June 2027. They have not passed Parliament. PROPOSED. [Latham & Watkins](https://www.lw.com/en/insights/mhra-publishes-draft-amendment-to-the-uk-medical-devices-regulations), accessed 2026-10-31. 2. **A national commission set out 44 recommendations for regulating AI in UK health on 10 September 2026**, including staged "L-plate" authorisations and continuous real-world monitoring. Government says it will "carefully consider" them. PROPOSED. [GOV.UK press release](https://www.gov.uk/government/news/independent-commission-led-by-nhs-doctors-sets-out-blueprint-to-accelerate-safe-ai-adoption-in-healthcare) (OBSERVED), accessed 2026-10-31. 3. **The MHRA's test for "is it a medical device?" (29 July 2026):** a tool that only transcribes, summarises or drafts letters is not a device, but one that supports diagnosis, treatment or prevention, or "take[s] automated action such as placing orders without clinician review", is. AGREEMENT (guidance). [MHRA news](https://www.gov.uk/government/news/mhra-clarifies-regulatory-status-of-ambient-voice-technologies-used-in-the-nhs) (OBSERVED), accessed 2026-10-31. EDGE Clinic triage, prescribing safety and retinal screening fall on the device side of that line. 4. **The NHS clinical safety standards DCB0129 and DCB0160 are being rewritten, with AI named as a reason.** The consultation ran 29 June to 11 September 2026. The current versions still apply. PROPOSED (revision); AGREEMENT (current versions). [NHS England consultation page](https://www.engage.england.nhs.uk/patient-safety/national-review-of-clinical-risk-management-standa/) (OBSERVED), accessed 2026-10-31. 5. **The EU AI Act's high-risk rules for AI inside medical devices now apply from 2 August 2028**, moved back by the "Digital Omnibus on AI", Regulation (EU) 2026/1744, which came into force on 27 July 2026. AGREEMENT. [Hunton](https://www.hunton.com/privacy-and-cybersecurity-law-blog/eu-digital-omnibus-on-ai-enters-into-force) (OBSERVED, law-firm summary), accessed 2026-10-31. 6. **In the US, FDA's lifecycle guidance for AI-enabled devices is still a January 2025 draft.** The final PCCP (predetermined change control plan) guidance is in force from December 2024. PROPOSED and AGREEMENT respectively. [RAPS](https://www.raps.org/resource/fda-s-device-center-releases-guidance-agenda-for-fy-2027.html); [Ropes & Gray](https://ropesgray.com/en/insights/alerts/2024/12/fda-finalizes-guidance-on-predetermined-change-control-plans-for-ai-enabled-device), accessed 2026-10-31. 7. **The big AI labs are now in consumer health.** OpenAI launched ChatGPT Health in January 2026, and Anthropic launched Claude for Healthcare in January 2026. Both are US-first and connect personal records. CLAIM. [Fortune](https://fortune.com/2026/01/07/openai-launches-chatgpt-health-in-a-push-to-become-a-hub-for-personal-health-data); [Anthropic](https://www.anthropic.com/news/healthcare-life-sciences), accessed 2026-10-31. 8. **Evidence that "a clever model" is not "a safe front door":** in a randomised study of 1,298 UK participants, people who used LLMs did not choose better than controls, even though the models alone named conditions correctly 94.9% of the time. CLAIM (peer-reviewed). [University of Oxford](https://www.ox.ac.uk/news/2026-02-10-new-study-warns-risks-ai-chatbots-giving-medical-advice), accessed 2026-10-31. 9. **No standard exists yet for an AI agent's identity, mandate and delegation in health.** The OpenID Foundation names these gaps and is drafting extensions. PROPOSED. [OpenID Foundation](https://openid.net/new-whitepaper-tackles-ai-agent-identity-challenges/), accessed 2026-10-31. 10. **There is no open, clinical-grade drug-interaction source.** The US NLM (National Library of Medicine) switched off its free interaction API around 2 January 2024. The BNF (British National Formulary) needs a licence for use inside systems. DrugBank's full data is non-commercial only. OBSERVED/CLAIM. [NLM RxNav](https://www.lhncbc.nlm.nih.gov/RxNav/news/APIUpdate.html); [Pharmaceutical Press licensing](https://www.pharmaceuticalpress.com/?p=146134) (OBSERVED); [DrugBank FAQ](https://dev.drugbank.com/guides/faqs), accessed 2026-10-31. --- ##### 1. Competitors ###### 1.1 Digital front doors and primary care platforms (UK first) | Product (owner) | What it offers | What it does with AI or agents | Label | Source | |---|---|---|---|---| | **NHS App** (NHS England) | The government says the app will be the "front door to the health service". The 10 Year Health Plan (July 2025) targets a "full front door to the entire NHS" by 2028. | The plan proposed "My NHS GP", an AI tool to route people to the right care. It is reported to reach 200,000 patients in year one, with full rollout by April 2028. No confirmed 2026 go-live found. | AGREEMENT (policy); PROPOSED (My NHS GP) | [GOV.UK oral statement](https://www.gov.uk/government/speeches/fit-for-the-future-health-and-social-care-secretarys-statement); [Full Fact tracker](https://fullfact.org/government-tracker/nhs-app-improvements/); [Digital Health](https://www.digitalhealth.net/2025/07/what-does-the-10-year-plan-mean-for-digital/), accessed 2026-10-31 | | NHS App + Rapid Health Smart Triage | A GP booking and triage tool that works inside the NHS App. | Reported as available to more than a million patients in April 2026. | CLAIM | [Digital Health](https://www.digitalhealth.net/2026/04/ai-tool-allows-a-million-patients-to-book-appointments-via-nhs-app/), accessed 2026-10-31 | | NHS App messaging shift | NHS App leadership has described the app as "not the only" service, moving from "front door" towards "companion". | n/a | CLAIM | [Digital Health](https://www.digitalhealth.net/2026/03/nhs-app-to-move-from-front-door-to-companion/), accessed 2026-10-31 | | **Accurx** | Patient messaging, online triage, video consultations and Scribe. | Accurx says Scribe (built with Tandem) has summarised over one million consultations across nearly 3,000 practices, and that 1,350 practices run "total triage" on Accurx. It advertises "AI-enhanced Patient Triage". | CLAIM | [Accurx blog](https://www.accurx.com/post/whats-new-in-2026-for-gp); [Accurx benefits](https://www.accurx.com/benefits), accessed 2026-10-31 | | Accurx pricing | Per-unit prices on the G-Cloud government catalogue. For example, Scribe is £0.50 down to £0.40 per unit, depending on volume. | n/a | CLAIM (vendor price list) | [G-Cloud pricing document](https://assets.applytosupply.digitalmarketplace.service.gov.uk/g-cloud-14/documents/708675/281422867192066-pricing-document-2026-08-10-1029.pdf), accessed 2026-10-31 | | **EMIS** (Optum UK, now TPG) | EMIS Web, the GP clinical record system, and EMIS-X (cloud). | EMIS Scribe, an ambient scribe for EMIS Web and EMIS-X. EMIS is on NHS England's AVT (ambient voice technology) registry. No autonomous agents found. | CLAIM | [EMIS April 2026 newsletter](https://www.emishealth.com/news-insights/april-2026-newsletter); [EMIS January 2026 newsletter](https://www.emishealth.com/news-insights/january-2026-newsletter), accessed 2026-10-31 | | EMIS ownership | Reported that TPG's acquisition of Optum UK, including EMIS, closed on 13 March 2026. | n/a | CLAIM (single secondary source) | [healthcare.digital](https://www.healthcare.digital/single-post/the-strategic-transformation-of-uk-primary-care-digital-infrastructure-tpg-s-acquisition-of-emis-an), accessed 2026-10-31 | | **TPP SystmOne + Airmid** | SystmOne is a clinical record system. Airmid is a patient app with booking, video, messaging and wearable readings (Apple Health, Google Fit), and it supports NHS login. | TPP says it is "embedding AI and Genetic data" for early disease detection. No ambient scribe found. | CLAIM | [TPP Airmid](https://www.tpp-uk.com/products/airmid); [TPP news](https://tpp-uk.com/news/taking-healthcare-to-the-next-level/), accessed 2026-10-31 | | **Doctolib + Medicus** | Doctolib (France) acquired Medicus, a UK GP software company, in May 2026, and plans to invest over £100m in the UK. | Positioned as an "AI-powered GP platform". | CLAIM | [BusinessCloud](https://businesscloud.co.uk/news/doctolib-acquires-medicus-plans-to-invest-100m-in-uk/); [TechMarketView](https://www.techmarketview.com/ukhotviews/archive/2026/05/06/doctolib-backs-medicus-to-challenge-gp-it-duopoly), accessed 2026-10-31 | | **Huma** (+ eConsult, iPlato) | Huma bought eConsult (digital triage, reportedly used by over 1,800 GP practices) in October 2024 and launched Huma Workspace. Its platform holds FDA 510(k) Class II (2023) and Health Canada Class II (2025) clearance as SaMD (software as a medical device). | Health Canada licence allows AI algorithms for screening, diagnosis and decision support. Huma acquired Aluna (remote spirometry) in May 2025. | CLAIM | [Huma newsroom](https://www.huma.com/newsroom/huma-acquires-econsult-and-launches-huma-workspace-for-health-systems); [Digital Health](https://digitalhealth.net/2024/10/huma-acquires-econsult-to-become-end-to-end-tech-platform); [Huma, Health Canada](https://huma.com/resources/huma-therapeutics-receives-health-canada-class-ii-medical-device-license-for-digital-health-platform); [MobiHealthNews](https://www.mobihealthnews.com/news/huma-receives-510k-class-ii-clearance-its-disease-agnostic-platform), accessed 2026-10-31 | | **Livi / Kry** | Video GP capacity sold to NHS commissioners. CQC (Care Quality Commission) lists Livi UK's location as "Archived", and Wikipedia says Livi was being rebranded Kry in France and the UK. | Not found. | UNKNOWN (current UK status) | [CQC](https://www.cqc.org.uk/location/1-10716659568/inspection-summary); [Wikipedia](https://en.wikipedia.org/wiki/Livi_(company)), accessed 2026-10-31 | | **Push Doctor** | Video GP. Square Health bought its business and assets; CB Insights dates the deal to December 2021. | Not found. | UNKNOWN (current status) | [Healthcare & Protection](https://healthcareandprotection.com/square-health-buys-push-doctor/amp/); [CB Insights](https://www.cbinsights.com/investor/push-doctor), accessed 2026-10-31 | | **Doctaly** | Paid face-to-face GP appointments with NHS GPs. A separate brand, "Doctaly Assist", offers GP access via WhatsApp. No 2025–2026 news found. | Not found. | UNKNOWN | [BusinessCloud](https://businesscloud.co.uk/news/platform-for-private-consultations-with-nhs-gps-crowdfunds/); [Endole](https://open.endole.co.uk/insight/brand/167752-doctaly-assist), accessed 2026-10-31 | | **Cera** | Home care with AI. Raised about $150m (debt and equity) in January 2025. Says it works in 150 of 382 UK local authorities and 29 of 42 ICBs (integrated care boards). | Cera claims up to 70% fewer emergency admissions. TechCrunch reported that some NHS partnerships were wrongly listed on Cera's website. | CLAIM (disputed in part) | [Med-Tech News](https://med-technews.com/news/latest-medtech-news/Ceras-ai-led-social-care-model-to-potentially-save-nhs-and-uk-government-1-billion-per-year); [City AM](https://www.cityam.com/?p=2297543); [TechCrunch](https://techcrunch.com/?p=1632310), accessed 2026-10-31 | ###### 1.2 AI clinical intelligence vendors | Vendor / product | What it offers | Agents? | Regulatory position | Label | Source | |---|---|---|---|---|---| | **OpenAI – ChatGPT Health / "Health in ChatGPT"** | Launched 7 January 2026 as a separate space where users connect medical records (via b.well) and wellness apps. OpenAI dated a relaunch to 23 July 2026, for US users only. | Consumer assistant. By default it asks before it uses connected records. | OpenAI says it "does not replace qualified medical care". Connected health data is not used to train foundation models. | CLAIM | [Fortune](https://fortune.com/2026/01/07/openai-launches-chatgpt-health-in-a-push-to-become-a-hub-for-personal-health-data); [Fierce Healthcare](https://fiercehealthcare.com/ai-and-machine-learning/openai-launches-chatgpt-health-connect-data-health-apps-medical-records); [OpenAI](https://openai.com/index/health-in-chatgpt/) (page blocked automated fetch), accessed 2026-10-31 | | **Anthropic – Claude for Healthcare** | Launched January 2026. Connectors for the CMS (US Centers for Medicare & Medicaid Services) Coverage Database, ICD-10 and the NPI (National Provider Identifier) Registry, plus a FHIR skill and a prior-authorisation template. Consumer connectors (HealthEx, Function, Apple Health, Android Health Connect) are in beta. | Agent skills. | Described as "HIPAA-ready". Anthropic says it does not train on user health data. | CLAIM | [Anthropic](https://www.anthropic.com/news/healthcare-life-sciences); [MobiHealthNews](https://www.mobihealthnews.com/news/anthropic-launches-claude-healthcare); [The Register](https://www.theregister.com/2026/01/12/claude_anthropic_healthcare/), accessed 2026-10-31 | | **Microsoft – Dragon Copilot** (formerly Nuance DAX) | Ambient documentation. Generally available in the UK since September 2025 and hosted in a UK Azure data centre. | A marketplace of partner apps and agents inside Dragon Copilot. | Microsoft says DCB0129 is complete and assessed by a UK Clinical Safety Officer. | CLAIM | [Phoenix Software](https://www.phoenixs.co.uk/resources/blog/microsoft-dragon-copilot-is-now-generally-available-in-the-uk/); [Microsoft UK](https://www.microsoft.com/en-gb/health-solutions/clinical-workflow/dragon-copilot), accessed 2026-10-31 | | **Microsoft 365 Copilot for NHS England** | June 2026 agreement covering Microsoft 365 Copilot and Copilot Studio, so NHS England and trusts can build their own agents (for example, for complaints and FOI requests), governed through "Agent 365". | Administrative agents, not clinical. | n/a | CLAIM | [Microsoft](https://news.microsoft.com/source/2026/06/08/nhs-england-accelerates-ai-adoption-with-microsoft-365-copilot-to-improve-service-delivery-reduce-costs-and-create-more-time-for-care/); [TechMarketView](https://www.techmarketview.com/ukhotviews/archive/2026/06/08/nhs-england-scales-microsoft-copilot-adoption), accessed 2026-10-31 | | **Google – AMIE and MedGemma** | AMIE is a research diagnostic dialogue agent, now extended to disease management. MedGemma is a family of open-weight medical models. | AMIE is multi-agent (a dialogue agent plus a management agent). It is still research. | Not a product. Studies are running with Beth Israel Deaconess and Included Health. | CLAIM | [Google blog](https://blog.google/innovation-and-ai/models-and-research/google-research/amie-for-disease-management-in-nature); [Google Research I/O 2025](https://www.research.google/blog/google-research-at-google-io-2025), accessed 2026-10-31 | | **Abridge** | Ambient documentation. In August 2026 it expanded its "clinical intelligence agent" with decision support to more than 300 health systems. | Yes, for clinicians. | US-focused. | CLAIM | [HLTH](https://hlth.com/insights/news/abridge-expands-clinical-intelligence-ai-across-300-health-systems), accessed 2026-10-31 | | **Tortus, Heidi, Tandem** (UK ambient scribes) | Ambient scribes. A tracker reports Tortus as UKCA Class IIa (24 June 2026) and Tandem as EU MDR Class IIa. Heidi's status conflicts between sources. | Documentation, plus coding suggestions. | Mixed; see the regulator section. | UNKNOWN (single commercial blog family, with internal conflicts) | [iatroX tracker](https://www.iatrox.com/blog/ai-scribe-regulatory-status-tracker-2026), accessed 2026-10-31 | | **NHS England AVT supplier registry** | A self-certified registry that is not a commercial framework and not an endorsement. It listed 19 suppliers in January 2026. | n/a | Suppliers must complete DTAC, plus other criteria. | CLAIM | [HTN](https://htn.co.uk/2026/01/19/nhs-england-lists-19-suppliers-meeting-criteria-for-ambient-voice-registry/); [Digital Health](https://www.digitalhealth.net/2026/01/nhs-england-launches-supplier-registry-for-ai-scribing-tech/), accessed 2026-10-31 | ###### 1.3 Clinical decision support and drug-interaction data | Source | What it is | Licence / access | Label | Source link | |---|---|---|---|---| | **BNF / BNFC** (British National Formulary, and the version for children) | The UK's standard prescribing reference, with A–Z interaction monographs. Free to NHS users via NICE (National Institute for Health and Care Excellence), UK only. | Non-NHS bodies need a licence from the BNF Partners. Merging BNF content into third-party systems needs a direct licence. RPS (Royal Pharmaceutical Society) offers a data feed or API. | OBSERVED | [Pharmaceutical Press licensing](https://www.pharmaceuticalpress.com/?p=146134); [NICE BNF UK only](https://www.nice.org.uk/bnf-uk-only), accessed 2026-10-31 | | **Stockley's Drug Interactions** | Monographs covering more than 5,000 interactions, with evidence, mechanism and management. Accessed by subscription through MedicinesComplete. | Subscription. | CLAIM | [SPS (NHS Specialist Pharmacy Service)](https://sps.nhs.uk/articles/drug-interactions-resources-to-support-answering-questions/), accessed 2026-10-31 | | **FDB Multilex / OptimiseRx** (First Databank UK) | FDB calls Multilex the UK's most widely used drug database and e-prescribing decision support (more than 10,000 sites). It covers drug–drug interactions, sensitivity and duplicate-therapy checks. OptimiseRx gives point-of-care prompts in EMIS Web. | Commercial. A 2026 NHS contract was awarded without competition because only FDB could supply it. | CLAIM | [SNOMED International marketplace](https://snomed.org/snomed-ct-marketplace/first-databank-uk-ltd.); [Find a Tender](https://www.find-tender.service.gov.uk/procurement/ocds-h6vhtk-06c6f3), accessed 2026-10-31 | | **NLM RxNav drug interaction API** (US) | A free US government interaction API. | Scheduled for shutdown on or about 2 January 2024. | OBSERVED (NLM notice) | [NLM](https://www.lhncbc.nlm.nih.gov/RxNav/news/APIUpdate.html); [DrugBank blog](https://blog.drugbank.com/nih-discontinues-their-drug-interaction-api/), accessed 2026-10-31 | | **DrugBank** | A drug knowledge base with interactions. | Open vocabulary data: CC0-1.0. Full data: CC-BY-NC-4.0 (academic, non-commercial). Commercial use needs a separate agreement. Several release pages show academic downloads "temporarily unavailable". | CLAIM | [DrugBank FAQ](https://dev.drugbank.com/guides/faqs); [DrugBank 5.1.22](https://go.drugbank.com/releases/5-1-22), accessed 2026-10-31 | | **DDInter 2.0** | An academic open-access database: 2,310 drugs and 302,516 drug–drug interaction records (NAR, Nucleic Acids Research, 2025). | "Open-access", but no data licence found. | UNKNOWN (licence) | [NAR paper](https://academic.oup.com/nar/article/53/D1/D1356/7740584); [DDInter 2.0](https://ddinter2.scbdd.com/), accessed 2026-10-31 | | **Medscape interaction checker** | Not researched in depth. No primary source checked. | — | UNKNOWN | — | **Gap for EDGE Clinic:** the brief asks for "an open clinical-grade drug-interaction source". On this evidence, none exists today. The clinical-grade sources (BNF, Stockley's, FDB) are licensed. The open ones (DrugBank open data, DDInter) are either non-commercial, of unclear licence, or not presented as clinical-grade. The free US government API was withdrawn. Sources as above, accessed 2026-10-31. ###### 1.4 Retinal AI vendors | Vendor / product | What | Status | Label | Source | |---|---|---|---|---| | **LumineticsCore** (Digital Diagnostics, formerly IDx-DR) | Autonomous detection of more-than-mild diabetic retinopathy. It was the first autonomous AI diagnostic authorised by the FDA (US Food and Drug Administration), via De Novo. | US authorised. | CLAIM (vendor) | [Digital Diagnostics](https://www.digitaldiagnostics.com/resources/insights/lumineticscore-engineered-by-digital-diagnostics-early-detection-of-diabetic-retinopathy-using-ai/), accessed 2026-10-31 | | **EyeArt** (Eyenuk) | Detects mild and vision-threatening diabetic retinopathy. FDA cleared in 2020, with more cameras added in 2023. | US cleared. | CLAIM | [Insight News](https://www.insightnews.com.au/fda-clearance-expands-eyenuk-ai-system-to-automatically-detect-diabetic-retinopathy/), accessed 2026-10-31 | | **AEYE-DS** (AEYE Health) | Reported as cleared on the Optomed Aurora handheld camera and the Topcon NW400. | Date not confirmed. | CLAIM | [ai2.work](https://ai2.work/blog/handheld-ai-eye-screening-clears-fda-inside-the-retina-shift), accessed 2026-10-31 | | **NHS Diabetic Eye Screening Programme (DESP)** | About 2.2 million screening appointments a year. The UK National Screening Committee consulted on AI in diabetic eye screening (May 2026). A Birmingham procurement framework favours AI as a first-pass triage, with specialist graders reviewing. | Outcome not found. No vendor confirmed in NHS DESP. | PROPOSED | [AOP](https://www.aop.org.uk/ot/news/2026/05/13/uk-national-screening-committee-consults-on-use-of-ai-in-diabetic-eye-screening); [University of Birmingham](https://www.birmingham.ac.uk/news/2026/birmingham-research-helps-shape-how-nhs-procures-ai-based-diabetic-eye-screening); [SGUL AIDED study](https://www.sgul.ac.uk/about/our-institutes/population-health/projects/aided-study), accessed 2026-10-31 | | **Cascader** (Moorfields, UCL, Topcon) | A new company for AI on eye disease and "oculomics" (reading signs of other diseases, such as heart disease or dementia, from retinal images), with access to over 35 million images. | Early stage. | CLAIM | [Moorfields](https://www.moorfields.nhs.uk/about-us/news-and-blogs/news/new-company-launched-to-advance-ai-driven-eyecare), accessed 2026-10-31 | | UK optometry pilots (Toku, Heart-Eye and others) | Small UK pilots of retinal AI for cardiovascular risk. | Pilots only. | CLAIM | [LOCSU case studies](https://locsu.co.uk/wp-content/uploads/2025/12/CVD-Case-Studies-AI-Powered-Retinal-Imaging-for-Cardiovascular-Risk-Assessment.pdf), accessed 2026-10-31 | ###### 1.5 What the competitors leave open, and bare minimum versus ambitious These gaps come from the tables above, with sources as listed there. - **Nobody joins the steps "front door → detection → social prescribing referral".** The NHS App is national but is a router. Accurx, eConsult (Huma) and Rapid Health triage GP demand. Cera is home care. Retinal AI is a single test. EDGE Clinic's "come about anything" scope has no direct UK rival found. UNKNOWN whether "My NHS GP" will cover non-medical needs. - **AI in competitors is mostly ambient documentation**, which the MHRA now says is not a device when it only summarises ([MHRA](https://www.gov.uk/government/news/mhra-clarifies-regulatory-status-of-ambient-voice-technologies-used-in-the-nhs), accessed 2026-10-31). Patient-facing agents that act (book, refer, check prescriptions) are rare and regulated. - **None of the vendors found publishes an agent identity, mandate or audit model that a regulator could re-run.** Microsoft's "Agent 365" governance is administrative only ([TechMarketView](https://www.techmarketview.com/ukhotviews/archive/2026/06/08/nhs-england-scales-microsoft-copilot-adoption), accessed 2026-10-31). | | Bare minimum (to be allowed in the NHS) | Ambitious (where EDGE Clinic could lead) | |---|---|---| | Front door | NHS login, DTAC, DCB0129, DSPT, signposting only (not a device) | An agent that triages and acts under a mandate the patient can see, with every step replayable by a regulator | | Detection | Buy a cleared retinal AI and keep a human grader in the loop | Multi-signal detection with staged ("L-plate") authorisation and a public post-market safety record | | Prescribing safety | License BNF or FDB through the EHR supplier | Contribute to an open, versioned interaction knowledge base with provenance | | Agents | No autonomous actions; human signs off every action | Agent identity, mandate, consent and escalation recorded as FHIR Consent, Provenance and AuditEvent | --- ##### 2. Regulators and rules ###### 2.1 United Kingdom | Body / rule | What a new engineer needs to know | Status and dates | Label | Source | |---|---|---|---|---| | **MHRA – UK MDR 2002** (UK Medical Devices Regulations) | Software with a medical purpose is a medical device. Post-market surveillance rules (SI 2024/1368) apply to devices placed on the GB market from 16 June 2025. | In force since 16 June 2025 | AGREEMENT | [GOV.UK](https://www.gov.uk/government/news/first-major-overhaul-of-medical-device-regulation-comes-into-force-across-great-britain), accessed 2026-10-31 | | MHRA – draft Medical Devices (Amendment) Regulations 2026 | New pre-market rules and an "international reliance" route, which lets devices authorised in Australia, Canada or the US register in GB. Software-specific provisions are reportedly not updated yet. | Draft published 8 May 2026. Expected to be adopted December 2026 and to come into force June 2027, subject to Parliament. | PROPOSED | [Latham & Watkins](https://www.lw.com/en/insights/mhra-publishes-draft-amendment-to-the-uk-medical-devices-regulations); [Fieldfisher](https://www.fieldfisher.com/en/insights/new-pre-market-requirements-for-medical-devices-mhra-publishes-draft-legislation---what%e2%80%99s-changed); [Qualio](https://www.qualio.com/blog/mhra-s-2026-devices-reform-what-changes-and-when), accessed 2026-10-31 | | MHRA – AVT (ambient voice technology) guidance | Transcription, summarising or drafting letters alone: not a device. Supporting diagnosis, treatment or prevention, or taking automated action without clinician review: a device. | Published 29 July 2026 | AGREEMENT (guidance, does not change law) | [MHRA news](https://www.gov.uk/government/news/mhra-clarifies-regulatory-status-of-ambient-voice-technologies-used-in-the-nhs) (OBSERVED); [MHRA PDF](https://assets.publishing.service.gov.uk/media/6a68daaff23ff62c51c1a7a3/MHRA_Guidance_on_Ambient_Voice_Technology-Enabled_Products.pdf), accessed 2026-10-31 | | MHRA – PCCP guidance for AI devices (UK) | Reported plan to issue draft guidance on predetermined change control plans by end 2026. A PCCP is a plan, agreed in advance, for how a model may change after approval. | Not published | PROPOSED | [Qualio](https://www.qualio.com/blog/mhra-s-2026-devices-reform-what-changes-and-when), accessed 2026-10-31 | | MHRA – AI Airlock | A regulatory sandbox for AI as a medical device. Phase 2 had 7 products, including note-taking and eye disease detection, and ran April 2025 to May 2026. DHSC (Department of Health and Social Care) funds £1.2m a year for 2026–2029. | Phase 2 reports out July 2026. Phase 3 not confirmed. | AGREEMENT (programme); UNKNOWN (Phase 3) | [GOV.UK collection](https://www.gov.uk/government/collections/ai-airlock-the-regulatory-sandbox-for-aiamd); [Phase 2 report](https://www.gov.uk/government/publications/ai-airlock-sandbox-phase-2-programme-report); [MHRA blog](https://medregs.blog.gov.uk/2026/06/09/advancing-ai-regulation-in-healthcare-insights-from-ai-airlock-phase-2/), accessed 2026-10-31 | | National Commission into the Regulation of AI in Healthcare | 44 recommendations, including staged authorisations ("L-plates"), continuous real-world monitoring, a public searchable safety database, and stronger MHRA enforcement. | Report 10 September 2026. GOV.UK says a "formal response to follow". One law firm says government accepted all 44 on 6 October 2026. | PROPOSED; UNKNOWN (acceptance) | [GOV.UK](https://www.gov.uk/government/news/independent-commission-led-by-nhs-doctors-sets-out-blueprint-to-accelerate-safe-ai-adoption-in-healthcare) (OBSERVED); [Report](https://www.gov.uk/government/publications/national-commission-into-the-regulation-of-ai-in-healthcare-recommendations-for-a-future-regulatory-framework); [CMS](https://cms.law/en/gbr/legal-updates/from-pre-market-to-lifecycle-the-national-commission-s-44-recommendations-on-ai-regulation-in-healthcare); [Ropes & Gray](https://www.ropesgray.com/en/insights/viewpoints/2026/10/102o5fp/ai-medical-devices-in-great-britain-what-the-mhras-response-to-the-national-com), accessed 2026-10-31 | | **NHS England – DCB0129** (the manufacturer's clinical risk management) and **DCB0160** (the deployer's) | Mandatory clinical safety standards for health IT in England. They require a named Clinical Safety Officer, a hazard log and a safety case. | Current versions in force. Revision consultation ran 29 June–11 September 2026, with AI named as a driver. | AGREEMENT (current); PROPOSED (revision) | [NHS England consultation](https://www.engage.england.nhs.uk/patient-safety/national-review-of-clinical-risk-management-standa/) (OBSERVED); [DCB0129 standard](https://adultsocialcare.standards.nhs.uk/published-standards/clinical-risk-management-its-application-in-the-manufacture-of-health-it-systems); [DCB0160 standard](https://standards.nhs.uk/published-standards/clinical-risk-management-its-application-in-the-deployment-and-use-of-health-it-systems), accessed 2026-10-31 | | **DTAC** (Digital Technology Assessment Criteria) | The NHS buyer's baseline checklist. It covers clinical safety, data protection, technical assurance, interoperability, and usability and accessibility. | New form from 24 February 2026, with about 25% fewer questions. The old form was retired on 6 April 2026. | AGREEMENT | [NHS Innovation Service](https://innovation.nhs.uk/news/updated-nhs-england-digital-technology-assessment-criteria-dtac-form-and-guidance/); [HTN](https://htn.co.uk/2026/02/25/digital-technology-assessment-criteria-refreshed-to-create-simpler-more-trusted-pathway-for-nhs-digital-innovation/), accessed 2026-10-31 | | **DSPT** (Data Security and Protection Toolkit) | An annual self-assessment that any organisation accessing NHS patient data must complete. Trusts and ICBs (integrated care boards) use a version aligned to the NCSC (National Cyber Security Centre) CAF (Cyber Assessment Framework). IT suppliers were reported to stay on the non-CAF version for 2025/26. | 2025/26 deadline 30 June 2026. Version 9 (2026/27) changes not checked. | AGREEMENT; UNKNOWN (2026/27 supplier detail) | [DSPT help](https://www.dsptoolkit.nhs.uk/Help/64); [Dionach](https://dionach.com/data-security-and-protection-toolkit-dspt-2025-2026-caf/), accessed 2026-10-31 | | NHS England – ambient scribing guidance and AVT registry | Guidance first published April 2025. NHS England withdrew its earlier statement that summarisation tools are likely devices, to match the MHRA. The registry launched January 2026. | Updated July 2026 | AGREEMENT (guidance) | [NHS England guidance](https://www.england.nhs.uk/long-read/guidance-on-the-use-of-ai-enabled-ambient-scribing-products-in-health-and-care-settings); [HTN](https://htn.co.uk/2026/07/29/mhra-clarifies-regulatory-medical-devices-status-of-ambient-voice-technologies/), accessed 2026-10-31 | | **CQC** (Care Quality Commission) | Registers providers of "regulated activities". A 2022 scope update says AI suppliers that analyse and report on screening images may need to register for "diagnostic and screening procedures" if no CQC-registered provider is doing so. | No AI-specific 2025–2026 CQC guidance found | AGREEMENT (2022 scope); UNKNOWN (newer) | [Osborne Clarke](https://osborneclarke.com/insights/updated-cqc-guidance-has-implications-health-technology-and-wellbeing-industries); [NHS digital regulations guide](https://www.digitalregulations.innovation.nhs.uk/regulations-and-guidance-for-developers/all-developers-guidance/regulated-activities-check-if-you-need-to-register-with-the-care-quality-commission-cqc/pdf/), accessed 2026-10-31 | | **NICE – ESF** (Evidence Standards Framework) for digital health technologies | Sets evidence tiers in proportion to risk, and includes AI and data-driven tools. A further revision is expected to follow MHRA reforms. Early Value Assessments let technologies be used while evidence is gathered. | ESF in use; next update pending | AGREEMENT; PROPOSED (update) | [Fieldfisher](https://www.fieldfisher.com/en/insights/new-changes-to-evidence-standards-framework-and-digital-health-technologies); [JMIR viewpoint](https://mhealth.jmir.org/2025/1/e67435); [ISPOR poster](https://www.ispor.org/heor-resources/presentations-database/presentation-cti/ispor-europe-2026/poster-session-5-4/learning-while-using-evidence-expectations-under-nice-s-reformed-healthtech-program), accessed 2026-10-31 | | **ICO** (Information Commissioner's Office) – UK GDPR, AI guidance | A DPIA (data protection impact assessment) is needed for AI processing health data at scale. AI inferences about health can count as "special category" data. | AI guidance updated March 2023; a 2026 refresh was not confirmed | AGREEMENT | [ICO guidance](https://www.ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/artificial-intelligence/guidance-on-ai-and-data-protection); [Hunton](https://www.hunton.com/privacy-and-cybersecurity-law-blog/uk-ico-issues-updated-guidance-on-ai-and-data-protection); [NHS DPIA step](https://digitalregulations.innovation.nhs.uk/regulations-and-guidance-for-developers/all-developers-guidance/how-to-comply-with-the-uk-gdpr-as-a-developer/step-2-do-a-data-protection-impact-assessment-dpia/pdf/), accessed 2026-10-31 | | Data (Use and Access) Act 2025 – automated decision-making | Most reforms came into force on 5 February 2026. Solely automated significant decisions are now restricted mainly where special category data (such as health) is involved, so health stays strict. | In force 5 February 2026. ICO ADM guidance was consulted on in 2026. | AGREEMENT; PROPOSED (ICO guidance) | [GOV.UK](https://www.gov.uk/guidance/data-use-and-access-act-2025-data-protection-and-privacy-changes); [Travers Smith](https://www.traverssmith.com/knowledge/knowledge-container/uks-data-protection-reforms-take-effect-a-new-era-for-automated-decision-making/); [Kennedys](https://www.kennedyslaw.com/en/thought-leadership/article/2026/the-data-use-and-access-act-2025-commencement-dates-and-planned-guidance-for-2026), accessed 2026-10-31 | | **Caldicott Principles / National Data Guardian (NDG)** | Eight principles for using confidential patient information, listed as legally required in the NHS digital regulations service. The NDG's 2024/25 report mentions input to AI scribe guidance. | Principles in force. No NDG AI-specific guidance found. | AGREEMENT; UNKNOWN (AI guidance) | [NHS digital regulations](https://digitalregulations.innovation.nhs.uk/regulations-and-guidance-for-adopters/all-adopters-guidance/complying-with-the-uk-gdpr-steps-1-7-an-introduction/step-7-follow-the-8-caldicott-principles/pdf/); [NDG annual report](https://assets.publishing.service.gov.uk/media/6942a76f9273c48f554cf506/NDG_annual_report_2024-25_v1.0_-_17.12.25.pdf), accessed 2026-10-31 | | **HRA** (Health Research Authority) | HRA approval covers project-based research in the NHS in England and Wales. Service evaluation may not need ethics review; use the HRA decision tool. CAG (Confidentiality Advisory Group) approval may be needed for identifiable data without consent. | In force | AGREEMENT | [HRA approval](https://www.hra.nhs.uk/approvals-amendments/what-approvals-do-i-need/hra-approval/); [HRA: what approvals](https://www.hra.nhs.uk/approvals-amendments/what-approvals-do-i-need/), accessed 2026-10-31 | ###### 2.2 European Union | Rule | What it means for medical AI | Dates | Label | Source | |---|---|---|---|---| | **EU MDR** (Medical Device Regulation 2017/745) | The device rules in the EU. Its classification decides whether the AI Act's high-risk rules apply. | In force | AGREEMENT | [EUR-Lex](https://eur-lex.europa.eu/eli/reg/2017/745/oj), accessed 2026-10-31 | | **EU AI Act** (Regulation 2024/1689) | Risk-based AI law. AI that is a medical device, or a safety component of one, and needs notified-body assessment is "high-risk". | Original high-risk date for Annex I products was 2 August 2027, now moved (next row) | AGREEMENT | [EUR-Lex](https://eur-lex.europa.eu/eli/reg/2024/1689/oj); [OpenRegulatory on MDCG 2025-6](https://openregulatory.com/mdcg/mdcg-2025-6), accessed 2026-10-31 | | **Digital Omnibus on AI** (Regulation 2026/1744) | Moves the high-risk deadlines. Annex III (standalone) to 2 December 2027; Annex I (AI in regulated products such as medical devices) to 2 August 2028. Transparency obligations start 2 August 2026, and some content-marking duties 2 December 2026. | In force 27 July 2026 | AGREEMENT | [Hunton](https://www.hunton.com/privacy-and-cybersecurity-law-blog/eu-digital-omnibus-on-ai-enters-into-force) (OBSERVED, law firm); [Gibson Dunn](https://www.gibsondunn.com/eu-ai-act-omnibus-agreement-postponed-high-risk-deadlines-and-other-key-changes/), accessed 2026-10-31 | | **MDCG 2025-6** (FAQ from the Medical Device Coordination Group and the AI Board) | Explains how the MDR and the AI Act fit together. The MDR class drives AI Act applicability, and the notified body assesses both together. | Published June 2025 (exact date disputed) | AGREEMENT (guidance) | [OpenRegulatory](https://openregulatory.com/mdcg/mdcg-2025-6), accessed 2026-10-31 | | **EHDS** (European Health Data Space, Regulation 2025/327) | Patient access to records, EHR (electronic health record) system requirements and secondary use of data. | In force 26 March 2025. Main application 2027–2031 in phases. | AGREEMENT (phased) | [Kennedys](https://www.kennedyslaw.com/en/thought-leadership/article/2026/the-european-health-data-space-is-in-force-implications-for-healthcare-medtech-and-life-sciences); [A&O Shearman](https://www.aoshearman.com/en/insights/ao-shearman-on-life-sciences/european-health-data-space-regulation-sets-requirements-for-electronic-health-record-systems), accessed 2026-10-31 | | **Product Liability Directive** (2024/2853) | Software and AI become "products" under strict liability. Courts can presume a product is defective where technical complexity makes proof too hard. | Member states must transpose it by 9 December 2026 | AGREEMENT (transposition due) | [Covington Inside Privacy](https://www.insideprivacy.com/uncategorized/eu-member-states-begin-rolling-out-new-product-liability-rules/); [Hogan Lovells](https://www.hoganlovells.com/en/publications/eu-introduces-comprehensive-digitalera-product-liability-directive), accessed 2026-10-31 | ###### 2.3 United States | Rule | What it means | Status | Label | Source | |---|---|---|---|---| | FDA – Clinical Decision Support (CDS) software guidance | Revised 6 January 2026 without a comment period. It relaxes oversight of some tools that recommend a single option to a clinician. | Final guidance (replaces 2022) | AGREEMENT | [King & Spalding](https://www.kslaw.com/insights/articles/fda-updates-general-wellness-and-clinical-decision-support-guidance-documents); [Arnold & Porter](https://www.arnoldporter.com/en/perspectives/advisories/2026/01/fda-cuts-red-tape-on-clinical-decision-support-software), accessed 2026-10-31 | | FDA – General Wellness guidance | Revised 6 January 2026. Adds criteria for non-invasive sensors, such as wrist-worn blood pressure estimates. | Final | AGREEMENT | [King & Spalding](https://www.kslaw.com/insights/articles/fda-updates-general-wellness-and-clinical-decision-support-guidance-documents), accessed 2026-10-31 | | FDA – PCCP for AI-enabled device software functions | Lets a maker pre-agree how a model will change: description of the changes, a modification protocol and an impact assessment. | Final, December 2024 | AGREEMENT | [Ropes & Gray](https://ropesgray.com/en/insights/alerts/2024/12/fda-finalizes-guidance-on-predetermined-change-control-plans-for-ai-enabled-device), accessed 2026-10-31 | | FDA – AI-enabled device software functions: lifecycle and submissions | Draft from January 2025. Finalising it is a top priority on the FY2027 (US fiscal year 2027) agenda. | Draft | PROPOSED | [RAPS](https://www.raps.org/resource/fda-s-device-center-releases-guidance-agenda-for-fy-2027.html); [CAP](https://www.cap.org/advocacy/latest-news-and-practice-data/fda-releases-new-artificial-intelligence-draft-guidance), accessed 2026-10-31 | | FDA – AI-enabled device list | FDA lists over 1,600 AI-enabled devices and says the list is not comprehensive. A commentator notes none of the authorised devices it cites are generative or agentic. | Live list | OBSERVED (page exists); CLAIM (count, commentary) | [FDA list](https://www.fda.gov/medical-devices/artificial-intelligence-enabled-medical-devices/list-artificial-intelligence-enabled-medical-devices); [Clearwater](https://clearwatersecurity.com/transcripts/agentic-ai-in-healthcare-navigating-regulatory-uncertainty-and-building-governance-that-lasts/), accessed 2026-10-31 | | **HIPAA** (Health Insurance Portability and Accountability Act) Security Rule update | The proposed overhaul (published January 2025) is not final. HHS (Department of Health and Human Services) moved the target to July 2027. The current rule still applies. | Proposed | PROPOSED | [Fierce Healthcare](https://www.fiercehealthcare.com/health-tech/feds-push-back-hipaa-security-rule-overhaul-july-2027); [Davis Wright Tremaine](https://www.dwt.com/blogs/privacy--security-law-blog/2026/07/hhs-updates-hipaa-rulemaking-timeframes), accessed 2026-10-31 | | ONC (Office of the National Coordinator for Health IT) – Interoperability Standards Platform AI appendix | Describes MCP (Model Context Protocol) over FHIR, A2A (agent-to-agent) standards, human sign-off on writes, and audit logging that links AI requests to users, agents and organisations. Says identity and delegation are "still evolving". | Reference content, no binding status | OBSERVED | [ONC ISP](https://isp.healthit.gov/appendix-vi-ai-interoperability), accessed 2026-10-31 | ###### 2.4 Worldwide | Rule | What | Status | Label | Source | |---|---|---|---|---| | WHO guidance on LMMs (large multi-modal models) for health | More than 40 recommendations, including that governments assign a regulator to approve LMMs used in care, and independent post-release audits for large-scale deployments. | Published January 2024 | AGREEMENT (non-binding guidance) | [WHO](https://www.who.int/publications-detail-redirect/9789240084759), accessed 2026-10-31 | ###### 2.5 Applying the MHRA test to EDGE Clinic (research reading, not legal advice) | EDGE Clinic function | Likely device? | Why | Source | |---|---|---|---| | Signposting to a social prescribing service, with no clinical claim | Probably not (UNKNOWN until assessed) | It does not "support diagnosis, treatment or prevention" | [MHRA](https://www.gov.uk/government/news/mhra-clarifies-regulatory-status-of-ambient-voice-technologies-used-in-the-nhs), accessed 2026-10-31 | | Triage, including online A&E (accident and emergency) routing | Likely yes | It supports diagnosis or decisions about urgency | same | | Retinal screening AI | Yes | A diagnostic or screening purpose | same; [FDA LumineticsCore example](https://www.digitaldiagnostics.com/resources/insights/lumineticscore-engineered-by-digital-diagnostics-early-detection-of-diabetic-retinopathy-using-ai/), accessed 2026-10-31 | | Prescribing safety (interaction checks) | Likely yes | It supports treatment decisions | [MHRA](https://www.gov.uk/government/news/mhra-clarifies-regulatory-status-of-ambient-voice-technologies-used-in-the-nhs), accessed 2026-10-31 | | Any agent that books, orders or refers without clinician review | Yes | "take automated action such as placing orders without clinician review" | same (OBSERVED quote) | ###### 2.6 Date line (in force versus coming) | Date | What | Label | Source | |---|---|---|---| | 26 March 2025 | EHDS in force (applies in phases from 2027) | AGREEMENT | [Kennedys](https://www.kennedyslaw.com/en/thought-leadership/article/2026/the-european-health-data-space-is-in-force-implications-for-healthcare-medtech-and-life-sciences), accessed 2026-10-31 | | 16 June 2025 | GB post-market surveillance rules apply | AGREEMENT | [GOV.UK](https://www.gov.uk/government/news/first-major-overhaul-of-medical-device-regulation-comes-into-force-across-great-britain), accessed 2026-10-31 | | 6 January 2026 | FDA CDS and wellness guidance revised | AGREEMENT | [King & Spalding](https://www.kslaw.com/insights/articles/fda-updates-general-wellness-and-clinical-decision-support-guidance-documents), accessed 2026-10-31 | | 5 February 2026 | UK DUAA automated decision-making reforms in force | AGREEMENT | [Travers Smith](https://www.traverssmith.com/knowledge/knowledge-container/uks-data-protection-reforms-take-effect-a-new-era-for-automated-decision-making/), accessed 2026-10-31 | | 6 April 2026 | Old DTAC form retired | AGREEMENT | [NHS Innovation Service](https://innovation.nhs.uk/news/updated-nhs-england-digital-technology-assessment-criteria-dtac-form-and-guidance/), accessed 2026-10-31 | | 27 July 2026 | EU Digital Omnibus on AI in force | AGREEMENT | [Hunton](https://www.hunton.com/privacy-and-cybersecurity-law-blog/eu-digital-omnibus-on-ai-enters-into-force), accessed 2026-10-31 | | 29 July 2026 | MHRA AVT guidance | AGREEMENT | [MHRA](https://www.gov.uk/government/news/mhra-clarifies-regulatory-status-of-ambient-voice-technologies-used-in-the-nhs), accessed 2026-10-31 | | 2 August 2026 | EU AI Act transparency obligations | AGREEMENT | [Hunton](https://www.hunton.com/privacy-and-cybersecurity-law-blog/eu-digital-omnibus-on-ai-enters-into-force), accessed 2026-10-31 | | 9 December 2026 | EU Product Liability Directive transposition deadline | AGREEMENT | [Covington](https://www.insideprivacy.com/uncategorized/eu-member-states-begin-rolling-out-new-product-liability-rules/), accessed 2026-10-31 | | End 2026 | MHRA draft PCCP guidance (planned) | PROPOSED | [Qualio](https://www.qualio.com/blog/mhra-s-2026-devices-reform-what-changes-and-when), accessed 2026-10-31 | | June 2027 | UK Medical Devices (Amendment) Regulations expected to apply | PROPOSED | [Latham & Watkins](https://www.lw.com/en/insights/mhra-publishes-draft-amendment-to-the-uk-medical-devices-regulations), accessed 2026-10-31 | | July 2027 | HIPAA Security Rule final (target) | PROPOSED | [Fierce Healthcare](https://www.fiercehealthcare.com/health-tech/feds-push-back-hipaa-security-rule-overhaul-july-2027), accessed 2026-10-31 | | 2 December 2027 | EU AI Act Annex III high-risk rules apply | AGREEMENT | [Hunton](https://www.hunton.com/privacy-and-cybersecurity-law-blog/eu-digital-omnibus-on-ai-enters-into-force), accessed 2026-10-31 | | April 2028 / 2028 | NHS App "full front door" and My NHS GP full rollout (targets) | PROPOSED | [Full Fact](https://fullfact.org/government-tracker/nhs-app-improvements/); [Digital Health](https://www.digitalhealth.net/2025/07/what-does-the-10-year-plan-mean-for-digital/), accessed 2026-10-31 | | 2 August 2028 | EU AI Act high-risk rules for AI in medical devices apply | AGREEMENT | [Hunton](https://www.hunton.com/privacy-and-cybersecurity-law-blog/eu-digital-omnibus-on-ai-enters-into-force), accessed 2026-10-31 | --- ##### 3. Agent assurance gaps in health "Agent" here means software that acts with or for a patient (books, refers, checks, messages), not only one that answers questions. | Gap | Evidence that it is a gap | What exists today to build on | What EDGE Clinic would need | Label | |---|---|---|---|---| | **Agent identity** (who or what is acting) | The OpenID Foundation names "agent identity fragmentation", "user impersonation" and "recursive delegation" as gaps ([OIDF](https://openid.net/new-whitepaper-tackles-ai-agent-identity-challenges/)). A 2026 survey says no production-ready standard traces authorisation through multi-hop agent chains ([arXiv 2604.23280](https://arxiv.org/pdf/2604.23280)). Accessed 2026-10-31. | OAuth token exchange "on-behalf-of" flows, and the OIDF's draft OpenID Connect Authority Claims Extension ([OIDF at ITU-T](https://openid.net/oidf-presents-at-itu-t-workshop-on-digital-id-and-agentic-ai/), accessed 2026-10-31). The FHIR `Device` resource can represent an AI system ([HL7 AI Transparency](https://hl7.org/fhir/uv/aitransparency/2026Jan/artifacts.html), accessed 2026-10-31). | A first-class agent identity issued by EDGE ID and bound to a human or organisation principal | PROPOSED (standards); OBSERVED (gap) | | **Patient identity and verification** | NHS login does not do session management or logout, and does not replace PDS (the Personal Demographics Service) ([NHS login, archived](https://tnaqa.mirrorweb.com/ukgwa/20260505142011mp_/https://digital.nhs.uk/developer/api-catalogue/nhs-login), accessed 2026-10-31). | NHS login uses OpenID Connect, with P0/P5/P9 verification levels (P9 means checked against photo ID) | Use NHS login P9 for record access, and keep the agent session separate | AGREEMENT | | **Mandate** (what the agent may do, for whom, until when) | A health-system preprint warns of "tool permissions that persist beyond the original use case" and "unclear accountability" ([arXiv 2601.15630](https://arxiv.org/pdf/2601.15630), accessed 2026-10-31). ONC says identity, authorisation and delegation are "still evolving" ([ONC ISP](https://isp.healthit.gov/appendix-vi-ai-interoperability), accessed 2026-10-31). | SMART on FHIR scopes ([SMART App Launch 2.2](https://hl7.org/fhir/smart-app-launch/apps.html), accessed 2026-10-31). No health standard for agent mandates. | A machine-readable mandate (scope, purpose, expiry, revocation) checked on every action | OBSERVED (gap) | | **Consent** | The OIDF names "consent fatigue" ([OIDF](https://openid.net/new-whitepaper-tackles-ai-agent-identity-challenges/), accessed 2026-10-31). OpenAI asks before using connected records by default ([OpenAI via Fierce](https://fiercehealthcare.com/ai-and-machine-learning/openai-launches-chatgpt-health-connect-data-health-apps-medical-records), accessed 2026-10-31). The UK's solely-automated-decision limits remain where health data is involved ([Travers Smith](https://www.traverssmith.com/knowledge/knowledge-container/uks-data-protection-reforms-take-effect-a-new-era-for-automated-decision-making/), accessed 2026-10-31). | The FHIR Consent resource (part of the FHIR R4 base specification, [HL7 FHIR R4 licence page](https://hl7.org/fhir/R4/license.html), accessed 2026-10-31). The Caldicott principles. | Consent captured per purpose and per agent, shown to the patient and revocable | AGREEMENT (law); OBSERVED (gap in practice) | | **Escalation to a human** | People using LLMs did no better than controls, with a "two-way communication breakdown" ([Oxford](https://www.ox.ac.uk/news/2026-02-10-new-study-warns-risks-ai-chatbots-giving-medical-advice), accessed 2026-10-31). ECRI's No. 1 hazard for 2026 is misuse of AI chatbots ([ECRI](https://home.ecri.org/blogs/ecri-news/misuse-of-ai-chatbots-tops-annual-list-of-health-technology-hazards), accessed 2026-10-31). The MHRA treats action "without clinician review" as device territory ([MHRA](https://www.gov.uk/government/news/mhra-clarifies-regulatory-status-of-ambient-voice-technologies-used-in-the-nhs), accessed 2026-10-31). | DCB0129 hazard logs. ONC's pattern of "write proposals" with human approval ([ONC ISP](https://isp.healthit.gov/appendix-vi-ai-interoperability), accessed 2026-10-31). | Hard-coded red-flag escalation, human sign-off on any clinical action, and user testing before launch (as the Oxford authors recommend) | OBSERVED (gap) | | **Liability** | No UK AI liability law was found. The Commission's report was summarised as patients wanting "clear accountability if things go wrong" ([GOV.UK](https://www.gov.uk/government/news/independent-commission-led-by-nhs-doctors-sets-out-blueprint-to-accelerate-safe-ai-adoption-in-healthcare), accessed 2026-10-31). In the EU, the new PLD (Product Liability Directive) makes software a product with strict liability ([Hogan Lovells](https://www.hoganlovells.com/en/publications/eu-introduces-comprehensive-digitalera-product-liability-directive), accessed 2026-10-31). Commentators say accountability for multi-step agent actions is "unsettled" ([Clearwater](https://clearwatersecurity.com/transcripts/agentic-ai-in-healthcare-navigating-regulatory-uncertainty-and-building-governance-that-lasts/), accessed 2026-10-31). | Device manufacturer liability, plus clinician duty of care. OpenMRS's licence puts responsibility on the clinician ([ScanCode](https://scancode-licensedb.aboutcode.org/openmrs-exception-to-mpl-2.0.html), accessed 2026-10-31). | A named accountable human or organisation for every agent mandate, written in the audit | UNKNOWN (UK law); AGREEMENT (EU PLD) | | **Audit a regulator can re-run** | The Commission recommends continuous real-world monitoring and a public searchable safety database ([GOV.UK](https://www.gov.uk/government/news/independent-commission-led-by-nhs-doctors-sets-out-blueprint-to-accelerate-safe-ai-adoption-in-healthcare), accessed 2026-10-31). The WHO recommends independent post-release audits ([WHO](https://www.who.int/publications-detail-redirect/9789240084759), accessed 2026-10-31). No standard for replaying an agent's decisions was found. | HL7 "AI Transparency on FHIR": Provenance links AI output back to the model, the input prompt and the human verifier. It is a ballot/continuous build, not yet published ([HL7 AI Transparency](https://hl7.org/fhir/uv/aitransparency/2026Jan/artifacts.html); [requirements](https://build.fhir.org/ig/HL7/aitransparency-ig/en/requirements.html), accessed 2026-10-31). FDA PCCPs for planned model change ([Ropes & Gray](https://ropesgray.com/en/insights/alerts/2024/12/fda-finalizes-guidance-on-predetermined-change-control-plans-for-ai-enabled-device), accessed 2026-10-31). | A content-addressed log of inputs, model version, prompt, tools called, mandate and human sign-off, so a regulator can replay a case | PROPOSED (standard); OBSERVED (gap) | | **Tool-calling standard** | MCP was reportedly donated to the Linux Foundation's Agentic AI Foundation in December 2025 ([Keragon](https://www.keragon.com/blog/medical-mcp-servers), accessed 2026-10-31). There is no official HL7 FHIR MCP profile, only community RFCs ([GitHub discussion](https://github.com/HeliosSoftware/hfs/discussions/231), accessed 2026-10-31). | MCP over FHIR with SMART auth (ONC pattern) | Wrap FHIR and NHS APIs as tools with per-mandate scopes | CLAIM / PROPOSED | | **Regulator view of generative and agentic devices** | None of the FDA-authorised AI devices cited by one commentator is generative or agentic ([Clearwater](https://clearwatersecurity.com/transcripts/agentic-ai-in-healthcare-navigating-regulatory-uncertainty-and-building-governance-that-lasts/), accessed 2026-10-31). The UK draft regulations do not yet update software-specific provisions ([Qualio](https://www.qualio.com/blog/mhra-s-2026-devices-reform-what-changes-and-when), accessed 2026-10-31). | The AI Airlock sandbox ([GOV.UK](https://www.gov.uk/government/collections/ai-airlock-the-regulatory-sandbox-for-aiamd), accessed 2026-10-31) | Engage the MHRA early (Airlock Phase 3, if it opens) | UNKNOWN | --- ##### 4. Reusable systems: open standards and open-source SPDX is the standard short-code for licences (for example `Apache-2.0`). Where there is no SPDX code, `LicenseRef-…` marks a custom licence. ###### 4.1 Standards and specifications | Item | What it is | Licence (SPDX) | Maturity | Fit for EDGE Clinic | Label | Source | |---|---|---|---|---|---|---| | **HL7 FHIR R4** | Fast Healthcare Interoperability Resources: the web API data standard for health records (Patient, Observation, MedicationRequest, Consent, Provenance, AuditEvent and so on). | `CC0-1.0`. SNOMED CT, LOINC and DICOM content are excluded and need their own terms. | Normative parts; widely used | Core data model and API | OBSERVED | [HL7 FHIR R4 licence](https://hl7.org/fhir/R4/license.html), accessed 2026-10-31 | | **FHIR UK Core** | UK profiles of FHIR R4 (HL7 UK and NHS England). STU2 (standard for trial use 2) is release 2.0; 2.0.2 is the latest seen. | UNKNOWN (not checked) | STU (trial use) | Required shape for NHS integrations | CLAIM | [HL7 UK](https://www.hl7.org.uk/2024/05/28/fhir-uk-core-release-2-0-stu2/); [Simplifier](https://simplifier.net/guide/uk-core-implementation-guide-stu2), accessed 2026-10-31 | | **SMART on FHIR – App Launch 2.2** | OAuth 2.0 / OpenID Connect profile for apps and backend services that access FHIR. | Spec: UNKNOWN. Reference JS client: `Apache-2.0` (OBSERVED). | STU 2.2, current published | The authorisation layer for agents (scopes) | OBSERVED (client licence) | [HL7 SMART App Launch](https://hl7.org/fhir/smart-app-launch/apps.html); [client-js LICENSE](https://raw.githubusercontent.com/smart-on-fhir/client-js/master/LICENSE), accessed 2026-10-31 | | **HL7 AI Transparency on FHIR** | Provenance, labels and model cards for AI-generated FHIR data. | UNKNOWN (an HL7 IG; not checked) | Ballot / continuous build, "not an authorized publication" | The best fit for replayable AI audit | CLAIM | [HL7](https://hl7.org/fhir/uv/aitransparency/2026Jan/artifacts.html); [build](https://build.fhir.org/ig/HL7/aitransparency-ig/en/requirements.html), accessed 2026-10-31 | | **SNOMED CT (UK Edition)** | Clinical terminology. dm+d codes are SNOMED CT codes. | `LicenseRef-SNOMED-CT-UK-Affiliate`: you accept both the UK National Release and the International affiliate licences via TRUD (NHS England's terminology download service). Fees and commercial terms were not confirmed. | Mature | Mandatory for UK clinical coding | OBSERVED (licence pages exist); UNKNOWN (commercial terms) | [TRUD UK licence](https://isd.hscic.gov.uk/trud/users/guest/filters/0/licence/10); [TRUD International licence](https://isd.hscic.gov.uk/trud/users/guest/filters/0/licence/4), accessed 2026-10-31 | | **dm+d** (Dictionary of Medicines and Devices) | The NHS standard list of medicines and devices, from the NHSBSA (NHS Business Services Authority), downloaded from TRUD. | Open Government Licence. Version not stated in the source; `OGL-UK-3.0` is likely but UNKNOWN. | Mature, weekly | Medicine identity for prescribing safety (identity, not interactions) | CLAIM | [NHSBSA](https://www.nhsbsa.nhs.uk/pharmacies-gp-practices-and-appliance-contractors/nhs-dictionary-medicines-and-devices-dmd/snomed-ct-and-dmd); [NHSBSA dm+d](https://nhsbsa.nhs.uk/pharmacies-gp-practices-and-appliance-contractors/dictionary-medicines-and-devices-dmd), accessed 2026-10-31 | | **openEHR** | A clinical data modelling standard (archetypes and templates). | Specification repo: `CC-BY-SA-3.0` | Mature, strongest in Europe | Optional long-term record store; FHIR is the better fit for UK integration | OBSERVED | [openEHR specifications-RM LICENSE](https://raw.githubusercontent.com/openEHR/specifications-RM/master/LICENSE), accessed 2026-10-31 | | **IHE profiles** (Integrating the Healthcare Enterprise) | Integration profiles, such as audit trails and document sharing. | Documents: `CC-BY-4.0`. Patents are handled separately, on reasonable-and-non-discriminatory terms. | Mature | Audit and identity patterns | CLAIM | [IHE Europe IP](https://ihe-europe.net/intellectual-property); [IHE General Introduction](https://profiles.ihe.net/GeneralIntro/ch-11.html), accessed 2026-10-31 | | **OMOP CDM / OHDSI vocabularies** (Observational Medical Outcomes Partnership Common Data Model; Observational Health Data Sciences and Informatics) | A research data model and vocabularies for population analytics. | Vocabularies: `Apache-2.0`, with some licensed vocabularies (for example FDB) needing proof of a licence. CDM spec: UNKNOWN. | Mature | Detection analytics and evaluation, not care delivery ("should not be used for individual patient healthcare") | CLAIM | [HL7 THO OMOP](https://build.fhir.org/ig/HL7/UTG/en/NamingSystem-OMOP.html), accessed 2026-10-31 | | **MCP** (Model Context Protocol) | An open protocol for connecting agents to tools and data. | UNKNOWN (not checked) | Emerging | The tool layer for agents over FHIR | CLAIM | [ONC ISP](https://isp.healthit.gov/appendix-vi-ai-interoperability); [Keragon](https://www.keragon.com/blog/medical-mcp-servers), accessed 2026-10-31 | ###### 4.2 Open-source software | Project | What it is | Licence (SPDX) | Maturity | Fit | Label | Source | |---|---|---|---|---|---|---| | **HAPI FHIR** | Java FHIR server and library | `Apache-2.0` | Mature, widely used | A self-hosted FHIR store and validator | OBSERVED | [LICENSE.txt](https://raw.githubusercontent.com/hapifhir/hapi-fhir/master/LICENSE.txt), accessed 2026-10-31 | | **Medplum** | TypeScript FHIR-native platform (server, SDKs, React components) | `Apache-2.0` | Production. ONC certification (dated 31 December 2025) is claimed, but disputed for self-hosted copies. | Fastest route to a FHIR backend with auth and bots | OBSERVED (licence); CLAIM (certification) | [LICENSE.txt](https://raw.githubusercontent.com/medplum/medplum/main/LICENSE.txt); [Mindbowser](https://www.mindbowser.com/open-source-ehr/), accessed 2026-10-31 | | **OpenMRS** | Open-source medical record system, strong in low- and middle-income countries | `MPL-2.0` plus a Healthcare Disclaimer (`LicenseRef-OpenMRS-Exception-to-MPL-2.0`) | Mature | Global and offline settings, not NHS primary care | OBSERVED (MPL-2.0); CLAIM (disclaimer) | [LICENSE](https://raw.githubusercontent.com/openmrs/openmrs-core/master/LICENSE); [ScanCode](https://scancode-licensedb.aboutcode.org/openmrs-exception-to-mpl-2.0.html), accessed 2026-10-31 | | **EHRbase** | openEHR clinical data repository | `Apache-2.0` | Production | Only if the openEHR route is chosen | OBSERVED | [LICENSE](https://raw.githubusercontent.com/ehrbase/ehrbase/develop/LICENSE), accessed 2026-10-31 | | **OHDSI WebAPI / ATLAS** | OMOP analytics tools | `Apache-2.0` | Mature | Population detection analytics and research | OBSERVED | [WebAPI LICENSE](https://raw.githubusercontent.com/OHDSI/WebAPI/master/LICENSE); [Atlas LICENSE](https://raw.githubusercontent.com/OHDSI/Atlas/master/LICENSE), accessed 2026-10-31 | ###### 4.3 NHS national services (reuse by integration, not by code) | Service | What | Licence | Access and maturity | Fit | Label | Source | |---|---|---|---|---|---|---| | **NHS login** | Patient identity (OpenID Connect). Verification levels P0, P5 and P9. Partners must meet the identity standard DAPB3051. | n/a (service; partner onboarding) | Live, national | Patient sign-in and verified identity | CLAIM (archived NHS page) | [NHS login, archived](https://tnaqa.mirrorweb.com/ukgwa/20260505142011mp_/https://digital.nhs.uk/developer/api-catalogue/nhs-login), accessed 2026-10-31 | | **GP Connect** (Access Record: Structured) | FHIR API to read structured GP records. Working version 1.5.0. | n/a (service) | Live, but the service page says there is "no longer an onboarding route" for new developers. It needs SCAL (the supplier conformance assessment list), a CSO (clinical safety officer) and a hazard log. | Would be high value; access is UNKNOWN for a new supplier | CLAIM / UNKNOWN | [NHS England DSIC wiki](https://nhse-dsic.atlassian.net/wiki/spaces/DCSDCS/pages/12723519489/GP+Connect+Access+Record+Structured), accessed 2026-10-31 | ###### 4.4 Recommended reuse stack (research suggestion only) - **Data and API:** FHIR R4 + UK Core on HAPI FHIR or Medplum (both `Apache-2.0`, OBSERVED above). - **Terminology:** SNOMED CT UK and dm+d from TRUD (licences above). - **Auth:** NHS login for patients, plus SMART scopes for agents. - **Audit:** FHIR Provenance and AuditEvent following the AI Transparency on FHIR IG. - **Analytics:** OMOP / OHDSI for detection evaluation. - **Interactions:** license BNF or FDB until an open, clinical-grade source exists (sources in 1.3). --- ##### 5. Open questions (UNKNOWN) to resolve next - Has the government formally responded to the National Commission? GOV.UK and Ropes & Gray differ. [GOV.UK](https://www.gov.uk/government/news/independent-commission-led-by-nhs-doctors-sets-out-blueprint-to-accelerate-safe-ai-adoption-in-healthcare); [Ropes & Gray](https://www.ropesgray.com/en/insights/viewpoints/2026/10/102o5fp/ai-medical-devices-in-great-britain-what-the-mhras-response-to-the-national-com), accessed 2026-10-31. - When will the revised DCB0129/DCB0160 be published, and what will they require for AI? [NHS England](https://www.engage.england.nhs.uk/patient-safety/national-review-of-clinical-risk-management-standa/), accessed 2026-10-31. - Is there any GP Connect onboarding route for a new supplier? [NHS England DSIC wiki](https://nhse-dsic.atlassian.net/wiki/spaces/DCSDCS/pages/12723519489/GP+Connect+Access+Record+Structured), accessed 2026-10-31. - What was the outcome of the UK NSC consultation on AI in diabetic eye screening? [AOP](https://www.aop.org.uk/ot/news/2026/05/13/uk-national-screening-committee-consults-on-use-of-ai-in-diabetic-eye-screening), accessed 2026-10-31. - What are the licences for UK Core, the SMART App Launch spec, MCP, DDInter and the OMOP CDM spec? These were not checked in primary sources. - What is the current UK status of Livi/Kry, Push Doctor and Doctaly? See the 1.1 sources. - Medscape's interaction checker and the Google ARDA retinal AI were not researched. ### research/clinic/ENGINE-MEDPLUM.md `research/clinic/ENGINE-MEDPLUM.md` · sha256 `5ca30ff729e16b58a32d711d9de691340451e4934811c84962e90e42c88326a1` #### EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record Built 1 Nov 2026 by ATBAN Claude-Jegudiel 00:5531102026-1 (service-builder lane). Code: `runtime/domains/clinic.py` (the domain, bound into `runtime/adapters.ADAPTERS` as `runtime/domains/bank.py` is), `runtime/plugs/medplum.py` (HTTP and OAuth), `tests/test_clinic_domain.py` (unit tests with a fake FHIR server; `LiveSmoke` against the real server when `MEDPLUM_SMOKE=1`). Every statement below is marked OBSERVATION (measured, with how and when), AGREEMENT (a decision taken here) or CLAIM. ##### 1. The engine (OBSERVATION, jude-1, 1 Nov 2026 04:5x–05:1xZ) - Medplum `5.2.13-15a306f` runs as podman container `jude-engine-medplum` in pod `jude-medplum-engines` with `jude-engine-medplum-db` (PostgreSQL) and `jude-engine-medplum-redis`; systemd units `jude-engine-medplum{,-db,-redis}.service` (enabled). Installed by Codex-Barachiel 20:4330102026-1 (`/opt/jude-reusable-engines/CONNECT-EXISTING-ENGINES.md`, `medplum-installation.json`, image `medplum/medplum-server@sha256:0c3f4ca2…684fd2`). Loopback only: `http://127.0.0.1:18904`; `/healthcheck` answers `{"ok":true,"version":"5.2.13-15a306f","postgres":true,"redis":true}`. Configuration under `/etc/jude-reusable-engines/` (root only; never read into chat). - Stop seen during this lane: at 04:48:05Z a `podman pod stop` (pid 1428339) stopped the three engine pods; systemd brought the Medplum units back at 04:51:42Z. After that restart the database container's `/dev/shm` was `drwx------ root` and every PostgreSQL backend failed with `could not open shared memory segment "/PostgreSQL.…": Permission denied` (Medplum answered 400 on every FHIR call, `/healthcheck` still said ok). Fix applied 05:0xZ: `chmod 1777 /dev/shm` inside `jude-engine-medplum-db`; FHIR calls answered 200 at once. Cause of the `podman pod stop` itself: not determinable from the journal (no sudo line names it); the engines' owner should confirm. The shm mode will revert on the next container start unless the pod's infra container or the unit sets it: an open item for the engine lane. - The seeded super-admin login still accepts Medplum's published first-run credentials (checked 04:5xZ: `POST /auth/login` 200). That is a security finding for the engine lane: rotate it. This lane did not change it (not its file) and does not use it at run time. ##### 2. Authentication (AGREEMENT: how; OBSERVATION: that it works; no values here) - Medplum Project `EDGE Clinic` (`strictMode: true`) created 1 Nov 2026 through the super-admin FHIR API, with one `ClientApplication` `edge-clinic-adapter` and its `ProjectMembership` (`/auth/me` as the client: project `EDGE Clinic`, profile `ClientApplication`; never the Super Admin project). `POST /admin/projects` answers 404 on this build; the FHIR route was used. - The adapter authenticates with OAuth 2.0 **client credentials** (`POST /oauth2/token`, `grant_type=client_credentials`, client id and secret in the form body; the HTTP Basic form answers 400 on this build). The token (3600 s) is cached in the process and renewed 60 s before expiry; a 401 triggers one fresh token and one retry. - The credential lives at `/etc/x-x/plugs/medplum.env` (`root:root 0640`; there is no `x-x` user or group on jude-1 and `/etc/x-x/plugs/griffin.key` does not exist yet, so the Griffin precedent could not be copied literally; the web service runs as root and reads it). Keys: `MEDPLUM_BASE_URL`, `MEDPLUM_PROJECT_ID`, `MEDPLUM_CLIENT_ID`, `MEDPLUM_CLIENT_SECRET`. The same names in the environment override the file (tests use that with fake values). No value is ever written to a record, a reply, the specification or Git. Without the file `medplum.bound()` is False and a live tenant's operation answers `no_answer` with `PLUG_REQUIRED` (UNKNOWN in the engine), never a pretended write. ##### 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`) Clinic has 4,151 receiver-labelled operations. 624 are bound to Medplum; 3,527 stay with the estate record or another receiver. (A count is live: `python3 -c "from runtime.domains import clinic, json; print(json.dumps(clinic.coverage()))"`.) | FHIR R4 resource | Bound operations | Estate entities and units (from `spec/receivers.json` labels `estate-record::`) | |---|---:|---| | Appointment (+ Location, Schedule, Slot when a time is chosen) | 89 | `record` of the booking units: book-an-appointment, book-for-the-person-i-care-for, book-a-pathway-test, book-from-the-referral-queue, book-an-interpreter, book-transport, rebook-or-cancel-an-appointment, offer-a-cancellation-slot, hold-a-slot-for-a-tertiary-handover, check-in-an-arriving-patient, check-in-on-arrival, front-door-check-in, confirm-attendance-and-prepare, attend-an-appointment-with-the-person, handle-a-did-not-attend, receive-a-did-not-attend-notice, record-attendance-outcome, add-an-appointment-to-my-calendar, schedule-the-scan, see-my-day; `feature_appointment` | | ServiceRequest | 88 | `referral`, `feature_referral`; referral and order units (make-an-urgent-same-day-referral, refer-on-a-suspected-cancer-pathway, refer-to-a-crisis-team, refer-to-tertiary-care, refer-to-the-mental-health-stream, refer-to-the-multidisciplinary-team, social-prescribing-referral, ask-for-advice-and-guidance, order-a-diagnostic-test, order-a-direct-access-test, order-a-genomic-test, request-a-second-read, request-a-second-read-or-discrepancy-review, request-a-second-opinion, track-a-referral, triage-a-referral, triage-a-mental-health-referral, close-a-referral, vet-an-imaging-request) | | Observation | 82 | `result`, `feature_result`, `feature_observation`; take-observations, record-a-point-of-care-urine-result, record-a-wound-or-skin-check, measure-a-nodule-and-its-growth, share-a-wearable-signal, wearable-signals, take-an-online-early-symptom-test, run-an-online-early-symptom-test, review-a-real-result, authorise-a-result, tell-the-patient-a-result, explain-a-result-to-the-patient, explain-a-result-in-plain-language, receive-a-critical-result-for-my-patient, phone-a-critical-result, view-my-results, view-results-in-scope, chase-missing-results, retinal-screening, grade-a-retinal-image | | Consent | 77 | `consent`, `feature_consent`; give-or-withdraw-consent, give-consent-on-behalf, take-informed-consent-for-research, apply-a-national-data-opt-out, record-sample-consent-version, set-my-data-sharing-choices, share-clinic-data-with-my-insurer, share-a-medicines-list-with-the-clinic, share-images-with-another-trust, approve-a-data-release | | CarePlan | 57 | `safety_plan`, `plan`, `feature_careplan`; write-a-care-plan, maintain-the-suicide-prevention-plan, follow-up-an-action-plan, set-a-surveillance-plan, discharge-from-the-pathway, discharge-from-the-stream, track-my-cancer-pathway, record-decision-to-treat | | QuestionnaireResponse | 36 | `assessment`, `answer`, `feature_questionnaire`; complete-a-pre-clinic-questionnaire, wellbeing-questionnaires, complete-a-mental-health-check-in, complete-a-risk-formulation, run-a-pre-assessment, score-and-interpret-a-response, act-on-a-high-risk-answer, symptom-check-and-triage, online-urgent-triage | | MedicationRequest | 33 | `feature_prescription`, `feature_medicationrequest`; prescribe, prescribe-under-the-pathway-protocol, order-a-repeat-medicine, order-repeat-medicines-on-behalf, review-a-repeat-request, clinically-check-a-prescription, check-a-chemotherapy-prescription | | Flag | 29 | `crisis_alert`, `flag`; flag-a-critical-finding, receive-a-crisis-alert, receive-a-crisis-alert-for-the-person, record-a-safeguarding-concern, raise-a-safeguarding-concern-from-a-response | | Patient | 25 | `feature_patient`; register-a-walk-in-patient, register-a-child-as-a-patient, update-patient-demographics | | Communication | 21 | message-my-care-team, send-a-results-ready-notice, send-appointment-letters-and-reminders, send-a-check-in-to-a-caseload, make-a-follow-up-call | | MedicationStatement | 18 | `medicine_list`, `feature_medication`; reconcile-medicines, reconcile-medicines-at-admission, view-my-medicines-list | | Specimen | 15 | take-a-blood-or-urine-sample, receive-and-book-in-a-sample, reject-an-unlabelled-sample, track-sample-custody | | Encounter | 10 | `check_in`, `feature_encounter`; record-a-clinic-consultation, hold-a-remote-session | | DiagnosticReport | 10 | `feature_diagnosticreport`; report-histology, issue-a-genomic-report | | MedicationDispense, Immunization, RelatedPerson, AllergyIntolerance, AdverseEvent | 5 each | dispense-a-medicine; give-a-vaccination; nominate-a-carer-or-proxy; record-an-adverse-drug-reaction; report-a-side-effect | | MedicationAdministration 4, Condition 3, Procedure 2 | | administer-a-medicine; `feature_condition`; `feature_procedure` | Operation classes covered, each with the receiver contract (create 159, read 30, change 383, remove 5, send 12, decide 27, plan 8 = 624, measured by `clinic.plan()`; `runtime/receivers.effective_class` still turns a "remove" of a workflow `record` into a change, as the estate record does): - **create** POST, then GET; confirmed with the Medplum id (the journey carries it on as `id` and `_id`). - **read** GET by any id in the inputs that is this resource in this tenant; else a search `_tag=` (newest first, 50). - **change** GET, merge, PUT; `expected_version`/`version` becomes `If-Match: W/""` and a stale one is `rejected VERSION_CONFLICT` (checked before the write and honoured by the server: 412). A change with no existing resource creates one (a journey's first touch), as the estate record does. The operation's verb moves the status: cancel/withdraw/ revoke, reject, did-not-attend, discharge/close/complete, hold, check-in/arrive, attend, amend, confirm/sign/submit/book/send/share/ authorise/approve/agree/pick-a (table `VERBS` and `STATUS`). - **remove** DELETE, then GET must answer 410 Gone; a cancelled Appointment frees its Slot first. - **send, decide, plan** like change, and the status lands on the resource's done state (`STATUS[resource]["done"]`); the outputs carry `state` and `state_field`. - **Confirmed** only when the read-back equals what was written (meta and narrative aside); **rejected** with `refusal_code` on any 4xx (OperationOutcome code upper-cased, else `INVALID_RESOURCE`, `NOT_FOUND`, `VERSION_CONFLICT`, `RECEIVER_FORBIDDEN`, `RATE_LIMITED`); **no_answer** on 5xx, on a transport failure (`TIMEOUT`, `UNREACHABLE`), on a read-back mismatch, or when no credential is bound. How inputs become FHIR: every input is kept verbatim in the extension `https://judiverse.tech/fhir/StructureDefinition/estate-inputs` (sorted JSON), so the read-after-write compares the whole input set exactly; on top, fields the unit declares (`data..fields` types) or names are placed structurally: a `datetime`/`date` field (or any value with a time of day) becomes `Appointment.start/end` (30-minute slot), `Observation.effectiveDateTime`, `Consent.dateTime`, `ServiceRequest.authoredOn`, `Immunization.occurrenceDateTime` and so on; a numeric input becomes `Observation.valueQuantity`; `name`, `date_of_birth`, `nhs_number` (10 digits, system `https://fhir.nhs.uk/Id/nhs-number`) fill `Patient`; the unit's label is the `code.text`/`serviceType`/`title`. Every resource carries `meta.tag {system: https://judiverse.tech/fhir/tenant, code: }` and `identifier {system: https://judiverse.tech/fhir/estate, value: //}` (a single identifier on AdverseEvent and QuestionnaireResponse, which are 0..1 in R4). The person behind the actor (or an agent's mandate owner) is one `Patient` per tenant found by identifier `https://judiverse.tech/fhir/edge-id` = `:`, created on first use, and referenced as subject/patient/participant. Booking a time: an Appointment that receives a start creates (once per tenant and clinic name) a `Location`, a `Schedule` whose actor is that Location, and a busy `Slot` for the 30 minutes, and references the Slot; cancelling frees the Slot. FHIR invariant app-3 ("only proposed or cancelled appointments can be missing start/end") is honoured: a confirm before a time is chosen leaves the Appointment `proposed`, never `booked` (OBSERVED: the live server rejects the alternative with 400 INVARIANT). ##### 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons) - `record` of the 496 Clinic units not in the map (219 capability units that are rotas, audits, incidents, governance, trials, genomics pipelines and the like, and 277 research, parity and problem units; 995 operations together) and every `case` (659: the workflow case is the estate's own, not a clinical resource); `evidence_query` (606) and the `identity_claim`s (28) are the estate's evidence and identity, not FHIR. - `feature_fhir` (579), `feature_screen`, `feature_chart`, `feature_manage`, `feature_ios`, `feature_web` and the other `feature_*` entities that are competitor-feature extractions, not clinical entities. A `feature_` is bound only when `` is a FHIR resource (Patient, Appointment, Observation, Condition, Encounter, Procedure, Immunization, CarePlan, CareTeam, MedicationRequest, AllergyIntolerance, DiagnosticReport, DocumentReference, Questionnaire(Response), Referral, Result, Prescription, Medication, Specimen, Consent, Flag, Communication, RelatedPerson, Slot, Schedule, Location). - The 112 operations whose receiver is an `external:` slot (NHS services, voice telephony, forecasting models…). - `session` (clinic sessions and therapy sessions share one word), `escalation`, `trusted_contact`, `concern`, `review`, `outcome` and the other single-digit entities: candidates (Schedule, Flag, RelatedPerson…) once a unit by unit reading says so. - Tenant isolation is by tag inside the one Project `EDGE Clinic`, enforced by the adapter on every read and write. Medplum's own isolation is one Project (and one ClientApplication) per tenant; that is the next step and needs a provisioning route (`POST /admin/projects` is 404 on this build; the super-admin FHIR route works). - Binary attachments (images, documents) are not sent: `DocumentReference.content` carries a title only. NHS connectivity (PDS, e-RS, GP Connect) is not inferred from FHIR availability, as the engine lane's note says. ##### 5. Measured round trips (OBSERVATION, `LiveSmoke`, jude-1, 1 Nov 2026 05:1xZ, tenant `smoke-d14def5f`, loopback) 20 resource types, each create → decide → stale-version change (412 expected) → read → remove (DELETE, 410 read-back): 0 failures, 237 HTTP calls in 7.0 s wall (unittest), raw per-call figures in `/tmp/clinic-smoke.json` on jude-1 (dated, not in Git). | Call | Groups (type) | Median of per-type medians | Largest per-type median | |---|---:|---:|---:| | `POST /oauth2/token` | 1 | 63.0 ms | 63.0 ms | | `GET /fhir/R4//` | 23 | 11.2 ms | 27.0 ms | | `POST /fhir/R4/` | 24 | 44.5 ms | 153.0 ms (Observation) | | `PUT /fhir/R4//` | 21 | 38.9 ms | 412.2 ms (one outlier) | | `DELETE /fhir/R4//` | 23 | 33.4 ms | 716.4 ms (one outlier) | A create as the receiver sees it (POST + read-back GET) is therefore about 55–60 ms on the host; a change (find GET + PUT + read-back GET) about 60–65 ms; a first touch adds the subject Patient (search + POST) once per person. The earlier run at 05:0xZ (199 calls, 0 failures) gave the same order: GET 7–16 ms, POST 29–116 ms, PUT 26–75 ms per type. Not measured: throughput under load, the public path (there is none: loopback only), or cold starts. ##### 6. Demo tenant and the live journey walk (OBSERVATION) - The demo tenant never touches Medplum: `_demo(engine, tenant)` routes to the estate record on the demo engine and the outputs say `SIMULATED: demo tenant; the clinic record is simulated and Medplum is not touched ()` with `fhir: {resource, simulated: true}` (unit test `test_demo_tenant_never_touches_medplum`: zero HTTP calls). - `tools/journeys_http.py --demo --only /tmp/clinic-medplum-journeys.txt --shard 1/1` on the live release 7a2a0ea, 05:1xZ: the 123 `#human` journeys of the 123 mapped units: `{"journeys": 123, "pass": 123, "plug": 0, "held": 0, "fail": 0, "seconds": 44.0}` (`ledger/evidence/journeys-demo-rerun-1.jsonl`). Note: the running service does not yet import this domain (section 7), so that walk proves the demo path through the estate record, which is also the path the domain takes for the demo tenant. ##### 7. Registration (what the running site needs; one line in a file this lane does not own) `runtime/server.py`, inside `_domains()`, after the bank import: from runtime.domains import clinic # noqa: F401 - binds the Clinic record to Medplum (FHIR R4) Shipped once through the deploy helper at 05:15:03Z (record `/opt/x-x/deploy/ledger/deploy/2026-11-01T051503Z-edit.json`, health live). Until that line is in the release the live tenants' Clinic operations keep the generic estate record; the demo tenant is unchanged either way. `python3 -m unittest tests.test_clinic_domain` (14 tests, fake server) passes on the Mac and on jude-1; `MEDPLUM_SMOKE=1 … LiveSmoke` passes on jude-1 (root, for the credential file). ##### 8. Open items (CLAIMs for the owners named) 1. Engine lane: rotate the seeded super-admin password; make the database container's `/dev/shm` mode survive a restart; name the cause of the 04:48:05Z `podman pod stop`. 2. Coordinator: add the registration line (section 7) and ship. 3. This domain, next: one Medplum Project per tenant; `Binary` upload for images and documents; the remaining single-digit entities after a unit-by-unit reading; `Schedule` capacity for clinic sessions (cancel-a-clinic-session, plan-weekly-capacity). ### research/clinic/MOBBIN-FLOWS.md `research/clinic/MOBBIN-FLOWS.md` · sha256 `84731ec9e69069c058137d22e1cf546c361eb5180f29e4423ba54290954709f9` #### EDGE Clinic research: real product flows from Mobbin **What this is.** Real screens and user flows from existing apps, recorded by Mobbin (a library of real app and website screenshots, grouped into step-by-step flows). It complements `REFERO-FLOWS.md` in this folder. It shows the floor that people already expect from a health front door, and good practice worth borrowing from other industries. **Statement class and date.** Every item below is an OBSERVATION of a real product, seen in Mobbin on 2026-10-31. The descriptions are ours, written from the preview images Mobbin returned (an evenly spaced sample of each flow's screens, not always every screen). Product names belong to their owners. Image links are not kept because they expire; each item carries its `mobbin_url`. **How it was gathered.** Searches by app name plus task, on iOS and web, about five flows per search: appointment booking, symptom checking, test results, medication reminders, mental health check-ins, health questionnaires, wearables and readiness scores, sharing health data, telehealth, food logging, emergency and crisis help, and local community activity (as a stand-in for social prescribing). ##### Coverage: what Mobbin held and what it did not - **Held and returned:** Ada (symptom assessment, condition detail, past assessments), Headspace (anxiety and stress check-ins, therapy booking on web), Apple Health (sharing, data sources and access, profile), Oura (readiness and a medical AI chat), WHOOP (labs), MyFitnessPal (meal logging), Zocdoc (doctor booking), GoodRx, Hers and Hims (medication reminders, treatment questionnaires), Ro, Superpower and Bevel (lab results and intake), Alan (consultations, with Livi as the video provider inside it), and single screens from Wysa, Bloom, Garmin Connect and Amazon's health assistant. - **Searched for but not returned:** NHS App (the booking search returned Zocdoc, IKEA, Crate & Barrel and Fresha instead) and Doctolib (the web search returned Selfridges, Square and Headspace instead). No NHS, GP practice or other UK public health service appeared in any result. - **Not searched by name in this pass:** Babylon, Healthily, Calm. Livi appeared only as the video provider inside Alan. - **Not found at all:** any social prescribing service, any link worker or referral to community groups, any flow where a GP or clinician receives what the person filled in, and any flow where an AI agent acts for a person or organisation with its own identity. ##### By task ###### Booking an appointment - **Zocdoc, iOS, "Booking an appointment"** (10 screens). A doctor's page with a "new patient" tick box, a choice of location and tomorrow's slots as buttons. The review step lets you book for yourself or "someone else", shows a red inline error when the address is missing, and asks for insurance, visit reason, a phone number and optional notes for the practice, with a tick box certifying the payment method. The confirmation offers "Add to calendar" and "Prepare for appointment". OBSERVATION, 2026-10-31. https://mobbin.com/flows/fcd6461e-af19-4ca9-9864-8b21c286888c - **Zocdoc, iOS, "Making a doctor appointment"** (12 screens). In-person or video visit toggle, slots grouped by day, then a gate: "This provider only sees in-network patients", with the choice of photographing the insurance card or choosing it by hand, and accepting the terms. Afterwards the home screen shows the upcoming appointment with "Almost done! Upload a photo of your ID". OBSERVATION, 2026-10-31. https://mobbin.com/flows/26c741b6-0231-4040-b5e9-890f0b6db100 - **IKEA, iOS, "Booking an appointment"** (15 screens, retail). A plain explanation of the session and what to prepare beforehand, a calendar with green dots on days that have space, time slots labelled with the time zone, a contact form that says why the phone number is needed (reminders), and a confirmation with a booking reference and all details. OBSERVATION, 2026-10-31. https://mobbin.com/flows/64df5892-fc18-4edb-ba02-8b365b63bebe - **Crate & Barrel, iOS, "Booking an appointment"** (14 screens, retail). A three-step progress bar, separate tick boxes for marketing texts and for appointment reminders, a calendar with an available and not available key, and a confirmation that explains "What's next", "What to expect" and "What you'll get". OBSERVATION, 2026-10-31. https://mobbin.com/flows/8e862fc0-87f2-44f0-a201-787bb47e5d6e - **Fresha, iOS, "Booking an appointment"** (22 screens, beauty). Date strip and time list, a review with price and fees, the cancellation and no-show charges stated before payment, then a "Confirmed" page with add to calendar, directions, message the venue and venue details. OBSERVATION, 2026-10-31. https://mobbin.com/flows/7121fd1b-35ef-48da-87d5-aacbd1ba81a8 - **Headspace, web, "Scheduling an appointment"** (23 screens, therapy). Steps shown at the top: check in, verify insurance, schedule. "Here's your estimated cost" confirms the insurer and lists things to know (pay nothing today, the most it could cost). Time ranges show how many clinicians are free and the time zone. The final page lists next steps, including "complete your pre-appointment assessment to help your therapist prepare". OBSERVATION, 2026-10-31. https://mobbin.com/flows/b46dac23-c0f1-4b3f-b506-ef8fea7c14d7 - **Selfridges, web, "Making an appointment"** (17 screens, retail). A four-step stepper with the appointment summary always visible on the right, a week grid by morning, afternoon and evening, a phone field that says it will only be used for a reminder 48 hours before, two separate consent tick boxes, and a confirmation with a reference, print and add to calendar. OBSERVATION, 2026-10-31. https://mobbin.com/flows/e5cf516c-6b42-4bb9-b0c2-291623dc0124 - **Square, web, "Booking an appointment"** (12 screens). A running summary while you add services, a checkout that holds the slot with a countdown, a cancellation timeline ("cancel before" a date), and a thank-you page with add to calendar, reschedule, cancel and book again, plus "Next steps: we'll call you". OBSERVATION, 2026-10-31. https://mobbin.com/flows/6992da62-212e-4947-991d-588142829394 ###### Symptom checking - **Ada, iOS, "Starting a symptom assessment"** (17 screens). One question per screen (name, then date of birth as day, month and year boxes), with "Previous" and a close button, and an "About Ada and Privacy" link on every screen. Symptom search with suggestions and a help icon per symptom. It ends: "No more questions... The report is not a medical diagnosis. If in doubt, seek advice from a healthcare professional." The report lists possible causes with the most urgent first, in a red box: "Needs immediate attention: seek emergency care now". Each cause shows which of your symptoms matched and "48% of people with your symptoms and health background had this condition". OBSERVATION, 2026-10-31. https://mobbin.com/flows/dc2c5dc4-6695-44fd-b21a-3cd4e7601c9a - **Ada, iOS, "Searching symptoms"** (3 screens). Typing "headache" gives suggestions and "Symptoms found: 3", each with a short plain description and a help icon. OBSERVATION, 2026-10-31. https://mobbin.com/flows/4d5fb715-f25b-47a7-b4cc-1fd4bcfe1505 - **Ada, iOS, "Condition detail"** (7 screens). A condition page with sections that open and close (symptoms, diagnosis, treatment, risk factors, outcomes), a statement that the list "is not a diagnosis and not exhaustive", and "Download report". OBSERVATION, 2026-10-31. https://mobbin.com/flows/ce7b8167-97a3-47da-8a93-00c89d64c098 - **Ada, iOS, "Health assessments"** (8 screens). A profile with basic information, health background and the list of past assessments. OBSERVATION, 2026-10-31. https://mobbin.com/flows/0c0ecbbf-82cb-4dee-af6b-353de04791ce - **Ada, iOS, "Home"** (3 screens). The home screen with one main action, "Start symptom assessment", and a news card about changes to how you access your data. OBSERVATION, 2026-10-31. https://mobbin.com/flows/35a15c59-17c0-4872-a99d-610cc2cec951 ###### Questionnaires and intake - **Ro, iOS, "Onboarding"** (20 screens). Choose a goal, then a progress bar over height and weight, a "select all that apply" list of heart conditions, and treatment preferences that include "Not sure yet" and "Another option not listed". It ends "You've qualified" and "What's next: a provider will review your health information". OBSERVATION, 2026-10-31. https://mobbin.com/flows/c40ee07f-972a-43b2-83b4-666d7f6df86e - **Hers, iOS, "Complete treatment questionnaires (mental health)"** (29 screens). A consultation with a segmented progress bar. Answers carry a plain description (for example "Fair: I have friends, but I don't feel close to or supported by them"). A loading screen ticks off "gathering your responses, calculating your scores, preparing your results". A free-text box asks "anything else you want your provider to know?" with a tick box for "No, I've provided all relevant information". OBSERVATION, 2026-10-31. https://mobbin.com/flows/60448996-a8dd-4632-9b46-4ab745af1a4e - **Hers, iOS, "Completing treatment questionnaires (weight loss)"** (18 screens). Includes a screening question about eating disorders with the reason it is asked, marketing charts mixed into the questions, and a clear refusal screen: "you are not eligible... we recommend that you see a provider in person", with a way back. OBSERVATION, 2026-10-31. https://mobbin.com/flows/061e04f8-11e2-49c9-99ba-409a707e5f25 - **Superpower, iOS, "Taking lifestyle quiz"** (11 screens). "Let's understand your day-to-day: this helps us customise your plan", then questions on exercise and work, each with a one-line reason ("your job plays a big role in daily stress"). It ends by selling more tests. OBSERVATION, 2026-10-31. https://mobbin.com/flows/48c850ab-006f-4536-a9bc-1d4033c46c9c ###### Mental health check-ins - **Headspace, iOS, "Tracking anxiety levels"** (6 screens). A progress area with stress and anxiety tabs, an empty chart that says "Start tracking your anxiety over time with bi-weekly check-ins", and a seven-question check-in ("Over the last two weeks, how often have you been bothered by..." with four answers). After a check-in the button greys out with "Opens in 13 days". OBSERVATION, 2026-10-31. https://mobbin.com/flows/79577cba-b6b5-487c-ae41-5416af934661 - **Headspace, iOS, "Tracking monthly stress reflection"** (6 screens). Ten questions, then a dated result: a score on a bar split into low, moderate and high with the number ranges shown, framed gently ("Think of this score as a way to get to know yourself a little better"), with suggested content and a link to questions and answers. OBSERVATION, 2026-10-31. https://mobbin.com/flows/ce032b9a-cb9f-42f9-b407-19a2796c6961 - **Headspace, iOS, "Profile"** (7 screens). Personal statistics, a streak described kindly ("Meditate on a schedule that feels good"), the check-in panel and activity history. OBSERVATION, 2026-10-31. https://mobbin.com/flows/8d8bd1bc-e2c4-4a04-b7f6-5ba7233321b6 ###### Test results - **Bevel, iOS, "Record detail"** (6 screens). Health records filtered by type (lab report, medications, imaging), a nudge that some markers are due for a refresh, and uploaded documents "ready for review". A lab report opens with a plain summary of what is high, then each marker with its range and value. One marker carries an orange warning: "This biomarker may be incorrect. Please review the value and update it if needed", with Edit. OBSERVATION, 2026-10-31. https://mobbin.com/flows/bcb74637-8d36-4b58-8e62-99e88312cb98 - **Superpower, iOS, "Biomarker detail"** (3 screens). A list of results with small range bars, then one marker marked "Out of range", a chart over time with the healthy band shaded, latest result next to the target range, what it is, why it matters, what you can do, and a list of sources. OBSERVATION, 2026-10-31. https://mobbin.com/flows/0a31c382-c9dc-4f06-b681-f8bceee8e991 - **WHOOP, iOS, "Labs summary"** (5 screens). A test offer, a summary ring with "optimal, sufficient, out of range", a promise of "clinician-reviewed insights", and an empty state for a marker: "No lab results here yet: upload past lab results or schedule a test". OBSERVATION, 2026-10-31. https://mobbin.com/flows/b7fceedb-b811-45ff-94f6-2760d2c4f382 - **Ro, iOS, "Lab history"** (2 screens). Settings that include lab history, Apple Health and privacy settings; the empty lab history is used to sell tests. OBSERVATION, 2026-10-31. https://mobbin.com/flows/aebafd34-bd72-4d0c-bed0-b96557c23662 ###### Medication reminders - **GoodRx, iOS, "Scheduling a daily med reminder"** (10 screens). "Add reminder details": a name, a time, days of the week, and switches for which medicines this reminder covers, with "you'll get a push notification on the days and time you select". A separate card offers refill reminders. OBSERVATION, 2026-10-31. https://mobbin.com/flows/dc7306f5-fc70-4af5-a256-95484befdc1b - **GoodRx, iOS, "Daily med reminders"** (2 screens). The list of reminders with Edit and "Schedule new reminder". OBSERVATION, 2026-10-31. https://mobbin.com/flows/efcc6f42-2a09-45b6-951d-2810aae5c0eb - **Hers, iOS, "Setting up a reminder"** (7 screens). The treatment page shows reminders on or off, refills left, a message from the provider team and how to take the medicine. The reminder sheet allows several times a day, and its small print says the reminder information "will not be shared with your Care Team. Contact your provider if you have questions." OBSERVATION, 2026-10-31. https://mobbin.com/flows/1396d89d-d0d5-4822-9df1-1211c80cec04 - **Hims, iOS, "Setting a reminder"** (7 screens). Next refill date, a plain reason why a refill is not yet allowed, a reminders tile and "Snooze for 1 month". OBSERVATION, 2026-10-31. https://mobbin.com/flows/56b3a643-67eb-4930-a326-46b5f76dea0c ###### Wearables, scores and data sharing - **Oura, iOS, "Readiness"** (6 screens). A daily score with a short, human message ("above your average today, how do you feel?"), a seven-day line, contributors labelled "Pay attention", "Fair" or "Good", and overnight heart rate and heart rate variability charts. OBSERVATION, 2026-10-31. https://mobbin.com/flows/b071a02f-88d4-4ee2-9a4e-ea9e9422bb56 - **Oura, iOS, "Readiness"** (10 screens). "Your first Readiness Score!" explains what the score means and that it improves as a baseline is learned; key measures as tiles; and "Recovery in progress: you've turned off Rest Mode... continue to take it easy", with a marker showing Rest Mode was on. OBSERVATION, 2026-10-31. https://mobbin.com/flows/cd0b2e3e-e462-4d9d-ba87-f0c214fe47ca - **Oura, iOS, "Daily readiness data"** (2 screens). A newer home with readiness, sleep, resilience and activity in one row and the message "your activity goal today is your baseline". OBSERVATION, 2026-10-31. https://mobbin.com/flows/1a062e29-239c-4971-94cd-292b42dfcc7c - **Apple Health, iOS, "Share with someone"** (4 screens). An explainer in three short points (you're in control; they get a dashboard and notifications; only a summary is shared, encrypted, and you can stop at any time), "Share with Someone" or "Ask Someone to Share", and an honest error: "Unable to share health data: this person does not have a device that can view your data". OBSERVATION, 2026-10-31. https://mobbin.com/flows/f5de1f34-8e2a-4567-840d-eac966c7ef88 - **Apple Health, iOS, "Sharing"** (3 screens). "Share with your doctor", apps and research studies with access, and "No one is sharing data with you". OBSERVATION, 2026-10-31. https://mobbin.com/flows/00d9fca9-9ce6-4ba4-8a91-e3141ce6e29d - **Apple Health, iOS, "Data sources & access"** (3 screens). For one kind of data (steps): which apps may read it, which research studies may read it (both "None"), and which sources write it, in priority order, with an explanation of how one source is chosen. OBSERVATION, 2026-10-31. https://mobbin.com/flows/e19317bd-987a-4ace-bd7d-42048e3786b1 - **Apple Health, iOS, "Profile"** (3 screens). Medical ID, health checklist, health records, and privacy lists for apps, research studies and devices, with "Export All Health Data" and the time it last synced. OBSERVATION, 2026-10-31. https://mobbin.com/flows/40639fc3-6978-4b39-99bd-cd51db256933 - **Apple Health, iOS, "Searching apple health"** (4 screens). Browse by category (heart, medicines, mental wellbeing, symptoms, vitals), add a health records account, search results. OBSERVATION, 2026-10-31. https://mobbin.com/flows/a50e2c67-b7f5-4b5b-94f5-6ed71c1d224e - **Garmin Connect, iOS, screen.** Abnormal heart rate alerts with high and low limits, each with a plain note of when the alert fires and who it is not suitable for. OBSERVATION, 2026-10-31. https://mobbin.com/screens/65376645-1229-4e72-9733-94a8837147e4 ###### Talking to a clinician, and AI health assistants - **Alan, iOS, "Talk with medical team"** (3 screens). "Talk with our medical team 24/7 by chat or video". The chat tab offers an AI ("Ask Mo") to "skip the wait", and lists specialties with expected reply times ("reply in about 30 minutes"), with a footer "Is it an emergency? Call 15". The video tab shows general medicine delivered by Livi ("6am to midnight, included, includes prescriptions") and therapy, with "Help is available: if you're in distress or having suicidal thoughts, reach out now". OBSERVATION, 2026-10-31. https://mobbin.com/flows/a633ff3f-4a5f-4a82-9c9f-3420c9ab2c78 - **Alan, iOS, "Consult"** (4 screens). A daily wellbeing page with small actions (walk, breathe, meditate), programmes run with physiotherapists and therapists, and topic tiles (mind, sleep, nutrition and more). OBSERVATION, 2026-10-31. https://mobbin.com/flows/5a2fbc6d-a598-4651-bac9-a191a52a02b3 - **Oura, iOS, "Ask a doctor"** (2 screens). A "Medical chat" where an AI advisor from a named partner says it will use your ring data and that you can bring in a doctor; example questions; "Add a doctor: will reply after 7pm"; footer "Messages from the AI are informational. For medical advice, add a physician. For emergencies, call 911." A paid offer says physicians see your ring data and records. OBSERVATION, 2026-10-31. https://mobbin.com/flows/ad69f6c5-15bc-413f-bd86-8b459b972fc1 - **Amazon Shopping, iOS, screen.** A health AI in beta that names itself, says "If this is an emergency, dial 911", offers "Talk to a provider", and shows a "Thinking" state with a stop button. OBSERVATION, 2026-10-31. https://mobbin.com/screens/ca1c14f1-cf34-4b27-b041-d8e30ebb332d ###### Crisis and urgent help - **Wysa, iOS, screen.** A chatbot confirms "your safety plan has been created" and reminds the person to press SOS whenever they cannot cope; buttons to check the plan and add a therapist. OBSERVATION, 2026-10-31. https://mobbin.com/screens/cd4930f5-31e0-4c34-86a3-a6679e342cf7 - **Bloom, iOS, screen.** A "Get help" sheet with a crisis line button and a local hotline button. OBSERVATION, 2026-10-31. https://mobbin.com/screens/2cf0f6a8-c1fb-4919-8ab3-e53826be7387 - **inDrive, iOS, screen (transport).** An SOS page with the person's location on a map, the details the police will need, and one large call button. OBSERVATION, 2026-10-31. https://mobbin.com/screens/3f075a2f-0493-4d53-a905-00549b8b2fa8 ###### Food and lifestyle logging - **MyFitnessPal, iOS, "Scanning a meal"** (4 screens). Today's calories and macros, quick ways to log (search, barcode, voice, photo of the meal), a scanned list of foods each with delete and "Add more", then "Log breakfast" and a "Breakfast logged!" message. OBSERVATION, 2026-10-31. https://mobbin.com/flows/e5e452d7-f91e-4939-9fa0-59ea87b7d47a - **MyFitnessPal, iOS, "Log meal"** (7 screens). Search, saved meals and foods, "copy previous meal", and a hint to swipe to log yesterday's meal again. OBSERVATION, 2026-10-31. https://mobbin.com/flows/d58bea22-7f26-48ef-a6de-134d9183c5ad ###### Local community activity (nearest thing to social prescribing) - **Meetup, iOS, screen.** Groups near a place with distance chips (2, 5, 10 miles), categories such as social activities and hobbies, and member counts. OBSERVATION, 2026-10-31. https://mobbin.com/screens/303b5f13-3a13-4b20-809c-f356d6993248 - **Nextdoor, iOS, screen.** Neighbourhood groups (for example help with housing) with member counts, and local posts and polls. OBSERVATION, 2026-10-31. https://mobbin.com/screens/9fa1c528-5530-4f11-bc24-0244ea7279fe - **Lex, iOS, screen.** A local feed filtered by events, new here, groups and friends; posts tagged "For free"; a group card with "Join now". OBSERVATION, 2026-10-31. https://mobbin.com/screens/36bffdb3-8d25-413f-9698-596d80c9c31a - **adidas Running, iOS, screen.** Challenges with participant counts and end dates, upcoming events ("starts in 28 days") and communities. OBSERVATION, 2026-10-31. https://mobbin.com/screens/e11d5383-ce5b-4038-b90a-39a9bf820ad7 ##### What to match - **Urgency first, in plain words.** Ada puts "seek emergency care now" at the top of the report in red; Alan, Oura and Amazon keep an emergency line visible in their health chats. EDGE Clinic must do at least this on every screen where a person describes a symptom. - **"This is not a diagnosis", said once, clearly, at the right moment** (Ada at the end of questions and on every condition page), not buried in terms. - **One question per screen, with a way back and the privacy link always visible** (Ada), and a reason given for sensitive questions (Hers, Superpower). - **Show what matched and how sure.** Ada lists which of your symptoms matched and a plain percentage of similar people. Superpower lists the sources behind its advice. - **Let people correct the record.** Bevel flags a value that may be wrong and lets the person edit it. Clinic results and imported data need the same. - **Spacing check-ins on purpose.** Headspace locks the check-in with "Opens in 13 days" and frames the score kindly with its ranges visible. - **Say who sees what.** Apple Health shares only a summary, can be stopped at any time, and lists every app and study that can read each kind of data. Hers says the reminder data is not shared with the care team. - **Bookings that explain themselves.** State what to prepare (IKEA), why each piece of contact data is needed (IKEA, Selfridges), separate consent for reminders and for marketing (Crate & Barrel, Selfridges), the cost and cancellation rules before payment (Fresha, Headspace, Square), and the next steps after booking (Headspace, Crate & Barrel, Square). - **An honest "not for you" route.** Hers tells ineligible people to see a provider in person instead of leaving them stuck. - **Booking for someone else** is a first-class choice (Zocdoc "Someone else"). ##### What none of them do - **No AI agent is ever a user in its own right.** Every AI here (Ada, Alan's "Mo", Oura's advisor, Amazon's assistant, Wysa) is a feature inside one company's app, acting for the signed-in person. None has its own identity, none shows whose behalf it acts on, what it is allowed to do, or a record of what it did. No flow lets a family member's agent, a carer's agent or a link worker's agent book, ask or fill in a form with the person's permission. - **No hand-off to the GP or NHS.** Nothing sends a symptom report, check-in score or wearable trend to a named clinician with the person's consent, and nothing shows the person what the clinician received or did with it. Ada offers "Download report" only. - **No social prescribing.** No app suggests a local group, walk, class or service as a response to how someone feels, and none involves a link worker. Community apps (Meetup, Nextdoor) know nothing about health. - **No organisation as a user.** No practice, charity or community group appears as an account that receives referrals or reports back. - **Early detection stays inside one product.** Wearable apps show trends and alerts (Oura, Garmin) but none explains when a trend should lead to seeing a clinician, or offers to share that trend with one. - **Confidence is rarely shown.** Apart from Ada's percentage and Bevel's "may be incorrect", nothing says how sure the system is or what it is based on. - **No UK public sector examples.** NHS App and Doctolib were not in the results, so UK identity checks (NHS login), GP booking and repeat prescriptions are not covered here. ### research/clinic/REFERENCE-CLINIC.md `research/clinic/REFERENCE-CLINIC.md` · sha256 `15a2191dce36d765e97476f1cce07f41db7aceaae72eeb876d200a68be3fc33f` #### EDGE Clinic: the reference flows the application entry copies **Statement class.** OBSERVATION of real products in Mobbin, recorded 31 October 2026 in `MOBBIN-FLOWS.md` in this folder; this file names which of them the Clinic's entry screen is built from (Owner, 1 November 2026: layouts come from the references, the only difference being the estate's added controls or journeys). ##### The entry (`/clinic`) - **Ada, iOS, "Home"** (Mobbin flow 35a15c59-17c0-4872-a99d-610cc…, 3 screens): one main action on the home, "Start symptom assessment", above a short card. The Clinic's home keeps that shape: one primary action (Check your symptoms, `clinic/symptom-check-and-triage`), then the designed tasks as tiles. - **Ada, iOS, "Starting a symptom assessment"** (17 screens): one question per screen with Previous and a close control, a privacy link on every screen. The Clinic's journey screens keep one question per screen (tools/typed_steps.py) and the help control. - **Zocdoc, iOS, "Booking an appointment"** (10 screens): a provider page, a choice of location, tomorrow's slots as buttons, a review step that lets you book for someone else. `clinic/book-appointment` follows this order. ##### What the estate adds, and only that - The designed tasks as tiles under the primary action: front door check-in, medication safety check, mental health care and crisis, online urgent triage, person-held health record, retinal screening, social prescribing referral, symptom check and triage, wearable signals, wellbeing questionnaires (the Clinic's bespoke units in `design/figma/FRAMES.json`). - Every other Clinic capability as the same tile beneath, so nothing is hidden and nothing is listed as prose. - Jude's help behind the help control; evidence behind the evidence control (glass box). ### research/clinic/REFERO-FLOWS.md `research/clinic/REFERO-FLOWS.md` · sha256 `5e49c9f08d65dba650a6b054e0e69181acc87774897c42ce1f19a7a24b628bc0` #### EDGE Clinic research: real product flows from Refero **What this is.** Real screens and user flows from existing products, recorded by Refero (a library of real app and website screenshots, each journey described step by step). They show the floor (what users already expect) and good customer experience from any industry worth borrowing. **Statement class.** Every flow is an OBSERVATION by Refero of a real product, searched 2026-10-31. Descriptions are Refero's wording. Product names belong to their owners. **How it was gathered.** Semantic searches for this application's tasks on iOS and web. Results from other industries are kept on purpose: they are often the best customer experience to learn from. Refero does not hold many incumbents (and few government, enterprise or NHS systems), so this is a sample, not a census. **Products in this sample:** Ada (20), Fresha (15), District (14), WhatsApp (14), Doctronic AI (13), TidyCal (9), ALO (7), Cohere (6), Care (6), Lovable (5), Willow (5), Poppy (5), Stardust (5), Omio (5), Breathwrk (5), Airtable (5), Maze (5), Napper (5), Structured (5), Meta AI (4), Airbnb (4), Preply (4), Podimo (4), Train Fitness (4), ElevenMusic (4), GoFundMe (3), Wabi (3), Foodllama (3), Parallel (3), Raycast (3), Longevity Deck (3), GlossGenius (3), Delphi (3), fal (3), Tinder (3), Sunsama (3), Calendly (3), Craft (3), Tripsy (3), Peerlist (2), Genie (2), TIDAL (2), Pool (2), Google Gemini (2), Fingerprint (2), The Outsiders (2), Trip (2), Anam (2), Understory (2), komoot (2), Haptic (2), Seed (2), Nextdoor (2), Acuity Scheduling (2), Cal (2), IKEA (2), Cushion (2), Leonardo.Ai (2), Adobe Photoshop (2), Orbit (2), Wispr Flow (2), Vocabulary (2), OpenAI (2), n8n (2), Intercom (2), Cofounder (2), Kit (2), Blackbird (2), Loowner (2), Ground News (2), Kin (2), The Leap (2), ElevenReader (2), On (2), Instacart (2), Netflix (2), Unwind (2), Clay (2), MonoDesk (2), New Balance (2), Duolingo (2), Typeform (2), Kinhive (2), Givingli (2), Melius (2), Chance AI (2), Dropbox (2), Rork (2), Mozi (2), time2book (2), Introspect (1), Google Translate (1), Cashcaded (1), Queue (1), Subo (1), Asos (1), Apple Image Playground (1), Spyglass (1), PayPal (1), Gumroad (1), Pi (1), Binge (1), Jasper AI (1), Allbirds (1), Deezer (1), Acne Studios (1), Saint Laurent (1), West Elm (1), ElevenLabs (1), Weekrise (1), Voicenotes (1), Synthesia.io (1), Telegram (1), Comet (1), FocusFlight (1), Granola (1), Vinyls (1), Substack (1), RAD Weather (1), Remy (1), 222 (1), Pinterest (1), Shazam (1), Fuse (1), Riverside (1), Resend (1), Rox (1), Butter (1), BYBORRE® (1), Linear (1), Instagram (1), UNHINGED (1), Twitch (1), ChatGPT (1), Microsoft Teams (1), pillowtalk (1), Sora (1), GO Club (1), Teal (1), DailyArt Magazine (1), LinkedIn (1), Slite (1), Polaroid (1), Volv (1), Mercury (1), Farfetch (1), Manus (1), Prose (1), Trip Way (1), Joi (1), X (1), B150 (1), Zendesk (1), TikTok (1), Sky Guide (1), MyMind (1), Runway (1), Apple Maps (1), Athleats (1), Verse (1), Exoplan (1), Miles (1), HYPE (1), Marathon (1), Make (1), Homerun (1), TwelveLabs (1), Сontra (1), [untitled] (1), Weather mini (1), Navia (1), Tripadvisor (1), Uber (1), Brink (1), Uber Eats (1), Cursor (1) ##### Task: AI agent setup permissions **Leonardo.Ai: Profile Settings Management** (web, 2 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/12914) > This is a settings screen in Leonardo.ai focused on personal profile management. The entry point is the user already being logged in and opening Settings, with the left sidebar showing Personal Settings and Workspace Settings. The user can review profile information, choose interest tags, confirm adult content access, and optionally delete the account. The exit point is after changes are saved and reflected in the profile section with confirmation feedback, or after the user leaves settings. Steps: View Profile Settings → Update Preferences **WhatsApp: Create Community Setup** (ios, 6 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12677) > Happy path for creating a new community on mobile. The user enters a community name, customizes the community image from the Edit control, selects either an emoji or sticker with a background color, then returns to the main creation screen with the chosen avatar and name ready to submit. **Structured: Structured AI Onboarding** (ios, 3 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3285) > A three-step onboarding carousel presented as a modal on a mobile device that introduces the "Structured AI" feature. Each slide contains a top header with the feature name, a horizontal progress indicator, an illustration of the feature, a bold headline, a short explanatory blurb, and a primary 'Continue' CTA. Controls available across slides: a Close (X) in the top-right, a Skip action, and a back arrow on later slides to navigate backwards. Entry: modal opens in front of the main app (user already in the app). Exit: user advances through slides with Continue, can Skip or Close to dismiss the modal; final Continue proceeds past the onboarding (next screen not shown). Assumption: mobile device; modal invoked to introduce a new feature; user can navigate forward/back or dismiss at any time. Steps: Intro — Speed Up Planning → Auto-Generate Your Schedule → Digitize Your Analog Planner ##### Task: AI chat assistant **Doctronic AI: Passwordless AI Doctor Sign-In** (web, 8 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13706) > The flow begins on Doctronic’s public landing experience, where an unauthenticated visitor can start a chat or access account options. The user chooses email-based login, enters an email address, receives a six-digit verification code by email, and submits the code. After successful verification, the user is taken to the authenticated AI health assistant screen, ready to ask a health question. The user is assumed to be logged out and able to access the entered email inbox. **Doctronic AI: Sleep Concern Assessment** (web, 13 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13694) > A logged-in user named Connie starts a conversation with Doctoric AI by entering a health concern about recurring nighttime waking. The system displays an emergency disclaimer, processes the message, and requires acceptance of the Terms of Service and AI-provider disclaimer before continuing. The user then provides age and sex information, receives follow-up questions about additional symptoms and wake-up patterns, and answers questions about health history. The flow ends with the assistant asking about current medications, with the user remaining in the conversation. **Google Gemini: Voice Query to Suggestions** (ios, 6 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/3207) > Happy-path: A signed-in user interacts with the Gemini conversational assistant on a mobile device. Steps: Conversation (idle with assistant intro) → Start Voice Input (listening) → Live Transcription (partial) → Full Transcription / Submit → Assistant Responding (audio playing) → Answer Rendered (suggestions with media) **WhatsApp: Meta AI Voice Chat Setup** (ios, 6 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12673) > Happy path for starting a voice conversation with Meta AI. The user enters the Meta AI profile/settings area, opens the voice feature, reviews the informational consent-style modal, and continues into a live voice session. **Meta AI: Save Recipe Instructions** (ios, 2 screens, AI Tool) [Refero](https://refero.design/flows/13345) > A user views an assistant-generated response containing ingredients and a 20-minute pasta cooking method. The user copies the response using the copy action beneath the message. A confirmation toast appears stating that the content was copied. The flow starts with an existing assistant response and ends with the response available on the device clipboard. Steps: Review Recipe Response → Copy Recipe ##### Task: API keys developer settings **Willow: Willow onboarding intro** (ios, 3 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/13227) > A short onboarding carousel for Willow shown to a new or returning user before entering the main experience. The user can swipe through several promotional intro screens and tap the primary Get started button to continue. **Poppy: Recommended Connection Setup** (ios, 4 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/12177) > Happy-path onboarding/settings sequence for a logged-in user named Connie. The user opens the profile/settings area, chooses a recommended connection type (Fitness), is shown a system Health permission request, and returns to the recommendation screen after granting access. Entry is from the profile/settings menu; exit is back to the recommendation step with the selected connection highlighted and a Next button available. Steps: Profile and Settings → Connection Recommendations → Health Permission Request → Connection Selected **komoot: New user onboarding and consent setup** (ios, 39 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8281) > Happy-path onboarding for a first-time mobile (iOS) user. Entry: app launch -> privacy/tracking consent banner. User can Accept All, Refuse, or Customize preferences. If user customizes they can toggle categories and open vendor details. After consenting, the flow presents sign-in options (social or email). Choosing email leads to email validation then account creation (password). Post-signup the user progresses through product onboarding cards that request push notifications and access to device integrations (Apple Health). The flow ends with device connections (optional) and a completion screen, then the main map/routes home screen. Steps: Privacy notice (top-level) → Customize preferences (categories visible) → Category list (scroll) → Analytics expanded (vendor row visible) → Vendor info modal (Firebase) → Sign-in options → Enter email (initial) → Email validation error → Email accepted (proceed) → Create account (password) → Onboarding hero / intro → Onboarding card: Navigation → Onboarding card: Notifications → iOS push permission dialog → Device integrations list → Health Access (iOS modal) → Integration connected state → Onboarding completion → Main map / routes home → Optional: Connect additional devices (post-onboarding) **Poppy: Support and Privacy Settings** (ios, 6 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12194) > A settings/support area presents account and privacy-related options, including feedback, data retention, privacy policy, terms, account deletion, experimental features, and sign out. Selecting certain items opens full-screen informational panels that explain what data Poppy collects, how it is used, and how long it is stored. Entry appears to be from an account/settings screen; exit is via the close button or back navigation. Steps: Support Menu → Data Overview → Location Details → Email Details → Health Details → Control and Deletion **RAD Weather: Ambient Weather Connection** (ios, 5 screens, Weather) [Refero](https://refero.design/flows/13803) > The flow starts from an “Add Uplink” provider selection screen and continues through the Ambient Weather credential setup. The user selects Ambient Weather, enters a personal API key, and tests the connection. The screenshots show the connection request loading and ending with a provider rejection error. The user is assumed to be on an iPhone and able to access their Ambient Weather API key. Credentials are described as being stored locally in Apple Keychain. Steps: Provider Selection → Enter API Key → Test Connection → Connection Check → Credential Rejection ##### Task: CRM contacts pipeline **222: Assign Close Friends** (ios, 11 screens, Lifestyle, Dating & Relationships) [Refero](https://refero.design/flows/9320) > Happy path: The user refines the app's compatibility model by answering a series of persona-style prompts and assigning a real friend to each prompt. From the Close Friends screen the user taps a prompt's "choose a friend" input, picks or filters a contact from their phone list, and triggers an in-app share (pre-populated message with a referral/preview) to invite that friend to take the compatibility test. A confirmation modal appears before the assignment is finalized. Each confirmed assignment increments a progress counter (e.g., 0/5 → 1/5). Steps: Close Friends — prompt overview → Contact list with keyboard → Filter to a specific contact → Select contact to send referral → Message compose with referral preview → Confirmation modal (explain & confirm) → Next prompt — pick another friend → Message compose for second contact → Progressed question cards → Close Friends summary → Return to People tab / flow complete ##### Task: access review permissions audit log **TIDAL: Track Report Reason Selection** (ios, 3 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/13140) > A mobile modal for reporting a track, starting with a searchable list of report reasons and ending with a selected reason shown after search. The user opens the report dialog, searches or browses categories, and selects the matching reason. Steps: Browse Report Reasons → No Match Search State → Filtered Result **Apple Image Playground: Report Generated Image** (ios, 9 screens, AI Tool, Design & Media Editing) [Refero](https://refero.design/flows/12798) > A user reviews a generated image in Image Playground, opens the overflow menu, and chooses Report a Concern. They select concern categories, optionally add more information, review what will be included in the report, accept the data-sharing prompt, and submit. The report then submits successfully and returns them to the image editor, where the generated image remains visible with the usual controls. Steps: Image Review → Action Menu → Concern Category → Personal Characteristics Check → More Information → Submit Report → Data Consent → Submitting → Return to Editor **fal: Recent History Preview Visibility** (web, 2 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/13542) > The user opens the Recent History area while authenticated and reviews recent API requests across endpoints. Results are displayed in a four-column card grid with generated previews, request metadata, timestamps, status codes, prompts, and actions such as Playground, Copy prompt, and Model Analytics. The user can use the Show preview toggle to switch between visible generated images and hidden preview placeholders. The flow ends with the history list displayed in the selected preview state; pagination controls and the site footer remain available below the results. Steps: Review History With Previews → Hide History Previews **District: Event Feedback Submission** (ios, 5 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12916) > Happy-path mobile feedback journey after attending an event. The user enters a rating screen for a specific event ('Sip and Act at 3 Art House'), selects an overall sentiment, optionally rates individual aspects, adds a text comment, and submits. Steps: Overall Rating → Detailed Ratings → Comment Entry → Submission Confirmation → Completion State **Cursor: Security Audit Assistant** (web, 18 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9922) > Happy path: From the Agents landing the user composes a task card, selects a GitHub repository, configures model(s) and multipliers, optionally attaches an image, then starts the task which spins up a cloud agent. The agent appears in the left agent list and the UI navigates to a three-column agent workspace showing the activity log (center) and a file-changes pane (right). The agent runs commands (visible as generated commands / progress) and produces a final security audit report with findings and code snippets. Steps: Agents Landing — Task Composer → Repository Picker Open → Repository Search → Repository Selected → Open Model Selector → Enable Multiple Models → Set Model Multipliers → Compose Task Instruction → Attach Image Attempt → Image Preview Modal → Task Created — Agent Listed → Open Agent Conversation → Agent Starting — Generating Commands → CLI Execution Attempt → Agent Working — Additional Steps → Audit Report — Commits Reviewed → Critical Finding Detail → Final Summary and Checks ##### Task: account closure **Care: Close Membership Account** (web, 3 screens, Job Search, Lifestyle) [Refero](https://refero.design/flows/12240) > Happy path for closing a Care.com membership. The user starts on a membership management screen while signed in, chooses to close their account, reviews cancellation options and reason selection, then completes the closure and reaches a confirmation screen. Steps: Membership Overview → Cancellation Review → Closure Confirmation ##### Task: apply for loan credit **Introspect: Friend decision response** (ios, 3 screens, Finance) [Refero](https://refero.design/flows/11952) > Happy path for a recipient responding to a friend’s product decision request on mobile. Steps: Decision Request → Sending State → Completed State **Lovable: Plans and credits management** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12301) > Happy path for managing a workspace subscription from the Plans & credits area. Entry is the workspace settings sidebar on the Plans & credits screen. The user can review current plan status, compare Pro/Business/Enterprise options, and initiate an upgrade or credit purchase. If upgrading, the flow continues to a Stripe-style checkout where the user enters payment details and confirms the subscription. Exit is the updated Plans & credits screen showing the new active plan and refreshed credit balance. Steps: Current Plan Overview → Checkout and Payment → Updated Subscription Status **Asos: Customer Care Search & No-Results** (ios, 2 screens, Marketplace) [Refero](https://refero.design/flows/2940) > Mobile customer care entry that surfaces a prominent search box and categorized FAQ topics. Entry: user opens the Customer Care section (mobile browser). The primary path shown is: user enters a query in the help search (example: "visor") and is taken to a search results screen. Exit points: refine search and resubmit, select a FAQ category or question, tap header navigation (menu, account, basket), or close the help overlay. Assumptions: user is on a mobile browser (portrait), may be signed in or out (header shows account icon but no explicit sign-in state), and has basic browsing permissions. Steps: Customer Care landing → Search results (empty state) **District: Apply Offer Code** (web, 3 screens, Entertainment, Food & Drink, Travel & Booking) [Refero](https://refero.design/flows/13007) > The user is on an Offers screen during checkout or booking. They can either manually enter a coupon/promo code in the input field at the top or choose from a list of suggested offers below. The entry point is the offers panel opened from a broader booking/payment context, and the exit point is either a successful offer application or an inline validation error if the entered code is invalid. **Lovable: Help and Resources Center** (web, 5 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12206) > This is a help and support entry experience for Lovable. The user starts on a public support/resources landing page with navigation, a central help search prompt, resource cards, FAQs, and a build prompt at the bottom. The user can also open a centered search modal, enter a query or numeric text, and review an AI-generated response with suggested follow-up chips. Assumption: the user is on desktop and not required to be logged in to access the visible help content. Steps: Help Landing Page → Search Modal Opened → Query Entered → Search Suggestions → Assistant Response ##### Task: book GP appointment **District: Salon Service Booking** (ios, 8 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12874) > Happy path for booking a salon service and reaching payment method selection. The user starts on a salon profile, chooses to book a service, selects one service and an available date/time slot, reviews the appointment details, and opens the payment method picker. **District: Event Booking Confirmation** (ios, 5 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12872) > Happy-path ticket booking for an event, starting on a booking review screen and ending on a confirmed order with an accessible QR code ticket. **Doctronic AI: Appointment Payment Confirmation** (web, 2 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13700) > The user is on the final payment step of a doctor appointment booking. The appointment details are shown alongside payment options. The user can use saved payment information, enter card details manually, or choose another payment method. In the shown path, the user selects saved payment information, reviews the populated card and billing fields, and can confirm the booking for $39.00. The flow starts at the payment step and ends when the user submits the booking confirmation. The screenshots do not show the post-submission result. **District: Movie Booking Support Request** (ios, 7 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12840) > A user opens a movie-support conversation that already contains a booking summary. They confirm the booking, are prompted to describe the issue, open the attachment menu, and grant photo-library access. A booking confirmation containing a QR code and ticket details is shown as supporting context. The user types a request to replace cheese popcorn with salt, then sends the message with the booking image attached. The flow starts in an existing support chat and ends with the issue submitted to the support assistant. **Fresha: Salon appointment booking** (ios, 9 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4973) > Happy-path booking of a single salon service with optional handling for fully-booked dates. Entry: user browsing/search results and opening a provider card. Exit: appointment is confirmed and shown in a booking detail screen. Assumptions: user is signed in (booking flows show payment and confirmation), has location enabled (distance shown), and is using a mobile device. Steps: Search results — provider card → Provider services list → Select services → Select time — fully booked state → Select time — date with availability → Choose time slot → Pre-visit question → Review and confirm → Confirmation & booking details **Fresha: Add Appointment to Calendar** (ios, 3 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4974) > Happy-path: a signed-in mobile user views a confirmed appointment and chooses to add it to a calendar. Steps: Appointment details / Confirmed → Calendar selection modal → Device event editor (prefilled) **Fresha: Search & Book — Venue Listings** (ios, 2 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4966) > User performs a location-aware search for a service (Massage) with a selected date/time. The top of the screen shows a map and a persistent search bar (query + date/time). Below are filter chips (Sort, Price, Type) and a scrollable list of venue cards. Each card shows an image, name, rating, distance, available services with prices, and tappable time-slot buttons for booking. Happy path: user reviews a venue’s service and taps a time-slot to proceed into the booking flow. Steps: Venue list with time slots → No-results / Clear search state **Fresha: Refine search and pick time** (ios, 19 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4958) > User starts on a results list for a previously-entered query (example: "Barbering") and opens the search/filter modal to refine the search. From the modal the user can change the treatment (Treatment or venue), location, date and time, or browse categories/recents. The user drills into each filter: selects a treatment type, picks or types a new location (including using current location), selects a date from a calendar, and chooses a time window (preset or custom). After applying filters the modal returns to the results list which updates to show venues matching the query and filters with available time slots. Assumptions: mobile (iOS-style) device; user is either unauthenticated or authentication state is not relevant to the flow; user navigates via modal sheets that stack and return to the search overlay and then to the results list. Entry: results list (search active). Exit: updated results list with applied filters and visible available times. Steps: Results list (initial) → Search / Filter modal (root) → Treatment search (open) → Treatment search — no match state → Treatment results (matches and venues) → Location picker (root) → Location search — no match → Location search — suggestions → Search modal updated (location chosen) → Date picker (open) → Date selection (select month / date) → Time picker (open) → Time selection — preset and custom → Time selection — picker list → Search modal (filters applied preview) → Top categories (browse) → Results list (filtered) → Select time slot / Book → Back navigation and re-open search **Omio: Booking retrieval by reference** (ios, 4 screens, Travel & Booking) [Refero](https://refero.design/flows/5242) > Happy path: A user on the 'Your bookings' screen opens a "Retrieve booking" modal, enters a booking reference and the email used for the booking, and submits to fetch the booking so it appears in their upcoming bookings. Steps: Bookings list (Upcoming) → Retrieve booking modal (empty) → Enter details and submit → Lookup error (no booking found) **Longevity Deck: Add activity to daily timeline** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/9895) > Happy path: From the day timeline (Yesterday) the user taps an empty timeslot (11) to add an activity. The app opens an "Add Activity at 11" picker with search, category filters, and a scrollable list of activities. The user picks an activity (optionally adjusting duration or amount in the inline numeric control) and taps the circular + control to add it. The picker closes (or the list dismisses) and the selected activity appears as a colored card in the 11 slot on the timeline. Entry: day timeline with empty slot. Exit: timeline updated with new activity card. Assumptions: user is signed in and on the daily timeline view, editing is allowed, device shown is a phone in dark mode. Steps: Day timeline with empty slot → Activity picker — top of list → Activity picker — select activity → Updated timeline with activity card **Omio: Booking retrieval via Help Center** (ios, 4 screens, Travel & Booking) [Refero](https://refero.design/flows/5247) > Mobile Help Center flow for retrieving a booking or searching support content. Steps: Customer Help Center landing → Search field focused → Inline empty search feedback → Full search results – empty state **Fresha: Appointment cancellation and management** (ios, 4 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4975) > Happy-path: A logged-in mobile user views the details of a scheduled appointment, opens the management sheet to reschedule or cancel, confirms cancellation, and returns to the appointment details now showing a Cancelled state with options to rebook. Steps: Appointment details (Confirmed) → Manage appointment (sheet) → Cancel confirmation → Appointment details (Cancelled) **Genie: Weekend Plan Assistant** (ios, 7 screens, AI Tool, Social Networking) [Refero](https://refero.design/flows/12431) > A mobile assistant conversation starts on a home/chat screen with suggested prompt chips and a message composer. The user types a planning request that mentions a friend, then sends it. The assistant responds by looking up the friend’s profile, generating a location-based recommendation, and showing nearby place cards. The flow assumes the user is already in the assistant experience and can type, mention people, and review suggestions. It ends with the assistant providing a drafted plan and a follow-up question about timing. Steps: Assistant Home → Draft Message with Mention → Ready to Send → Message Sent → Profile Lookup → Recommendation Results → Plan Draft and Follow-up **District: Upcoming Movie Reminder** (ios, 4 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12870) > Happy path for setting a reminder on an upcoming movie and optionally choosing preferred theatres. The user enters from the Upcoming movies listing screen, where movie cards show a reminder icon. After tapping a movie’s reminder control, a confirmation state appears and may prompt theatre preference selection. The user can open a theatre picker modal, select up to three theatres in the current city, save preferences, and return to the listing. **Fresha: Treatment / Venue search** (ios, 3 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4959) > User opens a search modal (overlayed on a map) to find a treatment or venue for "Barbering." The happy path: user enters a query (or picks a recent/category), refines optional filters (location/date/time), then selects a treatment or a venue from the suggested list. Steps: Search & filters → Type and refine → Review matching items **Fresha: Treatment & Category Search** (ios, 3 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4960) > Mobile (iOS) search & filter flow for finding a treatment or venue. Steps: Search & filter modal → Treatment or venue selector (with suggestions) → No-match / Top categories **Airbnb: Photographer booking** (ios, 4 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6436) > Happy-path: A mobile user discovers a photographer listing, taps Reserve, picks a package, date, and time, then confirms booking and payment. Entry: provider detail card (listing page) with service cards and a sticky Reserve CTA. Exit: user reaches the Confirm & Pay screen with booking summary and payment method (ready to review/complete payment). Assumptions: Mobile (iOS-style) interface; user has access to a saved payment method (a card is visible on the confirm screen); availability is shown per-date and per-package. The UI uses a modal sheet for scheduling and horizontal time-chip selectors for available time slots. Steps: Provider listing & Reserve CTA → Schedule modal — pick date and time → Schedule modal — time confirmed → Confirm and pay **Omio: Ticket access and booking management** (ios, 2 screens, Travel & Booking) [Refero](https://refero.design/flows/5231) > Happy path: A mobile iOS user opens the "Your journey" area and selects the Ticket tab to access their ticket and booking details. From this screen the user can: add the ticket to Apple Wallet, copy the booking number, or open management options (change booking, view cancellation options, view confirmation). Steps: Ticket overview → Copy confirmation toast **Airbnb: Guidebook creation and place recommendation** (ios, 20 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6448) > Happy-path: Host navigates from the Listing editor -> Arrival guide section to open Guidebooks, creates/edits a guidebook (Maya's Guidebook), attaches a listing, edits the guidebook cover image using the photo library, then adds a place recommendation (searches for a restaurant), categorizes it and adds images and a tip. Steps: Arrival guide overview → Guidebooks list → Guidebook editor (empty) → Choose listings modal (initial) → Select listing → Save listing selection → Guidebook editor (after listing linked) → Edit cover modal – pick image → Request photo library permission → Grant photo library access → Photo library picker → Cover updated preview → Add content modal → Place search UI → Select a place from results → Categorize place (required) → Add tip and images for place → Save place recommendation → Preview guidebook item → Return to Guidebooks list **Poppy: Create Calendar Event** (ios, 6 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12189) > This is a mobile assistant-style scheduling experience. It starts on a home screen with shortcut actions and a microphone-based command entry, then moves into event creation after the user chooses Create and Event. The user can pick a quick date option or enter a custom title, then refine the scheduled date and time on a calendar-style editor before confirming. Steps: Assistant Home → Choose Create Task → Select Event Type → Set Event Timing → Enter Event Details → Adjust Calendar Event **WhatsApp: Schedule Call and Invite Participants** (ios, 8 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12683) > A mobile calls interface opens on the Calls tab. The user can access recent calls, shortcuts for calling and scheduling, and create a new scheduled call. The flow starts from the Calls screen, moves into scheduled call creation, then into contact selection and chat confirmation, and ends back on the Calls screen with the new upcoming call visible. Steps: Calls Home → Calls Menu → Scheduled Calls Empty State → Schedule Call Details → Send To Recipients → Recipient Confirmed → Chat Event Confirmation → Calls Upcoming State **TidyCal: Therapy Session Booking** (web, 6 screens, Business, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/12503) > Happy path for a therapy provider booking site. The user starts on a therapist’s public landing page, reviews available services, selects a session type, chooses a date and time from either a calendar or list view, enters name and email in a confirmation dialog, and receives a booking confirmation screen. Steps: Therapist Landing → Service Selection → Calendar Availability → Booking Details → Enter Contact Details → Booking Confirmed **Acne Studios: Private Appointment Booking** (web, 3 screens, Marketplace) [Refero](https://refero.design/flows/4360) > User reaches the 'Book a private appointment' page (desktop layout). The page shows a left-hand form and a large image on the right. Country/region and store are shown as dropdown selectors at the top of the form, followed by Name, Email, Phone (optional) and Requested date & time (optional). The primary CTA at the bottom is 'REQUEST APPOINTMENT'. Steps: Open booking form → Fill contact details → Submit request and see confirmation **Acuity Scheduling: Consultation Booking** (web, 5 screens, Business, Scheduling Tools) [Refero](https://refero.design/flows/9866) > Happy-path booking: an unauthenticated desktop user selects a service, chooses a date and time from a calendar, fills required contact fields, confirms the booking, and arrives at a confirmation screen with calendar export and reschedule/cancel options. Steps: Select Appointment Type → Choose Date & Time → Enter Contact Information → Review Filled Information → Appointment Confirmed **Acuity Scheduling: Create and publish appointment type** (web, 8 screens, Business, Scheduling Tools) [Refero](https://refero.design/flows/9857) > Happy path: A signed-in user managing appointment offerings starts on the Appointment Types list, chooses to create a new individual appointment type, completes details (name, description, duration, confirmation message), adds an image, selects calendars and optional resources/forms, then saves the new appointment type. Steps: Appointment Types list → Details — initial → Details — completed → Image & Calendar selection → Image uploaded → Forms, Resources & Additional Options → Saving — loading → Appointment Types list — confirmation **Fresha: Salon appointment booking** (web, 14 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/2379) > Happy-path booking of a single service (Gel Manicure) at a listed salon (AMA Brixton). Steps: Salon detail (services overview) → Choose booking option → Select services (list) → Service details modal → Service selected (summary visible) → Select professional (grid) → Professional confirmed → Select time (calendar + slots) → Date view and slot selection → Full time list (scroll view) → Change/select professional from time screen (modal) → First-visit question → First-visit selected (example: No) → Review & confirm (payment entry) **Saint Laurent: Store Appointment Booking** (web, 6 screens, Marketplace) [Refero](https://refero.design/flows/501) > Happy path: a shopper uses the store locator to find a nearby Saint Laurent boutique, opens the store details, and requests a private in-store appointment via a right-side booking panel. Steps: Store locator results → Store detail (Grant Avenue) → Open appointment panel → Select date — no hours available → Select alternate date and time → Submit appointment — error response **TidyCal: Create Booking Invite** (web, 6 screens, Business, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/12502) > Happy path for creating a booking invite from the Bookings section. The user starts on the Bookings screen, opens the Create a booking panel, selects a booking type, timezone, date, time, and guest details, then sends an invite. After submission, the modal closes, a success toast appears, and the new invite is shown in the Invited list with its status and actions. Steps: Invited Bookings Empty State → Booking Details Panel → Date and Time Selection → Guest Details Completed → Invite Sent Confirmation → Invited Bookings List **West Elm: Design Appointment Scheduling** (web, 10 screens, Home & Decor, Marketplace) [Refero](https://refero.design/flows/1674) > Happy-path summary: A desktop user requests a free Design Crew appointment. They start on the service overview, pick an appointment type (in-home, virtual, or in-store), select a local store (if applicable), choose a date and time from an availability grid, enter contact information, add project details, and receive an on-screen booking confirmation with a reference and follow-up questions. Steps: Appointment Type Selection → Find Local Store → Date & Time - Availability Overview → Select Time Slot → Change Date (Calendar) → Enter Contact Information → Info Required Modal → Confirm Contact Details → Add Project Details → Booking Confirmation & Follow-up **GlossGenius: Create multi-service appointment** (web, 11 screens, Business) [Refero](https://refero.design/flows/9972) > Happy path for creating a multi-service appointment from the weekly schedule on desktop. Entry: user is on the weekly calendar for a location (Laboratory) and a target date is visible (Feb). The user selects a time slot or the global + button to open a right-side "New appointment" drawer, chooses a client and provider, adds services (which updates the end time), and books the appointment via a confirmation modal. Exit: the appointment appears on the calendar (block with hashed buffer) and a success toast is shown. Assumptions: user is authenticated and has permission to create appointments, the interface is desktop with sufficient horizontal space for a right-side drawer and central calendar, provider list and client list are prepopulated. Steps: Weekly calendar view (initial) → Select start time → Open New appointment drawer → Choose client → Open services selector → Select multiple services → Review assembled appointment → Open booking confirmation modal → Toggle notification and confirm → Appointment created (side panel) → Calendar updated with appointment **Cal: Meeting scheduling** (web, 6 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3497) > Happy-path booking: A visitor opens a host's public booking listing (desktop, not signed-in) and sees a list of offered meeting types (durations and descriptions). The visitor selects a meeting type (30 Min Meeting), picks an available date from a month calendar, chooses an available time slot, fills contact information and optional notes, and confirms the booking. Steps: Meeting types list → Confirm meeting type selection → Month calendar — pick a date → Choose time slot → Booking details form → Confirmation summary **Fresha: Group appointment and waitlist booking** (web, 30 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/2375) > Happy-path booking attempt for a group appointment (up to 3 guests) starting from a salon’s listing. Steps: Salon overview & services → Booking type chooser → Select services – master list → Service options dialog (Lash Lift) → Add additional services (e.g., Level 5) → Add guests and services summary → Edit services for a guest → Confirm selected service(s) → Join the waitlist – date/time entry → Join the waitlist – time range picker → Add multiple preferred times → First visit intake question → Review & confirm waitlist – payment and notes → Waitlist error – unavailable → User attempts to exit booking → Exit confirmation decision → Booking type chooser (alternate view) → Service options selection confirmation → Multiple preferred time rows shown in summary → Final error state and guidance **Doctronic AI: Licensed Doctor Appointment Booking** (web, 25 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13699) > The user begins with a saved Doctoronic medical report and chooses to continue to a licensed doctor. They select a payment method, provide the state where the appointment will occur, choose an immediate or scheduled appointment time, enter patient and contact details, and reach the final booking confirmation and payment step. The flow assumes the user has a report ready to share and may have saved contact or payment information. The visible sequence ends on the payment screen before a completed booking confirmation is shown. **TidyCal: Booking Types Management** (web, 3 screens, Business, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/12493) > This is a booking types management screen in a scheduling admin area. The user can view a list of booking types, filter them by status, enable or disable a type, open an actions menu for each item, and share a booking page link through a share panel. Steps: Booking Types List → Item Actions Menu → Share Booking Link **TidyCal: Therapist Booking Landing Page** (web, 2 screens, Business, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/12486) > A therapist’s public booking landing page is shown in two visual theme variations. The user is assumed to be a site owner or editor previewing the page in an editing environment, with access to theme, typography, color, and save controls. The page begins with a hero introduction, then scrolls through credibility stats, bio, service offerings, testimonials, FAQs, and a final booking call to action, ending with footer/social link placeholders. The primary exit point is the booking button, which leads to scheduling, while the editor can also save the design changes from the top bar. Steps: Page Preview and Booking Entry → Theme Variation Preview **Fresha: Search availability by treatment, location, date, time** (web, 8 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/2323) > User starts on their Appointments screen (desktop, logged-in — profile avatar and an upcoming confirmed appointment are visible). From here they use the persistent search bar at the top to find new availability: they open the Treatment selector and pick "Tattoo & piercing", open Location and choose a saved address (Home), open the Date picker and choose a specific day (Dec 23), then set a custom time window (16:00–21:00). After applying filters and using Search, the interface displays search results on a split layout: a scrollable list of venues and available time slots on the left and an interactive map on the right. Steps: Appointments overview → Choose treatment category → Open location selector → Open date picker → Select a calendar date → Open time filter presets → Set custom time range → View filtered search results **Cal: Request reschedule for booking** (web, 5 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3493) > Happy-path: A signed-in user views their Upcoming bookings list, opens a booking's Edit dropdown, chooses "Request reschedule," types an optional reason in a modal, and submits the request. Entry: Bookings > Upcoming (desktop, authenticated user). Exit: modal closes and a transient confirmation toast is shown: "Reschedule request sent." The interface guides the user with an Edit dropdown → Request reschedule modal that explains the effect (cancel meeting and notify scheduler) and provides an optional free-text reason. Steps: Bookings list (Upcoming) → Open Edit dropdown → Request reschedule modal (empty) → Enter reason and submit → Confirmation toast **IKEA: Interior design appointment booking** (web, 9 screens, Marketplace, Home & Decor) [Refero](https://refero.design/flows/8270) > A desktop booking flow for IKEA's online interior design services. Steps: Service landing → Enter postcode → Choose service → Service details (expanded) → Select date (calendar) → Confirm timeslot (alternate state) → Contact & project details (form start) → Contact form (filled state) → Booking confirmation & payment prompt **TidyCal: Booking Page Setup and Scheduling** (web, 5 screens, Business, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/12499) > Happy path for configuring and sharing a booking page, then reviewing available booking types and availability. The flow appears to start from a dashboard/booking setup area after the account is already live, with the user signed in as the business owner. It exits on a booking-type preview/scheduling screen where the user can choose a date and time or continue editing shared booking settings. Steps: Setup Checklist → Availability Calendar → Theme Selection → Availability Review → Booking Type Review **Doctronic AI: Appointment Document Upload** (web, 4 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13702) > A logged-in user opens the Appointments area and views an upcoming video appointment with booking in progress. They open the Files section to review the file-upload prompt, add documents, and wait for the upload to complete. The uploaded PDF files then appear in the appointment's file list, where each file has a remove control. ##### Task: budget spending insights **Wabi: Submit Personal Analysis** (ios, 8 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12052) > This is a mobile analysis experience where the user enters a personal feeling, dream, or life situation, then submits it for interpretation by a panel of analysts. Steps: Prompt Entry → Text Composed → Submit Analysis → Processing State → Results Begin → Analyst Responses → Detailed Reading → Expanded Analysis View **Pool: Place Details to Directions** (ios, 2 screens, Utility) [Refero](https://refero.design/flows/11980) > Happy path starts from a saved or viewed place detail screen for Citizens Bank Park, showing venue information, actions like call/website/tickets, and a directions button. The user can open navigation from the place detail view, which transitions into the maps interface focused on Citizens Bank Park and nearby points of interest. Steps: Place Details → Map Results **Spyglass: Export and Rate Analysis** (web, 3 screens, Business, Inspiration Board) [Refero](https://refero.design/flows/13461) > A user reviews a generated advertising analysis in the Spyglass workspace and interacts with the response action toolbar. The sequence begins with the completed response visible, then shows a PDF download confirmation, a hover tooltip for the positive-feedback control, and finally the positive-feedback control selected. The user appears to be logged in and viewing an existing conversation or research workspace. The flow ends with the response retained in the conversation and feedback recorded. Steps: Review Generated Analysis → Identify Helpful Response Control → Submit Positive Feedback **GoFundMe: Search Results Category Filter** (web, 3 screens, Fundraising) [Refero](https://refero.design/flows/11734) > A logged-in user on GoFundMe searches for “paris” from the top search field and browses result types using the category filter. The initial results show nonprofit organizations with a secondary location filter. The user opens the category dropdown and changes the result type to Profiles, which updates the results grid to people profiles matching the same query. Entry is the global search/results view; exit is either selecting a result or refining the search/filter. Steps: Nonprofit Search Results → Category Dropdown → Profile Search Results **Foodllama: Nutrition Insight Completion** (ios, 7 screens, Health & Wellness) [Refero](https://refero.design/flows/11779) > This is a guided content experience in the Foodllama mobile interface. The user starts on a home/dashboard screen showing a featured insight card and a progress summary for today’s insights. Tapping or swiping into the card opens a full-screen lesson sequence with multiple steps, each presenting one nutrition tip with illustration, title, short explanation, and a progress indicator plus a Leave action. The sequence ends on a completion screen with celebratory confetti, a trophy graphic, and options to tell friends or finish. After completion, the user returns to the dashboard, where the insight count updates. Assumptions: user is already inside the app, on a mobile device, and viewing personalized insight content. Steps: Insight Card Dashboard → Intro Tip Screen → Carb Tip Screen → Strength Tip Screen → Completion Celebration → Return Dashboard → Updated Insight State **Binge: Viewing Statistics Breakdown** (ios, 3 screens, Entertainment) [Refero](https://refero.design/flows/12972) > A signed-in user views a personal viewing breakdown screen, then opens a display filter to switch between All, Movies, and Shows. The screen updates the chart and stats based on the selected category. Entry is from a stats or profile area; exit is via back navigation. Steps: Overview Stats → Display Filter Menu → Filtered Stats View **District: Restaurant Q&A Assistant** (ios, 7 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12846) > Happy path for an in-context restaurant question assistant. The user starts from a restaurant detail context and asks about Hotel Rishi. The interface shows the current topic, displays the user’s question as a highlighted prompt, and returns an AI-generated response with related action cards (Food Menu, Bar Menu, Beverage Menu). The user can rate the answer, ask a follow-up by text or voice, or submit a new query such as finding nearby breakfast options. The flow ends with a refreshed response and the follow-up input ready for another question. **Meta AI: Weekend Activity Suggestions** (web, 9 screens, AI Tool) [Refero](https://refero.design/flows/12751) > User starts on the assistant home screen, enters a query about free things to do nearby for the weekend, and waits while the system searches and composes an answer. The response appears in the main conversation area with a right-side panel that can switch between Steps and Sources. From there, the user can review how the answer was assembled, inspect cited sources, and hover or click source items to preview the referenced content. Steps: Home Prompt → Searching State → Steps Panel Opened → Draft Answer Displayed → Citation Hover → Sources Tab Opened → Source Preview Card → Image Preview → Source Link Hover **Cohere: Market overview chat** (web, 3 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3856) > Happy path: A user composes a natural-language prompt in the Cohere Playground chat area, optionally configures JSON mode/schema and advanced parameters from the right-hand panel, sends the prompt, and reads the generated market overview. Entry: user lands on the Playground with Chat selected and an empty chat editor. Exit: the user has a multi-section, generated response (market size, drivers, innovations, challenges) visible in the chat and the editor ready for follow-up. Assumptions: user is on a desktop viewport, logged in, and has access to model/parameter controls on the right-hand panel. Steps: Prompt composition & schema setup → Submit prompt and view error / initial generation → Review expanded assistant response ##### Task: business account onboarding company verification **Peerlist: Create Company Profile** (web, 6 screens, Job Search, Business, Social Networking) [Refero](https://refero.design/flows/1102) > Happy path: A logged-in user (with permission to create a company) initiates creating a company profile via a modal. They choose an organization profile type, complete the profile form (display name, website, tagline, location, handle, size, etc.), then trigger an email verification step. After entering the verification code and accepting terms, the profile is created and the user is taken to the newly created company profile overview. Steps: Choose profile type → Company profile form (empty) → Company profile form (filled) → Open verify work email modal → Enter verification code → Company profile overview ##### Task: calendar scheduling meeting **Genie: Chat Reminder Setup** (ios, 3 screens, Social Networking, AI Tool) [Refero](https://refero.design/flows/12426) > A chat-style assistant screen where the user types a reminder request, sends it, and receives confirmation that a reminder has been created for tomorrow at 9:00 AM. The flow begins in an active conversation with the keyboard open and ends with the assistant confirming the reminder and offering to save reference photos to a vault. Assumptions: user is already in the conversation, signed in, and has permission to create reminders. Steps: Compose Reminder → Message Sent → Reminder Confirmed **Cofounder: Daily Meeting Summary Test** (web, 5 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/7900) > Happy path: a user viewing a configured scheduled automation called "Daily Meeting Summary Email" starts a manual test of the scheduled trigger, watches the test run initialize, and inspects the run timeline as integrations (Calendar) and internal reasoning steps are executed. Steps: Plan view — Automation details → Test confirmation modal → History view — Test run initializing → History view — Run processing progress → Run timeline — Integration and reasoning details ##### Task: care plan **Poppy: Poppy Account Setup and Subscription** (ios, 42 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/12176) > A first-time user sets up Poppy on an iPhone. The user enters a phone number, verifies it with an SMS code, provides a preferred name, selects the areas where Poppy can help, and configures notification behavior. The user is then shown Sprout and Bloom subscription options, selects the Bloom annual plan, and completes the purchase through Apple Pay. After confirmation and a short preparation state, the user reaches the personalized Poppy dashboard. The screenshots show a successful paid onboarding path; the visible “Not now” option on the subscription screen is an alternative exit that is not followed. Steps: Welcome and Phone Entry → Phone Number Input → Submit Phone Number → Phone Verification Request → Enter Verification Code → Complete Verification → Name Entry → Confirm Name → Choose Support Areas → Confirm Support Preferences → Customize Daily Outreach → Sprout Subscription Option → Bloom Subscription Option → Compare Subscription Features → Confirm Bloom Annual Plan → Apple Pay Subscription Review → Authorize Subscription → Dismiss Purchase Confirmation → Prepare Personalized Experience → Personalized Dashboard **Ada: Symptom assessment - conversational** (ios, 31 screens, Health & Wellness) [Refero](https://refero.design/flows/4097) > Mobile, conversational symptom assessment starting from a 'New assessment' entry screen (user name shown). The user is guided through: confirming who the assessment is for, entering a main symptom (search with autocomplete), answering a short series of multiple-choice questions, optionally opening explanatory anatomy/help modals, and receiving a results report with ranked possible causes. Steps: Intro / Start assessment → Select assessment recipient → Prompt for main symptom → Symptom search modal (keyboard) → No results fallback → Autocomplete suggestions → Symptom detail and selection → Duration question → Orthostatic effect question → Open explanation modal (example) → Explanation modal — confirmation → Explanation modal (neck mobility) → Neck movement question → Reduced mobility illustration (detail) → Concentration question → Symptom trend question → Assessment complete — report being generated → Open report CTA → Results summary / possible causes → Detailed causes list and continue **Ada: Record a symptom entry** (ios, 12 screens, Health & Wellness) [Refero](https://refero.design/flows/4107) > Happy path: From a feedback prompt the user chooses to track symptoms, picks a symptom (Dizziness), answers a short guided questionnaire (trigger: standing up effect; description type), optionally sets current severity using a segmented color-coded slider and +/- controls, optionally adds freeform notes, saves the entry, then returns to a timeline where the new entry can be opened to view details. Steps: Feedback prompt → Choose symptom → Initial symptom questions → Select trigger effect → Select symptom description → Open severity control → Set severity (low) → Set severity (mid) → Set severity (high) → Open notes field → Enter free-text note → Save and view entry **Ada: Profile setup and symptom logging** (ios, 66 screens, Health & Wellness) [Refero](https://refero.design/flows/4095) > Happy path for adding a new user profile and logging a symptom (dizziness) into the symptom tracker. Steps: Home / Start assessment → Add profile — Intro & safety note → Ask for user name → Name input active → Ask sex assigned at birth → Why only two sexes (modal) → Ask date of birth → Ask smoking history → Ask high blood pressure history → Open symptom modal (Dizziness) → Dizziness — trigger & type → Dizziness — optional severity collapsed → Set severity — low value → Set severity — mid value → Set severity — high value → Optional notes — edit → Save symptom (network notice) → Symptom tracking timeline → Symptom detail summary → Finish / Done **Ada: Medical quality article walkthrough** (ios, 10 screens, Health & Wellness) [Refero](https://refero.design/flows/4104) > Happy path: A user finishes a symptom assessment and sees a high-priority recommendation ("Seek emergency care") with an inline "Tell me more" button. Tapping that link opens Ada’s external "Medical quality" article (about.ada.com) in a browser/webview. The user scrolls through the article: headline, supporting copy, visual illustrations, a "What to look for" section with accordion items (clinical governance, user experience, condition coverage, etc.), expands "Condition coverage" to view bullet points, continues to the "How Ada’s AI works" section, sees a CTA to contact the company, reviews the reference list (numbered DOIs), and reaches the footer with legal links and social icons. Steps: Symptom result with "Tell me more" → Article landing — Medical quality → Evidence section header → "What to look for" introduction → Accordion list of quality pillars → Expand 'Condition coverage' → How Ada's AI works → Contact CTA and Keep reading → Reference list (citations) → Footer and site navigation **Doctronic AI: Anonymous Health Consultation** (web, 6 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13696) > The user starts on a health consultation entry screen and submits a symptom description to an AI assistant in Anonymous Mode. The system displays privacy and emergency notices, processes the message, summarizes the concern, and asks the user to accept the Terms of Service before continuing. The flow begins with the user ready to ask a health question and ends with the consultation ready for the user to provide more information. **Doctronic AI: Start Health Consultation** (web, 2 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13693) > The flow begins on a logged-in health assistant home screen with a collapsed navigation rail. The user expands the navigation to view available sections while remaining on the New consult screen. From here, the user can enter a health question and submit it. The screenshots show the entry state and expanded-navigation state; no question has been submitted yet. **Breathwrk: Breathwrk Onboarding & Purchase** (ios, 35 screens, Health & Wellness) [Refero](https://refero.design/flows/7724) > Onboarding → account / purchase → first-use tutorial. Entry begins at a marketing-style welcome/splash that highlights benefits (sleep, stress, energy). The user is guided through a short educational carousel (value props + neuroscience), asked to choose goals, offered a quick lung calibration test, then presented with a premium subscription modal (7-day trial pricing). Native platform purchase and sign-in modals appear (App Store / Sign in with Apple). After confirmation the user sees a brief onboarding tutorial and lands on the main activity screen. Steps: Welcome Splash → Benefit Tags → Problem Statement → Quick-Time Benefit → Neuroscience Evidence → Goals Selection (intro) → Goals Confirmation → Outcome Graph + CTA → Lung Test Intro → Inhale Instruction → Measurement / Feedback → Subscription Promotion → Pricing Selection → Native Purchase Confirmation → Purchase Success → Account Sign-in Prompt → Sign in with Apple Modal → Loading / Transition → Quick Tutorial Intro → Main Home (Ready State) **MonoDesk: Conversational task planner** (web, 8 screens, Productivity, Project Management) [Refero](https://refero.design/flows/10852) > Happy-path: A logged-in user opens an in-product conversational assistant to request a personal productivity/growth plan. The assistant asks clarifying questions, offers quick-reply choices for focus areas, generates a week-based task plan, and then provides CTAs to add professional tasks or schedule the generated tasks into the product Planner (calendar). Entry: assistant chat screen. Exit: Planner view filtered to the user's selected project with the new tasks visible. Assumptions: user is authenticated, has at least one project (named Humanity) available, and the product can create tasks and update the Planner/calendar. Steps: Chat start / prompt → User submits goal → Assistant thinking → Assistant asks clarifying questions → Quick-reply focus selection → Assistant generates task plan → Task detail links and CTAs → Planner view confirmation and options **MonoDesk: Weekly planning assistant** (web, 5 screens, Productivity, Project Management) [Refero](https://refero.design/flows/10853) > Happy-path: A signed-in user opens the assistant home screen, uses a category chip to reveal suggested prompts, selects a suggested prompt which populates the message composer, sends the prompt, and receives a multi-section weekly plan with recommendations and follow-up action buttons. Steps: Assistant home / composer → Category suggestions opened → Prompt populated in composer → Assistant returns weekly plan → Recommendations and follow-up CTAs **Longevity Deck: First Launch Introduction** (ios, 5 screens, Health & Wellness) [Refero](https://refero.design/flows/9879) > This is a first‑time user onboarding carousel that introduces the app’s core features, asks the user to acknowledge a medical disclaimer, then lands the user on the app’s main entry screen where they can begin setup (upload a photo, browse protocols, or connect health data). Steps: Intro: Personal Story → Feature: Connect Health Data → Feature: Track Daily Timeline → Legal: Medical Disclaimer → Main Entry: Setup & CTAs **Givingli: New User Onboarding & Subscription** (ios, 45 screens, Lifestyle, Scheduling Tools) [Refero](https://refero.design/flows/8150) > Happy-path onboarding for a gifting app that collects a user's name, preferences, celebration types and recipient categories, then presents a premium subscription offer and completes an in‑app purchase. Steps: Welcome / Splash → Name Input (first name) → Name Input (last name) → Welcome Greeting → Why are you here? (multi-select) → Selected reasons (chips active) → Choose celebrations (grid, initial) → Celebrations — some selected → Celebration options continued (scroll) → Recipients / Gift list categories → Recipient selection confirmation → Premium upsell landing → Choose plan (yearly selected default) → Plan selection swap (monthly highlighted) → Loading / requesting purchase → Platform purchase sheet → Purchase success confirmation → Home screen (post onboarding) → Subscription UI states (alternate) → Platform confirmation details **Structured: AI Meal Planner** (ios, 5 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3288) > Happy path: A user opens the 'Structured AI' modal to create a plan (example: a meal plan). The modal greets the user, offers horizontal suggestion chips and a prominent input field with voice and camera affordances. The user selects or types a prompt, then taps Generate. The UI returns a generated plan preview area with a 'Your plans' summary, an 'Accept All' action, trust notice about AI accuracy, and a list of created calendar-style 'Meal Plan' items. Steps: Open Structured AI and view suggestions → Browse suggestion chips and focus input → Type a prompt → Generate plan → Review generated items **Mozi: Share Plans via Messages** (ios, 4 screens, Social Networking, Travel & Booking) [Refero](https://refero.design/flows/4147) > User is on the My Plans screen (mobile, logged-in) and wants to share their upcoming plans with others. From the My Plans list the user opens a "Share my plans" modal, chooses to share, selects Messages from the system share sheet, and lands in the Messages composer with a preview card of their plans inserted. Steps: My Plans list → Share options modal → iOS share sheet → Messages composer with preview **Weather mini: Trip Planning and Save** (ios, 7 screens, Weather) [Refero](https://refero.design/flows/13105) > Happy path for creating a multi-city trip from the Trip overview, entering dates and stops, using AI/search suggestions, generating trip details, and saving the plan back to the main trip list. Steps: Trip Overview → Initial Trip Setup → Search Suggestions → Processing Entry → Trip Draft with Stops → City Weather Details → Saved Trip List **Mozi: Create a trip plan** (ios, 16 screens, Social Networking, Travel & Booking) [Refero](https://refero.design/flows/4137) > Happy-path: A logged-in user on the app Home screen taps the central + to create a new travel plan. The New plan sheet opens as a modal with fields for city, dates, description and a tentative toggle. The user opens the city selector, searches for and selects a city (chooses San Francisco), grants location permission when prompted, then sets a date range via the calendar, adds a short description (24-character limit), marks the plan as tentative, and saves. The flow ends on the city detail screen showing the newly created upcoming plan card and a confirmation CTA. Assumptions: user is signed in, device is iOS (modal and system location prompt shown), user has network access for location search. Steps: Home (empty state) → New plan form (empty) → Plan location (search) → Plan location (search ready) → Location permission prompt → Plan location (current location enabled) → Plan location (search results) → Plan location (select city) → New plan (city filled) → Calendar (month view) → Calendar (select end date) → Calendar (range saved) → New plan (dates filled) → Description entry (character limit) → Tentative toggle → Save plan and city confirmation **Structured: Structured Planner Download** (ios, 4 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3286) > Happy path: A user on the Structured AI home screen taps a promotional CTA to obtain a printable planner sheet. The CTA opens an external link (Linktree) listing Weekly / Monthly / Annual versions. The user selects a version, confirms the download in the browser prompt, and the PDF opens in the device's PDF viewer where they can share, save, or print it. Steps: Home - Download CTA → Link hub - Choose Version → Download Confirmation → PDF Preview **Structured: Handwritten Planner Extraction** (ios, 5 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3293) > Happy-path: A user scans a handwritten weekly planner and reviews AI-extracted calendar items. From a modal titled "Structured AI" the user can preview extracted tasks, open a larger scan preview, add custom instructions, regenerate the extraction, and then review the updated list. Steps: Initial extraction preview → Detailed scan preview → Provide custom instructions → Regeneration / processing → Updated extraction results **Napper: Baby schedule & invite** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/8812) > Happy-path: A parent edits a baby's schedule and invites a caregiver. Steps: Schedule settings → Share baby profile (invite) → Caregiver added / list view **Structured: AI-powered day planner** (ios, 21 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3287) > Happy path: From the daily timeline the user launches a task-creation flow, chooses the "Plan your day with AI" suggestion, provides a natural-language description of their plans (typed or voice), grants required system permissions (speech recognition, microphone, calendar/tasks access), and lets the Structured AI generate a prioritized agenda. The user reviews the AI-proposed tasks, accepts (or edits/discards) items, and adds a selected item into the calendar. Steps: Daily timeline (entry) → New Task sheet → Structured AI — intro screen → Structured AI — suggestions carousel → Provide freeform plan (typed) → Start speech transcription (optional) → Grant microphone permission → Share calendar/tasks permission → AI processing — prioritizing → AI processing — optimizing → Review AI suggestions summary → Accept / expand suggestions → Item actions — Add / Edit / Discard → Add Task — set time and duration → AI list after edits → Discarding a suggestion → Finalized agenda → Voice recording states → Generate AI agenda → Per-item inline actions and final edits **Care: Care.com Member Login** (web, 6 screens, Job Search, Lifestyle) [Refero](https://refero.design/flows/12227) > A returning Care.com user starts on the public homepage, enters the login experience, provides an email address and password, and is taken to their member dashboard. The final dashboard shows the user’s existing job, upgrade prompts, caregiver discovery options, educational content, and account navigation. The flow assumes the user is not currently logged in and has valid credentials. Steps: Care.com Homepage → Email Entry → Password Prompt → Password Entered → Member Dashboard → Dashboard Actions **Sunsama: AI-Assisted Task Capture** (web, 7 screens, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/11601) > A logged-in user is viewing the Backlog and opens the Sunny assistant from the floating action button. The user selects the Braindump context, sends a natural-language request about planning a trip to Tokyo, and receives confirmation that Sunny added a corresponding task to the backlog for the next month. The flow begins in the Backlog workspace and ends with the new task visible in the appropriate time bucket. Steps: Open Sunny Assistant → Choose Braindump Context → Start Conversation → View Assistant Prompt → Enter Travel Idea → Send Travel Request → Confirm Backlog Task **Airtable: Project intent and company details** (web, 3 screens, Project Management, Collaboration Tools, Scheduling Tools, Paperwork, Productivity) [Refero](https://refero.design/flows/1542) > Happy path: A user starts the Airtable Cobuilder onboarding screen to describe the project they want to build. They select one or more purpose tags, optionally write a short description, open the company details modal from the bottom-left control, fill optional company fields, and then press Next to continue to the next onboarding step. Steps: Select purpose and add description → Open company details modal → Fill company info and continue **Airtable: AI-guided base generator** (web, 15 screens, Project Management, Collaboration Tools, Scheduling Tools, Paperwork, Productivity) [Refero](https://refero.design/flows/1540) > Happy-path: A signed-in user with access to an empty Airtable workspace uses the Cobuilder (AI) flow to generate a new base. Steps: Empty workspace → Start method chooser → Cobuilder: initial prompt → Select use-case tags → Company details modal open → Fill company details → Open color picker → Color plan limitation tooltip → Theme applied preview → Generation begins — creating tables → Generation continues — creating pages → Generation status with artifacts count → Review screen: 'How does this look?' → Creating specific object (modal overlay) → Generated base preview **Navia: Content Plan Assistant** (web, 6 screens, Business, Productivity) [Refero](https://refero.design/flows/14325) > The flow starts on Liam’s dashboard with the Navia assistant panel open on the right. The user selects a suggested question about how the content plan works, reviews Navia’s response, then selects a second suggested question about what happens after 30 days and reads the answer. The flow ends with the assistant remaining open and ready for another question. The user is assumed to be signed in, viewing the dashboard on a desktop-sized screen, and permitted to access the assistant. No account changes or external navigation occur. Steps: Open Dashboard Assistant → Select Plan Question → Read Plan Explanation → Choose Follow-up Topic → Wait for Follow-up Answer → Read Renewal Explanation **Care: Pet Caregiver Search** (web, 32 screens, Job Search, Lifestyle) [Refero](https://refero.design/flows/12226) > A visitor starts on the Care.com home screen and selects the option to find a caregiver. They choose pet care, provide timing, pet count, service type, dates, hours, and pay preferences, then optionally add notes about the pets. After submitting the request, they choose between premium and limited access and arrive at a filtered pet-caregiver search results screen. The screenshots show a happy path for a user seeking one-time cat sitting, with no sign-in step visible. Steps: Care.com Home → Choose User Path → Select Care Type → Pet Care Selected → Care Timing → Pet Count → Cat Count Dropdown → Pet Count Updated → Service Type → Service Selected → Schedule Details → Schedule Updated → Pay Range → Pay Range Updated → Caregiver Notes → Notes Guidance → Notes Entered → Choose Access → Caregiver Search Results → Caregiver Profile Preview **Lovable: Plan Review and Approval** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12337) > A logged-in user is reviewing a proposed plan in a split-screen editor. The left side shows the task discussion and a plan card with a summary, while the right side shows the detailed plan content with a formatting toolbar. The user can inspect the proposal, optionally change the content type/format using the text style dropdown, and then either cancel or save and approve the plan. Steps: Plan Review → Text Style Selection → Approved Plan State **Cohere: Business Plan Outline Generation** (web, 2 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3827) > A single conversational session where a logged-in user submits a prompt to generate a business plan outline in French and views the formatted result. Steps: Open conversation and submit prompt → Expose copy control for generated content **Cohere: Business plan assistant interaction** (web, 3 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3829) > Happy-path summary: A signed-in user is inside a chat conversation and asks the assistant for a business-plan outline in French. The assistant returns a formatted outline in the conversation pane. The user has two PDFs staged in the message composer (not yet sent). The user hovers or clicks on the assistant message actions (e.g., Regenerate) to request a different response. After requesting regeneration, the assistant posts a new reply (in this case an apologetic message saying it couldn't find a plan in the provided documents). Steps: View generated business plan → Reveal message actions → Regenerated assistant reply **Care: Caregiver Review Submission** (web, 5 screens, Job Search, Lifestyle) [Refero](https://refero.design/flows/12231) > Happy path for leaving a review for a caregiver on Care.com. The user starts on a caregiver profile, opens the review form, selects an overall star rating, adds written feedback, answers whether they hired the caregiver through Care.com, submits the review, and reaches a confirmation screen with a call to action to continue browsing care options. Steps: Caregiver Profile → Start Review → Write Review → Review Ready to Submit → Submission Confirmed **Care: Pet Care Caregiver Search** (web, 12 screens, Job Search, Lifestyle) [Refero](https://refero.design/flows/12230) > Happy path for finding and narrowing pet care caregivers on Care.com. The user starts on the Pet care search results view with location, date, and basic search controls already set. They then refine the list using filter chips and dropdown panels (smart search, pay rate, pet type, rating, and professional skills), and the results list updates after each applied filter. The flow begins on the caregiver search results screen and ends on a filtered set of pet care profiles with a selected caregiver highlighted in the details panel, ready for contact. Steps: Results Overview → Smart Search Entry → Smart Search Applied → Pay Rate Filter → Pay Range Updated → Rate Filter Applied → Pet Type Selection → Pet Type Applied → Rating Filter → Rating Applied → Professional Skills Filter → Filtered Profile Review **time2book: Create and publish a plan** (web, 15 screens, Business, Fitness, Productivity) [Refero](https://refero.design/flows/6309) > Happy-path: A logged-in studio manager (desktop) opens Catalog > Plans, clicks New plan, completes the Add plan panel (general info, image, pricing type and amount, benefits, services with quantities), saves the plan, and returns to the Plans list where the new plan appears with a confirmation toast. Entry: Catalog > Plans (empty state). Exit: Plans list showing the created plan. Assumptions: user is authenticated with permission to manage catalog, there are existing services to attach to a plan, and file upload/image constraints are enforced by the interface. Steps: Plans list (empty state) → Add plan panel (General section) → Fill name, description and upload image → Pricing defaults (Membership selected) → Switch to single payment → Open validity dropdown → Enter price and choose validity → Add plan benefits → Open Services section → Service selector and quantity → Service list dropdown → Select first service → Add multiple services with quantities → Save plan and confirmation → Plans list (post-save) **Cohere: Business Plan Creation via Chat** (web, 6 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3791) > Logged-in user starts at the Cohere dashboard, navigates into the chat workspace and is shown an informational modal about multi-step tool use. They dismiss or accept the modal, explore prompt suggestion cards under "Just Chat," enter a custom prompt in the message composer asking for a business plan outline in French, submit it, and receive a formatted, multi-section French business plan in the conversation. Steps: Dashboard Landing → Onboarding Modal — Multi-Step Tools → Prompt Suggestions Panel → Compose and Submit Prompt → Generated Response — Part 1 → Generated Response — Continued / Completion **Tripadvisor: Personalized London trip planner** (web, 15 screens, Travel & Booking) [Refero](https://refero.design/flows/7068) > Happy-path: A user on the "Things to Do — London" listing triggers the site’s AI-powered trip planner and answers a short multi-step questionnaire (dates → trip type → interests). After submitting, the system generates personalized recommendations and presents them as categorized cards plus a map. The user can open item details in a slide-over, save selections, choose to create an itinerary, name it, and then view the newly created itinerary — a day-by-day timeline with saved items and an interactive map. Entry: user arrives at London "Things to Do" content (logged-in state not required by the UI shown). Exit: a named itinerary is created and displayed with saved items mapped to trip days. Steps: Start on destination recommendations → Select trip dates (calendar) → Confirm date range selection → Choose trip type and pet option → Select interests (chips) → Open 'Add other interests' modal → Enter custom interest text → Submit interest selections → Recommendations loading → View generated recommendations → Open item detail slide-over → Choose how to continue (save vs create itinerary) → Name your trip → View created itinerary (timeline + map) → Edit and extend itinerary ##### Task: case management **OpenAI: Prompt Evaluation Setup** (web, 13 screens, Business, Development Tools, AI Tool) [Refero](https://refero.design/flows/4223) > Happy path: A logged-in user opens the Evaluations area, creates a new evaluation by selecting a data source (Import Chat Completions / create data / upload file / custom), composes or imports test cases, selects a model to generate responses, adds one or more test criteria (e.g., valid JSON), reviews the evaluation configuration, runs it, and sees the run status and results saved in the evaluations list. Entry: Evaluations screen (empty or default). Exit: Saved evaluation entry and run result shown in the list. Assumptions: User is logged in to the project, has permission to run evaluations, has access to models referenced (e.g., gpt-4o-mini), and is on a desktop browser. Test data can be created manually or imported from stored completions or files. Steps: Evaluations list (empty) → Select data source → Import Chat Completions (instructions) → Create test data (blank form) → Create test data (sample content) → Create test criteria (empty) → Add testing criteria (chooser) → Configure 'valid JSON or XML' criterion → Criteria list updated → Review evaluation → Run in progress → Run completed (failed) → Evaluations list (with saved evaluation) **Cohere: Account registration and onboarding** (web, 15 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3790) > Happy-path desktop web flow for a new user who discovers the product marketing site, creates an account, completes an onboarding questionnaire, and arrives at the product dashboard. Steps: Marketing landing → Create account — entry form → Password strength — weak → Password strength — improving → Password strength — good → Password strength — strong → Verify your email → Onboarding — name collection (step 1/3) → Onboarding — name confirmed → Onboarding — role selection (step 2/3) → Role selected → Onboarding — use case selection (step 3/3) → Use cases selected with custom note → Dashboard — welcome → Dashboard — modules and docs ##### Task: code repository pull request **Blackbird: Referral Invite** (ios, 2 screens, Food & Drink) [Refero](https://refero.design/flows/9024) > A mobile referral/promotional modal that shows a scannable QR code, an explanatory headline and copy about the reward, the user's referral code, and two primary actions: Copy Code and Share Link. Steps: Referral Promo Display → Copy Confirmation **Ground News: Invite Friends - Referral Code** (ios, 2 screens, News) [Refero](https://refero.design/flows/7580) > Logged-in user lands on the Invite Friends screen (accessed from Profile tab). The screen displays an explanation of the referral promotion, the user's unique invite code in a bordered input, a Copy control, and a large Share Referral Code button. Primary actions: copy the code to clipboard or open sharing. Secondary content: promotion terms and a link to Terms & Conditions. Exit points: user copies the code, shares the code (system share sheet or similar), navigates back via the back arrow, or taps another bottom nav item. Steps: Invite Screen (Initial) → Copy Confirmation **Leonardo.Ai: Copy Generation API Code** (web, 4 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/12888) > Logged-in user on Leonardo.ai’s AI Creation screen generates or reviews an audio asset, opens the item actions menu, chooses the option to view the Generation API Code, and then copies that code from a modal. Steps: AI Creation Workspace → Actions Menu → Generation API Code Modal → Copied Confirmation **Raycast: Question search and select** (ios, 2 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/10193) > User is on a mobile "Questions" screen that shows a searchable list of question titles. The header shows a rounded search input with placeholder text and a numbered list of question rows; each row has a right chevron indicating it leads to details. Happy path: user taps the search field to open the keyboard, types a query, the list filters in real time to matching questions, then the user taps a row to navigate to that question's details (or taps the X control to dismiss). Steps: Questions list (initial) → Search with keyboard and filtered results ##### Task: compliance dashboard audit **Lovable: Generated Landing Page Review** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12309) > This flow shows a user reviewing an AI-generated SaaS landing page inside a split workspace. The user starts from a prompt describing the desired landing page, then inspects the generated result in a live preview on the right while the prompt/result card remains on the left. The user can bookmark the output to history and optionally give feedback via reaction controls. The flow appears to assume the user is already signed in or has access to the workspace and preview environment. It ends with the generated landing page visible and a confirmation message that the result was saved/bookmarked or feedback was submitted. Steps: Preview Generated Design → Bookmark Result → Submit Feedback **Foodllama: Nutrition and body stats review** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/11789) > Happy path starts on the main Foodllama dashboard, where the user can review daily calorie intake, macro distribution, streak status, and weekly trend data. From there, the user taps a profile/avatar area to open a detailed stats view with weight, calorie goal, and BMI information, then can move between avatar/profile variations and return to the dashboard. Steps: Dashboard Overview → Detailed Stats View → Avatar Selection State → Dashboard with Selected Avatar ##### Task: connect bank account open banking **Fingerprint: Find and open account-related documentation** (web, 4 screens, Business, Development Tools) [Refero](https://refero.design/flows/11143) > Happy path: a user lands on the Fingerprint Docs home screen, opens the site-wide search/Ask AI modal, searches for "account" (or a related term), and selects an "Account Sharing" / "Account Takeover" article. The flow ends with the full documentation article loaded, including code examples and step-by-step instructions. Steps: Docs Home / Discover → Search Modal (empty) → Search Results → Documentation Article ##### Task: contract management **Queue: Subscription Cards List** (ios, 3 screens, Entertainment, News) [Refero](https://refero.design/flows/6239) > A mobile (iPhone) list of subscription cards where each card represents a recurring item (podcast, news subscription, website). Steps: List — default state → Reveal destructive action → List — updated / scrolled view **Google Gemini: Report legal content from Gemini chat** (ios, 18 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/3187) > Happy path: A signed-in user starts in a Gemini conversation, opens the message actions menu on an assistant reply, and chooses to report the content. The app links to Google Legal Help. The user selects a legal reason, taps Create request, completes the online ‘Report a legal removal issue’ form (country, personal details, contact email), attaches screenshots/files from the device photo picker, provides a written explanation, confirms the perjury declaration, types a signature (digital), and submits. The flow ends with a submission confirmation screen telling the user they will receive a confirmation email. Assumptions: user is signed in (avatar visible), using a mobile browser/webview on iOS, has the required evidence files in their photo library, and can receive email confirmations. Steps: Gemini conversation - message actions visible → Message options menu → Legal Help landing → Select reason to report → Create request CTA → Form – country selector (tap) → Form – country list (choose) → Form – fill identity fields → Allegedly infringing content section → Photo picker – browse library → Photo picker – confirm selection → Files attached in form → Explain why content is unlawful → Perjury declaration checkbox → Signature and date → Submit request → Submission success → End state / wait for confirmation **Delphi: Expert Legal Consultation** (web, 19 screens, AI Tool) [Refero](https://refero.design/flows/13982) > Happy path begins in the Discover area, where a user searches for advice about arguing a first court case. The system returns matching expert profiles and excerpts. The user opens a conversation with Jason Jones, reviews his contextual response, chooses suggested follow-up questions, attaches a PDF, asks for feedback, records and sends a voice message, and continues the conversation based on the expert’s responses. The flow ends with another assistant response and additional suggested questions. The user is assumed to be signed in, using a desktop browser, and permitted to search experts, start conversations, upload files, and record audio. **Cohere: Document-based conversation review** (web, 3 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3832) > A desktop chat workspace where a user interacts with an AI assistant to extract or summarize information from uploaded documents. Steps: Conversation view with attachments visible → Activate collapse control for left panel → Expanded conversation (left panel collapsed) ##### Task: course enrolment university **WhatsApp: Add Community Member** (ios, 5 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12678) > Happy path in a community management area. The user starts on the Community info screen as the community owner, opens Add members, selects one contact, confirms the add, and returns to Community info where the member count updates and a success message appears. **ALO: Prenatal Program Detail** (web, 2 screens, Fitness, Lifestyle) [Refero](https://refero.design/flows/12377) > A logged-in user opens a wellness program detail screen for a prenatal yoga series. The entry point is the program landing page with a hero image, program title, instructor, and tabs for Overview and Classes. The user can review the description, preview the series video, inspect instructor information, browse the full list of classes, view similar programs, and add a comment. In the second state, the user is shown as in-progress with completion tracking and a Stop Program control, indicating they have already started the series. The flow ends when the user leaves the program page, starts or continues a class from the series, schedules a class, or posts a comment. Assumption: the user is signed in and has access to the program content on desktop web. Steps: Program Overview → Program In Progress ##### Task: credit score report **Lovable: Referral Invite Modal** (web, 2 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12350) > A centered referral modal opens over a dimmed dashboard/sidebar view. The user is shown a shareable invite link and QR code with instructions for earning credits through referrals. The intended entry point is a referral/share action from within the product, and the exit point is either closing the modal or copying the invite link. Steps: Referral Details → Link Copied State **Willow: Referral Card Sharing** (ios, 2 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/13241) > A logged-in user views a referral card screen for sharing an invite link or QR code. The entry point is likely from a rewards/referral section, and the exit points are back navigation, copying the referral link, downloading the card, or opening the system share sheet. The screen assumes the user already has an active account and referral code available. Steps: Referral Card View → Copy Confirmation **TIDAL: Artist Biography** (ios, 2 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/13136) > A music biography screen shows a long, scrollable artist article in a dark theme. The user enters from an artist or music profile context and can read the biography, follow linked artist/song names, and scroll through the full text. A back button returns to the previous screen. A persistent mini player stays docked at the bottom, allowing playback control while reading. Steps: Biography Reading → Biography Continued ##### Task: dispute transaction chargeback **PayPal: Open a transaction dispute** (web, 15 screens, Finance) [Refero](https://refero.design/flows/7409) > End-to-end happy-path for reporting a problem against a transaction from the activity/transaction detail screen. Steps: Transaction detail → Choose issue category → List items — empty → List items — filled → Pick issue — primary options → Pick issue — subtype selection → Provide other details — empty → Date picker → Provide other details — filled → Seller contact — asked → Seller contact — selection → Add documents — empty → Add documents — uploaded + notes → Review entered info → Confirmation / Case filed **Gumroad: Process sale refund** (web, 4 screens, Business, Marketplace) [Refero](https://refero.design/flows/9247) > Seller views the Sales list, opens an individual sale in a right-hand details panel, attempts to refund the purchase using a Refund control which asks for confirmation, and the system returns an error banner reporting insufficient balance. Steps: Sales list → Sale details panel → Refund confirmation modal → Insufficient balance error **Trip: Canceled Booking Refund Support** (web, 8 screens, Travel & Booking) [Refero](https://refero.design/flows/12086) > Happy path starts on a canceled hotel booking details screen for a logged-in Trip.com user. The user reviews cancellation and refund information, then opens Hotels & Homes Service Chat from the booking context to ask about refund timing and what to do if the money has not returned. The chat shows an automated response with refund details, the user follows up with an additional question, and the system returns more guidance about contacting the hotel if needed. The flow ends in the chat with support information visible; no booking changes are made. Steps: Canceled Booking Details → Service Chat Opens → Refund Question Sent → Refund Answer Returned → Follow-up Question → Follow-up Processing → Refund Escalation Guidance → Additional Deposit Advice ##### Task: doctor video consultation **District: Restaurant Question Assistant** (ios, 5 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12847) > Happy-path mobile assistant experience for asking a restaurant-related question and continuing with a follow-up. Entry is a restaurant detail or assistant overlay with an active query context (“Asking about Hotel Rishi”). The user asks a question by voice, the system transcribes and processes it, shows an answer with suggested next actions, and keeps a persistent follow-up input available. The flow ends when the user has a ready-to-send follow-up request, such as asking for nearby breakfast options. **Ada: Dizziness results and condition detail** (ios, 5 screens, Health & Wellness) [Refero](https://refero.design/flows/4103) > Happy-path: a mobile user who completed a symptom assessment for dizziness views the results list, opens the top suggested condition (Concussion), and reads detailed, structured information via expandable sections. Steps: Results — Dizziness Summary → Condition Preview — Concussion → Full Detail — Overview → Expand Section — Risks → Multiple Sections — Read Diagnosis & More **Ada: Condition Search and Results** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/4125) > User opens the Condition Library (on a mobile device, likely signed-in since profile-specific recommendations appear) to find information about health conditions. The happy path: user scans the curated lists (Most read / Mentioned in your latest assessments), taps the search field, types a query, and either selects a result or sees an empty-state message if there are no matches. Steps: Library listing → Search focused (keyboard visible) → Search submitted — no matches **Ada: Pain behind eye question** (ios, 2 screens, Health & Wellness) [Refero](https://refero.design/flows/4101) > Happy path: A mobile user opens a new assessment and is presented with a single symptom question: "Do you have pain behind the eye?" The user can either answer directly (Yes / No / I don't know) or tap the inline help link to view a full explanation (a modal/overlay) showing a short definition and an anatomical illustration. From the help overlay the user can return to the question using a prominent confirmation button or indicate they still don’t understand. Steps: Question: Do you have pain behind the eye? → Info overlay: Pain behind the eye **Ada: Neck mobility question and clarification** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/4102) > Happy path: A user working through a new medical assessment is asked a single symptom question about neck mobility. From the question screen they can either answer directly (Yes / No / I don't know) or open an inline help modal labeled “What does this mean?” The help modal provides a short definition and anatomical illustrations (front and side). The user indicates they understand the explanation and returns to the question to answer. Entry: user is mid-way through a newly started assessment. Exit: modal closed and question answered (moves to the next assessment step). Assumptions: user is on a mobile device (portrait), has already started a “New assessment” session, and can interact with modal overlays and standard controls. No special permissions or account state are required based on visible UI. Steps: Neck mobility question → Clarification modal — definition → Clarification modal — additional illustration **Ada: Condition lookup and article view** (ios, 5 screens, Health & Wellness) [Refero](https://refero.design/flows/4126) > Mobile Condition Library — happy path: user opens the Conditions tab (bottom navigation), lands on the Condition Library list and either browses the pre-populated sections (Most read / Mentioned in your latest assessments) or activates the search field to find a specific condition. Typing a query shows live search results; selecting a result opens a condition article with an overview and expandable sections (Risks, Symptoms, Diagnosis, Treatment, Prevention). Steps: Library landing → Search activated → Partial query / live results → Browse full results → Condition detail **Doctronic AI: Clinical Summary Feedback** (web, 4 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13701) > A user reviews an AI-generated clinical summary and provides product feedback. The flow begins on the summary view, where the user can rate how useful the summary was. The user then enters written feedback in an improvement form and submits it. After submission, the system confirms that the feedback was received. A chat input remains available throughout for asking additional questions. **Parallel: Research Chat with Sourced Answers** (web, 8 screens, AI Tool) [Refero](https://refero.design/flows/9073) > A desktop Chat API research interface where a logged-in user composes web-research questions, receives summarized answers with source links, and can change the chat model to prioritize speed or factual citations. Steps: Chat home / Prompt templates → Submit initial question → Compose follow-up → Follow-up visible while waiting → Open model selector → Submit a data-focused follow-up with new mode → Receive detailed, sourced results → Consolidated summary and source list **WhatsApp: Switch to Video Call** (ios, 4 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12665) > Happy path for an ongoing 1:1 call with Samuel on a mobile device. The user starts on an active voice call, chooses to switch to video, confirms the change in a modal, and then reaches the video call state. **WhatsApp: Active Call Controls** (ios, 2 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12663) > An active voice/video call screen with a contact named Samuel. The user is already in an ongoing call; the screen shows the call timer, a large contact avatar placeholder, quick actions on the right, and a bottom control bar. **WhatsApp: Video Message Editing** (ios, 2 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12646) > A mobile media composition screen for preparing a short video before sending it in a chat. The user enters from a selected video, can edit it with tools like trim/crop/rotate/text and AI restyle, optionally mute or toggle GIF/video mode, add a caption, choose a recipient, and then send. **ElevenLabs: Style selection & generate video** (ios, 2 screens, AI Tool) [Refero](https://refero.design/flows/10737) > Mobile screen where the user previews a text/voice card and chooses a background style before creating a video. Entry: user arrives with a prepared text/voice clip (voice shown as 'Rob') and a card preview. Primary controls: close (top-left), three background style swatches (Multicolor, White, Black), and a primary 'Generate video' CTA. Exit: user taps Generate video to initiate video creation. Assumption: user is already at the create/preview step (logged-in state not shown); device is a smartphone. Steps: Preview with Multicolor selected → Preview with White selected **WhatsApp: Send Recorded Video** (ios, 11 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12650) > A user is viewing an existing conversation with Samuel and sends a newly recorded video. The user opens the attachment menu, launches the camera, switches to video mode, grants microphone access when prompted, records a short clip, reviews it, and sends it to the conversation. The sent video then appears in the chat and can be opened for playback. The user is assumed to be on a mobile device with the conversation already open. Camera access appears available; microphone access requires approval. **WhatsApp: Group Voice Conversation** (ios, 6 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12680) > Happy path for joining and using a group voice conversation inside a community chat titled "General." The entry point is an existing chat thread where the user can read messages, add members, or start talking. The user begins by holding the talk control, then transitions into a live voice room connection where participants are shown and call controls appear. The exit point is the expanded in-call voice interface with connected participants and controls for speaker, mute, reactions, and ending the session. **WhatsApp: WhatsApp Call Transition** (ios, 9 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12661) > Happy path for a WhatsApp conversation that transitions from chat to audio call and then to video call. **Weekrise: Embedded support chat conversation** (web, 7 screens, Collaboration Tools, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/1296) > Happy-path interaction with an embedded live-chat widget on a desktop landing screen. Steps: Open chat / initial greeting → Compose and send question → Email capture prompt → Offline / slow-reply fallback → Record voice message → Upload in progress → Audio playback available **Anam: Live Conversation Session** (web, 4 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12945) > This flow begins in a live conversation experience, first showing a call startup/loading state, then an active video call with a remote speaker and a self-view thumbnail. During the call, the user can open device/audio controls, give feedback, or end the call. The flow ends on a post-call confirmation screen with options to build an interactive avatar or submit feedback. Steps: Call Starting → Active Call → Microphone Settings → Call Ended **Delphi: Digital Persona Voice Call** (web, 5 screens, AI Tool) [Refero](https://refero.design/flows/13986) > The flow begins in an existing conversation with a digital representation of Jim Goetz. The user initiates a voice call using the green call control, waits while the call connects, and then interacts with Jim’s digital mind by voice. The call displays mute and end controls, shows the conversation as it progresses, and ends on a call summary state with options to view the transcript, share the call, return to chat, or start another call. The user is assumed to be in an active conversation and authorized to call this digital persona on a desktop device. **Voicenotes: Voice query to notes** (web, 5 screens, Collaboration Tools, AI Tool) [Refero](https://refero.design/flows/6977) > User opens the 'Ask' modal to query their notes using a voice recording. The modal shows an Ask/Create toggle, a notes filter (All notes), a prompt area labeled "Ask a question" with a mic icon, and a prominent floating Record control. Happy path: user opens the Ask interface, starts recording, stops the recording (Done), the system transcribes the audio and displays the recorded audio/transcription as a message bubble, and then the assistant posts a textual reply referencing the user's notes. Steps: Open Ask modal → Start recording → Stop recording → Transcription / audio message appears → Assistant reply displayed **Synthesia.io: Generated Video Feedback** (web, 8 screens, AI Tool) [Refero](https://refero.design/flows/13748) > A user reviews a generated video inside a workspace and completes a sequential feedback questionnaire. The video remains visible in the main viewer while a dark survey card presents one question at a time in the lower-left corner. The user rates overall video quality, avatar realism, voice audio, graphics and visuals, and pacing and transitions using five-point scales, then enters one suggested improvement in a text field and submits it. The flow begins with the video open in the Videos area and ends after the improvement response is entered and submitted. Steps: Overall Video Quality → Avatar Realism → Voice Audio Quality → Graphics and Visuals → Pacing and Transitions → Improvement Suggestion → Review Written Feedback → Submit Feedback **Doctronic AI: Consultation Sharing** (web, 3 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13695) > A user is viewing an ongoing health consultation and opens the sharing dialog to share the current message or the entire consultation. The user keeps the default “Message Only” option, chooses “Copy Link,” and receives confirmation that the link was copied to the clipboard. The sharing dialog indicates that links expire after five days and can be regenerated. The user is assumed to be in an active consultation on a desktop-sized device. **Telegram: In-Chat Voice Call** (web, 4 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/571) > User is in a one-to-one conversation with a contact (Henry) on a desktop messaging client. From the open chat the user initiates a call using the header call control. A centered call card appears showing 'Ringing...' while the call is being placed and exposes inline controls (camera, screen, mute, decline). When the call connects a green active-call bar and a running timer appear; the call card updates to show the elapsed time and an End control. After ending the call the conversation is returned to and the chat thread (and left-hand conversation list) contains outgoing-call messages that show the call durations. Assumptions: user is signed in and in the specific one-to-one chat; device is desktop with enough screen width to show two-column layout (conversation list + chat). Entry: open chat with contact. Exit: chat updated with call records and user back to messaging view. Steps: Open Conversation → Ringing Call Card → Connected Call (Active) → Return to Chat with Call Record **Cushion: Direct Message Conversation** (web, 4 screens, Communication Tools, Collaboration Tools) [Refero](https://refero.design/flows/12528) > A workspace messaging interface where the user starts a new direct conversation, sends an initial message, then continues the thread with a reply and an attached image. Steps: New Chat Composer → Initial Message Draft → Conversation Thread → Reply With Attachment ##### Task: expense management receipts **Cashcaded: Receipt Sharing Preparation** (ios, 2 screens, Finance) [Refero](https://refero.design/flows/14136) > The user opens a Share sheet for a purchase record and can copy the displayed social-media tag before sharing the receipt image. The flow begins on a mobile device with a share modal already open over a purchase detail screen and ends with confirmation that the tag was copied. The user appears to be viewing an existing receipt or transaction; no login, permissions, or account state are visible. The final Share button remains available but is not shown as selected. Steps: Share Preview → Copy Confirmation **Understory: Manage categories** (web, 4 screens, Business, Scheduling Tools) [Refero](https://refero.design/flows/12710) > Logged-in desktop admin interface on the Experiences section. The user starts on an empty Manage categories screen, opens a modal to create a category, enters a category name, and saves it. After saving, the modal closes and the new category appears in the list with an experience count and overflow actions. Steps: Empty Categories State → New Category Modal → Category Name Entered → Categories List ##### Task: eye test vision **Google Translate: Lens Text Translation** (ios, 2 screens, Reference, Utility) [Refero](https://refero.design/flows/12142) > A user is viewing a document through Google Lens with text recognized and translated from English to Spanish. The screen shows the live camera/document view with translated text overlays, a language selector at the top, and a bottom sheet offering actions for the translated text. The user is assumed to be on a mobile device, already inside Google Lens, with camera access granted. The flow starts on the translated document view and ends after the user dismisses or responds to the feedback prompt. Steps: Translated Document View → Feedback Confirmation **Ada: Select primary symptom** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/4099) > Happy path: A user starts a new assessment and is prompted to enter the symptom that troubles them most. Tapping the input opens a modal-style symptom search with the keyboard. As the user types, autocomplete suggestions appear. The user selects a suggestion to view a symptom detail card, then confirms selection with a ‘Select symptom’ button. Steps: Start assessment — symptom prompt → Symptom search modal (empty) → Autocomplete suggestions → Symptom detail and confirm **Nextdoor: Rate Ask Answer** (web, 2 screens, Communication Tools) [Refero](https://refero.design/flows/11439) > A logged-in neighbor reviews an AI-generated answer in the Ask section and rates it as helpful. The user starts on an answer about places to learn hobbies, selects the thumbs-up feedback control, and receives a confirmation toast. The answer remains visible, and the selected feedback state is highlighted. Steps: Review Answer → Feedback Confirmed **Preply: Pre-lesson device check** (web, 11 screens, Education) [Refero](https://refero.design/flows/9213) > Happy-path: A learner on a scheduled lesson landing screen runs through an in-browser device check (camera, microphone, speakers, network) before entering the classroom. Steps: Lesson lobby with test prompt → Open audio & video test modal → Recording in progress → Visual confirmation prompt → Audio playback (speaker) confirmation → Test your devices overview → Network and device start screen → Network/device test in progress → Test results — success → Checklist with checks completed → Return to lesson lobby (ready state) **Preply: Pre-class device check** (web, 6 screens, Education) [Refero](https://refero.design/flows/9214) > Happy path: A learner arrives at a scheduled lesson landing screen and runs the built-in audio & video diagnostics before entering the classroom. From the landing area the user opens a modal that guides them to record a short (5-second) test video, watches a recording progress indicator, then confirms two simple checks: can they see themselves and can they hear the playback. The user can rerun the test and then view the consolidated "Test your devices" panel which shows device check items (Camera and mic, Network and device). Steps: Lesson landing with test prompt → Start test modal → Recording in progress → Playback – visual confirmation → Playback – audio confirmation → Device checklist overview **DailyArt Magazine: Image-based quiz with results** (web, 7 screens, Education, Magazines & Newspapers, Marketplace) [Refero](https://refero.design/flows/3501) > Happy-path: A desktop user navigates from a Quizzes listing page to a specific quiz article. The article shows a hero image, intro copy and a sequence of 10 image-based multiple-choice questions. Each question displays a large image, a short question label, a set of radio-button answers and a primary action button that advances to the next question. After the final question the primary action changes to reveal results. Steps: Quizzes listing → Quiz article — start & first question → Question 2 — answer selection → Mid-quiz questions → Late quiz question (8/10) → Final question and reveal action → Results / answers summary **Duolingo: Practice Word Question** (web, 8 screens, Education) [Refero](https://refero.design/flows/11827) > Happy-path practice sequence in the Duolingo English Test practice area. The user starts on the Practice dashboard, launches a practice test, waits through a loading state, answers a timed word-recognition question, receives either incorrect or correct feedback, advances through the question, and returns to the Practice dashboard with progress updated. Steps: Practice Dashboard → Loading State → Word Decision Question → Incorrect Feedback → Next Question → Correct Feedback → Practice Complete → Practice Dashboard Return **TikTok: Visual product search camera** (ios, 3 screens, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/2973) > Mobile visual-search entry point that lets a user take a photo or upload an existing image to find visually similar products. Entry: user opens the visual search camera view. Primary actions: take a photo (shutter), upload from photo library (Upload), or read contextual help (Help). Secondary controls include Flash toggle and a Close (X). Help appears as a two-page bottom sheet with short guidance for taking or uploading images. Exit: after capturing or selecting an image the flow should proceed to a results/search step (not shown). Assumptions: device has camera/photo permissions, product is shown centered on a plain background, user is on a mobile device. Steps: Camera / Visual Search → Help — Take a photo → Help — Upload a photo **Chance AI: Visual Product Lookup** (ios, 7 screens, AI Tool) [Refero](https://refero.design/flows/7365) > Happy-path: A user discovers the "Search by Seeing" feature on the app home feed, opens the in-app camera experience (AI Snaps), selects a product recognizer snap, points the camera at a product, captures an image, and is taken to a generated result page containing a review, photo gallery, price cards, expert commentary and suggested follow-up questions. Entry: Home feed. Exit: Detailed result/article card with share and follow-up options. Assumptions: User is on a mobile device, has camera access, and has network connectivity. No special permissions or account state are visible/required in the screenshots. Steps: Home — Search by Seeing → Camera — AI Snap Selector → Camera — Capture → Results — Article Header → Results — Visual Gallery & Price Cards → Results — Expert Review Content → Results — Actions & Suggested Questions **Sky Guide: Tonight's Best — Observation Mode Selector** (ios, 3 screens, Education, Reference) [Refero](https://refero.design/flows/8478) > Happy-path: User opens the "Tonight's Best" list (mobile, portrait) and sees a scrollable list of recommended celestial objects. Each row shows an icon, object name, object type (star, planet, etc.) and visibility window. A prominent center pill at the bottom shows the current observation mode (e.g., "Naked Eye"). Tapping that pill opens a small modal with three instrument options (Telescope, Binoculars, Naked Eye). Choosing an option updates the center pill and refreshes the list to show objects appropriate for that instrument. Entry: user already in the app and viewing "Tonight's Best." Exit: user either scrolls the updated list, taps the back arrow to return to the previous screen, or reopens the instrument picker to change mode. Assumptions: mobile device in portrait; user has permission to view location-based visibility times; the list content is filtered client-side and updates immediately when mode changes. Steps: Tonight's Best List (Naked Eye) → Observation Mode Picker → Tonight's Best List (Telescope) **Podimo: Search Episodes in Show** (ios, 4 screens, Entertainment) [Refero](https://refero.design/flows/12262) > Mobile show episode list with an in-show search experience. Entry is from a show’s episode list, where the user taps the search icon to open a dedicated search screen for that show. The user types a query in the episode-title field and the list filters in real time. Results count updates as matches change, and an empty state appears when no episodes match. Steps: Episode List → Search Entry → No Results → Filtered Results **Chance AI: Visual Snap Interpreter** (ios, 8 screens, AI Tool) [Refero](https://refero.design/flows/7364) > Happy-path: A signed-in mobile user with camera and photo-library permissions starts on the app home / discovery screen, taps the camera/Explore action, selects an AI Snap (Destiny Mirror) from the filter carousel, either takes a live photo or picks one from the photo library, lets the app analyze the image (face/keypoint detection), and receives a generated interpretive report ("Today's card: The Silver Seraph") with expandable sections, suggested follow-up questions, and share/export actions. Steps: Home / Discover → Live Camera with Snap Carousel → Photo Library Picker → Image Loading & Face Keypoint Overlay → Preview with Snap Applied → Initial Result / Contextual Prompt → Full Report with Sections → Export & Follow-up Interaction **MyMind: Visual Reference Search** (web, 3 screens, Productivity, Reference, Inspiration Board) [Refero](https://refero.design/flows/13313) > A user starts on a personal visual search dashboard with content grouped across images, books, web pages, and videos. They enter a search term, first producing an empty state for "00000," then replace or refine the query with "or" to view matching content. The flow assumes the user is already in the search workspace and can access indexed personal or curated references. It ends on a results view showing tags and a masonry-style collection of books, artwork, design references, and video content. Steps: Search Dashboard → Empty Search Results → Browse Matching Results **Dropbox: Video Review and Viewing Controls** (web, 4 screens, Collaboration Tools, Business, Productivity) [Refero](https://refero.design/flows/13044) > A user is viewing a video in a file preview/player and uses the top toolbar and player controls to inspect the media. The entry point is a paused video in the standard viewer, with options like Open in, Edit, Transcript, Share, comments, and tools. The user can open the tools menu to show comments, show transcript, hide tools, or switch to full screen. The main interaction shown is toggling the viewer chrome so the video can be watched with fewer distractions. The flow ends with the video remaining ready to play, either in the compact viewer or full-screen mode, with tools hidden or shown as needed. **Runway: Video feedback submission** (web, 3 screens, AI Tool) [Refero](https://refero.design/flows/1368) > Happy-path: The user reviews a freshly generated short video in a preview player, is prompted to rate the result, and chooses to send structured feedback via a modal. Steps: Preview and rating prompt → Feedback modal — choose issues → Feedback modal — submit **Melius: Skincare Hero Image Generation** (web, 3 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/14070) > The user reviews an AI-generated commercial skincare product image inside a canvas workspace. The flow starts with a conversational prompt referencing an attached product image, continues through a source image node and generated result, and ends with the user reviewing the output and optionally rating the response. The user appears to be signed in and working in an existing canvas on a desktop device. No additional permissions or account setup are shown. Steps: Prompt and Reference → Generation Ready → Result Feedback **Dropbox: Video Playback Review** (web, 4 screens, Collaboration Tools, Business, Productivity) [Refero](https://refero.design/flows/13043) > A user is viewing a video in a file/media preview screen and changes playback speed while the clip continues playing. The entry point is an opened video file in a viewer with standard controls. The user remains in the same viewer throughout, using the speed control and timeline to review the content. **Apple Maps: Place Details and Street View** (web, 3 screens, Map & Navigation) [Refero](https://refero.design/flows/12472) > A user opens a specific place listing in Maps, reviews its details in a side panel, and then switches to Street View to inspect the venue’s exterior and surrounding street context. The entry point is a selected place marker or search result on the map; the exit point is the place shown in Street View with the option to close or continue exploring. Steps: Place Details Opened → Street View Preview → Return to Map Context **Rork: Analytics Review and Date Range Change** (web, 4 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12160) > Happy path for viewing analytics in a dark-mode web dashboard. The user is already signed in and on the Analytics section of the workspace. They review activity metrics, change the date range using a dropdown, briefly see a loading state while the data refreshes, and then return to the updated analytics view. The screen also shows a revenue integration card with a connect action, but no connection is completed in this sequence. Steps: Analytics Overview → Date Range Menu → Loading State → Updated Analytics View **Fingerprint: Inspect Event JSON Payload** (web, 3 screens, Business, Development Tools) [Refero](https://refero.design/flows/11153) > Happy path: A logged-in user (desktop) viewing an event record opens a modal that displays the raw JSON payload. The JSON viewer supports switching between two versions (v3/v4), searching within the payload, showing result counts, and navigating matches. The user can copy/download the payload or close the modal. Steps: Open JSON modal → Search for specific ID (no results) → Search for boolean value (matches found) **Rork: Model Catalog Search** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12165) > A logged-in user is on the AI Cloud section of a project dashboard and views a catalog of available models with pricing, context, and capability indicators. The user can browse by category tabs or search the model list. In the happy path, the user types a query that filters the list and then clears or changes the query to view matching models again. The flow starts on the model catalog screen and ends with the filtered results displayed or with the full list restored after search refinement. Steps: Model Catalog → No Results State → Filtered Results ##### Task: health onboarding profile **GoFundMe: Profile Sharing** (web, 3 screens, Fundraising) [Refero](https://refero.design/flows/11725) > Logged-in user profile page on GoFundMe with a share-profile modal opened from the profile screen. Entry is the profile page; exit is either closing the modal or copying the profile link. Steps: Profile Overview → Share Profile Modal → Copied Confirmation **Ada: Profile setup wizard** (ios, 12 screens, Health & Wellness) [Refero](https://refero.design/flows/4096) > Happy-path onboarding to add a user profile before starting a symptom assessment. Steps: Home — Start assessment → Add profile — Intro / safety tip → Ask for name (empty) → Ask for name (typing) → Sex at birth selection → Gender explanation modal → Date of birth prompt (picker shown) → Date of birth selected → Smoking status → High blood pressure question → Diabetes question → Completion / Start assessment **Ada: Profile health editor** (ios, 17 screens, Health & Wellness) [Refero](https://refero.design/flows/4111) > Happy path: A signed-in mobile user opens the Health profiles list, selects a profile (Tony), and navigates through profile sections to view and edit Basic information, Personal information (height/weight), Health background (yes/not answered/no), Medication (add/edit/delete), and Allergies (add/delete). Steps: Health profiles list → Profile sections menu → Basic information → Personal information (entry) → Personal information (data entry) → Health background → Add medication (empty) → Add medication (typing) → Medication list updated → Medication item actions → Add allergy (empty) → Add allergy (typing) → Allergies list updated → Add multiple allergies → Allergy item actions → Final allergies list → Return to profile sections **Ada: Delete health profile** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/4124) > Happy path: a signed-in user opens the Profiles tab, selects a secondary health profile (Sophie), chooses the Delete health profile action from that profile's settings screen, confirms deletion in a modal, and is returned to the Profiles list where the deleted profile no longer appears. Steps: Profiles list → Profile details → Delete confirmation → Updated profiles list **Doctronic AI: Healthcare Subscription Signup** (web, 4 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13691) > A four-step subscription signup modal appears over a medical blog article. The user starts by entering an email address, proceeds to provide a phone number for health updates, and submits the registration. The flow ends with a confirmation message. The user is assumed to be browsing the article and not necessarily logged in. The modal can be dismissed at any point using the close icon. **ALO: Wellness Club Registration and Personalization** (web, 21 screens, Fitness, Lifestyle) [Refero](https://refero.design/flows/12365) > Happy path from the public ALO Wellness Club marketing experience to a newly created member account and personalized content dashboard. The user starts logged out, reviews the wellness offering, selects Join Now, enters account details, accepts the required terms, and creates an account. After signup, the user can personalize recommendations through six questions covering goals, schedule, styles, fitness level, equipment, age, and gender preferences. The user then reaches a logged-in dashboard showing account status, progress metrics, library shortcuts, and content recommendations. The final dashboard may include a short guided walkthrough with tooltips. Steps: Membership Landing → Account Creation → Completed Account Form → Personalization Welcome → Personalization Choice → Goals Questionnaire → Goals Selected → Schedule Questionnaire → Schedule Selected → Style Preferences → Fitness Level → Fitness Level Selected → Equipment Questionnaire → Equipment Selected → Personal Details → Personal Details Selected → Member Dashboard Welcome → Progress Walkthrough → Personal Picks Walkthrough → Personalized Content Dashboard **Napper: Confirm personal details** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/8766) > Happy-path review step in an onboarding/onboard-like flow where the user verifies and fills missing personal details before proceeding. Entry: user reaches a summary screen titled "Is this correct?" with several fields (Name, Birthday, Gender, Nursing, Height, Weight). Exit: user either updates one or more fields using inline edit or add controls, confirms a selection via a modal/bottom-sheet picker, and returns to the summary with the field(s) updated — then may continue by tapping the Next CTA. Assumptions: user is already partway through onboarding (progress indicator visible) and is on a mobile device; the UI shows editable and addable fields and uses a bottom-sheet picker for selection. Steps: Summary / Confirm screen → Open gender picker (bottom sheet) → Select option and confirm → Summary updated with chosen value **Athleats: First-time sign-up and onboarding** (ios, 23 screens, Health & Wellness, Food & Drink) [Refero](https://refero.design/flows/11281) > Happy-path onboarding for a first-time iOS user who signs up using Sign in with Apple, completes a short profile (photo + username), answers a multi-step preference questionnaire, is presented with a Pro subscription upsell, completes payment via Apple Pay, and lands on the authenticated home screen. Steps: Marketing / Welcome → Sign in with Apple (permissions) → Sign in with Apple (loading) → Sign in complete → Sign Up — Basic profile → Add profile picture options → Photo Library picker → Photo selection applied → Username validation (empty) → Enter username → Sign Up — Continue enabled → Onboarding — Current goal → Onboarding — Dietary preferences → Onboarding — Referral source → Pro upsell modal — feature comparison → Apple subscription confirmation → Purchase success confirmation → Finalizing onboarding → Authenticated Home — Welcome → First-time user landing state **Verse: New user onboarding and profile creation** (ios, 33 screens, Social Networking, Design & Media Editing) [Refero](https://refero.design/flows/3955) > Happy-path new-user onboarding that guides a user from launch -> phone verification -> username claim -> short preference quiz -> automated profile assembly -> final profile reveal. Steps: Splash / Get started → iOS tracking permission → Phone number entry (empty) → Phone number entry (typing) → Enter 6-digit code (empty) → Code input attempt (invalid) → Code entry success (loading) → Claim username (empty) → Username validation error → Select star sign (carousel) → Choose aesthetic (grid) → Select favorite movie/TV genre → Select current hobby interests → Processing: analyzing your vibe → Profile reveal (confetti) → First-time profile (overview) → Edit profile CTA and placeholders → Bottom navigation visible → Profile content populated from choices → Onboarding complete **Exoplan: Health & Calendar Onboarding** (ios, 10 screens, Health & Wellness, Scheduling Tools) [Refero](https://refero.design/flows/7342) > This is an onboarding sequence that asks the user to grant two platform-level permissions (Health data & Calendar access) so the product can provide personalized focus scheduling and health-based insights. Steps: Intro: Health Insights → Health Permission Sheet (off) → Health Permission Sheet (on) → Intro: Connect Calendar → Calendar Access System Modal → Onboarding Transition / Loading → Intro: Notifications → Health Data Missing Message → Main Calendar (empty) → Final App Launch / Splash **Miles: First-time onboarding and Health permission setup** (ios, 5 screens, Health & Wellness) [Refero](https://refero.design/flows/6704) > Happy path for first-time user setup on iOS. Entry: user launches the Miles app for the first time. The app shows a sequence of stacked modals guiding the user to start and optionally connect to Apple Health, then a short interactive tutorial, and finally reveals the main tracker screen in its initial (empty) state. Steps: Welcome modal → Apple Health sync prompt → iOS Health permission sheet → Quick tutorial card → Main tracker (initial state) **District: Edit Profile Settings** (ios, 5 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12811) > A logged-in user edits profile information on a mobile screen, including contact details, birthday, optional personal details, and invoice billing state. The user starts on the Edit Profile screen, makes changes in text fields and date pickers, then submits the form with Update profile. The flow assumes the user already has an account and access to this screen; phone number appears read-only while other fields are editable or selectable. Steps: Profile Form → Birthday Date Picker → Month and Year Selection → Date Confirmation → Updated Profile State **HYPE: Profile Completion and Newsletter Selection** (ios, 7 screens, News, Lifestyle) [Refero](https://refero.design/flows/12244) > Happy path starts on the Hypebeast account profile completion form in a mobile browser and ends on the newsletter preferences screen after a successful profile update. Steps: Profile Details Entry → Personal Information Filled → Gender Menu Open → Gender Selected → Display Name Validation → Profile Submitted → Newsletter Preferences **The Outsiders: First Launch Setup** (ios, 12 screens, Health & Wellness) [Refero](https://refero.design/flows/10292) > This is a linear, first-launch onboarding sequence for a health-focused mobile product running on iOS. Steps: Hero / Branding → Brand Definition Slide → Name Input (empty) → Name Input (typing) → Privacy Reassurance → Connect to Health Prompt → Health Access Modal (initial) → Health Access Modal (confirm) → Reminders / Notifications Offer → Notification Permission Alert → Final Onboarding Slide → Main Dashboard (Today) **Airbnb: Profile Prompts** (ios, 4 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6410) > Happy path: A signed-in user is asked to answer at least one short personal prompt to enrich their profile. The entry point shows a horizontal, scrollable set of prompt cards with an overall instruction, an answered-count (0/10), and disabled Next. The user taps a card to open it full‑screen, types a response in a short text field (40 character limit), closes the keyboard, and returns to the card list where the answered card shows a preview and a checkmark. The answered-count increments (1/10) and the Next button becomes active so the user can continue. Assumptions: user is authenticated and has permission to edit their profile; device is a mobile phone; text entry uses the system keyboard; responses are locally validated for length before being saved. Steps: Prompts selection → Open prompt detail → Enter response → Confirm answer and enable next **Haptic: First-run setup and permissions** (ios, 16 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/2100) > Happy-path onboarding for a habit-tracking app (Haptic). Entry: first-run welcome carousel. User optionally skips or progresses through 3 steps: (1) choose habit categories to track, (2) enable Health access to import activity, (3) start adding events. The flow includes a Pro upsell if user attempts to add more than the free limit, an App Store subscription purchase sheet, an access/permission step for the iOS Health app, and a success confirmation. Exit: main timeline screen with the app installed and ready to add events. Steps: Welcome screen → Onboarding carousel — Step 1 card → Category selector (initial view) → Category selector with Pro upsell banner → Pro subscription modal → App Store purchase sheet (confirm) → Purchase success animation / confirmation → Category selector — selections persisted → Onboarding carousel — Step 2 card (Health) → iOS Health access sheet (initial) → iOS Health access sheet (allowed state) → Onboarding carousel — Step 3 card → Main timeline (post-onboarding) → Category selection — deep scroll state → Category selection — final review → Health toggles final state & return **Napper: Confirm profile details** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/8767) > Happy path: user reviews a pre-filled child profile during onboarding and adds missing physical measurements (height, weight, nursing) before proceeding. Steps: Review profile summary → Open height input modal → Enter numeric height → Save and return to summary **Doctronic AI: Email Signup and Verification** (web, 11 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13692) > The flow begins on Doctronic’s public landing screen, where a new visitor selects Sign up. A signup modal offers email or social registration; the shown path uses email. The user enters an email address, requests a verification code, retrieves the code from email, and confirms it. After the account is created, the user can optionally provide a name. The flow ends in the authenticated Doctronic experience, where the user is greeted by name and prompted to describe their health question. The user is assumed to be logged out and using a desktop browser with access to the submitted email inbox. **Peerlist: Account creation and initial profile setup** (web, 16 screens, Job Search, Business, Social Networking) [Refero](https://refero.design/flows/1081) > Happy-path new-user onboarding: an anonymous visitor lands on Peerlist's marketing entry and chooses to join. The flow proceeds: sign up (email + password or Google), email verification, initial profile creation via a modal that can autofill from LinkedIn, photo upload/crop, basic personal details and bio, add skills, then land on the user's dashboard/profile with initial content blocks and a profile completion widget. Entry: landing hero (not-logged-in). Exit: user is logged in and sees their profile/dashboard with ~30% completion. Assumption: user is not logged in and has access to email used during signup. Steps: Landing / Claim username → Sign up (empty form) → Sign up (credentials entered) → Email verification prompt → Invalid code feedback → Valid code entered → Welcome modal — LinkedIn autofill start → Username suggestion / conflict messaging → Username validated → Profile details form → Photo cropper (open) → Photo cropper (adjust) → Profile form populated → Add skills modal (suggestions) → Skills selected and confirmed → Profile dashboard (initial state) **Kinhive: Onboarding & First-Time Setup** (web, 16 screens, Communication Tools) [Refero](https://refero.design/flows/6813) > End-to-end new user onboarding for Kinhive starting from the marketing landing to a logged-in dashboard. Entry: visitor on the marketing/home landing (email CTA). Assumption: user is not signed in. Primary path: user enters email → completes account creation (choose Member or Founder, email, password) → completes step-by-step profile onboarding (name, location, optional LinkedIn import, skill selection, photo + bio, social opt-in) → automatic account setup → in-app first-time prompts (create project) → lands on home dashboard. Exit: user is authenticated and sees the main dashboard with onboarding to-dos and project prompt. Steps: Marketing landing / email CTA → Create account (blank) → Create account (founder selected) → Basics: name & location (empty) → Basics: name & location (filled) → Optional LinkedIn import → Skill selection (choices) → Skill selection (selected state) → Profile photo & bio entry → Photo crop / adjust modal (open) → Photo crop / finalize (confirmation) → Profile photo & bio (completed) → Social feature opt-in → Account setup (progress) → Create project modal (founder prompt) → First-time dashboard / Home **Parallel: Personalized job onboarding** (web, 18 screens, Job Search) [Refero](https://refero.design/flows/1036) > Happy path: A new visitor on the homepage clicks a prominent "Start Matching" CTA, triggers a sign-up modal and begins a 5-step onboarding sequence that collects interests, values, job-search status, profile details (including photo and resume), and account credentials. After completing the steps the user sees a welcome modal and is routed to their personalized dashboard which shows profile completeness and curated content. Steps: Homepage CTA → Sign-up modal → Interests - question → Interests - selected state → Values - question → Values - selected state → Job search status - question → Job search status - selected → Profile details form → Upload photo modal (crop) → Upload photo modal (adjust) → Photo upload progress → Profile form with uploads → Account credentials - empty → Account credentials - filled → Account creation processing → Welcome modal → Authenticated dashboard (profile completeness) **ElevenMusic: Artist Profile Setup** (web, 2 screens, Music Streaming, Entertainment, AI Tool) [Refero](https://refero.design/flows/12555) > A centered onboarding modal guides a user through becoming an artist. The entry point is an in-context prompt shown over a blurred background. The user can enter an artist name, optionally add a profile photo, navigate back, skip the setup, or continue to the next step. The exit point is either proceeding to the next onboarding step via Continue or dismissing/skipping the setup from this modal. Steps: Artist Name Entry → Profile Ready to Continue **Marathon: Basic profile setup** (web, 2 screens, Business, Productivity, Project Management, Collaboration Tools, Scheduling Tools) [Refero](https://refero.design/flows/4841) > Happy path: a new-user onboarding step that collects basic profile details. Steps: Profile setup — empty state → Profile setup — populated state **Make: Signup, onboarding, and dashboard setup** (web, 17 screens, Business, Productivity) [Refero](https://refero.design/flows/10956) > Happy path: A new user arrives on the marketing landing page, starts account creation, verifies their email, signs in, completes a multi-step onboarding questionnaire (discovery, experience, skillset, company size, AI intent, tech stack) and is then routed into the initialized organization dashboard. Steps: Marketing landing → Sign up (initial) → Sign up (form filled) → Email verification notice → Sign in (post-verification) → Sign in (credentials entered) → Onboarding - Discovery → Onboarding - (duplicate view) → Onboarding - Experience → Onboarding - Skillset → Onboarding - Company size → Onboarding - AI intent → Onboarding - Tech stack selection → Tech stack - selections visible → Workspace initializing (partial load) → Loading spinner → Organization dashboard (ready) **Homerun: New user trial signup and onboarding** (web, 24 screens, Productivity, Job Search, Communication Tools) [Refero](https://refero.design/flows/1263) > This is a happy-path new user sign-up and first-time onboarding sequence for Homerun. The user starts at the marketing home hero, chooses which product to try (ATS or HR), completes a multi-step sign-up form (social sign-in or manual name/email), provides company and contact details, verifies their email, answers a short onboarding survey (how they heard about the product, company size, current hiring setup, timeline, roles), and lands in the product where an in-app guided onboarding modal collects learning preferences. Steps: Marketing hero - product selection → Choose product to start → Sign-up form - entry → Sign-up with prefilled fields → Company details collection → Company details - filled → Email verification prompt → Onboarding - acquisition channel question → Onboarding - acquisition selection confirmation → Onboarding - company size → Onboarding - current hiring tools → Onboarding - hiring timeline → Onboarding - timeline confirmation → Onboarding - roles hiring for → In-product welcome modal → Welcome modal - selection → Guidance preference question → Guidance preference - selected → Onboarding confirmation message → Authenticated dashboard - first run state **ElevenReader: Author profile creation** (web, 2 screens, Books, AI Tool) [Refero](https://refero.design/flows/9196) > This is the "Author profile" step of a multi-step publishing/onboarding sequence. The screen lets an authenticated user create or edit their author profile by uploading an author image and entering basic metadata (display name, language, website, short bio). Steps: Author profile — empty image placeholder → Author profile — image uploaded **TwelveLabs: Signup to Upload Onboarding** (web, 17 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/5560) > Happy-path new user journey from marketing landing to an authenticated product dashboard. Steps: Marketing hero / entry → Create account — enter email → Email populated / continue → Set password — initial → Password validation feedback → Password meets requirements → Verify your email prompt → Email verified confirmation → Profile collection — initial → Profile — reason dropdown → Profile — role dropdown → Profile — industry dropdown → Profile — form completed → Onboarding modal — Marengo intro → Onboarding modal — Pegasus features → Onboarding modal — Responsible AI → Authenticated dashboard — Upload videos **Сontra: Freelancer profile management** (web, 3 screens, Job Search) [Refero](https://refero.design/flows/4876) > Logged-in freelancer (desktop) viewing and managing their public profile. Steps: Profile overview (initial) → Profile state update (rating shown) → Final profile state (confirm) **[untitled]: Account onboarding and profile setup** (web, 21 screens, Music Streaming) [Refero](https://refero.design/flows/5931) > Happy path: a first-time user lands on the product homepage, chooses a sign-in method (Google/Apple/phone), completes phone verification, selects an avatar and username, optionally adds a profile photo, steps through a short feature tour, optionally adds a recovery email, and arrives at an empty-project dashboard where they can add audio. Steps: Marketing Hero / Entry → Sign-in method selection → Phone number entry (empty) → Phone number entry (filled) → Verification code prompt → Code entry / signing in → Avatar picker — intro → Avatar picker — selection (blue) → Avatar picker — selection (pink) → Username entry (empty) → Username entry (filled) → Add profile image (placeholder) → Add profile image (uploaded) → Feature tour — page 1 → Feature tour — page 2 → Security & finish prompt → Recovery email modal (empty) → Recovery email modal (filled) → Check your email confirmation → Empty-project dashboard ##### Task: health record medical history **Anam: Session History Export** (web, 2 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12949) > This is an authenticated analytics/admin interface on the Sessions section, with the History tab selected. The user can review a paginated table of recorded sessions, filter or search the list, and export the sessions as CSV. The flow starts on the Sessions history screen and ends after the user downloads the session list, confirmed by a toast notification. Steps: Review Sessions → Export Sessions **Pi: Export personal chat history** (web, 2 screens, AI Tool) [Refero](https://refero.design/flows/4708) > Happy path: a logged-in user navigates to Profile > Manage history on a desktop browser, sees an explanation about portability and a primary CTA to download their personal conversation (PI) history, taps the CTA to start export, and the UI updates to a "Retrieval initiated" state showing that the export is being prepared and an estimated ready date. Steps: Manage history — Download prompt → Retrieval initiated — In-progress status **Ada: Manage symptom tracking history** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/4108) > User is on the mobile Symptom Tracking timeline viewing entries grouped by date. From this list the user opens an action menu for a specific symptom and chooses to delete the tracking history for that symptom. The UI presents a bottom sheet with actions, then a confirmation modal. If the user confirms deletion, the symptom entries are removed and the screen shows an empty state with instructions to start tracking again. Assumptions: mobile (iOS) device, user is on the Symptom Tracking screen and has permission to manage their own tracked entries. Steps: Symptom timeline (list view) → Action sheet (bottom options) → Delete confirmation modal → Empty state after deletion **Doctronic AI: Summary Feedback** (web, 2 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13698) > A user has reached the end of a medical assessment or consultation summary. The summary card remains visible with options to continue, review the Assessment and Plan, or open the SOAP Note. Below it, the user can rate how useful the summary was. After submitting a feedback rating, the interface replaces the rating controls with a confirmation message and an option to leave a Trustpilot review. The flow starts on the completed summary screen and ends after feedback confirmation, with an optional external review action. The user is assumed to be viewing the summary on a desktop browser and has already completed the preceding consultation or assessment. **Parallel: Chat API Research Query** (web, 2 screens, AI Tool) [Refero](https://refero.design/flows/9074) > Happy-path: A signed-in desktop user runs a research query in the Chat API interface, reviews a synthesized answer with numbered source citations, and then clears the conversation to start a new query. Steps: Research answer with sources → Cleared chat / start new query **Ada: Add medication to profile** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/4120) > Happy path: A logged-in mobile user adds a medication to their profile. Entry is from the 'Your medication' area (plus button) which opens an 'Add a new drug' screen. The user types the drug name, optionally toggles 'I am taking this permanently', and taps Save. After saving the app returns to the medication list where the new drug appears. The user can then go back to the main profile menu. Steps: Add a new drug (empty state) → Enter name & enable save → Medication list (confirmation) → Profile / Settings return **Ada: Browse and read medical article** (ios, 6 screens, Health & Wellness) [Refero](https://refero.design/flows/4109) > Happy path: a user opens an Articles list, scans article cards, selects an item and reads the full article in an in-app web view. Steps: Articles list → Article header (hero) → Intro and section header → Content sections (symptoms) → Further sections & tips → References, writer and reviewer **Comet: History search and filter** (ios, 4 screens, AI Tool, Utility) [Refero](https://refero.design/flows/10941) > User is viewing the History section (bottom navigation selected) to find a previously visited item. Entry: user opens History from the bottom nav. In the happy path the user types into the search field to filter items and selects a matching history entry. Exit points: selecting a history item (to open it), tapping the close button (X) to dismiss History, or clearing the search to return to the full list. Assumptions: user is on a mobile device (iPhone iOS keyboard visible), has at least one history item, and the history list and search operate locally in the client UI. Steps: History list open → Search field focused → Search with no matches → Filtered results shown **FocusFlight: History scenario filter** (ios, 4 screens, Productivity) [Refero](https://refero.design/flows/8610) > A mobile "History" sheet that lists past items (presented like ticket cards) with three top-level filter controls: Status, Scenario, and Sort By. The user opens the Scenario control (anchored popover) to select one or more scenario tags from pill-shaped chips. The popover includes a "Clear All" action and shows chips in unselected/selected states. Selecting chips updates the Scenario control to show a count (e.g., "2 scenarios") and filters the list below. Steps: History list (default) → Open Scenario popover → Select scenario chips → Apply filters and return to list **Ada: Manage allergies** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/4122) > Happy path: a logged-in user (profile labelled "Tony") opens their Allergies section, chooses to add a new allergy, types the allergen name, saves it, and returns to the allergies list which shows the newly added item. Steps: Profile — Your information → Add new allergy — empty input → Add new allergy — enter text → Your allergies — item added **Granola: Search Chat History** (ios, 3 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14161) > The user starts in an open conversation and opens the Chat History panel. They select the Search field, enter a query, and submit it using the keyboard search action. The history list is then replaced with an empty-state message because no chats match the entered value. The flow begins with an existing conversation visible and ends in the Chat History search results state. The user is assumed to be on a mobile device with access to stored chat history; authentication and permissions are not shown. Steps: Open Chat History → Enter Search Query → View Search Outcome **Foodllama: Weight Progress Overview** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/11794) > A weight-tracking screen that shows the user’s predicted goal date and current weight trend. The flow starts on a dashboard-style summary card and allows the user to change the chart time range using a dropdown. Steps: Progress Summary → Time Range Menu → Updated Trend View **Delphi: Add User Memory** (web, 3 screens, AI Tool) [Refero](https://refero.design/flows/13995) > A signed-in user opens a participant details panel for Samson Ware and adds personal context for Delphi to remember. The panel begins with an empty user-information prompt and an Add control. After the first item is added, it appears as a saved row and a follow-up prompt becomes available for another note. The flow ends with the first note saved and the second prompt ready for input. The user appears to have permission to manage this participant’s information; no device-specific behavior is visible. **fal: Search Recent Endpoints** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/13544) > The user is reviewing recent API requests and wants to filter the history by endpoint. From the Recent History view, the user opens the endpoint selector, searches for an endpoint identifier, receives a no-results message for an unmatched query, then enters "gpt" to see matching endpoints. The flow ends with matching endpoint options displayed; no endpoint selection is shown. The user is assumed to be viewing the history screen with access to recent requests. Steps: Open Endpoint Filter → Search Without Match → Find GPT Endpoints **Doctronic AI: Health Record File Deletion** (web, 3 screens, Health & Wellness, AI Tool) [Refero](https://refero.design/flows/13763) > The user is viewing Connie’s Health Record with no connected providers and three uploaded files. They select the delete control for “I parked my car.pdf,” review a confirmation dialog, and confirm deletion. The file is removed from the list and a success notification appears. The user is assumed to be signed in and authorized to manage their health record. Steps: Health Record Files → Delete Confirmation → Deletion Success **fal: Recent History Display** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/13545) > The flow starts on a Recent History screen showing previously generated image requests in a card-based grid. The user changes the history presentation from a visual grid to a list with previews, then disables previews to create a more compact text-focused list. The flow assumes the user can access Recent History and that existing request records are available. It ends with the same history records displayed in compact list mode; no request data is changed. Steps: Grid History → List With Previews → Compact List ##### Task: insurance claim health **Wabi: Service Access Settings** (ios, 3 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12043) > This mobile settings experience lets the user enable or disable data connections for several services. Steps: Service Toggles → Health Permission → Settings Updated **Haptic: Enable Health Access** (ios, 3 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/2128) > User is prompted during onboarding to allow the app (Haptic) to access their iPhone Health data so the app can import activities (sleep, steps, weight). Steps: Onboarding: Health Access Prompt → System Health Access Modal — Initial → System Health Access Modal — Enabled **ALO: Content Search and Results** (web, 5 screens, Fitness, Lifestyle) [Refero](https://refero.design/flows/12416) > Happy path for searching content within ALO Wellness Club. The user starts from a profile/landing view with a global search field, types a query, reviews autocomplete suggestions or category previews, and then lands on a full search results page. Steps: Landing Search → No Matches State → Search Suggestions → Refined Suggestions → Full Results Page **Longevity Deck: Comment interactions** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/9887) > This is a product discussion section for an 'Air Purifier' product. The user sees the product hero at the top, a prominent medical-disclaimer card, and a chronological list of user comments below. Primary interactions available on this screen are: liking/unliking a comment, opening an action menu for a comment (three‑dot), and entering a new comment using the input at the bottom. Entry into this screen is from the product area (back button returns out). Exit is via the back control at top-left. Assumptions: the user is on a mobile device and appears to be authenticated (avatar shown and ability to like/report). Steps: Comments list (initial state) → Open comment actions → Reaction toggled (unlike) **Omio: Fare selection and Omio Flex insurance** (ios, 2 screens, Travel & Booking) [Refero](https://refero.design/flows/5182) > Mobile booking step where the user selects a fare tier and can optionally add the Omio Flex "cancel for any reason" insurance. Entry: user arrives from a previously chosen route/date (Barcelona → Madrid, Wed Apr 9, 1 adult). Primary actions: choose one fare card (only one selectable), review fare terms, and toggle or review the Omio Flex add-on. Exit: user either keeps the default fare selection (and may or may not have the insurance toggled) and continues the booking, or opens the Omio Flex detail screen and then returns with the insurance decision. Assumptions: mobile (iOS) booking flow, single passenger shown, the UI shows incremental price deltas for fare selections and insurance per passenger. Steps: Select fare and toggle Omio Flex → Omio Flex details **komoot: Device Connection - Apple Health** (ios, 4 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8347) > Happy path: A signed-in iOS user is prompted to connect a device/service (Apple Health) so Komoot can sync activities and routes. Steps: Device & Services List → Health Access Prompt (initial) → Enable Permissions & Confirm → Connected State & Continue **Brink: Enable Health Integration** (ios, 5 screens, News) [Refero](https://refero.design/flows/10923) > Happy path: while a user is in an active Walking Mode session, they open the Health Integration modal, choose to connect to Apple Health, grant the permission, and return to the running session with a confirmation banner showing the connection is active. Steps: Active Walk — Health Not Connected → Health Integration Overview → Connect Apple Health — Permission → Connecting (Progress State) → Active Walk — Health Connected **Uber Eats: Promotion Claim and Confirmation** (web, 4 screens, Food & Drink, Food Delivery) [Refero](https://refero.design/flows/440) > Happy path: a signed-in user on the desktop Discover/Home view sees a promotional banner, opens a short promo card, claims the offer, reviews the full terms, acknowledges them, and receives a confirmation that the promotion was claimed. Steps: Home / Discover - Promotional Banner → Promo Summary Modal → Promo Details Dialog → Confirmation Toast and Home Refresh **Trip: Attraction Discount Claim** (web, 4 screens, Travel & Booking) [Refero](https://refero.design/flows/12093) > Happy-path entry begins on a Trip.com attraction detail page for Warner Bros. Studio Tour London, showing the hero image, rating, tabs, tickets, reviews, photos, and nearby content. The user opens a discounts modal, reviews available promo codes, and claims one or more codes. After claiming, the modal updates button states from Claim to Claimed and shows validity details. Steps: Attraction Detail Page → Discounts Modal → Claimed Discount State → Return to Attraction Page ##### Task: learning management course progress **ALO: Class Detail and Playback** (web, 4 screens, Fitness, Lifestyle) [Refero](https://refero.design/flows/12371) > Logged-in class detail page for a yoga/video workout. Entry is the class landing page with a large video player at the top, class metadata, related content, and comments. The user can play the class, adjust playback or quality, view class details, browse related classes, and read or add comments. The flow ends when the user stays on the class page, starts or continues the video, or interacts with supporting content such as comments or recommendations. Steps: Class Overview → Playback Settings → Video Playing → Class Page with Playback **ALO: Personalized Fitness Dashboard** (web, 3 screens, Fitness, Lifestyle) [Refero](https://refero.design/flows/12369) > Happy path for a logged-in user on a fitness/training dashboard. Entry is the home dashboard with preferences and progress widgets visible, plus an onboarding-style tooltip system. The user can dismiss or advance tips, then browse recommended classes from the personalized section. Exit is not shown, but the likely endpoint within this view is opening a class, browsing categories, or saving content. Steps: Dashboard Overview → Preferences Guidance → Personal Picks ##### Task: medication reminder **Breathwrk: Create Reminder** (ios, 2 screens, Health & Wellness) [Refero](https://refero.design/flows/7732) > Happy path for creating a new breathing/exercise reminder from within the app. Steps: Add Reminder (initial) → Add Reminder (selection adjusted) **Loowner: Item reminder creation** (ios, 7 screens, Marketplace) [Refero](https://refero.design/flows/6598) > Happy path: A user opens an item detail (Nike AIR Zoom Pegasus), selects Add reminder, chooses either a preset interval or a custom date/time, optionally edits the reminder message, grants notification permission if requested, saves the reminder, and sees a success confirmation then an updated item detail showing the created reminder. Steps: Item detail — view → Choose preset or custom → System notification permission → Reminder editor modal → Save / confirmation → Item detail — reminder shown → Empty / secondary states **Kin: Memory edit & reminder setup** (ios, 38 screens, AI Tool) [Refero](https://refero.design/flows/5432) > Happy-path walkthrough: A user starts a new chat with Kin, shares personal preferences which are parsed into highlighted entities and stored as encrypted memories. The user inspects a stored entity (node), edits its label (triggering validation and a meaning-change warning), reviews the proposed updates, and approves them—updates are applied across related memories. Later in the same conversation Kin suggests a recurring reminder (weekly weigh-in). The user grants microphone permission, records/inputs confirmation, reviews an automatically generated reminder card, approves it, and the reminder is saved. Steps: Onboarding greeting → User composes preference message → Send message → System parses and confirms → Open node detail → View related memory → Initiate node edit → Edit node label → Detect potential meaning change → Review changes → Conversation continues → Request microphone permission → In-conversation recorder → Send confirmation → Review reminder card → Reminder saved → Record or type subsequent answers → Background recording lifecycle → Conversation summary card → End state — reminder and memory persisted **Netflix: Set release reminder** (ios, 3 screens, Entertainment, Video Streaming) [Refero](https://refero.design/flows/10593) > Happy path: A signed-in mobile user browsing the "New & Hot" section finds a "Coming Soon" title and taps the "Remind Me" CTA on a listed item (movie "Smile 2"). The app shows a top confirmation banner/modal that states the reminder is set and gives a relative time ("in 19 days"). After dismissal, the list item updates to a persistent "Reminder Set" state (button text and icon change). Steps: Item detail in list (Remind Me CTA) → Confirmation overlay (All set!) → List item updated (Reminder Set state) **District: Manage movie reminders** (ios, 2 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12879) > A logged-in mobile user views a list of saved movie reminders on the Manage reminders screen. Each reminder card shows poster art, title, release date, and city; some entries also show a theatre list or a Select theatres action. The user can either edit theatre selections or remove a reminder. The entry point is the reminders list screen, and the exit state shown is after removing one reminder, with a confirmation toast indicating the reminder has been deleted and notifications will stop for that movie. Steps: Reminder List → Removal Confirmation **Breathwrk: Create Reminder** (ios, 10 screens, Health & Wellness) [Refero](https://refero.design/flows/7729) > Happy-path: A logged-in mobile user opens the Reminders area, taps + to create a new custom reminder, chooses an exercise type (from Automatic to a specific exercise 'Calm'), sets the time using a native time-wheel picker, selects days of the week, and saves. Steps: Reminders List → Add Reminder (initial) → Exercise Selector (list) → Select 'Calm' Exercise → Exercise Selection Confirmed → Time Picker (initial) → Time Picker (adjust) → Add Reminder (time & days updated) → Save Reminder → Reminders List (updated) **Loowner: Item Reminder Setup** (ios, 6 screens, Marketplace) [Refero](https://refero.design/flows/6600) > Happy path: A logged-in user opens the Reminders area, switches to the Past tab (initially empty), and starts creating a new reminder. The user selects a product from a list, proceeds to a details modal where a default date/time (30 days from now) and a message are prefilled, optionally edits those values, saves the reminder, sees a success confirmation, and is returned to the Reminders list where the new reminder appears. Entry: Reminders screen (Past tab). Exit: Reminders list updated with the newly created reminder. Assumptions: user is authenticated and has at least one product/thing to choose from; the device can display modal overlays and confirm dialogs. Steps: Reminders (Past tab, empty state) → Select item to remind → Change selection (optional) → Reminder details modal → Save confirmation → Reminders list updated **Linear: Issue Reminder Setup** (ios, 3 screens, Collaboration Tools, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/6618) > Happy-path for setting a reminder on an issue from the issue detail view. Steps: Issue detail (initial) → Reminder options sheet → Confirmation and updated issue **Kin: Create and schedule reminder** (ios, 9 screens, AI Tool) [Refero](https://refero.design/flows/5439) > Happy path: a logged-in mobile user (greeting: "Welcome back, Mira") navigates from the home screen to the Reminders section, chooses to create a custom reminder, enters reminder text, enables recurrence, sets a time and one or more repeat days, and saves the reminder. Steps: Home — quick actions → Reminders list → Create reminder — modal (initial) → Create reminder — entering text → Create reminder — text entered → Recurring controls revealed → Time picker — select time → Recurring days selected → Save and return to Reminders list **Instagram: Set event reminder from tagged photo** (ios, 3 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/2865) > Happy path: A logged-in mobile user viewing a tagged photo discovers an event call-to-action (Remind me), taps it, reviews event details in a bottom sheet, enables a reminder, and receives a confirmation toast. Steps: Tagged post — Remind CTA → Event details modal → Reminder confirmation **Poppy: Create Location Reminder** (ios, 14 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12188) > The user starts on Poppy’s assistant dashboard and opens the Create shortcut. They choose the Reminder type, select or edit a suggested reminder prompt, configure it to trigger when arriving at a saved place, and choose Home as the location. On the final step, they keep the suggested name “Spend Check,” configure the recurring reminder’s end date, and create it. The dashboard updates to show five active items. The user then opens Inbox to review saved reminders. The screenshots show an authenticated user on a mobile device; some intermediate screens represent selectable template and date-picker states. Steps: Assistant Dashboard → Choose Creation Type → Enter Reminder Prompt → Select Commute Template → Select Spend Check Template → Choose Reminder Trigger → Enter Arrival Location → Confirm Home Location → Name Reminder → Open End Date Picker → Select Reminder End Date → Adjust End Date → Reminder Created → Review Reminder Inbox **Breathwrk: Set Breathing Reminder** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/7731) > Happy path: A logged-in user opens the Add Reminder sheet to create or edit a breathing reminder, enables the reminder, chooses which breathing exercise variant, sets a time with a native time picker, selects days of the week, and saves. Entry: user navigates to Add Reminder (sheet/modal) from the reminders list or settings. Exit: user saves the reminder and is returned to the previous screen (reminders list or parent view). Assumptions: user is authenticated, the app has permission to schedule notifications (or the user will be prompted elsewhere), and the device uses a wheel-style time picker (mobile). Steps: Add Reminder (main sheet) → Open Time Picker → Adjust Time → Review & Save Reminder **UNHINGED: Add person note with reminder and category** (ios, 11 screens, Health & Wellness) [Refero](https://refero.design/flows/6341) > Happy path: From the People tab the user opens a contact (Harry) and adds a new note tied to that person. In the note editor they type text, add a reminder (pick date, time, and repetition), choose a category, then save. Steps: Person profile — Notes empty → New note editor (empty) → Typing note content → Activate reminder calendar → Navigate months and choose date → Open time picker → Set repetition → Confirm reminder and return to editor → Open category selector → Category selected and note ready → Save and view note in list **Unwind: Reminder setup** (ios, 8 screens, Health & Wellness) [Refero](https://refero.design/flows/5839) > This is the onboarding reminder setup happy path. The flow starts on a post‑signup 'Congratulations' screen where the user can immediately set up reminders. Selecting 'Set up reminders' navigates to a 'Daily Reminders' modal with three frequency choices (No reminders, Twice a day, Custom). Choosing Custom reveals an 'Add Custom Reminder' area where the user can add one or more daily reminder times using an iOS-style time wheel picker and Save. After each save the new reminder appears in the custom list (label + time). The user can add multiple custom reminders, then exit back to the onboarding screen and finish by tapping 'Done'. Assumptions: user is in onboarding (just completed a setup), has notification permission available or will be prompted elsewhere, and is on a mobile device with a native time picker component. Steps: Congratulations / Quick actions → Daily Reminders — initial → Daily Reminders — Custom selected → Time picker — choose time → Custom list — one reminder added → Time picker — add second time → Custom list — multiple reminders → Return to Congratulations / Finish **Cushion: Set Post Reminder** (web, 3 screens, Communication Tools, Collaboration Tools) [Refero](https://refero.design/flows/12519) > Happy path: a user opens a post in a workspace channel, uses the post’s overflow/more-actions menu, and selects a reminder option. The entry point is the post detail view in the channel. The exit point is a confirmation toast indicating the reminder has been set. Assumptions: the user is already signed in, has access to the channel, and is viewing the post on desktop. Steps: Open Post Menu → Choose Reminder Time → Reminder Confirmation **Netflix: Title detail and reminder** (web, 2 screens, Entertainment, Video Streaming) [Refero](https://refero.design/flows/10511) > Happy path: A user browsing content opens a large, centered title detail overlay for a specific movie. The overlay shows a hero image, primary CTA to set a reminder, short synopsis, cast/genre metadata, and a “More Like This” grid. The user taps the Remind Me CTA and the control toggles to a confirmed state (“Reminder set”). Steps: Title detail (pre-reminder) → Reminder confirmed within modal **Calendly: Contact Note with Reminder** (web, 8 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/14033) > The user starts on an existing contact’s profile, opens the note composer, writes a note, optionally adds a reminder date, and saves it. The saved note then appears in the contact’s activity timeline with the selected reminder date and a completion control. The flow assumes the user is logged in, has access to the Contacts section, and has permission to add notes. It is shown on a desktop interface. **Craft: Reminder creation from document** (web, 5 screens, Productivity, Project Management, Paperwork) [Refero](https://refero.design/flows/7952) > Happy-path: The user opens the Notifications (Reminders) panel from a document view, chooses to create a reminder, provides a title and a scheduled time (using quick presets or custom), and saves it. Steps: Open Notifications (Reminders) → Start Create Reminder → Choose Reminder Time (Presets) → Enter Title and Confirm Time → Reminder Saved and Listed **Craft: Block Reminder Setup** (web, 7 screens, Productivity, Project Management, Paperwork) [Refero](https://refero.design/flows/7949) > Happy-path: A signed-in user editing a document locates a specific block (an image in this case), opens the block’s context menu, selects “Remind Me,” chooses a custom date and time in the calendar picker, and applies the reminder. Entry: user viewing the document with editing/selection enabled. Exit: the UI shows a reminder chip attached to the block and the reminder is stored as metadata on that block. Assumptions: user has edit permissions for the document, is on a desktop/larger screen (sidebar and right inspector visible), and the document supports block-level actions (context menu, reminders). Steps: Open document and select block → Open block context menu → Select Remind Me → Open custom date picker → Pick time slot → Apply the reminder → See reminder confirmation **Clay: Contact Reminder Composer** (web, 4 screens, Productivity) [Refero](https://refero.design/flows/5877) > Happy path where a user adds a timed reminder plus a short note to a contact's timeline. Steps: Open Timeline Composer → Open Date/Time Picker → Enter Reminder Note Text → Saved Reminder Appears in Timeline **Calendly: Create Meeting Reminder** (web, 10 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/14031) > A logged-in Calendly user starts from the Automations area, chooses an email reminder template, configures the event types it applies to, changes the trigger from 24 hours to 1 day before the event, reviews the email content, and saves the automation. The flow begins on the Automations landing screen and ends with the new automation visible in the user’s automation list. The user is assumed to have permission to create automations and is using a desktop browser. **Twitch: Set Stream Reminder** (web, 2 screens, Social Networking, Entertainment, Video Streaming) [Refero](https://refero.design/flows/13608) > The user searches for “discoelysium” and reviews matching Twitch results. They locate the DiscoElysium513 channel’s scheduled stream and select its reminder control. The interface confirms the reminder is set without leaving the search results. The flow assumes the user is viewing Twitch in a desktop browser and can access channel reminder controls. Steps: Search Channel Results → Reminder Confirmation **ChatGPT: Delete Saved Memory** (web, 3 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/6902) > Happy path: A signed-in user opens the Personalization > Saved memories modal, reviews a listed memory, chooses to delete it, confirms the deletion in a small confirmation dialog, and sees the saved memory cleared with a success toast and an empty list. Steps: Open Saved memories → Confirm deletion → Show success and update list **Craft: Manage In-Document Reminder** (web, 3 screens, Productivity, Project Management, Paperwork) [Refero](https://refero.design/flows/7951) > Happy path: From an open document, the user opens the Notifications tray, switches to the Reminders tab, inspects an overdue reminder tied to an item in the document, opens the reminder's action menu, and removes or resolves the reminder. Steps: Open Reminders — View Overdue Reminder → Open Reminder Actions Menu → Confirm Removal — Empty Reminders State **Microsoft Teams: Add meeting from chat to calendar** (web, 4 screens, Communication Tools, Scheduling Tools) [Refero](https://refero.design/flows/9716) > Happy path: A signed-in desktop user opens a channel/chat thread that contains a meeting invite card, chooses to add the meeting to their calendar, confirms the action in a modal, and receives a success toast. Entry: user is in the "Weekly Wrap-Up" chat. Exit: the event is added to the user's calendar and a transient confirmation is shown. Steps: Meeting invite card visible in chat → Reveal Add to calendar action → Confirm add to calendar modal → Success toast and final state **Clay: Set contact recurring reminder** (web, 6 screens, Productivity) [Refero](https://refero.design/flows/5880) > Happy path: From a person record, the user opens the command/controls to configure a recurrence (cadence) for reminders, selects a weekly cadence, picks a day of the week, and confirms the selection. Steps: Open person record and action menu → Open command/search modal for cadence → Inspect weekly cadence configuration → Open weekday dropdown → Select a different weekday → Confirm weekly cadence preview **New Balance: Coming Soon Alert Signup** (web, 3 screens, Marketplace, Sports) [Refero](https://refero.design/flows/11853) > Happy path for a product detail page where the item is not yet available. Entry is the product page for ABZORB 1890 Grey Days Shoes. The user can review product imagery and details, then use the launch alert signup to be notified when the item becomes available. Steps: Product Detail Page → Launch Alert Form → Subscription Confirmation **Kit: Manage Recommendations** (web, 4 screens, Business, Communication Tools) [Refero](https://refero.design/flows/2264) > Logged-in desktop web flow for managing creators in the "My Recommendations" list. Steps: Recommendations list (My Recommendations) → Open per-creator actions → Confirm stop recommending → Return to recommendations list ##### Task: mental health questionnaire mood check **Ada: Symptom assessment feedback and rating prompt** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/4106) > Happy-path: After completing a symptom assessment the user is asked a single-question feedback prompt (Was this symptom assessment helpful?). The user selects an option (Yes / No / I'm not sure). If the user selects the positive path the flow asks for optional written suggestions in a multi-line text area; the keyboard appears for typing and a Send button becomes available when text is entered (Skip is available to bypass text entry). After sending (or skipping), the user sees a thank-you/acknowledgement screen with a clear call-to-action to rate the product in the store and a secondary dismissal option. Steps: Quick feedback prompt → Request optional written feedback → Text entry and send → Acknowledgement and rating CTA **Stardust: Topic discovery — feelings & search** (ios, 2 screens, Health & Wellness) [Refero](https://refero.design/flows/5858) > Happy path: The user opens The Cyclepedia overlay and sees a search field, horizontal category chips (Feelings is selected), and a 2-column grid of large feeling tiles (Energized, Exhausted, Anxiety, Calm, Happy, Irritable, etc.). From here the user either: 1) browses and taps a feeling tile to view details, or 2) taps the search field, types a query, and picks a predictive suggestion. Steps: Feelings category grid → Search with predictive suggestions **TidyCal: Attendee Questions Setup** (web, 3 screens, Business, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/12491) > This is an edit screen for a booking type called "Family/Couples therapy session" focused on configuring attendee questions. The user enters this section from the booking type settings, can enable or disable questions, add manual questions, generate questions with AI, and save changes. Assumption: the user is already logged in with permission to edit booking types, on a desktop device. Steps: Questions Enabled → AI Question Generation → Multiple Questions Drafted **Meta AI: Chat Response Feedback** (web, 3 screens, AI Tool) [Refero](https://refero.design/flows/12737) > Logged-in chat interface with a left navigation sidebar and a conversation area. The user asks the assistant for an unusual fact, receives a response, then uses the message feedback controls to mark it positively. After the feedback action, the interface shows a confirmation that the feedback was submitted and the reaction state updates to indicate the response was liked. Entry is from an existing chat thread; exit is the completed feedback state on the assistant message. Steps: Assistant Reply → Positive Rating → Feedback Confirmation **Napper: Daily emotional check-in** (ios, 12 screens, Health & Wellness) [Refero](https://refero.design/flows/8806) > Happy‑path: A mobile daily check‑in where a logged‑in user (identified by name in header) records their current moods, rates their day, selects contextual reasons, and optionally writes a journal note. Entry: Home / Daily check‑in landing. Exit: Home updated with a saved check‑in summary and journal snippet. Assumptions: User is signed in (name shown), on a phone, has permission to write text. The UI uses a multi‑step modal/overlay with a top progress bar, neon‑outlined tile selection, a five‑point rating control, reason buttons, and a full‑screen text editor for journaling. Steps: Home — Daily check‑in landing → Pick moods — page 1 → Pick moods — selected state → Pick moods — page 2 → Rate your day — scale (empty) → Rate your day — selected → Choose reasons — context options → Choose reasons — multiple selected → Journal — new entry (keyboard up) → Journal — draft saved → Confirmation — check‑in saved → Home — updated summary cards **Napper: Daily Question Poll** (ios, 2 screens, Health & Wellness) [Refero](https://refero.design/flows/8808) > Happy path: a user opens the in-app "Today's Question" card, reads a large headline question, and selects one of three emoji-labeled answer cards. After selecting, the UI immediately shows aggregated results for each option (percentage + a horizontal progress bar). Steps: Question Display → Results Display **Train Fitness: Feed sentiment prompt** (ios, 2 screens, Health & Wellness, Social Networking) [Refero](https://refero.design/flows/7160) > Happy path: a signed-in mobile user opens the Feed tab and sees a large hero card asking for quick sentiment feedback (two prominent buttons: Good / Bad). The user taps one of the buttons and an ephemeral confirmation toast appears at the top of the screen reading “Feedback Submitted.” The feed content (activity post with metrics and muscle map) remains visible and interactive. Steps: Feed — Feedback Prompt → Feed — Confirmation + Post **Tinder: Safety Center Quiz** (ios, 14 screens, Dating & Relationships) [Refero](https://refero.design/flows/3456) > Happy-path: A logged-in user (identified by name/avatar in the header) opens the Safety Center, selects the "Online Dating Safety Quiz," reads a short intro modal, then proceeds through an 8-question multiple-choice quiz. Each question appears inside a full-screen card with two choices. After selecting an answer the UI immediately shows correctness feedback (Correct!/Incorrect!), highlights the chosen option, and displays a short explanatory message. The user advances through questions using a "Next question" button. At the end the app shows a summary card with a progress bar, brief wrap-up text and a CTA to view detailed guidelines. Steps: Safety Center (Guide list) → Quiz Intro Modal → Question 1 — Prompt → Question 1 — Incorrect feedback → Question 1 — Explanation + Continue → Question 2 — Prompt → Question 2 — Explanation + Next → Question 5 — Prompt (mid-quiz) → Question 6 — Prompt → Question 6 — Incorrect Explanation → Question 7 — Prompt → Question 7 — Explanation + Next → Question 8 — Final Prompt → Quiz Completion Summary **pillowtalk: Guided Mood Reflection** (ios, 8 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13497) > The user begins a mood and energy reflection by responding to a written prompt. They type a free-form journal entry, then choose either to finish or explore the entry further. Selecting the deeper exploration option opens a conversational reflection experience where the system generates an empathetic response. The user can request suggestions, review suggested reflection directions, and select one to receive additional guidance. The entry and generated responses remain visible in a transcript-style record, with an option to switch to an analysis view. The flow begins in a focused text-entry state and ends with an expanded reflective conversation and suggested grounding steps. Steps: Mood Reflection Prompt → Write Journal Entry → Choose Deeper Exploration → Receive Reflective Response → Open Suggestion Choices → Select Reflection Direction → Review Suggested Next Steps → Continue Guided Support **Podimo: Live Poll Participation** (ios, 2 screens, Entertainment) [Refero](https://refero.design/flows/12276) > A logged-in mobile user is viewing a content detail screen for Casa Jedet with comments and a live poll. The entry point is the content card/detail view; the exit point is either closing the screen or staying after voting. Steps: Poll Preview → Poll Results **Unwind: Session completion check-in** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/5840) > Happy-path: after finishing an Unwind session the user sees a congratulatory screen where they can optionally share the success, select how they are feeling (mood chips), attach a short free-form note, and finish by tapping Done. Steps: Completion summary → Open note editor → Enter note text → Save and review **Sora: Mood-based Feed Filter** (ios, 6 screens, AI Tool, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/7842) > Happy path: A signed-in user browsing the main content feed opens the feed sorting/filter menu, selects "Pick a mood (BETA)", types a short text describing the mood they want, submits it, and returns to the feed with a "Custom" mood filter applied. Steps: Feed (Latest) → Sort/Filter Menu → Pick a mood — Prompt → Type Mood Text → Submit Mood → Feed with Custom Tag **GO Club: Plan review & daily reflection** (ios, 5 screens, Health & Wellness) [Refero](https://refero.design/flows/6352) > Happy-path flow for a logged-in user viewing a weekly activity plan and submitting a quick daily reflection. Entry: user arrives at the Plan/Progress screen showing a 7‑day weekly progress chart and the Plan Breakdown for Week 1. Primary in-screen action is the prominent floating "Plan" control which launches an overlaid "Reflect on your today's journey" modal. In the modal the user reviews read-only daily metrics (steps, distance, calories, speed, standing hours, heart rate), selects a mood via an emoji slider, and submits. Exit: modal closes and user returns to the Plan/Progress screen. Assumptions: user is authenticated, sensors or health data access may be available but are presented read-only here, device has standard mobile viewport and supports modal overlays and touch gestures. Steps: Plan Overview → Open Reflection Modal → Review Metrics and Set Mood (neutral) → Adjust Mood to Positive → Submit Reflection and Return **Stardust: Archetype Quiz & Reveal** (ios, 11 screens, Health & Wellness) [Refero](https://refero.design/flows/5899) > Happy path: A signed-in mobile user opens their profile and chooses to discover their archetype. They tap Take the quiz, read a short intro (with an estimated time), then begin a short series of statement-based questions. Each question is answered by selecting a star on a 5-point scale (labels: DISAGREE to AGREE). A top progress bar updates as they respond; a back control allows returning to previous items. After the last question the app shows a loading/processing screen then reveals the user's archetype as a visual card plus a detailed, scrollable description and a CTA to Discover More. Steps: Profile entry - Archetype prompt → Intro modal — quiz summary → Question screen — statement → Rating feedback — selection confirmed → Navigation control visible → Subsequent question — continued rating → More questions — consistent interaction → Processing / calculating result → Result reveal — archetype card → Result details — scrollable description → Result summary with CTA **Tinder: Profile quiz setup (Ask Me About)** (ios, 6 screens, Dating & Relationships) [Refero](https://refero.design/flows/3449) > Happy path: A signed-in user opens the Edit Info screen and taps "Add quiz" on the "Going Out" row under the "ASK ME ABOUT" section. A modal bottom sheet presents a sequence of short multiple-choice questions (2 options per question). The user selects one option per question; the modal advances through all questions and then shows a compact results card summarizing the chosen answers. The user taps Done to close the modal — the parent list updates to show the result summary next to the "Going Out" row. Entry: Edit Info screen (logged-in user). Exit: Bottom sheet closed, Edit Info list updated with quiz summary. Assumptions: user is signed in and allowed to edit profile fields; device is a phone (vertical modal bottom sheet); selections are single-choice per question and recorded locally or persisted when Done is tapped. Steps: Edit Info — Ask Me About list → Question 1 — Select where you can be found → Question 2 — Dress style preference → Question 3 — Arrival timing → Question 4 — Exit strategy → Finalize — Close results and update list **Teal: Work-style assessment** (web, 21 screens, Productivity, Job Search) [Refero](https://refero.design/flows/2730) > Happy-path: A signed-in user on the platform launches a short, 17-question work-style survey from a company/dashboard view and completes it. Each question is presented as a bipolar adjective pair (e.g., Competitive — Cooperative) with a five-position slider. The user moves the slider (or taps a position), then navigates forward with Next; from question 2 onward a Previous button appears. The final step asks the user to choose one of three career-growth card options and press Complete. On completion the user is taken to a results dashboard that shows a video, a textual summary of the user's work style, a 360 invite CTA, a right-hand 'Dimensions' sidebar visualizing each bipolar dimension as miniature sliders, and expanded content sections (Career Insights, Start & Finish, Results & People, Sequence Pattern, Job Searching). Steps: Dashboard — Work Styles entry → Question 1 — Competitive vs Cooperative (initial) → Question 1 — Competitive vs Cooperative (selected) → Question 2 — Laid-back vs Assertive → Question 3 — Peaceful vs Defiant → Question 4 — Agreeable vs Pioneering → Question 5 — Animated vs Soft-spoken → Question 6 — Cautious vs Adventurous → Question 7 — Convincing vs Restrained → Question 8 — Self-controlled vs High-spirited → Question 10 — Spontaneous vs Planned → Question 11 — Methodical vs Enthusiastic → Question 12 — Skeptical vs Trusting → Question 13 — Light-hearted vs Serious → Question 14 — Impatient vs Patient → Question 15 — Lenient vs Strict → Question 16 — Questioning vs Accepting → Question 17 — Career growth options (card selection) → Results — Work Style dashboard → Results — Expanded details and CTAs **WhatsApp: Create and Review Poll** (web, 7 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/13192) > Happy path for creating and sharing a poll in a chat, then viewing the live results and detailed vote breakdown. **New Balance: Post-Purchase Survey** (web, 7 screens, Marketplace, Sports) [Refero](https://refero.design/flows/11861) > A post-purchase survey modal appears on top of an order confirmation/order summary screen. The user starts from a completed purchase state, then rates their satisfaction, optionally provides a reason and additional feedback, and finishes by dismissing a thank-you confirmation. Assumption: the user is already on the order confirmation page after checkout, with a modal overlay available and no special permissions required. Steps: Satisfaction Rating → Rating Selected → Reason Prompt → Rating Explanation Filled → Purchase Motivation Prompt → Purchase Motivation Answered → Thank You Confirmation **Duolingo: Friend update reaction** (web, 4 screens, Education) [Refero](https://refero.design/flows/11829) > Happy path starts on the user’s profile dashboard and moves to a friend updates view where the user reacts to a friend’s achievement post. Steps: Profile dashboard → Friend updates → Reaction picker → Reaction confirmed **Typeform: AI question generation** (web, 9 screens, Productivity, Communication Tools) [Refero](https://refero.design/flows/5586) > Happy path: A logged-in user working in a workspace uses the form content editor to auto‑generate survey questions via an AI helper. From the content editor they open a "Create with AI" modal, choose a general template, then walk through a three-step guided input (describe form goal, add topics, add business info). After submitting, the system shows a loading/"crafting" state and then inserts generated questions into the form editor. Steps: Form editor (initial) → Create with AI - entry → Describe your form (intro) → Describe your form (filled) → Add important topics (empty) → Add important topics (filled) → Add business information → Generating (loading) → Generated questions inserted **Intercom: Agent feedback survey** (web, 5 screens, Business, Communication Tools) [Refero](https://refero.design/flows/647) > Happy path: a logged-in user on the Fin AI Agent Overview is presented with an Intercom-powered feedback modal asking for input about AI-generated answers. The modal walks the user through three quick screens: an intro, a five-point satisfaction rating, and an optional free-text justification, then shows a confirmation. Steps: Survey intro → Satisfaction rating → Rating selection confirmation → Qualitative comment → Completion confirmation **LinkedIn: Help Center Feedback Submission** (web, 4 screens, Business, Social Networking, Job Search) [Refero](https://refero.design/flows/342) > Happy-path: A not-signed-in desktop user lands on the LinkedIn Help center home and opens the small feedback widget in the right column. The widget expands into a short multi-step survey (radio options). The user answers the questions, navigates between questions using Next, and submits answers. The survey responds with a confirmation screen and a Close action that returns the user to the Help center landing page. Assumptions: user is on desktop, not signed in (Sign in button visible), and has access to the Help center landing content. Steps: Help center landing (feedback widget visible) → Question 1 — satisfaction rating → Question 2 — primary purpose → Submission confirmation **Kinhive: Daily Prompt Response** (web, 4 screens, Communication Tools) [Refero](https://refero.design/flows/6866) > Happy-path: A signed-in user on the Home dashboard clicks the Daily Prompt card to answer a one-question prompt. A modal dialog opens where the user types a required response, optionally toggles whether the answer is shown on their profile, and submits or skips. On successful submit the modal closes and the Daily Prompt card updates to confirm the user has answered today. Steps: Home dashboard → Prompt modal (empty) → Prompt modal (filled) → Dashboard — prompt acknowledged **Calendly: Add Session Type Question** (web, 5 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/14049) > An authenticated scheduling user edits an existing routing form. Starting with a form containing a required Name field, the user adds a radio-button question, enters the question text and three answer choices, and returns to the completed form view. The system autosaves the changes and keeps the user on the Form tab. The flow begins on the form editor and ends with the updated form ready for preview or progression to the Next step. The user appears to have permission to edit the form and is using a desktop interface. **IKEA: Chatbot post-conversation survey** (web, 5 screens, Marketplace, Home & Decor) [Refero](https://refero.design/flows/8274) > Happy path: A user browsing the IKEA 'Contact us' section interacts with the site chat widget after a conversation has ended to submit a feedback survey. Steps: Contact page with post-chat survey → User selects a star rating → User enters written feedback → Submit survey and show success modal → Acknowledge and return to chat state **OpenAI: Submit feedback on assistant message** (web, 6 screens, Business, Development Tools, AI Tool) [Refero](https://refero.design/flows/4197) > Happy-path for submitting feedback on a single assistant response in the Playground chat. Steps: View assistant response → Consent prompt appears → Allow consent and open feedback entry → Enter feedback text → Submit feedback → Final confirmation state ##### Task: prescription refill repeat **Subo: Add Subscription** (ios, 4 screens, Finance) [Refero](https://refero.design/flows/8581) > Happy path for adding a recurring subscription (Apple Music) to the user's list. Steps: Blank Add form (initial) → Open Card Editor → Select Color and Icon → Complete Form & Confirm **Podimo: Referral Link Share** (ios, 2 screens, Entertainment) [Refero](https://refero.design/flows/12260) > A mobile share overlay for a referral or promotional link. The user enters from an existing Podimo screen and can dismiss the overlay with the close button or share/copy the link using quick actions such as Copy link, WhatsApp, Instagram Stories, or Messages. In the happy path shown, the user taps Copy link and receives confirmation that the link was copied to the clipboard. Steps: Share Options → Copy Confirmation **Train Fitness: Referral Link Sharing** (ios, 2 screens, Health & Wellness, Social Networking) [Refero](https://refero.design/flows/7137) > Happy path: a logged-in user opens the Rewards and Referrals screen to share their personal referral URL and claim rewards. The screen presents a promotional hero, the user's referral link in a bordered input area with a copy icon, and a prominent "Share Now" button. The user can either copy the link to clipboard (copy icon) or start a share action (Share Now). A transient confirmation toast appears after copying. Steps: Referral Landing → Copy Confirmation **Willow: Friend Referral Invite** (ios, 2 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/13240) > Logged-in mobile profile screen with a referral bottom sheet opened from the account area. The user can either invite friends through a share/invite action or copy a referral link. If the user taps copy, a confirmation toast appears and the sheet remains open. Steps: Referral Offer Sheet → Copy Confirmation **ElevenMusic: Invite Friends Referral** (ios, 2 screens, AI Tool, Entertainment, Music Streaming) [Refero](https://refero.design/flows/12445) > A logged-in user views an invite-friends panel from within Account Settings and can either copy a referral link or trigger the invite action. The flow starts with the referral modal open over the settings screen and ends after the link is copied, showing a success toast while keeping the modal in view. Steps: Referral Modal Open → Copy Confirmation **The Leap: Referral sharing center** (web, 2 screens, Business, Entertainment) [Refero](https://refero.design/flows/4569) > Referral center for a logged-in merchant where the user obtains and copies their personal referral/store link to share. Steps: Referral overview and copy affordances → Confirm link copied **The Leap: Referral Link Copy and Share** (web, 2 screens, Business, Entertainment) [Refero](https://refero.design/flows/4568) > Logged-in merchant visits the Referrals area (left navigation). The screen shows referral KPIs (total referrals, paid referrals, total earned) and a prominent CTA to copy a personal referral link. Happy path: user clicks Copy referral link (or the inline `Copy to share` button), the app copies the referral URL to the clipboard and shows a success toast. Steps: Referral overview → Copy confirmation **Seed: Supplement Purchase and Cart Review** (web, 2 screens, Health & Wellness, Marketplace, Lifestyle) [Refero](https://refero.design/flows/12732) > Happy path for viewing a supplement product detail experience and adding the item to cart from a right-side cart drawer. Steps: Product Detail Review → Cart Drawer Review **ElevenReader: Referral link sharing** (web, 2 screens, Books, AI Tool) [Refero](https://refero.design/flows/9188) > Happy path: A signed-in desktop user navigates to the Explore section and sees a promotional "Refer a friend" card with a referral URL and a Copy link button. The user copies the link and receives immediate visual confirmation. Steps: Explore — Referral card visible → Copy link confirmation **GlossGenius: Sales refund handling** (web, 6 screens, Business) [Refero](https://refero.design/flows/9978) > Happy-path: A logged-in user on desktop opens the Payments -> Sales history list, selects a completed appointment to view details in a right-hand panel, opens the appointment's action menu, issues a full refund via an 'Issue refund' modal, confirms success, and then sees the transaction updated with a refund line. Steps: Sales list → Completed appointment panel → Overflow action menu → Issue refund modal → Refund completed confirmation → Updated transaction details **Butter: Share referral modal** (web, 2 screens, Design & Media Editing) [Refero](https://refero.design/flows/3035) > A desktop modal that lets a user share a referral/gift link with a friend. Entry: user triggers a "Gift/Share" action which opens this modal. The modal displays a referral message, a prefilled share link, a 'Copy Link' button, and a close (X) control. Exit: user closes the modal or leaves it open after copying the link. Steps: Referral modal (initial) → Referral link copied confirmation **ALO: Refer a Friend Sharing** (web, 3 screens, Fitness, Lifestyle) [Refero](https://refero.design/flows/12381) > Logged-in user opens a referral sharing dialog from the site, copies an invite link or chooses another sharing method, then closes the dialog or continues sharing. Steps: Home Screen → Invite Link Modal → Copied Confirmation **ElevenMusic: Invite Friends Referral** (web, 3 screens, Music Streaming, Entertainment, AI Tool) [Refero](https://refero.design/flows/12593) > Happy path starts in the ElevenMusic Studio dashboard with the account menu open. The user selects the invite-friends option, which opens a centered referral modal over a blurred background. The user copies the invite link and sees a confirmation state on the copy control. Steps: Studio Dashboard Menu → Referral Dialog → Copy Confirmed **On: Order Details Review** (web, 2 screens, Marketplace, Fitness, Lifestyle) [Refero](https://refero.design/flows/13833) > The flow begins on a logged-in customer’s account Orders section, displaying the details of a placed order. The user reviews the order number, delivery estimate, item, status timeline, shipping address, payment summary, and available cancellation action. In the shown interaction, the user selects the copy icon beside the order number. The system copies the order number and confirms the action with a toast notification. The page also includes recommended products and the site footer below the order information. Steps: Order Details → Order Number Copied **Instacart: Referral invite** (web, 2 screens, Food & Drink, Marketplace, Food Delivery) [Refero](https://refero.design/flows/10080) > A post-order referral invite modal presented as an overlay on the order details screen (desktop). Entry: user taps an "Invite"/referral CTA from the order details (not shown) or the system auto-displays the referral modal after order completion. The modal offers a $10-for-you/$10-for-friend referral incentive, shows a large illustration, and provides two primary actions: share via a selected channel (WhatsApp in the screenshots) or copy the referral code/link. Assumptions: user is signed in and on a desktop browser; the modal is an overlay blocking interaction with the page behind it. Exit: user closes the modal via the Close control after sharing or copying the code. Steps: Referral modal — initial state → Referral modal — success/confirmation **Sunsama: Create Weekly Task Recurrence** (web, 5 screens, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/11590) > The user opens an existing task titled “Lunch with Lisa,” accesses its actions menu, and configures it to repeat every week on Thursday at 2:00 PM. The user saves the recurrence settings, and the task detail view confirms that the recurring schedule has been created. The flow assumes the user is already viewing the task detail modal and has permission to edit the task. Steps: Task Details → Open Repeat Settings → Choose Recurrence Pattern → Set Recurrence Time → Recurrence Confirmation **Understory: Add Date and Location** (web, 7 screens, Business, Scheduling Tools) [Refero](https://refero.design/flows/12705) > This is a scheduling setup step within an editor for an experience. The user starts on the Add date & location section, configures a date/time block, optionally sets repeat rules through a modal, and returns to the main editor to continue with the next step. Steps: Date and Location Setup → Repeat Rule Dialog → Custom Recurrence Settings → Recurrence End Date Picker → Selected Recurrence End Date → Repeat Rule Confirmed → Updated Date and Location **GlossGenius: Referral link copy and confirmation** (web, 2 screens, Business) [Refero](https://refero.design/flows/9982) > Happy path: A logged-in desktop user (business account) opens the Referrals area from the left navigation and lands on a referrals overview. The hero presents a personalized referral link and a prominent Copy control. The user copies the link and receives immediate visual confirmation (button state change and toast). The page includes supporting content under “Here’s how it works” explaining sharing options, benefit details, and community access. Steps: Referral overview (initial) → Copy confirmation **Cofounder: Referral Link & Copy** (web, 2 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/7905) > This is a two-step happy-path for the Referral Program screen on a desktop device. Steps: Referral Overview → Copy Confirmation **ElevenMusic: Track Published Confirmation** (web, 2 screens, Music Streaming, Entertainment, AI Tool) [Refero](https://refero.design/flows/12588) > A publishing confirmation modal appears after a music track is successfully published. The modal shows the track artwork, publishing label, track title, and a success message indicating the track was added to an album. The user can copy a shareable track link, dismiss the modal with Done, or continue to the album with View Album. Assumptions: the user is already signed in and has just completed a publish action on a desktop web experience. Steps: Publish Success → Copy Link Confirmed **Instacart: Configure out-of-stock substitution** (web, 3 screens, Food & Drink, Marketplace, Food Delivery) [Refero](https://refero.design/flows/10069) > Happy-path: A signed-in shopper on desktop is viewing a product detail for a 7-Eleven glazed donut and configures what should happen if the item is out of stock. Steps: Product detail / item instructions → Out-of-stock modal (options) → Select specific replacement and save **BYBORRE®: Full-Repeat Sample Checkout** (web, 4 screens, Home & Decor) [Refero](https://refero.design/flows/7551) > Happy-path: A designer in the textile design workspace reviews a finished design and initiates ordering a full-repeat sample. From the editor they open a "Details of your sample" modal, review material, size, color swatches and artwork layer count, check a required confirmation about repeatability, and confirm. The confirmed sample is added to the cart where the user sees a thumbnail, line-item details (size, material, price), can remove or change quantity, and proceed to Checkout. Entry: completed design in editor; Exit: user lands on Cart with sample(s) added. Assumptions: user is on a desktop-class viewport in the design workspace with a prepared design to order; the UI indicates an active design session (save controls visible). Logged-in state is not assumed or required from the screenshots. Steps: Design Editor — Sample Preview → Details Modal — Review Sample → Details Modal — Confirm Repeatability → Cart — Sample Line Items & Summary ##### Task: procurement tender bid **Pool: Refero MCP Search** (ios, 2 screens, Utility) [Refero](https://refero.design/flows/11986) > Happy path starts from a preview/card view of a Refero MCP result shown in a mobile viewer. The user can either dismiss the preview or tap the primary call to action to explore the MCP content. Selecting the CTA opens a Google search results page for the query “Refero MCP for AI agents,” showing related results including the Refero site and a GitHub result. Assumptions: the user is on a mobile device, has network access, and is not required to be signed in to view the results. Steps: Preview and Explore → Search Results ##### Task: referral to specialist **Willow: Refer a Friend** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/13239) > Happy path starts on the user’s Profile screen, where the user sees account details and a Refer a friend option. Tapping that option opens a referral modal with invite benefits and actions. From there, the user can either open the native share sheet to send an invite or copy the invite link. The flow ends when the share sheet is opened for sharing, or when the referral modal is closed and the user returns to Profile. Steps: Profile → Referral Offer Modal → Native Share Sheet → New Message Composer → Referral Modal with List **Omio: Refer a Friend invite** (ios, 5 screens, Travel & Booking) [Refero](https://refero.design/flows/5236) > Happy path: a logged-in user on the app home screen discovers a “Refer friends” promotional card, opens the promo, reviews the program details, then taps the CTA to share their unique referral link via the device’s native share sheet and selects Messages to send a prefilled invitation to a contact. Steps: Home Search & Offers → Promo Modal — Quick overview → Refer a Friend — Details → Native Share Sheet → Messages Composer (prefilled) **Ground News: Invite Friends — Referral Sharing** (ios, 3 screens, News) [Refero](https://refero.design/flows/7581) > Happy path: a logged-in iOS user opens the Invite Friends screen, views their personal referral code, and shares it with a friend. From this screen the user can copy the code to clipboard or open the system share sheet using the prominent "Share Referral Code" button. Choosing Messages from the share sheet opens the iOS Messages composer with a prefilled invite message (including download link and referral code) so the user can add a recipient and send. Steps: Invite Friends (Referral Landing) → Native Share Sheet → Messages Composer with Prefill **Willow: Referral Card Share** (ios, 3 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/13242) > This is a mobile referral/share flow for a user profile card. It starts on a profile card screen with actions to copy a link, download, or share. Choosing Share opens the system share sheet, where the user can pick an app or contact. In the shown path, the user selects Messages and lands in a new iMessage compose screen with the referral card preview attached and the recipient filled in. The user can then send the message. Assumption: the user is already signed in and viewing their own shareable card. Steps: Card Options → Share Sheet → Message Compose **Wispr Flow: Friend Referral Invitation** (ios, 3 screens, AI Tool, Utility, Productivity) [Refero](https://refero.design/flows/13683) > A logged-in user opens the Referrals area, reviews the offer and personal invite link, then shares the invitation through the device’s native share sheet. In the shown path, the user selects iMessage, which opens a prefilled message addressed to a contact. The flow ends with the message ready for review and sending; the screenshots do not show the message being sent or the recipient claiming the offer. **Blackbird: Referral — share $FLY reward** (ios, 5 screens, Food & Drink) [Refero](https://refero.design/flows/9023) > Happy path: a signed-in user on the Wallet tab opens the referral offer (Get $20), views a referral modal with a QR code and alphanumeric referral code, and then shares that referral via the iOS share sheet (Messages). Steps: Wallet Home → Referral Details (QR + Code) → Referral Share — action initiated → iOS Share Sheet → Message Composer — prefilled referral **Givingli: Refer & Share Invite** (ios, 3 screens, Lifestyle, Scheduling Tools) [Refero](https://refero.design/flows/8218) > Happy-path: A logged-in user on an iPhone opens the referral/invite screen, chooses to share their referral link (either via the main "Share your invite link" button or by tapping a "Gift $5" button beside a suggested contact), then uses the iOS share sheet to pick Messages. The Messages composer opens with a preview of the referral URL; the user can optionally edit and send the message. Steps: Invite & Contacts → iOS Share Sheet → Message Composer **Podimo: Referral Offer Share** (ios, 4 screens, Entertainment) [Refero](https://refero.design/flows/12257) > Happy path for sharing a promotional Podimo offer from the home screen. The user is on the app home view, opens a gift/referral promotion, and taps Share. The system opens the native iOS share sheet, then the user chooses Messages. This launches a new iMessage draft with the referral text and rich link preview prefilled, ready to send. Steps: Home and Promotion → Gift Offer Details → Share Sheet → Message Draft **Volv: Referral Invite & Reward Progress** (ios, 6 screens, AI Tool, News) [Refero](https://refero.design/flows/2166) > Happy path: Logged-in user opens the Rewards area to see the next reward (Stickers) and starts the referral process to unlock it. From the reward card they open the invite screen (Find your friends), optionally search/filter contacts, select a contact and send a prefilled iMessage/SMS containing their referral code and link. After sending, the Messages thread shows the sent message and the Rewards screen updates to reflect the newly referred friend. Assumptions: user is authenticated, on an iPhone with Messages available, and the app can surface a contact list or friends list for inviting. Steps: Rewards - Next Reward Card → Find Your Friends - Contact List → Search Contacts → Compose Message (prefilled) → Message Sent - Thread View → Rewards Updated **Melius: Send Referral Invitation** (web, 3 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/14098) > A logged-in user opens the Referrals section from account settings, reviews how the referral reward works, enters a recipient’s email address, and sends an invitation. The system confirms the invitation with a success notification and clears the email field. The flow begins in the Referrals settings modal and ends after the invitation is successfully sent. Assumptions: the user has permission to access account settings and referral invitations, and is using a desktop interface. Steps: Referral Details → Enter Recipient Email → Invitation Confirmation **Airbnb: Refer-a-host referral creation** (web, 5 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6009) > Happy path: A logged-in user visits the "Refer a host" landing hero, clicks Get started, picks which type of host they are referring (Home, Experience, or Service), generates a referral link, and immediately opens sharing options to distribute the link (copy, email, social, messaging). Entry: Referral landing hero while signed in (avatar visible). Exit: Referral link generated and shared (copy confirmation shown). Assumptions: desktop layout, user is authenticated, referral program is available for their account. Steps: Referral landing hero → Choose host type (modal) → Host type selected → Share referral options → Copy confirmation **Mercury: Refer a Contact via Email** (web, 5 screens, Business, Finance) [Refero](https://refero.design/flows/11080) > Happy-path: A logged-in user navigates to the Referrals area to invite a contact via email. From the referrals overview (shows bonus card, referral URL, and a table of referred companies), the user opens an "Invite via email" modal, enters or selects an email address, confirms the invite, and returns to the referrals page where a success toast is shown. Steps: Referrals overview → Invite modal (initial) → Typing recipient email → Recipient tokenized and confirm → Referrals overview (success state) **Sunsama: Referral Invitation** (web, 5 screens, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/11602) > A logged-in subscriber opens an invitation screen and chooses to invite people through a referral offer. The user selects the option to tell friends, enters an invitee’s email address and optional name, then sends the invitation. The system confirms that the invite was sent. A Close control remains available throughout the experience. Steps: Invitation Options → Referral Form → Enter Recipient Details → Send Invitation → Invitation Confirmation **Farfetch: Refer-a-friend invite** (web, 3 screens, Marketplace) [Refero](https://refero.design/flows/3553) > Logged-in customer lands on a "Refer a Friend" landing area that communicates a dual-sided discount offer (friend gets 10% off; referrer gets 10% off after spending threshold). Happy path: user reads the offer, taps the main CTA to participate, selects a sharing method or copies a personalized referral link, and the system copies the link and confirms it has been emailed to the user. Steps: Referral Offer Overview → Choose Invite Method → Share Confirmation **Manus: Send referral invite** (web, 4 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/10479) > Happy path: a logged-in user opens the main dashboard and uses the "Share Manus with a friend" control in the left sidebar to open a referral/invite modal. In the modal the user can copy an invite link, trigger a Facebook invite flow, choose one of the shown social share icons, or enter an email address and send an invitation. After submitting an email the system shows a transient success toast; the modal remains open and displays invitation counters and links for redeem/invitation history. Steps: Dashboard - open share control → Invite modal - view options → Invite modal - enter email → Invite modal - confirm send **Prose: Referral Program — Share & Confirm** (web, 3 screens, Marketplace, Health & Wellness) [Refero](https://refero.design/flows/6321) > Logged-in user accesses the Prose Referral Program from their account sidebar. The page presents a two-column layout: a large brand/product image on the left and referral controls on the right. The right column exposes the referral link (with a COPY button), an email composer to send invites, a prefilled message box, an optional reminder checkbox, and social share icons. Happy path: user either copies the referral link or enters recipient emails and clicks SEND. After sending, the page navigates to a confirmation state that shows a thank-you message, the same referral link (copyable), and a CTA to "Share Again". The top of the page also displays referral totals (Total Earned, Total Used, Current Balance). Steps: Share options / Compose start → Recipients added / Ready to send → Share confirmation ##### Task: savings pot goal **The Outsiders: Set Sleep Goal** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/10294) > Happy-path for setting a personal Sleep Goal from the Sleep section of a health app on an iPhone. Entry: user views Sleep details and taps the control to set or edit their sleep goal. The flow opens a modal overlay where the user changes the target hours using a prominent numeric picker and an associated horizontal slider. Visual feedback (striping, color, and a label) indicates whether the chosen value is below, within, or above the app's recommended range (7–9 hours). Exit: user taps Done to save/confirm the new sleep goal and returns to the Sleep overview. Assumptions: user is on an unlocked iPhone, in the Sleep area of a health app, and has permission to change sleep goal settings; this flow does not show any account authentication or cross-device sync steps explicitly. Steps: Sleep Overview (No Data) → Sleep Goal Modal — Default Value → Adjust Goal — Below Recommended → Adjust Goal — Above Recommended and Confirm **GoFundMe: Fundraiser Details and Organization Follow** (web, 3 screens, Fundraising) [Refero](https://refero.design/flows/11731) > A donor opens a fundraiser detail page for a dog’s medical costs. The main entry point is the public fundraiser page showing the title, story, updates, organizer, sharing tools, and donation summary. The user can donate or share from the page, and can also open the organization card modal to view Pasadena Humane details and choose to learn more or follow. The flow ends with the organization marked as followed in the modal state, or the modal dismissed back to the fundraiser page. Steps: Fundraiser Detail Page → Organization Profile Modal → Following Confirmation State ##### Task: scheduled payment standing order **Fresha: Salon service booking with deposit** (ios, 42 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4953) > Happy-path: A user discovers a salon, opens the Services tab, filters by a category, selects one or more services, reviews the selection and price, provides (or selects) a payment method for a required deposit, and then confirms the booking. Steps: Salon detail (header) → Services list (Featured) → Category switch (Skin Care) → Category switch (Hair) → Open Select services → Select services (list) → Multi-select services → Proceed to payment (Add card) → Add card (attempt save) → Session timeout (possible idle) → Select payment method (saved card) → Review and confirm (booking summary) → Pay deposit modal → Payment declined (insufficient funds) → Retry or change card → Notes input → Pay deposit attempt (small amount) → Payment declined (retry) → Leave booking confirmation → Abort or retry ##### Task: search everything **Substack: Global search with empty-note results** (ios, 3 screens, News, Social Networking) [Refero](https://refero.design/flows/4486) > Happy-path: a logged-in user on the Home feed opens the global search, types a query (numeric in this example), and inspects filtered results. Steps: Home feed — open search → Search input — type query → Results with filter chips — empty state ##### Task: symptom checker **Ada: Sign-in to start assessment** (ios, 5 screens, Health & Wellness) [Refero](https://refero.design/flows/4128) > Happy-path: A new user lands on an onboarding screen that prompts them to "Check your symptoms" and is given two primary CTAs: Create your account or Log in. Choosing Log in moves the user to a sign-in options screen where they can pick email or a social provider. If they choose email, they are shown an email/password form (with a 'show' toggle for the password). After entering credentials and tapping Log in the user is taken to the authenticated Home feed which greets the user by name and presents a "Start symptom assessment" CTA and a bottom navigation (Home, Profiles, Conditions). Entry point: Onboarding carousel (not authenticated). Exit point: Authenticated Home screen. Assumptions: user is on a mobile device (iPhone-style), initially not authenticated, and the screenshots represent the linear path of selecting Log in → Sign in with email → entering credentials → successful login. Steps: Onboarding — Check your symptoms → Sign-in Options → Email Login — Empty Form → Email Login — Prefilled Email → Authenticated Home — Greeting & CTA **Ada: Symptom search and selection** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/4098) > Happy-path summary: From the New assessment screen the user is prompted to enter the symptom that troubles them most. Tapping the rounded input opens a modal titled “Search for a symptom” and focuses the text field so the on-screen keyboard appears. The user types a symptom, taps the keyboard search/return key (or selects a suggested result if available), then a chosen symptom is applied to the assessment and the modal closes. Steps: New assessment — symptom prompt → Search modal — empty input → Search modal — query entered → Search modal — empty results **Stardust: Cycle Day Symptom Logging** (ios, 7 screens, Health & Wellness) [Refero](https://refero.design/flows/5861) > Happy path: a signed-in user views a cycle-day screen (top calendar + phase bar) and logs physical/mental symptoms for that specific date. The user taps suggested symptom icons from the grid, a bottom sheet opens to set intensity (when applicable), and they save — saved symptoms appear as rounded chips under "YOU'RE FEELING...". Steps: Cycle Day — Symptom Suggestions → Symptom Detail — Headache (Bottom Sheet) → Adjust Intensity — Preview → Save Symptom — Chip Added → Add Another Symptom — Back pain Detail → Save Additional Symptom — Multiple Chips → Review & Edit Logged Symptoms **Stardust: Custom symptom creation** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/5864) > Happy-path for creating and saving a custom symptom/tag from the daily symptom picker. Steps: Open symptom picker → Enter symptom name → Confirm name → Verify custom item **Stardust: Daily symptom management** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/5865) > Happy path: from a selected date in the calendar the user manages symptoms for that day. Steps: Daily symptom list (date view) → Symptom detail sheet → Updated daily list after removal **Seed: Personalized Supplement Recommendation** (web, 35 screens, Health & Wellness, Marketplace, Lifestyle) [Refero](https://refero.design/flows/12731) > The flow begins on Seed’s health-focused homepage. A visitor selects “Take the Quiz” and completes a personalized wellness questionnaire covering goals, gut symptoms, bowel frequency, mental clarity symptoms, and symptom frequency. The experience presents educational interstitials about ingredients and capsule delivery, then asks for an email address and optionally a phone number in exchange for product discounts. Based on the answers, the system recommends a routine containing Daily Synbiotic and Energy + Focus. The visitor can adjust the recommended products and continue to the Energy + Focus product detail page. The flow assumes the visitor is not signed in and is using a desktop browser. Steps: Health Homepage → Health Priorities → Health Priorities Selected → Gut Symptoms → Gut Symptoms Selected → Bowel Frequency → Gut Health Feedback → Mental Clarity Symptoms → Mental Symptom Selected → Symptom Frequency → Ingredient Education → Capsule Delivery Education → Email Discount Prompt → Email Submitted → SMS Discount Prompt → SMS Submitted → Discount Confirmation → Recommended Routine → Routine Adjustment → Energy + Focus Product **Jasper AI: Plagiarism checker help** (web, 2 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/5294) > Happy path: Within a content editor for a blog post, the user opens a centered help modal titled "Using the Plagiarism Checker" to read instructions. After reading, the user submits feedback using the modal's feedback controls. The system displays an inline confirmation message inside the modal. Steps: Open help modal → Submit feedback **WhatsApp: Help Center Search** (web, 4 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/13159) > A user opens the WhatsApp Help Center, browses support topics, and uses the search bar to find help articles. The entry point is the Help Center home screen with category navigation, topic cards, and popular articles. As the user types in the search field, autocomplete suggestions appear. If the query returns no matches, the interface shows an empty-state message. If the query matches content, a results list is shown with article titles and snippets. The flow assumes the user is not logged in to a chat product area; this is a public support site with language selection available. **Allbirds: Rate Assistant Response** (web, 3 screens, Marketplace) [Refero](https://refero.design/flows/14330) > A shopper is browsing a filtered product listing with the Wearallbirds AI assistant open. The assistant suggests that the current catalog does not contain products matching the shopper’s request and offers alternative browsing actions. The shopper provides negative feedback on the response by selecting the thumbs-down control. The flow begins with the assistant panel already open and ends with an inline confirmation that the response was disliked. The shopper is assumed to be on a desktop device; no sign-in or special permissions are visible or required. Steps: Review Assistant Response → Choose Negative Feedback → Feedback Confirmed **Deezer: Report lyrics problem** (web, 3 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/10696) > Happy path: while a track is playing with synchronized lyrics visible, the user opens a "Report a problem" dialog, selects one reason for the issue, and submits the report. Steps: Lyrics view (player) → Report modal — initial state → Report modal — selection made **Raycast: In-Chat Palindrome Explanation** (web, 6 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/10114) > Happy path: a user asks the in-app AI assistant to provide a palindrome-checking function and explanation (explicitly without using built-in reverse). The user composes and submits a prompt, the assistant shows a thinking state, then returns a structured reply that includes a thought-process note, multiple approaches, complexity analysis, a most-optimized code example, and a recommendation summary. Steps: Review complexity analysis and rationale → Reveal or invoke action controls → Submit prompt and wait for response → Receive structured explanation and thought process → View optimized code example → Read complexity summary and recommendation **Raycast: AI Prompt-to-Code Assistant** (web, 6 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/10113) > Happy path: A user visits the Raycast AI Lite landing view, types a programming prompt into the central input, selects the default model, and submits. The assistant returns a structured response in sections: an explicit 'Thought Process' card, then a titled answer with implementation examples (Python, Java) and a complexity analysis. The UI keeps the input area available under each response for follow-ups. Steps: Landing with Empty Prompt → Enter Prompt → Assistant Thought Process Appears → Implementation Example — Python → Implementation Example — Java → Complexity Analysis and Rationale ##### Task: test results **Vinyls: Turntable calibration** (ios, 4 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/8932) > Happy path: From the app home, the user opens the "Calibrate your turntable" shortcut, reads brief instructions, places a vinyl record on the turntable and starts the calibration check. The app listens/measures while showing a loading state and a prominent "Stop checking" button. When measurement completes it displays the detected RPM and a success message, plus options to retry or indicate the record uses a different (45 RPM) speed. Steps: Home — Shortcuts → Calibration Instructions → Measuring (Active) → Result — Detected RPM **Adobe Photoshop: Color balance adjustment** (ios, 6 screens, Design & Media Editing) [Refero](https://refero.design/flows/4801) > Happy-path: From the image editor canvas the user opens the "Add adjustment layer" modal, selects "Color balance", then uses the Color Balance control sheet to adjust Cyan/Red, Magenta/Green and Yellow/Blue sliders for a selected tonal range (Shadows, Midtones, Highlights). The image preview updates in real time while the user drags sliders. The user can toggle Preserve Luminosity, refine values, and then confirm with the check button to apply the adjustment layer and close the control sheet. Steps: Open adjustment type modal → Color balance sheet (defaults) → Adjust sliders (initial changes) → Refine tonal and channel adjustments → Toggle Preserve Luminosity and test result → Confirm or dismiss adjustment **Orbit: Notifications settings and test** (ios, 3 screens, Finance) [Refero](https://refero.design/flows/10000) > User is on the Notifications settings screen (accessed from the Settings tab). The screen shows four primary controls: an Enable Notifications toggle, a Notification Time row (shows current reminder time), a Send Test Notification action, and a Send Notification Logs action. In the happy path the user taps Send Test Notification while notifications are enabled: the UI shows a short countdown/progress state, then a success indicator once the test has completed. Steps: Notifications settings — overview → Send test — sending state → Send test — success state **Wispr Flow: Action Button Dictation Setup** (ios, 14 screens, AI Tool, Utility, Productivity) [Refero](https://refero.design/flows/13685) > This flow helps a user configure Flow’s Quick Dictation to Clipboard shortcut and assign it to the iPhone Action Button. The user starts in Flow Settings, opens the Action button shortcut setup, follows an onboarding walkthrough, adds the provided iCloud Shortcut, assigns it through iOS settings, and tests dictation. The flow ends with a confirmation screen explaining how to use the shortcut. The user is assumed to be on a compatible iPhone with access to iCloud Shortcuts and permission to change Action Button settings. **Remy: Remy first‑launch onboarding** (ios, 61 screens, Health & Wellness, AI Tool) [Refero](https://refero.design/flows/9761) > A first‑time launch onboarding and subscription flow for a sleep/health coach called “Remy”. Steps: App Tracking Transparency prompt → Launch / splash background → Opening headline → Narrative line → Narrative eye imagery → Phone silhouette cue → Messages are appearing → First chat bubble appears → User reply / input → Assistant 'eyes' reveal → Data & test summary → Assistant offering welcome gift → Offer detail (7 days for free) → Double‑click confirmation hint → App Store subscription sheet → Purchase success confirmation → Notification permission prompt → Conversational wrap‑up → Home tiles reveal → Completed onboarding state **Orbit: Notification Preferences — Set Reminder Time** (ios, 4 screens, Finance) [Refero](https://refero.design/flows/9999) > Happy path: A signed-in user in the app's Settings tab reviews notification preferences, opens a time picker to change the daily reminder time, confirms a new time, and returns to the Notifications settings screen with the updated time displayed. Steps: Notifications Settings Overview → Open Time Picker → Select New Time → Confirm Updated Time **Vocabulary: Custom Mix Builder** (ios, 6 screens, Education) [Refero](https://refero.design/flows/3803) > Happy path: A user opens the Categories screen, taps Make your own mix, uses an in-screen builder to filter/select word sources (All, Favorites, Collections, My words), parts of speech, topics, roots, tests and difficulty levels, then saves the selection and returns to a word/practice view where chosen words are presented. Steps: Categories overview → Mix builder — top level → Mix builder — parts of speech & topics → Mix builder — topic list selection → Mix builder — additional filters and collections → Practice view — single word card **District: Add Items to Favorites** (ios, 4 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12849) > A mobile overlay for adding items to a Favorites collection. The user starts from a Favorites screen and opens an Add items panel. They can search across item types such as restaurants, movies, events, stores, and activities, then select an item from the results. The panel can be dismissed with the close button, and the keyboard appears when the search field is active. **Vocabulary: Vocabulary practice drill** (ios, 5 screens, Education) [Refero](https://refero.design/flows/3822) > Mobile vocabulary practice flow (happy path). Entry: user starts a vocabulary practice session and is shown a single definition with three candidate words. The user selects one answer per question. The UI gives immediate, color-coded feedback (green = correct, red = incorrect) and updates a horizontal progress bar. The user repeats selections across multiple questions until the session ends. Exit: a completion card shows the final score and two primary actions: Review answers or Start again. Assumptions: single-answer multiple-choice interaction, mobile device, three choices presented per question, visual feedback is immediate and persistent until the next question or user action. A bottom control area contains a centered 'V' tile and a Search button (purpose not specified in screenshots). Steps: Question prompt → Immediate answer feedback → Next question with updated progress → Subsequent question and ongoing progress → Practice completion summary **Tinder: Manage Profile Photos** (ios, 3 screens, Dating & Relationships) [Refero](https://refero.design/flows/3441) > Happy-path: A signed-in mobile user edits their profile media by removing, replacing, or adding photos. Steps: Edit Media Grid → Photo Action Sheet → Updated Grid / Saved Edits **Pinterest: Image Text Editor** (ios, 4 screens, Reference, Inspiration Board, Social Networking, Lifestyle) [Refero](https://refero.design/flows/2524) > A single-screen image text editor where the user types a caption on top of a photo and adjusts styling in-place. Entry: user opens an image and invokes the text tool (keyboard appears). Core actions: type text, choose font/style and color from the toolbar, then adjust font size with a vertical slider while seeing a live preview. Exit: tap Done to apply the text or Cancel to discard. Steps: Enter text → Open size slider → Increase text size → Maximize and finalize layout **Adobe Photoshop: Layer blend and opacity edit** (ios, 8 screens, Design & Media Editing) [Refero](https://refero.design/flows/4783) > Happy path: user selects a photo layer, opens Layer properties, uses the Blend & opacity controls to try different blend modes and adjust opacity, previews the composite in the canvas, then applies the chosen settings. Steps: Select layer → Open Blend & opacity → Preview dark blend → Adjust opacity → Try Color Burn → Switch to Luminosity → Confirm blend settings → Resume editing / transform **Shazam: Location-based event filtering** (ios, 8 screens, Music Streaming, Utility) [Refero](https://refero.design/flows/9624) > Happy-path: User opens a list of concerts (initially showing results for the typed artist). They open the Location selector, try an invalid/unknown input (numeric) which returns no results, then enter a valid city name, select a suggested location, and return to the main list where results are refreshed for the chosen city. Entry: Find Concerts list with search field and active filters. Exit: Filtered event list for the selected city with a primary CTA to view all dates. Assumptions: user is on mobile (iOS-like chrome), location permission may be available ("Near Me" shown). User is performing a search for an artist and wants to limit results by location. Steps: Artist results (initial) → Location selector opened → User types an invalid/unknown query → No results feedback → Text query with suggestions → Location results list → Apply location — loading state → Filtered results displayed **Fuse: Adjust Slippage Tolerance** (ios, 5 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8742) > Happy path: A mobile user reviewing a small SOL→USDT swap opens the Slippage settings sheet, inspects the Dynamic option, toggles to Manual, adjusts the slippage percent to an acceptable value, saves the change, and returns to the Swap summary where the slippage value and the "Receive at least" estimate are updated. Steps: Swap summary → Slippage modal — Dynamic selected → Slippage modal — Manual selected (low value) → Slippage modal — Manual adjusted to acceptable value → Swap summary (updated) **Airtable: Configure and test send-email action** (web, 9 screens, Project Management, Collaboration Tools, Scheduling Tools, Paperwork, Productivity) [Refero](https://refero.design/flows/1636) > Happy path: From the Automations view the user opens an existing automation (Automation 1) that already has a trigger: "When a record is created" in the Tasks table. The user adds a "Send an email" action, fills the To/Subject/Message fields in the Properties panel, and uses the built-in Test Step to validate the action. An initial test fails due to a billing restriction for emailing non-collaborators; the user updates the recipient, reruns the test, sees a successful preview of the sent email, and reviews the stored test result. Entry: Automations list with an existing automation selected. Exit: Action configured and test run recorded (automation remains OFF). Assumptions: user is signed in with permission to edit automations, working on a desktop browser, the trigger is already defined and the automation itself is not turned on. Steps: Add action from action picker → Open Send email properties → View configuration fields → Fill email fields → Prepare test step → Handle test failure (billing restriction) → Update recipient and re-run test → View successful email preview → Review stored test results **Airtable: Email on Task Creation** (web, 2 screens, Project Management, Collaboration Tools, Scheduling Tools, Paperwork, Productivity) [Refero](https://refero.design/flows/1637) > Happy path: A user opens an automation editor, configures an action that sends an email when a new Task record is created, tests that action, and then activates the automation. Steps: Configure Send Email action → Activate automation **Preply: Placement test and speaking assessment** (web, 25 screens, Education) [Refero](https://refero.design/flows/9215) > Happy-path placement test that measures a user's English level through adaptive multiple-choice questions and two short spoken responses. Steps: Welcome / Start → Block intro: Questions 1-6 (A1) → Readiness prompt → Single-choice question display → Multiple questions on one page → Block completion / Next-level intro → Next-level question (word-chip selection) → Confirmation of selection → Final written block complete / Speaking intro → Speaking: overview and start → Recording in progress → Playback and delete controls → Second speaking question / Completion message → Results: Level summary → Mid-test question block (progressing) → Higher-difficulty question block → Feedback for incorrect selections → Near-completion encouragement → Final question block before speaking → Completion and review entrypoint **Preply: Vocab self-check and results** (web, 7 screens, Education) [Refero](https://refero.design/flows/9233) > Happy-path: A logged-in learner opens the Tests & quizzes section, launches the Vocab quiz, reads a short instruction modal, answers a series of Yes/No word-recognition questions while a progress bar advances, and then sees a results screen with score, estimated vocabulary size, and an option to add new words to their vocab list. Entry: user is signed in and on the Learn > Tests & quizzes area. Exit: user either adds words (leads to vocab management) or dismisses results and returns to the Tests & quizzes overview. Assumptions: desktop device, user has an existing account, the learner has permission to take quizzes and modify their vocab list. Steps: Tests & quizzes overview → Vocab quiz instructions modal → Quiz question (word prompt) → Quiz question (next prompt) → Quiz question (later prompt) → Results and follow-up CTA → Return to Tests & quizzes overview **Maze: Report step commenting** (web, 4 screens, Business, Communication Tools) [Refero](https://refero.design/flows/3674) > Happy path: A user views a single step inside a Maze report and opens the comment drawer to add feedback tied to that step. The user types a message, posts it, and sees the new comment appended to the comment thread (with author and timestamp). Steps: Report step display → Open comment drawer → Compose comment → Comment posted and thread view **Maze: Share test link** (web, 5 screens, Business, Communication Tools) [Refero](https://refero.design/flows/3666) > Happy path: A logged-in project member opens the Share area for a maze, selects the “Copy maze link” option, configures link options (tester tracking and unique-session), copies the public test link, and sees a confirmation and a new entry in the performance table. Steps: Share options overview → Link generation modal → Update link options → Copy confirmation → Performance table entry **n8n: AI summary automation** (web, 86 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9514) > Happy-path: A signed-in user opens their Personal workspace, selects or opens an 'AI Agent' workflow template, configures a data-fetch node (RSS), runs the workflow to inspect node output and logs, tests the AI summary node, then publishes the workflow version. Steps: Workspace templates and start → Open workflow editor (canvas overview) → Start execution (running state) → Execution success indicator → Open node parameter panel (Get World News) → View parsed response structure → Confirm repeated node output (detailed view) → Search within node output → Open Logs panel and inspect formatted output → View raw JSON output → Open Evaluations tab → Return to Editor and adjust layout → Initiate publish (open publish modal) → Add version description → Confirm publish (published modal) → Opt to not show published modal again → View production checklist → Canvas post-publish state (zoomed out) → Open Version History slideout → Return to Workflows list **Airtable: Record-created email notifier** (web, 27 screens, Project Management, Collaboration Tools, Scheduling Tools, Paperwork, Productivity) [Refero](https://refero.design/flows/1635) > Happy-path: user builds an automation that sends an email when a new task record is created. Entry: Automations workspace with an empty automation canvas. Exit: automation is configured, tested with a sample record, and turned on. Assumptions: user is signed in with access to the base and tables, has at least one table with records (Tasks), and permission to create/modify automations. Steps: Start automation - choose trigger → Trigger list → Trigger properties - require table → Reference data hint → Preview table grid → Change table to Tasks → Open record picker → Test step - selected record → Add action - open action palette → Add 'Send an email' action → Action properties - configure → Automation menu options → Turn automation on → Left panel - automation list → Open automation history → Revision history list → Revision snapshot modal → Open test automation pane → Choose test record → Run test and review results **Intercom: Create & Publish Custom Answer** (web, 15 screens, Business, Communication Tools) [Refero](https://refero.design/flows/669) > Happy path: A logged-in product/content manager navigates to the Custom Answers section, opens an existing draft answer, edits/inspects the answer canvas, uses the built-in tester to validate example questions (iterate until a question matches), confirms training examples, selects an audience for the answer, and saves/publishes the answer. Steps: Custom Answers list → Answer canvas overview → Zoom control hover → View menu (minimize/center) → Canvas collapsed → Open Test Answer modal → Enter a test question → Test result (no match) → Test result (match) after iteration → Editor stepper - language → Choose your questions (training progress) → Training examples list → Confirm steps and open audience → Choose audience preview → Attempt to close - unsaved changes modal **Maze: Delete Participant Session** (web, 4 screens, Business, Communication Tools) [Refero](https://refero.design/flows/3673) > Happy path: A researcher viewing test results selects a participant from the Participants list, opens their session details, deletes the session, confirms in a modal, and receives a success confirmation. Steps: Select participant → Invoke delete action → Confirm deletion → Show success and update list **Riverside: Test Recording, Upload & Playback** (web, 7 screens, Productivity, Entertainment, Collaboration Tools) [Refero](https://refero.design/flows/4665) > Happy path where a logged-in host prepares a two-person remote session, starts a test recording, waits for a short countdown, speaks a test phrase, waits while the clip uploads, and then reviews the recorded test with guidance and the option to retest or end. Entry: Host is in the studio lobby/preview with participant tiles and a primary Record control visible. Exit: Test clip is uploaded and playable; user either accepts the test or starts another test. Assumptions: user is signed into their account (header shows account controls), has granted camera and microphone permissions, is on a desktop or laptop browser with typical network access, and has at least one remote participant tile visible. Steps: Studio Preview → Record Options Menu → Recording Countdown → Test Recording Active → Uploading Recording → Test Guidance and Controls → Playback Review **Kit: Subject-line A/B testing** (web, 6 screens, Business, Communication Tools) [Refero](https://refero.design/flows/2248) > Happy path: A logged-in user composes a broadcast email, opens the Run an A/B test modal, adds a second subject variant, adjusts the A/B test duration, saves the variations and settings, and returns to the composer with the A/B test enabled. Steps: Compose broadcast → Open A/B test modal → Add subject variant → Open A/B settings → Adjust duration → Save and return to composer **Resend: Test Email from Broadcast Composer** (web, 5 screens, Development Tools, Communication Tools) [Refero](https://refero.design/flows/10619) > Happy-path summary: A signed-in user composes a broadcast (HTML editor on the left, rendered preview on the right), opens a "Send test email" modal, enters one or more recipient addresses, and sends a test. The modal provides a textarea for comma- or newline-separated addresses and action buttons (Send test email, Cancel). In this captured run the system attempts to send but returns a domain verification error via a toast. Steps: Broadcast composer (HTML + preview) → Test email modal — initial recipients → Test email modal — add additional recipients → Send action — in progress → Send blocked — domain verification error **Rox: Outreach email send sequence** (web, 3 screens, Business, Productivity, AI Tool) [Refero](https://refero.design/flows/10771) > Happy-path: A logged-in user opens a drafted outreach message for a selected account, reviews and optionally edits the message in an inline editor, optionally sends a test message, and then sends the final email. Steps: Draft review and suggested actions → Prepare or test send → Final send and confirmation **n8n: Evaluation Setup and Debugging** (web, 18 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9529) > Happy path: a user sets up the Evaluations workflow for an AI automation by wiring a test dataset, adding an Evaluation trigger, configuring dataset fetch parameters, adding a "Set outputs" action to write back outputs, and running tests in the editor. Steps: Evaluations checklist (start) → Open Editor — choose trigger → Select Evaluation trigger → Open Evaluation actions & triggers → Add 'On new Evaluation event' trigger → Configure dataset-fetch trigger (open modal) → Fill dataset parameters → Execute trigger test (error occurs) → Inspect error details → Open AI assistant session prompt → AI assistant analyzing error → AI assistant provides solution → Evaluations checklist updated (first step complete) → Add/Configure 'Set Outputs' evaluation action → Test 'Set Outputs' node by ID → Pin/unpin test data info → Checklist updated (second step complete) → Next steps or run in editor **Maze: Connect Live Site for Testing** (web, 11 screens, Business, Communication Tools) [Refero](https://refero.design/flows/3630) > Happy path: A user in the Build editor adds a 'Live Website Test' block to a test (maze), connects a target site (domain) by registering its name and URL, chooses how to install the code snippet (shares with a developer), copies the snippet, attempts verification, and saves the new site. Steps: Build editor (initial block selected) → Live Website Test configuration → Connect website — intro modal → Enter site name and URL (example) → Confirm site details (example updated) → Choose snippet installation method → Share code snippet → Verify website (pre-check) → Site saved (verification failed notification) → Return to Live Website Test (saved but unverified) → Block validation state and required fields ##### Task: therapy booking **Fresha: Search and filter results** (ios, 5 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4965) > User searches for massage services for a specific date/time and browses nearby venues in a combined list + map interface. Entry: search bar with query and date/time pre-filled. The listing shows venue cards with image, rating, address, services, prices and available time slots. From the listing the user opens the Price filter (modal slider), adjusts maximum price, applies the filter and the list updates (result count changes). The user can switch to a map view to see venues geographically. Assumptions: user is on a mobile device, already entered a search query and date/time in the top search bar, and has location enabled (distances shown). User is not required to be signed in for these browsing/filter interactions. Steps: Search results listing → Open Price filter → Adjust price slider → Apply price filter → Switch to map view **Fresha: Salon detail — booking & favorite** (ios, 2 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4949) > Mobile salon-detail screen showing a specific business (Poiz Beauty Salon). Entry: user is already viewing the salon’s detail screen (likely opened from a search or listing). The screen presents a large header image with carousel controls and three prominent action icons (back, share, favorite). Below the header is business metadata (name, verification badge, rating, review count, distance, neighborhood, opening hours), followed by a horizontal Services tab selector and a list of services. Each listed service shows title, duration, starting price and an inline “Book” button. A persistent primary CTA (“Book now”) at the bottom reveals the number of services available. Exit: user either books a service, favorites the salon, navigates back, or continues browsing services. Assumptions: this is a mobile portrait view; favoriting is available from the header heart icon; no authentication state or booking backend details are visible in the screenshots. Steps: Salon overview and services → Favorite confirmation **District: Save Theatre to Favourites** (ios, 4 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12828) > This is a movie showtimes listing screen for a selected film, with date tabs, filter/sort controls, theatre cards, showtime buttons, and a save/bookmark action for each theatre. The user enters from a film detail or search result and can browse theatres, open save options for a theatre, add it to an existing list or create a new list, then return to the listing with the theatre marked as saved and a confirmation toast. Steps: Showtimes Listing → Save Panel → Saved Confirmation → Saved State in Listing **Fresha: Discover and Select Service Time** (ios, 3 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4964) > User has searched for “Massage” with a selected date/time (shown in the header) and is browsing nearby venues in a mobile listing. The UI shows a map header, a control row with filters (filters, Sort, Price, Type), and a vertically scrollable list of venue cards. Each card includes an image, venue name, rating, distance, location, available services (name, duration, price) and selectable time chips. The happy path: user opens the search, optionally changes sorting, scans venue cards, and selects a service time to begin booking. Entry: search results list after a search query and date selected. Exit: user selects a time slot (leads to booking/checkout flow outside these screens). Assumptions: user is on a mobile device, likely signed in (bottom nav and booking actions visible), has location access (distances shown), and can tap controls and chips to interact. Steps: Search results — venue card → Sort options modal → Updated results and multi-service card **District: Swimming Slot Reservation** (ios, 4 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12875) > Happy path for booking a swimming session on mobile: the user starts on a swimming listings screen, filters and selects a pool, chooses a date, duration, and available time slot in a booking sheet, then proceeds to a review screen to confirm details and pay. **TidyCal: Hero Image Edit** (web, 4 screens, Business, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/12487) > A site editor view for a therapist profile/booking landing page. The user starts on the main page editor, opens the hero image editing modal, replaces the portrait image, optionally adjusts zoom/crop, and applies the change. The updated image is then reflected in the hero section of the page. Assumptions: the user is already in edit mode with permission to modify the page, and the device is a desktop browser. Steps: Landing Page Editor → Image Selection and Crop → Apply Image Update → Updated Landing Page Preview **TidyCal: Provider Account Setup** (web, 20 screens, Business, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/12482) > A new provider starts on TidyCal’s marketing landing screen and selects a Google-based signup path. After Google account authorization, the user completes a six-step setup sequence: profile details, calendar connection, meeting link, payment connection, service categories, and service description. TidyCal generates booking pages and prompts the user to create an account with email and password. After account creation, the user reaches the dashboard with booking links and growth actions. The screenshots show a desktop browser experience; some setup steps can be skipped using “I’ll do this later.”. Steps: TidyCal Landing → Profile Setup → Profile Details Entered → Calendar Connection → Google Account Selection → Google Sign-In Consent → Calendar Permissions → Permissions Confirmed → Meeting Link Setup → Meeting Link Entered → Payment Connection → Service Categories → Categories Confirmed → Service Description → Description Revised → Content Generation → Generated Booking Pages → Account Credentials → Account Submission → Provider Dashboard **Uber: Desktop ride booking with map pickup & scheduling** (web, 30 screens, Business, Real Estate) [Refero](https://refero.design/flows/451) > User begins on the Ride tab in a desktop web view and builds a ride request by selecting a pickup (via map or search), choosing a dropoff (or saved place), optionally adding stops, and scheduling a pickup time. The flow shows both happy-path interactions (pin placement, selecting suggested addresses, adding a saved place, setting date/time, route preview) and failure states (payment loading/error, reCAPTCHA). Entry: Ride tab is open and user is signed in (avatar visible). Exit: user either proceeds to a ride selection/payment step or an error/verification state (reCAPTCHA or payment error) blocks completion. Steps: Start ride - blank form → Choose pickup on map - place pin → Pin moved - confirm disabled → Pickup recognized by place name → Search typed into pickup field → Suggestions list - pickup options → Suggestion list (repeat state) → Dropoff selection - saved places visible → Choose a Place modal open → Add place modal - input form → Dropoff selected and route preview → Open scheduling panel → Date picker - choose date → Time slot selector open → Time adjusted → Proceed to ride selection (error state) → Verification step (reCAPTCHA) → Retry choose ride (still error/loading) → Add multiple stops - retry → Edge case: global map coordinates **TidyCal: Booking access and management** (web, 4 screens, Business, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/12483) > Happy path starts on a public marketing landing screen for TidyCal, where a new user can choose a sign-up method or navigate to login. After authentication, the user reaches the main dashboard and opens the Bookings section to review scheduled appointments. Steps: Landing and Sign Up → Login Form → Credentials Entered → Bookings Dashboard **time2book: Class reservation at studio** (web, 3 screens, Business, Fitness, Productivity) [Refero](https://refero.design/flows/6313) > Happy path: A user views a studio profile and weekly schedule, selects a class date/time, reviews the booking at checkout where "Pay in person" is presented, confirms the reservation, and sees a booking confirmation with details and a link to their schedule. Steps: Studio listing — schedule → Checkout — review and payment option → Booking confirmed — success summary **Fresha: Search results to venue details** (web, 6 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/2371) > Desktop search-to-book discovery: The user starts at a map-centric search results screen with a left-side list of matching venues and a large interactive map on the right. They can refine what they’re looking for using the top search bar (treatment, location, date, time) and an expanding left panel with treatment suggestions and categories. Selecting a different treatment/category updates the left list and map. Clicking a map pin shows a compact venue preview overlay; from there the user can open the venue’s full detail page which lists services, prices, booking buttons, reviews, address, hours and a mini-map. Steps: Initial search results (map + list) → Open treatment suggestions panel → Browse top categories → Filtered results for new category → Map pin selection and mini-card preview → Venue detail page **Care: Daycare Tour Booking** (web, 9 screens, Job Search, Lifestyle) [Refero](https://refero.design/flows/12237) > Happy path for a caregiver searching daycare options, viewing results in a split list/map layout, opening a provider detail panel, booking a tour through a multi-step modal, optionally adding a second daycare tour, and reaching a confirmation state. Steps: Search Results → Provider Details → Tour Scheduling → Tour Time Selection → Additional Daycare Suggestion → Alternative Time Choice → Second Time Confirmed → Booking Confirmation → Updated Search Results **District: Basketball Court Booking** (web, 4 screens, Entertainment, Food & Drink, Travel & Booking) [Refero](https://refero.design/flows/13015) > This booking journey starts on a basketball venue scheduling screen where the user selects a date, duration, time slot, and available court. After choosing a slot and confirming the court, the user proceeds to a booking review screen that summarizes the reservation, shows policies and offers, and presents the final amount before payment. Assumptions: the user is already logged in or identified, is booking on a desktop device, and has permission to reserve the shown court. ##### Task: triage questionnaire **Meta AI: Daily Trivia Quiz** (ios, 5 screens, AI Tool) [Refero](https://refero.design/flows/13383) > The user starts from a daily trivia quiz introduction and launches the July 28 edition. They answer a fixed set of three multiple-choice questions, receiving immediate correctness feedback and an explanation after each selection. Once all questions are answered, the user reviews the final score and can replay the set. The screens suggest the user is already in the host product and does not need to sign in. The sequence shows question 1, a post-answer state for question 1, a completed question 3 state, and the final results screen; the second question is not shown. **Slite: Knowledge-base question and sourced answer** (web, 4 screens, Paperwork, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/1706) > Happy path: A signed-in user opens the knowledge-base question composer, types a question, receives inline guidance if the query is too short, refines the question, and then gets a concise answer with an explicit source citation. Steps: Open question composer → Enter short query and get validation → Refine query → Receive answer with citation **Typeform: AI Lead Qualification Form Builder** (web, 10 screens, Productivity, Communication Tools) [Refero](https://refero.design/flows/5596) > Happy-path: A logged-in user creates a new lead-qualification form using the product's AI assistant, reviews and edits the generated questions and answer scoring, and returns to the workspace list with the new form created. Steps: Workspace listing (start) → Choose form creation method → AI creation modal - choose AI use case → Enter business information → Review / edit ideal customer profile → Generating questions (loading) → Form editor - review AI draft → Edit question and answer choices → Processing / preview overlay → Workspace listing (result) **Polaroid: Gift & Camera Recommendation Quiz** (web, 11 screens, Marketplace) [Refero](https://refero.design/flows/9471) > A desktop-first, multi-step product recommendation quiz launched from the Polaroid homepage hero. Entry: user clicks the START button in the 'Find the perfect Polaroid gift' hero. The quiz is presented as a 4-step questionnaire (progress indicator shown as X/4) with large selectable cards for each question. Each question uses visual option cards (radio or checkbox style) and a persistent footer with the global site header and footer. The user progresses by selecting options and pressing a Confirm button; the button is disabled until a valid selection is made. After the fourth question the site displays a brief loading state and then shows recommended product matches with match percentages and quick-actions (View Product / Add to Bag). Exit points: product detail page, add-to-bag/cart, or retry the quiz. Assumptions: desktop viewport, user is not required to be logged in, and the quiz is accessible from the site hero as an entry point. Steps: Quiz Entry (Hero) → Question 1 — Is this a gift? → Confirm Answer 1 → Question 2 — What kind of shots? → Confirm Answer 2 → Question 3 — Photo format preference → Confirm Answer 3 → Question 4 — Desired control level → Finalize & Loading → Results — Matched Products → Post-results CTAs ##### Task: wellbeing activities community **Wabi: Popular Category Browse** (ios, 3 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12069) > This is a mobile content-discovery screen titled "Popular". The user enters from a previous screen via the back stack and can browse a ranked list of templates or experiences grouped by category tabs. The user can switch between categories such as Top, Lifestyle, Health & Fitness, Entertainment, Family, and Education. Each selected category updates the list below to show matching items with thumbnail, title, short description, creator handle, and engagement counts. Assumption: the user is already on a browsing/discovery area and can tap a category chip or a list item to view more details or add it, though the destination is not shown here. Exit point is likely via back navigation or by opening a selected item. Steps: Top Category Listing → Health & Fitness Category → Education Category **Tripsy: Filtered Places List by Category** (ios, 3 screens, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/2356) > Mobile Places list (Quest Voyage) showing saved activities grouped by city. Entry: user opens the Places/Quest Voyage list on a mobile device. Happy path: user opens the Categories picker, selects a category (Amusement Park), and the list updates to show only matching activities. The interface shows chips for Category and Cities, a Sort control, city section headers with flags, activity rows (icon, title, address, optional date/note), and a persistent bottom bar displaying total activities and an add button. Exit: filtered list view with updated chip state and updated activity count. Assumptions: user is on a mobile device, viewing their saved/available places (logged-in state not shown), user has at least one activity that matches the category, and tapping an activity would open its details (detail view not shown). Steps: Places list - all categories → Categories picker open → Filtered list - category applied **Trip Way: Add suggested activity to trip** (ios, 7 screens, Travel & Booking) [Refero](https://refero.design/flows/8467) > Happy path: From an existing trip (Lakewood Retreat) the user opens the discover/search sheet, types a natural-language request (e.g., "Canoeing on Lake Louise"), receives an AI-generated suggestion card for a matching place, taps Add to create an activity prefilled with location and description, optionally edits fields, and saves it into the itinerary. Entry: trip details screen. Exit: updated trip timeline with the newly added activity. Assumptions: user is within a trip context (trip is open), user has permission to edit the trip, and device is a mobile phone with keyboard available. Steps: Trip overview → Open search overlay → Enter natural-language query → Loading and suggestion generation → Review suggestion card → Edit prefilled activity → Activity added to itinerary **Tripsy: Add New Activity** (ios, 2 screens, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/2334) > Mobile 'New Activity' modal that lets the user add a place or activity to their itinerary by browsing categories or searching by name. Steps: Browse categories and subcategories → Search results — no matches **Joi: Daily habit item actions** (ios, 3 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/4383) > Happy-path: user opens a daily habit view for a specific date (Feb 23, 2025). The screen shows a week date selector and a vertical list of habits/tasks for that day. The user taps a task row to reveal a bottom action sheet with choices (Complete, Edit, Remove). After choosing Complete the sheet dismisses and the task is visually marked done (checked icon + strikethrough). Steps: Daily habit list → Action sheet for selected habit → Item marked complete **Train Fitness: Workout Detail** (ios, 2 screens, Health & Wellness, Social Networking) [Refero](https://refero.design/flows/7141) > A logged-in user is viewing a single workout detail screen for a completed session. The screen shows the author, session title, three summary metrics (total time, volume, calories), and a scrollable workout log composed of exercise cards (thumbnail, title, sets). A persistent social action bar is shown over the bottom of the content with Like and Comment controls and counters. Steps: Workout Detail — default view → Social Action — like/comment **Breathwrk: Explore Search & Results** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/7738) > User is on the app's Explore tab (bottom navigation). The screen exposes a prominent search field at the top, category pills for quick filtering (Calm, Sleep, Awake, Balance, etc.), and content sections: Breathing Exercises, Daily Classes, and Guided Classes. Happy path: user taps the search field, types a query, and the list updates to show matching breathing exercises and guided classes. If nothing matches, the UI shows explicit "No Results" messaging. Entry: Explore tab open. Exit: user either clears the query or selects a result (cards are presented for selection/play). Assumptions: user is on a mobile device, logged-in state not required to see these lists, bottom Explore tab is active. Steps: Explore Landing → Search — No Results → Search — Results Displayed **Tripsy: Places — Quest Voyage** (ios, 2 screens, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/2360) > User is viewing a "Quest Voyage" Places module that lists activities/places grouped by city/country. The header exposes filter chips (Categories, Cities), sorting (Sort by Distance), and a Current Location control. Happy path: user opens the Places view, inspects a list of places (title, distance, snippet/address), and can tap a place to view details or use the + button to add a new activity. Alternative path shown: applying filters that yield no matches results in an empty state with a clear-filters CTA. Steps: Places List — Results → Filtered Empty State **Train Fitness: Workout Summary Editor** (ios, 2 screens, Health & Wellness, Social Networking) [Refero](https://refero.design/flows/7159) > Happy-path: A user reviews and finalizes workout metadata for a specific date (Saturday, Aug 16, 2025) on a mobile device. Entry: user opens the workout details for that date. The screen lists editable rows (Perceived Effort, Exercises, Workout Visibility, Duration & Data) and asks for quick feedback (thumbs up / thumbs down). Exit: user taps Done to finish and return to the previous screen. Assumptions: user is authenticated and viewing this on a mobile handset; rows with chevrons navigate to secondary screens or pickers; the main intent is to review/adjust metadata and provide quick feedback before leaving. Steps: Workout Details — initial → Workout Details — feedback selected **X: Community search and feed** (ios, 2 screens, Social Networking, Entertainment, News, Video Streaming) [Refero](https://refero.design/flows/2908) > Happy path: A signed-in mobile user views a community (The Design Sphere) feed, inspects pinned content and tags, then uses the community search to find posts mentioning a keyword ("saas"). Steps: Community landing — The Design Sphere → In-community search results **Maze: Participant recruitment purchase** (web, 9 screens, Business, Communication Tools) [Refero](https://refero.design/flows/3667) > Happy-path summary: From the project's Share area the user chooses the 'Hire panel participants' option to recruit external testers. The flow opens a multi-step modal with three main stages: Criteria (choose country, age, gender, and optional advanced filters), Settings (select number of responses and target devices), and Checkout (enter payment and confirm). A persistent right-hand order summary shows active participant pool size, selected criteria, settings, and cost that updates as the user changes inputs. Steps: Share options (recruit entry) → Loading panel filters → Criteria — basic targeting → Criteria — change country and filters → Advanced criteria selection → Settings — preset responses and device targets → Settings — custom responses and devices → Checkout — payment entry → Checkout — payment validation error **Nextdoor: Local Hobby Recommendations** (web, 3 screens, Communication Tools) [Refero](https://refero.design/flows/11438) > A logged-in neighbor opens the Ask section, reviews suggested local discovery topics, and selects a suggested prompt about places to learn a new hobby. The system processes the request, then displays a summarized answer based on neighbor posts along with sources, feedback controls, and suggested follow-up questions. The user can continue the conversation through the Ask input. Steps: Ask Discovery → Gathering Advice → Recommendation Summary **ALO: Class Finder Browsing** (web, 3 screens, Fitness, Lifestyle) [Refero](https://refero.design/flows/12370) > Happy path for browsing and narrowing a class catalog. The user enters the Class Finder from the Alo Wellness Club navigation, reviews a grid of class cards, and can refine results using left-side filters, removable filter chips, and a sort control. The user is assumed to be on desktop, already on the catalog experience, with no special permissions required. The flow exits when the user finds a class and would select a class card for more detail, or when filters narrow results to an empty state prompting further refinement. Steps: Class Catalog Browsing → No Results State → Refined Results List **WhatsApp: Create WhatsApp Community** (web, 8 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/13208) > A logged-in WhatsApp desktop user creates a new community, configures its name and icon, and lands in the newly created community with its default groups and announcements. Entry starts from the New chat area and the New community option. Exit ends in the community conversation view for the created community, ready to add members or start messaging. **WhatsApp: Add Member to Community Group** (web, 5 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/13209) > Happy path for adding a person to a community group in a messaging interface. The user starts in a community with a group chat open, opens the add-member dialog, selects one person from the contact list, confirms the addition, and sees success feedback once the person is added to both the community and the group. **B150: Community feed — browse categories** (web, 2 screens, Social Networking) [Refero](https://refero.design/flows/7975) > Two-step happy-path showing a logged-in user viewing a community feed and interacting with the left-hand community/category navigation. Entry: user is on the main community feed (logged in). The interface shows a three-column layout: left vertical navigation for communities and categories, center feed with posts (text + embedded video), and right rail with community manager, resources, leaderboard, and invites. The user inspects content and opens/expands the communities list to reveal more categories. Exit: user remains on the same feed with the left navigation expanded. Assumptions: user is authenticated, has access to multiple communities/categories, and the device is a desktop or wide-screen browser. Steps: Community feed (category selected) → Left navigation expanded (communities list) **Zendesk: Community discovery and search** (web, 2 screens, Business) [Refero](https://refero.design/flows/1359) > Happy path: An anonymous or signed-out user arrives at the Community landing page (desktop). From the landing page they can either browse topic tiles, view upcoming events, or run a free-text search using the prominent search field. Submitting a search opens a Search results page that lists matching community threads. The user can refine results using the left-side filters (By type and Topics) and navigate result pages via pagination. Steps: Community landing → Search results **On: Gear Recommendation Assistant** (web, 3 screens, Marketplace, Fitness, Lifestyle) [Refero](https://refero.design/flows/13832) > A visitor uses the on-site shopping assistant to request a product recommendation while browsing a fitness promotion. The assistant begins with an activity and gear-selection question. The user also rates the assistant experience as “Good,” submits the rating, and receives confirmation. The conversation then continues with a more specific question about gear for road running and hiking. The screenshots do not show the final recommendation or require an authenticated account. Steps: Start Recommendation Conversation → Rate Assistant Experience → Continue Activity Details ### research/clinic/REVERSIBILITY.md `research/clinic/REVERSIBILITY.md` · sha256 `6fcfbd1b6f0f60ccd8e1d90fc69e3f486b9fbc934bc6659ec22c6947f4e03614` #### Reversibility in health: the register Started 31 October 2026. What was done to a body cannot be undone, but its impact often can be, through a counter-process that reverses the harmful one. For every condition treated as irreversible, this register records: - the harmful process and the counter-process; - what starts the counter-process; - how much reversal has been measured, and where; - the window in which it works; - the evidence class. Evidence classes: - **OBSERVED:** a randomised trial or a cohort reported a measured reversal. - **HYPOTHESIS:** a plausible counter-process, not yet shown. - **OPEN:** no counter-process known yet; a research target. Every entry is information for clinicians and researchers, not advice to a patient. Decisions stay with an accountable clinician. ##### Reversals with measured evidence | Condition | Harmful process | Counter-process | What starts it | Measured reversal | Window | Evidence | |---|---|---|---|---|---|---| | Type 2 diabetes | Fat in liver and pancreas impairs insulin secretion and action | Removal of that fat restores first-phase insulin secretion | Intensive weight management: total diet replacement (825–853 kcal/day) for 12–20 weeks, then food reintroduction and maintenance | 46% (68 of 149) in remission at 12 months. More than a third at 24 months; 64% of those keeping off at least 10 kg. It returns with weight regain. | Diabetes under 6 years, not on insulin (trial population) | OBSERVED: DiRECT, cluster-randomised, Lancet 2018 and Lancet Diabetes & Endocrinology 2019. [Glasgow](https://www.gla.ac.uk/news/headline_639272_en.html), [WikiJournalClub](https://www.wikijournalclub.org/wiki/DiRECT) | | Fatty liver disease with fibrosis (NASH, now MASH) | Fat, inflammation and scarring in the liver | Fat clearance, resolution of inflammation, scar remodelling | Weight loss of at least 10% through lifestyle change | In those losing at least 10%: 90% NASH resolution and 45% fibrosis regression at 52 weeks | Mostly early fibrosis (89% at F0–F2). Advanced fibrosis needs weight loss held over a year. | OBSERVED: cohort of 293, Vilar-Gomez et al., Gastroenterology 2015 (association, not controlled). [Healio](https://www.healio.com/news/gastroenterology/20150624/lifestyleinduced-weight-loss-produces-improvement-in-nash-fibrosis), [Mount Sinai](https://scholars.mssm.edu/en/publications/weight-loss-through-lifestyle-modification-significantly-reduces--2/) | | Coronary atherosclerosis | Cholesterol-laden plaque builds in artery walls | Plaque regression as LDL falls | Very low LDL cholesterol: evolocumab added to a statin | Percent atheroma volume −0.95% against +0.05% on placebo. Regression in 64.3% against 47.3%, linear with achieved LDL down to 20 mg/dL. | Established disease on statins | OBSERVED: GLAGOV, randomised, JAMA 2016 (a surrogate measure; manufacturer-sponsored). [ACC](https://www.acc.org/Latest-in-Cardiology/Clinical-Trials/2016/11/14/00/18/GLAGOV) | | Heart failure with reduced ejection fraction | The ventricle enlarges and weakens | Reverse remodelling: smaller ventricle, stronger contraction | Sacubitril/valsartan, earlier and at higher dose | Ejection fraction +5 points at 6 months and +9.5 at 12 months; smaller ventricle and atrium | Earlier treatment after diagnosis does better (observational) | OBSERVED: PROVE-HF, single-arm, so not causal. [TCTMD](https://www.tctmd.com/news/new-prove-hf-data-may-reassure-clinicians-sacubitrilvalsartan-benefit-hfref), [ConsultantLive](https://www.consultantlive.com/view/james-januzzi-prove-hf-trial) | | Sight loss from diabetic macular oedema | Leaking vessels swell the retina's centre | Swelling resolves and vision returns | Anti-VEGF injections (aflibercept, ranibizumab, bevacizumab) | +10 to +12.8 letters at 2 years; +13 to +18 letters when starting vision was worse. Gains decline without ongoing care. | Centre-involved oedema; sustained follow-up | OBSERVED: DRCR.net Protocol T, randomised, 660 eyes. [Retina Today](https://retinatoday.com/articles/2015-mar/news), [Healio](https://www.healio.com/news/ophthalmology/20160415/antivegfs-improve-vision-at-2-years-in-eyes-with-diabetic-macular-edema), [AAO](https://aao.org/eyenet/article/drcr-net-five-year-outcomes-for-protocol-t) | | Osteoporosis | Bone resorption outpaces formation | Bone formation rises and density is rebuilt | Romosozumab for 12 months, then denosumab | Spine density +13.3% at 12 months and +17.6% at 24 months. New spine fractures −73%; non-vertebral fractures not significant at 12 months. | Postmenopausal osteoporosis | OBSERVED: FRAME, randomised, 7,180 women. [Amgen](https://www.amgen.com/newsroom/press-releases/2016/09/results-from-phase-3-frame-study-of-romosozumab-showed-significant-reductions-in-both-new-vertebral-and-clinical-fractures-in-postmenopausal-women-with-osteoporosis), [PMC](https://pmc.ncbi.nlm.nih.gov/articles/PMC9652294) | ##### Impact reversal for conditions that are part of who a person is For autism, ADHD, Down syndrome and intellectual disability the aim is not a cure: they are ways of being, not diseases. The aim is to reverse the specific impacts that cause harm, each one separately, so that the person keeps who they are and loses what hurts. Depression is different: remission itself is a measured outcome. Each row records impacts, not the person. Figures are to be verified against the primary trial or guideline before they are relied on. | Condition | Harmful impacts to reverse (not the person) | Counter-process | Evidence | |---|---|---|---| | Depression | Low mood, loss of interest, sleep and energy disturbance, functional loss | Remission: mood and function restored | OBSERVED that remission is achievable with psychological therapy, medication and exercise, singly or combined (NICE guideline NG222). Effect sizes per treatment: to verify against primary trials. | | Autism | Co-occurring anxiety; sleep problems; sensory overload; communication barriers in environments not built for autistic people | Anxiety reduced; sleep restored; environments and communication adapted (the impact reverses when the environment changes) | CLAIM, to verify: adapted CBT for anxiety in autistic people; melatonin for sleep onset (NICE CG170). Environment adaptation: HYPOTHESIS, to measure per person. | | ADHD | Impairment at school or work; accidents; emotional dysregulation; sleep problems | Symptoms and impairment reduced while treated; environment and routines adapted | CLAIM, to verify: medication reduces core symptoms in randomised trials (NICE NG87). The impact returns when treatment stops. | | Down syndrome | Treatable co-occurring conditions: hearing loss, hypothyroidism, obstructive sleep apnoea, heart defects; early-onset Alzheimer's risk | Each co-occurring condition found early and treated: hearing restored by aids or ear treatment, thyroid by thyroxine, sleep apnoea treated, heart repaired | OBSERVED for routine screening and treatment of co-occurring conditions (published health surveillance guidelines; to cite per condition). Alzheimer's risk: OPEN research target. | | Intellectual disability | Where the cause is metabolic or hormonal, the disability itself can be prevented; otherwise the impacts are barriers to learning, health and independence | Prevention where the cause is treatable; supported learning and adapted environments otherwise | OBSERVED: newborn screening with dietary treatment prevents intellectual disability in phenylketonuria (PKU), and screening with thyroxine prevents it in congenital hypothyroidism (NHS newborn blood spot programme). Other causes: impacts addressed by support, measured per person. | Measuring it: each impact gets its own measure (for example sleep onset time, anxiety score, hearing threshold, days in work or school), a baseline before, and the change after. The register reports which impacts reversed, for whom, and by how much. It never treats the person as the problem. ##### Research targets: conditions still treated as irreversible (OPEN) Each needs its harmful process and a candidate counter-process researched, then evidence, before any claim: - established dementia; - sensorineural hearing loss; - spinal cord injury; - loss of insulin-producing cells in type 1 diabetes; - advanced cirrhosis; - photoreceptor loss in inherited retinal disease; - brain tissue lost to stroke (function, not tissue, may return through rehabilitation); - chronic kidney disease beyond early stages. ##### How the register grows, deterministically 1. **List the conditions treated as irreversible:** from NHS conditions, NICE guidance and the Clinic capabilities. 2. **Record the harmful process** for each, from its clinical description. 3. **Name the candidate counter-process** (its opposite: clearance, regression, regeneration, remodelling, restoration of function) and what could start it. 4. **Search trial registries and the literature** for measured reversal of that counter-process. Record the effect size, window and evidence class. 5. **No evidence means OPEN:** it becomes a research target with the trial that would test it. Nothing is claimed until measured. ### research/clinic/SOCIAL-PRESCRIBING-AND-DETECTION.md `research/clinic/SOCIAL-PRESCRIBING-AND-DETECTION.md` · sha256 `6b41991afe790641488bd85df42d60896cf6e44fa5b82f31ed2db57d89716e87` #### EDGE Clinic: social prescribing, symptom checking, early detection, wearables and physical presence Research for EDGE Clinic, part of JudeAI | EDGE ID. Written 31 October 2026. EDGE Clinic is meant to be the front door before the GP (general practitioner, the family doctor). Anyone can come about anything, find out what the problem is, and be connected to the right service. That might be a clinical service, a screening test, or a non-clinical service such as debt advice, housing help or a walking group. It would use symptom checking, validated questionnaires, early detection (including retinal screening), wearables, edge hardware (devices that do the computing on the spot, not in a distant data centre), and physical presence: places and devices that come to the person. **This is research, not clinical or legal advice.** A qualified clinical safety officer, a regulatory specialist and a lawyer must check anything here before it shapes a product decision. **Companion file.** The competitor list, the wider rules and the reusable open-source systems are in `research/clinic/COMPETITORS-REGULATORS-REUSE.md`. This file does not repeat them in detail. --- ##### How to read the labels | Label | Meaning | |---|---| | **OBSERVED** | I read it myself, on the primary source page or document, during this research. | | **CLAIM** | A vendor, a charity, a trade paper, a news report or a study says it, and I saw it only in a search-result summary or a secondary page. I did not check it against the original. | | **AGREEMENT** | A law, regulation, standard, mandatory NHS (National Health Service) requirement or official guidance that is adopted and in force. | | **PREDICTION** | A target, plan, forecast or projection, including my own reasoning about what EDGE Clinic should do. It has not happened yet. | | **UNKNOWN** | Sources disagree, or I could not find a reliable answer. | - Every link was accessed on 2026-10-31. - Many CLAIM lines come from search-result summaries, because several primary sites (NICE, BMJ, UCL, NHS England news) blocked automated reading. Check the original before quoting a CLAIM figure in anything public. - My web search allowance ran out once during this research. After it reset I filled the waiting-list and obesity figures. The gaps that remain are listed in section 9. --- ##### 0. The twelve things to know first 1. **Social prescribing is now mass-scale in England.** An estimated 1.3 million patients were referred in 2023 alone, against an original target of 900,000 by 2023–24. OBSERVED. [Bu et al., Lancet Public Health 2025 (UCL Discovery PDF)](https://discovery.ucl.ac.uk/id/eprint/10214690/1/PIIS2468266725002178.pdf), accessed 2026-10-31. 2. **Its evidence is weaker than its popularity.** A 2024 meta-analysis of 12 trials (3,566 people) found "some improvements" but said "the quality of evidence remains poor". CLAIM. [CHCS summary of the review](https://www.chcs.org/resource-center-item/the-effectiveness-of-social-prescribing-in-the-management-of-long-term-conditions-in-community-based-adults-a-systematic-review-and-meta-analysis/), accessed 2026-10-31. 3. **Link workers are overloaded, and outcomes go unrecorded.** Almost 2 in 5 link workers hold more than 250 cases a year, which is NHS England's recommended maximum safe caseload. 24% never record patient outcomes. OBSERVED. [NASP Link Worker Survey 2025](https://socialprescribingacademy.org.uk/media/eznh1x3a/splw-survey-report-2025.pdf), accessed 2026-10-31. 4. **Symptom checkers are poor at naming the condition.** A systematic review found the first suggested diagnosis was right only 19% to 38% of the time, and triage advice was right 48.8% to 90.1% of the time. OBSERVED. [npj Digital Medicine 2022 review](https://pmc.ncbi.nlm.nih.gov/articles/PMC9385087/), accessed 2026-10-31. 5. **NHS 111 online is the reference service.** In 2024/25 it had 30.5 million sessions and 7.8 million completed triages. 9% of completed triages ended in "call an ambulance" and 11% in "go to an emergency or urgent treatment centre". OBSERVED. [NHS England 111 online flow diagram 2024/25](https://www.england.nhs.uk/statistics/wp-content/uploads/sites/2/2025/05/111-online-flow-diagram-2024-25.pdf), accessed 2026-10-31. 6. **AI grading in NHS diabetic eye screening is not yet recommended.** In May 2026 the UK National Screening Committee (UK NSC) consulted on new evidence. It found that some automated graders match a human first grader and "could be tested in real-world NHS services". OBSERVED. [UK NSC blog, 11 May 2026](https://nationalscreening.blog.gov.uk/2026/05/11/uk-nsc-consults-on-use-of-artificial-intelligence-in-the-diabetic-eye-screening-programme/), accessed 2026-10-31. 7. **Population screening for atrial fibrillation (AF, an irregular heart rhythm) is not recommended in the UK.** The review was last done in August 2019 and is waiting for the SAFER trial. OBSERVED. [UK NSC AF recommendation](https://view-health-screening-recommendations.service.gov.uk/atrial-fibrillation/), accessed 2026-10-31. 8. **Up to 5 million UK adults may have undiagnosed high blood pressure.** Finding and treating it could prevent over 125,000 heart attacks and strokes in the coming decade. OBSERVED (the British Heart Foundation's own estimate). [BHF, 25 July 2025](https://www.bhf.org.uk/what-we-do/news-from-the-bhf/news-archive/2025/july/slash-heart-attack-strokes-find-millions-undiagnosed-high-blood-pressure), accessed 2026-10-31. 9. **PHQ-9 and GAD-7, the standard depression and anxiety questionnaires, are free to use.** Pfizer holds the copyright and charges nothing. One expert source notes that the licence terms have "never been made explicit". CLAIM. [psyctc.org GAD-7 entry](https://www.psyctc.org/psyctc/glossary2/gad-7-general-anxiety-disorder-7/); [Wikipedia PHQ-9](https://en.wikipedia.org/wiki/PHQ-9), accessed 2026-10-31. 10. **Great Britain will keep accepting CE-marked medical devices long-term.** This was confirmed on 21 October 2026. The UKCA mark stays as the domestic route, and future primary legislation is promised for software and AI licensing. OBSERVED. [GOV.UK, 21 October 2026](https://www.gov.uk/government/news/government-confirms-indefinite-recognition-of-ce-marked-medical-devices), accessed 2026-10-31. 11. **The NHS clinical safety standards DCB0129 and DCB0160 are mandatory and being revised.** AI is named as a reason for the revision. The consultation ran from 29 June to 11 September 2026. OBSERVED. [NHS England consultation](https://www.engage.england.nhs.uk/patient-safety/national-review-of-clinical-risk-management-standa/), accessed 2026-10-31. 12. **The hospital waiting list in England was 7.4 million pathways (about 6.2 million people) at the end of August 2026.** 65.2% were waiting up to 18 weeks, against a 92% standard. OBSERVED. [NHS England RTT statistical press notice, 8 October 2026](https://www.england.nhs.uk/statistics/wp-content/uploads/sites/2/2026/10/Aug26-RTT-statistical-press-notice-PDF-389K-vjgRdG.pdf), accessed 2026-10-31. --- ##### 1. UK social prescribing **Social prescribing** connects people to non-clinical support for things that affect their health, such as loneliness, debt, housing, physical activity, arts and volunteering. A **social prescribing link worker (SPLW)** does the connecting. Most link workers are employed through **primary care networks (PCNs)**, which are groups of GP practices covering about 30,000 to 50,000 people, and funded by the **ARRS (Additional Roles Reimbursement Scheme)**. ###### 1.1 Scale | Fact | Label | Source | |---|---|---| | By the end of 2023, an estimated 9.4 million GP consultations in England had involved social prescribing codes. 5.5 million of them led to referrals. | OBSERVED | [Bu et al., Lancet Public Health 2025](https://discovery.ucl.ac.uk/id/eprint/10214690/1/PIIS2468266725002178.pdf), accessed 2026-10-31 | | An estimated 1.3 million patients were referred in 2023 alone. The paper says this "far exceed[s]" the target of 900,000 patients by 2023–24. | OBSERVED | [same paper](https://discovery.ucl.ac.uk/id/eprint/10214690/1/PIIS2468266725002178.pdf), accessed 2026-10-31 | | Who is offered it: 60% female and 23% from ethnic minority groups. The share from more deprived areas rose from 23% (2017) to 42% (2023). Refusal fell from 22% (2019) to 12% (2023). | OBSERVED | [same paper](https://discovery.ucl.ac.uk/id/eprint/10214690/1/PIIS2468266725002178.pdf), accessed 2026-10-31 | | The data cover 1.2 million patients in 1,736 GP practices (from CPRD, the Clinical Practice Research Datalink). The study was part-funded by NASP (the National Academy for Social Prescribing). | OBSERVED | [same paper](https://discovery.ucl.ac.uk/id/eprint/10214690/1/PIIS2468266725002178.pdf), accessed 2026-10-31 | | About 3,300 full-time-equivalent link workers work in primary care teams. A later NASP submission says "more than 3,500". | CLAIM | [NASP, via search summary](https://socialprescribingacademy.org.uk/resources/new-study-tracks-five-years-of-social-prescribing-growth/); [NASP consultation response](https://socialprescribingacademy.org.uk/media/tysbwu3w/nhs-consultation-nasp.pdf), accessed 2026-10-31 | | No official NHS England headcount of link workers for 2025 was found. | UNKNOWN | search summary of [NASP survey](https://socialprescribingacademy.org.uk/media/eznh1x3a/splw-survey-report-2025.pdf), accessed 2026-10-31 | ###### 1.2 What link workers actually do (NASP survey, February–March 2025, 411 respondents) All of the following are OBSERVED in the [NASP Link Worker Survey 2025](https://socialprescribingacademy.org.uk/media/eznh1x3a/splw-survey-report-2025.pdf), accessed 2026-10-31. - **Where referrals come from:** GPs 87%, other PCN staff 52%, self-referral 26%, pharmacists 9%. - **What people are referred for:** financial support 61%, mental health 60%, housing 55%. - **Main barrier:** "lack of provision/waiting lists". Cost is the main barrier to physical activity and arts. - **Caseloads:** 21% hold 300 or more cases a year. "Almost 2 in 5" hold more than 250, which the survey calls "the maximum safe caseload recommended by NHSE". Among link workers employed directly by PCNs, 46% hold 250 or more. - **How they work:** contact is mostly by phone (80%). 41% work remotely. Most support lasts 6–12 weeks (43%) or 3–6 months (28%). - **Retention:** 4 in 10 have considered leaving within the next year. Only 62% feel primary care teams value their role. - **Systems:** two-thirds track referrals in the GP clinical system (EMIS or SystmOne). The others named are Joy, Excel spreadsheets, in-house CRMs (customer relationship management systems), Access Elemental and Social RX. 4% have no tracking system at all. The exact percentage for each named product could not be read reliably from the PDF. (UNKNOWN for the per-product split.) - **Outcomes:** ONS4, the four Office for National Statistics wellbeing questions, is the most common outcome tool. Nearly 1 in 5 use no outcome tool. 24% never record outcomes. Only 4% report outcomes publicly. - **Coding:** 55% are familiar with SNOMED codes, the standard NHS clinical coding system. Only 43% are confident adding them. Only 51% are aware of the Social Prescribing Information Standard. - **What they asked for:** one respondent wanted "a dedicated, but simple case management system that had direct access to S1" (S1 is SystmOne). ###### 1.3 NHS England and government plans | Fact | Label | Source | |---|---|---| | The 10 Year Health Plan (July 2025) builds a "neighbourhood health service" with new roles, including Community Health Workers, and services such as debt advice in neighbourhood health centres. | CLAIM (NASP reading of the plan) | [NASP response to the 10 Year Health Plan](https://socialprescribingacademy.org.uk/resources/nasp-responds-to-10-year-health-plan-for-nhs/), accessed 2026-10-31 | | The PCN contract specification for 2026/27, which took effect on 1 April 2026, lists "changes to the minimum role requirements for Social Prescribing Link Workers in the ARRS" (Annex B.3). | AGREEMENT | [NHS England, Network Contract DES from April 2026](https://www.england.nhs.uk/long-read/primary-care-networks-network-contract-directed-enhanced-service-from-april-2026/), accessed 2026-10-31 | | What those 2026/27 minimum-role changes actually say. | UNKNOWN (not retrieved) | [NHS England, changes to the GP contract 2026/27](https://www.england.nhs.uk/long-read/changes-to-the-gp-contract-in-2026-27/), accessed 2026-10-31 | | Where link workers will sit in future workforce plans is listed as an open question. | CLAIM | [Westminster Health Forum agenda](https://www.westminsterforumprojects.co.uk/conference/Social-Prescrib-26), accessed 2026-10-31 | ###### 1.4 Outcome and cost evidence | Fact | Label | Source | |---|---|---| | NASP's November 2024 report pooled local evaluations from nine areas. Tameside and Glossop saw a 42.2% fall in GP appointments among 1,751 patients. Kent saw a 15.4–23.6% fall in A&E (accident and emergency) attendances among 5,908 patients. In Newcastle, secondary-care costs were 9% lower than a matched control group. | CLAIM (local before/after evaluations; mostly uncontrolled) | [NASP report page](https://socialprescribingacademy.org.uk/resources/new-report-shows-impact-of-social-prescribing-on-health-service-use-and-costs/); [Healthcare Conferences UK summary](https://healthcareconferencesuk.co.uk/news/2025-4-8/new-report-reveals-social-prescribing-reduces-strain-on-nhs-and-saves-costs), accessed 2026-10-31 | | NASP's 2023 rapid review gave a "social and economic return" of £2.14 to £8.56 for every £1 invested. | CLAIM | [NASP economic impact briefing](https://socialprescribingacademy.org.uk/media/bnjcszbz/nasp-briefing-economic-impact.pdf), accessed 2026-10-31 | | NASP's own briefing says evidence about the effect on secondary care was "inconclusive". | CLAIM | [NASP economic impact briefing](https://socialprescribingacademy.org.uk/media/bnjcszbz/nasp-briefing-economic-impact.pdf), accessed 2026-10-31 | | Long-term conditions meta-analysis: 12 trials, 3,566 people. "Some improvements", but evidence quality is "poor", with poor blinding and high dropout. | CLAIM | [CHCS summary](https://www.chcs.org/resource-center-item/the-effectiveness-of-social-prescribing-in-the-management-of-long-term-conditions-in-community-based-adults-a-systematic-review-and-meta-analysis/), accessed 2026-10-31 | | Sweden's health technology agency (SBU) found no low-to-moderate-bias systematic review of social prescribing for loneliness in people under 60. | CLAIM | [SBU enquiry response](https://www.sbu.se/en/publications/responses-from-the-sbu-enquiry-service/effects-of-social-prescribing-on-loneliness/), accessed 2026-10-31 | | A 2026 review says no systematic review has yet assessed the health economics of social prescribing internationally. | CLAIM | [Frontiers in Public Health 2026](https://public-pages-files-2025.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2026.1753435/pdf), accessed 2026-10-31 | **What this means for EDGE Clinic (PREDICTION).** The weakest links in social prescribing are the ones software can strengthen. They are referral tracking, consistent outcome measurement (ONS4 and the SNOMED codes), caseload triage and knowing what local services actually have space. The evidence gap is itself an opening, because a front door that records outcomes by default would produce the data the field lacks. ###### 1.5 Digital platforms for social prescribing | Platform | What is known | Label | Source | |---|---|---|---| | **Joy** | A digital social prescribing platform with case management for link workers. Leicester ICB (integrated care board) bought licences at 9p per registered patient. "Joy Connect" was rolling out to PCNs from September 2025. | CLAIM | [G-Cloud listing](https://www.applytosupply.digitalmarketplace.service.gov.uk/g-cloud/services/251287113942465); [Leicester council paper](https://cabinet.leicester.gov.uk/documents/s165854/Joy%20HWB.pdf), accessed 2026-10-31 | | **Access Elemental** | A referral and case management tool linked to EMIS. A third-party guide claims it covers 20 million+ patients and 4,400+ social prescribers. | CLAIM | [EMIS partner page](https://www.emishealth.com/partners/social-prescribing-tools/access-elemental-social-prescribing); [Plinth guide](https://www.plinth.org.uk/complete-guide/software-social-prescribing), accessed 2026-10-31 | | **GP clinical systems** (EMIS, SystmOne) | Used by two-thirds of link workers to track referrals. | OBSERVED | [NASP survey 2025](https://socialprescribingacademy.org.uk/media/eznh1x3a/splw-survey-report-2025.pdf), accessed 2026-10-31 | --- ##### 2. Symptom checkers and digital triage **Triage** means deciding how urgently someone needs care and where (999, A&E, urgent care, GP, pharmacy, self-care). **Diagnosis** means naming the condition. Tools are much better at the first than the second. ###### 2.1 The services | Service | What it is | Evidence and status | Label | Source | |---|---|---|---|---| | **NHS 111 online** | The NHS's own symptom triage, built on NHS Pathways and also inside the NHS App. For people aged 5 and over. | 2024/25: 30,464,500 sessions. 8,245,200 triages started and 7,794,000 completed. Median triage took 82 seconds. Outcomes: ambulance 704,600 (9%), emergency or urgent treatment centre 876,500 (11%), primary care 3,421,300 (44%), self-care 594,700 (8%), dental 452,100 (6%), urgent repeat prescription 1,404,200 (18%), clinical callback 279,900 (4%). 451,200 triages were abandoned. | OBSERVED | [NHS England flow diagram 2024/25](https://www.england.nhs.uk/statistics/wp-content/uploads/sites/2/2025/05/111-online-flow-diagram-2024-25.pdf); [NHS 111 page](https://www.nhs.uk/nhs-services/urgent-and-emergency-care-services/when-to-use-111/), accessed 2026-10-31 | | NHS 111 online, independent evaluation | NIHR (National Institute for Health and Care Research) multimethod study, 2023 | Evidence on safety and triage accuracy is "weak and variable". Online users fully followed the advice less often (67.5% vs 88%). Users had higher eHealth literacy, so "digital first" may widen inequalities. | CLAIM | [Turnbull et al., NIHR Journals Library](https://njl-admin.nihr.ac.uk/document/download/2038038); [Macquarie record](https://researchers.mq.edu.au/en/publications/a-multimethod-study-of-nhs-111-online/), accessed 2026-10-31 | | **Ada** (Ada Health, Germany) | Consumer app and enterprise "Ada Assess" | Ada Assess was certified Class IIa under the EU MDR (Medical Device Regulation) in December 2022. Ada's help page said the consumer app was Class I, with IIa pending. | CLAIM | [Ada announcement](https://about.ada.com/?p=10932); [Ada help page](https://Www.ada.com/help/is-ada-a-medical-device/), accessed 2026-10-31 | | **Healthily** (formerly Your.MD) | Consumer symptom checker and self-care app, founded in Oslo in 2013, now London-based. Raised a £23m Series A in 2020, led by Reckitt. | A New Zealand app directory lists the Your.MD app as discontinued in August 2024. Current status not confirmed. | CLAIM; UNKNOWN (current status) | [HTN, 2020](https://htn.co.uk/2020/10/05/london-health-tech-firm-raises-23m-series-a/); [Healthify NZ](https://healthify.nz/apps/y/yourmd-symptom-checker-app), accessed 2026-10-31 | | **Klinik** (Finland) | Online triage and patient-flow tool for GP practices, sold on G-Cloud | No independent accuracy evaluation found. | CLAIM; UNKNOWN (evidence) | [G-Cloud listing](https://www.applytosupply.digitalmarketplace.service.gov.uk/g-cloud/services/231000431749303), accessed 2026-10-31 | | **Patchs** | Online consultation and AI triage for GP practices | Registered with the MHRA (Medicines and Healthcare products Regulatory Agency) as a Class I device since October 2021. An impact study (at least 20 practices with AI triage vs 20 without) and an accuracy study were still recruiting in 2025. No accuracy results published. | CLAIM | [ISRCTN10744072](https://isrctn.com/ISRCTN10744072); [NCT07237919 listing](https://patient.info/research/trials/primary-care-nct07237919), accessed 2026-10-31 | | **eConsult** (owned by Huma since October 2024, see the companion file) | Online consultation forms with red-flag screening. Urgent symptoms are signposted to 999, A&E or 111. | Bristol evaluation of 36 practices: 70% of online consultations still led to a face-to-face or phone GP consultation, and staff cost per online consultation was higher than a face-to-face one. 14% of patients had heard nothing back after a week. eConsult's own claims of 60–70% "remote closure" are not independently verified. | CLAIM | [University of Bristol project](https://research-information.bris.ac.uk/en/projects/evaluating-an-online-consultation-system-in-gp-practices/); [eConsult blog](https://econsult.net/blog/the-red-flag), accessed 2026-10-31 | ###### 2.2 Babylon Health: why it failed | Fact | Label | Source | |---|---|---| | In 2018 Babylon said its AI scored 81% on MRCGP (Royal College of General Practitioners membership exam) questions, against a 72% average doctor pass mark. A Lancet letter by Fraser, Coiera and Wong said the evaluation "did not provide convincing evidence" that it beat doctors in real settings, and that it "might perform significantly worse". | CLAIM | [Digital Health](https://www.digitalhealth.net/2018/11/lancet-review-babylons-ai/); [MobiHealthNews](https://www.mobihealthnews.com/news/emea/researchers-question-babylon-claims-its-system-beat-doctors-rcgp-exam); [White Rose record](https://eprints.whiterose.ac.uk/id/eprint/138305/), accessed 2026-10-31 | | The advertising regulator (ASA) found some GP at Hand marketing misleading. In August 2022 Babylon cut GP at Hand in Birmingham because it was not profitable. | CLAIM | [Computing](https://www.computing.co.uk/news/4121796/babylon-health-future-nhs-goes-administration), accessed 2026-10-31 | | Babylon lost almost all of a $4.2bn valuation, was delisted from the New York Stock Exchange in July 2023, and its rescue merger with MindMaze collapsed. The UK business went into administration in August 2023, and GP at Hand was sold to eMed Healthcare. | CLAIM | [Sifted](https://sifted.eu/articles/babylon-administration-gp-at-hand-news); [UKTN](https://www.uktech.news/medtech/babylon-health-administration-20230831); [Digital Health](https://www.digitalhealth.net/2023/08/babylon-looks-to-sell-uk-business-amid-bankruptcy-fears/), accessed 2026-10-31 | | A 2022 review cites unverified media reports that Babylon's tool downplayed a breast lump and read a heart attack as a panic attack. | CLAIM (allegation) | [npj Digital Medicine 2022](https://pmc.ncbi.nlm.nih.gov/articles/PMC9385087/), accessed 2026-10-31 | **Lessons (PREDICTION, my reading of the above).** 1. Do not market accuracy you have not shown in real patients. 2. Do not depend on a funding model that loses money on high-need patients. 3. Do not let product claims run ahead of independent evaluation. 4. Do not grow faster than clinical governance. ###### 2.3 Accuracy studies, with numbers | Study | Design | Key numbers | Label | Source | |---|---|---|---|---| | Semigran et al., BMJ 2015 | 23 symptom checkers | Appropriate triage advice ranged from 33% to 78% across checkers. | CLAIM | [Harvard DASH](https://dash.harvard.edu/handle/1/17820908), accessed 2026-10-31 | | Gilbert et al., BMJ Open 2020 (Ada-affiliated authors) | 8 checkers against 7 GPs on 200 vignettes (written case scenarios) | Average first-diagnosis accuracy 26.1% (range 18–48%). Average triage accuracy 90.1% (range 80–97.8%). Ada says it came closest to the GPs. | OBSERVED (review table); CLAIM (Ada's framing) | [npj Digital Medicine 2022](https://pmc.ncbi.nlm.nih.gov/articles/PMC9385087/); [Ada summary](https://about.ada.com/?p=1185), accessed 2026-10-31 | | Hill, Sim and Mills, Medical Journal of Australia 2020 | 36 checkers, 48 vignettes | Correct diagnosis listed first 36% of the time, in the top 3 52%, in the top 10 58%. Correct triage 49% (95% CI 44–54%): emergency 63%, urgent 56%, non-urgent 30%, self-care 40%. Advice was "generally risk-averse". | CLAIM | [ECU repository](https://ro.ecu.edu.au/ecuworkspost2013/8431); [MJA news](https://www.mja.com.au/journal/2020/online-symptom-checkers-two-three-diagnoses-wrong), accessed 2026-10-31 | | Wallace et al., npj Digital Medicine 2022 (systematic review) | 10 studies, half with real patients | First-diagnosis accuracy 19–37.9%. Top-3 accuracy 33–58%. Triage accuracy 48.8–90.1%. One ophthalmic study found only 39% of emergencies triaged correctly. Most studies had unclear or high risk of bias. The authors say reliance "could pose significant patient safety hazards". | OBSERVED | [PMC9385087](https://pmc.ncbi.nlm.nih.gov/articles/PMC9385087/), accessed 2026-10-31 | | Riboli-Sasco et al., JMIR 2023 (systematic review) | 14 studies | 9 of 13 showed suboptimal triage accuracy. Accuracy varied greatly between checkers. | CLAIM | [PubMed 37266983](https://pubmed.ncbi.nlm.nih.gov/37266983/), accessed 2026-10-31 | | Oxford randomised study, 2026 (large language models) | 1,298 UK participants | Language models alone named conditions correctly 94.9% of the time, but people using them did no better than controls. | CLAIM (recorded in the companion file; the page blocked automated reading here) | [University of Oxford news](https://www.ox.ac.uk/news/2026-02-10-new-study-warns-risks-ai-chatbots-giving-medical-advice), accessed 2026-10-31 | ###### 2.4 What makes triage safe Each point below is drawn from the sources above. How it is combined into design rules for EDGE Clinic is my own reasoning (PREDICTION). 1. **Check for red flags first, before anything else.** eConsult and 111 online both screen for emergencies and signpost to 999, A&E or 111 at once. [eConsult blog](https://econsult.net/blog/the-red-flag); [Bristol project](https://research-information.bris.ac.uk/en/projects/evaluating-an-online-consultation-system-in-gp-practices/), accessed 2026-10-31. 2. **Under-triage (sending someone to less urgent care than they need) is the harm to design against.** Over-triage (more urgent than needed) costs money and capacity. The review names both. [npj Digital Medicine 2022](https://pmc.ncbi.nlm.nih.gov/articles/PMC9385087/), accessed 2026-10-31. Checkers tend to be "risk-averse". [ECU repository](https://ro.ecu.edu.au/ecuworkspost2013/8431), accessed 2026-10-31. 3. **Prove it on real patients, not only vignettes.** Most studies used vignettes entered by clinicians rather than by patients. [npj Digital Medicine 2022](https://pmc.ncbi.nlm.nih.gov/articles/PMC9385087/), accessed 2026-10-31. 4. **A good model does not guarantee a good outcome for people.** See the Oxford study above, and the lower rate of following advice online (67.5% vs 88%). [NIHR study](https://njl-admin.nihr.ac.uk/document/download/2038038), accessed 2026-10-31. 5. **Design for people who are not online.** "Digital first" may widen inequalities. [NIHR study](https://njl-admin.nihr.ac.uk/document/download/2038038), accessed 2026-10-31. 6. **Close the loop.** 14% of online consultation users had heard nothing after a week. [Bristol project](https://research-information.bris.ac.uk/en/projects/evaluating-an-online-consultation-system-in-gp-practices/), accessed 2026-10-31. 7. **Run formal clinical safety work under DCB0129 and DCB0160** (section 6), and monitor the tool after launch. --- ##### 3. Early detection ###### 3.1 The NHS diabetic eye screening programme (DESP) and AI graders | Fact | Label | Source | |---|---|---| | People aged 12 and over with diabetes are invited every 1 or 2 years. | OBSERVED | [UK NSC blog, May 2026](https://nationalscreening.blog.gov.uk/2026/05/11/uk-nsc-consults-on-use-of-artificial-intelligence-in-the-diabetic-eye-screening-programme/), accessed 2026-10-31 | | The UK NSC "does not currently recommend the incorporation of automated grading in the DES programme". It first looked at this in 2021, and the topic was resubmitted in 2024. | OBSERVED | [same](https://nationalscreening.blog.gov.uk/2026/05/11/uk-nsc-consults-on-use-of-artificial-intelligence-in-the-diabetic-eye-screening-programme/), accessed 2026-10-31 | | New evidence: most ARIAS (automated retinal image analysis systems) matched human graders at finding more severe disease. Most did not match them at identifying people with no disease, which means more false positives. Performance varied by manufacturer and setting, and in some cases by ethnicity or age. There were no new UK value-for-money studies. Some systems "could be tested in real-world NHS services". | OBSERVED | [same](https://nationalscreening.blog.gov.uk/2026/05/11/uk-nsc-consults-on-use-of-artificial-intelligence-in-the-diabetic-eye-screening-programme/), accessed 2026-10-31 | | The consultation closed on 3 August 2026. No updated recommendation was found by 31 October 2026. | OBSERVED (closing date); UNKNOWN (outcome) | [same](https://nationalscreening.blog.gov.uk/2026/05/11/uk-nsc-consults-on-use-of-artificial-intelligence-in-the-diabetic-eye-screening-programme/), accessed 2026-10-31 | | No product is "approved" for NHS England DESP grading. The UK NSC page names no products. | OBSERVED | [same](https://nationalscreening.blog.gov.uk/2026/05/11/uk-nsc-consults-on-use-of-artificial-intelligence-in-the-diabetic-eye-screening-programme/), accessed 2026-10-31 | | Scotland has used a rules-based autograder since 2011. Other UK programmes use human graders only. | CLAIM (from a 2025 Lancet Digital Health paper by committee members) | [Optician Online](https://www.opticianonline.net/content/news/uk-considers-introducing-ai-diabetic-screening-tool); [Gloucestershire repository](https://eprints.glos.ac.uk/14971/), accessed 2026-10-31 | | A 2020 evaluation in English screening found AI could triage images and could roughly halve the human grading workload. | CLAIM | [St George's, University of London](https://www.sgul.ac.uk/news/artificial-intelligence-could-help-prevent-sight-loss-in-people-with-diabetes-and-save-nhs-millions), accessed 2026-10-31 | **What this means for EDGE Clinic (PREDICTION).** EDGE Clinic cannot replace or bypass the NHS DESP. People with diabetes should still be sent to it. Retinal imaging at EDGE Clinic is best framed in one of two ways: as an evaluation run in partnership with an NHS programme, or as case-finding in people not yet in the programme, with any abnormal or ungradable image sent on to an NHS pathway. ###### 3.2 Atrial fibrillation (AF) | Fact | Label | Source | |---|---|---| | The UK NSC does not recommend AF screening. Last review: August 2019. Next review was estimated for 2023–24, after the SAFER trial. Reasons: it is unclear whether all AF types carry the same stroke risk, and treatment benefit for screen-detected AF is unknown. | OBSERVED | [UK NSC AF page](https://view-health-screening-recommendations.service.gov.uk/atrial-fibrillation/), accessed 2026-10-31 | | The SAFER trial plans more than 100,000 participants aged 70 and over, with about 5 years of follow-up. | CLAIM | [BMJ Open protocol paper](https://bmjopen.bmj.com/content/12/9/e065066); [ISRCTN16939438](https://www.isrctn.com/ISRCTN16939438), accessed 2026-10-31 | Wearable AF detection is in section 4. ###### 3.3 Hypertension (high blood pressure) | Fact | Label | Source | |---|---|---| | About 16 million UK adults have high blood pressure. Up to 5 million are undiagnosed. More than 125,000 heart attacks and strokes could be prevented in the coming decade by finding and treating it. | OBSERVED | [BHF, 25 July 2025](https://www.bhf.org.uk/what-we-do/news-from-the-bhf/news-archive/2025/july/slash-heart-attack-strokes-find-millions-undiagnosed-high-blood-pressure), accessed 2026-10-31 | | BHF elsewhere says up to 8 million people live with undiagnosed or uncontrolled high blood pressure. | CLAIM | [BHF Know Your Numbers](https://www.bhf.org.uk/yourbp), accessed 2026-10-31 | | The community pharmacy Hypertension Case-Finding Service has delivered almost 7 million blood pressure checks since October 2021. It targets people aged 40 and over with no diagnosis. | CLAIM | [Pharmaceutical Journal](https://pharmaceutical-journal.com/article/news/almost-7-million-blood-pressure-checks-delivered-in-community-pharmacies-since-service-began); [Community Pharmacy England](https://cpe.org.uk/national-pharmacy-services/advanced-services/hypertension-case-finding-service/), accessed 2026-10-31 | | In November 2024, GPs in England backed a motion to end pharmacy blood pressure checks. | CLAIM | [Pharmaceutical Journal](https://pharmaceutical-journal.com/?p=352559), accessed 2026-10-31 | ###### 3.4 Diabetes | Fact | Label | Source | |---|---|---| | Almost 6 million people in the UK live with diabetes. Over 4.7 million are diagnosed, and almost 1.3 million may have undiagnosed type 2 diabetes. | CLAIM | [Diabetes UK statistics](https://www.diabetes.org.uk/about-us/about-the-charity/our-strategy/statistics), accessed 2026-10-31 | | NHS Health Check uptake has never gone above 50%. A digital NHS Health Check was piloted in Norfolk, Medway and Lambeth from early 2025, aiming at about 1 million checks over 4 years. | CLAIM; PREDICTION (target) | [Digital Health](https://digitalhealth.net/2024/09/digital-nhs-health-checks-to-be-piloted-at-three-councils-in-2025); [GOV.UK press release](https://www.gov.uk/government/news/over-130000-people-to-benefit-from-life-saving-health-checks), accessed 2026-10-31 | ###### 3.5 Cancer symptoms | Fact | Label | Source | |---|---|---| | 58.7% of common cancers in England were diagnosed at stage 1 or 2, the highest on record (period ending August 2024). | CLAIM | [Euronews](https://www.euronews.com/health/2025/01/02/nearly-3-in-5-cancer-patients-in-england-are-now-diagnosed-at-an-early-stage-nhs-says); [Patient Safety Learning hub](https://www.pslhub.org/news/cancers-getting-diagnosed-at-earlier-stage-r7810/), accessed 2026-10-31 | | The NHS Long Term Plan target is 75% diagnosed at stage 1 or 2 by 2028. | PREDICTION (target) | [same Euronews article](https://www.euronews.com/health/2025/01/02/nearly-3-in-5-cancer-patients-in-england-are-now-diagnosed-at-an-early-stage-nhs-says), accessed 2026-10-31 | | In January 2023, 34% of cancers had no stage recorded. | CLAIM | [Nuffield Trust, cited in search summary](https://www.pslhub.org/blogs/entry/7852-cancers-getting-diagnosed-at-earlier-stage/), accessed 2026-10-31 | | The National Cancer Plan for England was published on 4 February 2026. It aims for 75% of people diagnosed in 2035 to survive five years, completion of lung screening rollout by 2030, and all cancer waiting time standards met by 2029. | PREDICTION (targets) | [LSHTM](https://www.lshtm.ac.uk/newsevents/news/2026/new-cancer-plan-aims-75-five-year-survival-england); [UCLH](https://www.uclh.nhs.uk/news/uclh-welcomes-national-cancer-plan), accessed 2026-10-31 | | In July 2026, 79.3% of people had cancer diagnosed or ruled out within 28 days of urgent referral, against an 80% target. | CLAIM | [Cancer Research UK](https://news.cancerresearchuk.org/2026/09/13/cancer-waiting-times-latest-updates-and-analysis/), accessed 2026-10-31 | ###### 3.6 Mental health and validated questionnaires | Questionnaire | Scoring | Risk items and actions | Licence | Label | Source | |---|---|---|---|---|---| | **PHQ-9** (Patient Health Questionnaire, 9 items, depression) | 0–27. Bands: 0–4 none or minimal, 5–9 mild, 10–14 moderate, 15–19 moderately severe, 20–27 severe. Original validation (580 people): 88% sensitivity and 88% specificity at a score of 10. | Item 9 asks about thoughts of being better off dead or of self-harm. Practitioner guidance says a positive answer needs immediate risk assessment, whatever the total score. | Copyright Pfizer, free access, "no permission required to reproduce, translate, display or distribute" (the line printed on the form). | CLAIM | [Wikipedia PHQ-9](https://en.wikipedia.org/wiki/PHQ-9); [findacode LOINC copyright line](https://www.findacode.com/loinc/44257-4--copyright-pfizer-inc-all-rights-reserved-developed-by-drs-robert-l-spitzer-janet-bw-williams-kurt-kroenke-and-colleagues-with-an-educational-grant-from-pfizer-inc-no-permission-required-to-reproduce-translate-display-or-distribute.html), accessed 2026-10-31 | | **GAD-7** (Generalised Anxiety Disorder, 7 items) | 0–21. Bands: 0–4 minimal, 5–9 mild, 10–14 moderate, 15+ severe. A cut-off of 10 gave 0.89 sensitivity and 0.82 specificity for GAD. | No self-harm item, so pair it with PHQ-9 item 9 or a dedicated risk question. | Copyright Pfizer, no charge. One expert says the licence terms "has never been made explicit". | CLAIM | [Wikipedia GAD-7](https://en.wikipedia.org/wiki/Generalized_Anxiety_Disorder_7); [psyctc.org](https://www.psyctc.org/psyctc/glossary2/gad-7-general-anxiety-disorder-7/), accessed 2026-10-31 | | Original GAD-7 paper | Spitzer, Kroenke, Williams and Löwe, Archives of Internal Medicine 2006;166(10):1092 | | | OBSERVED (citation) | [psyctc.org](https://www.psyctc.org/psyctc/glossary2/gad-7-general-anxiety-disorder-7/), accessed 2026-10-31 | | ONS4 (social prescribing outcome measure) | Four Office for National Statistics wellbeing questions | | | OBSERVED (most common tool among link workers) | [NASP survey 2025](https://socialprescribingacademy.org.uk/media/eznh1x3a/splw-survey-report-2025.pdf), accessed 2026-10-31 | Older printed copies of PRIME-MD (the earlier family the PHQ came from) carry "Reproduced with permission". Confirm the current terms on Pfizer's terms page before using these questionnaires commercially. The PHQ Screeners site links to [Pfizer's terms](https://www.pfizer.com/general/terms) but does not state them inline. CLAIM/UNKNOWN, accessed 2026-10-31. **Mental health crisis routes.** In England, anyone calling NHS 111 can "press 2" for urgent mental health support, rolled out nationally in 2024. CLAIM. [North London NHS FT](https://www.northlondonmentalhealth.nhs.uk/news/nhs-111-available-for-mental-health-across-north-london-5150); [Dereham Times](https://derehamtimes.co.uk/news/24548780.nhs-111-mental-health-crisis-line-set-national-roll-out), accessed 2026-10-31. Wales launched its own 111 press 2 in December 2022. CLAIM. [Welsh Government](https://www.gov.wales/nhs-111-press-2-big-step-forward-access-urgent-mental-health-support), accessed 2026-10-31. ###### 3.7 Cost of late diagnosis | Fact | Label | Source | |---|---|---| | Treating late-stage (stage 3 and 4) colon, rectal, lung and ovarian cancer costs the NHS about 2.5 times as much as early-stage treatment. Colon: £3,373 at stage 1 vs £12,519 at stage 4. Ovarian: £5,328 vs £15,081. Lung: £7,952 vs £13,078. Matching the best areas could save just under £210m, with 52,000+ people diagnosed earlier. (Incisive Health for CRUK, 2014. These are old prices.) | CLAIM | [Cancer Research UK, 2014](https://news.cancerresearchuk.org/2014/09/22/saving-lives-and-averting-costs-the-case-for-earlier-diagnosis-just-got-stronger/), accessed 2026-10-31 | | Undiagnosed high blood pressure leads to heart attacks and strokes that could be prevented (125,000+ in a decade). | OBSERVED | [BHF, July 2025](https://www.bhf.org.uk/what-we-do/news-from-the-bhf/news-archive/2025/july/slash-heart-attack-strokes-find-millions-undiagnosed-high-blood-pressure), accessed 2026-10-31 | | No current, UK-wide figure was found that costs late diagnosis across all conditions. | UNKNOWN | section 9 | --- ##### 4. Wearables and edge hardware ###### 4.1 What is clinically validated | Device or function | Evidence | Label | Source | |---|---|---|---| | **Apple Watch irregular rhythm notification (AF)** | Apple Heart Study (NEJM 2019): 419,297 participants, and 2,161 (0.52%) were notified. Positive predictive value (the share of alerts that were real) for simultaneous AF on an ECG (electrocardiogram) patch was 0.84. Of the returned patches, 34% confirmed AF. | CLAIM | [Stanford Medicine](https://med.stanford.edu/news/all-news/2019/11/through-apple-heart-study--stanford-medicine-researchers-show-we.html); [Catalogue of Bias](https://catalogofbias.org/?p=10338), accessed 2026-10-31 | | **Fitbit irregular heart rhythm detection** | Fitbit Heart Study (Circulation 2022): 455,699 enrolled, 4,728 (1%) notified. Positive predictive value 98.2% for simultaneous AF. AF was present in 32.2% of notified people who wore a patch. Reported sensitivity about 68%. The study did not test whether screening improved outcomes. | CLAIM | [Mass General](https://www.massgeneral.org/news/press-release/fitness-trackers-may-diagnose-potentially-irregular-heartbeats); [UMass repository](https://repository.escholarship.umassmed.edu/entities/publication/5153dbdc-9498-427e-9da4-84987a47b82c), accessed 2026-10-31 | | **Apple Watch hypertension notifications** | FDA (US Food and Drug Administration) cleared in September 2025 for adults aged 22+. Sensitivity 41.2%, specificity 92.3%. Labelling says no notification does not rule out hypertension. A January 2026 editorial in Hypertension said it is not suitable for large-scale screening. | CLAIM | [9to5Mac](https://9to5mac.com/2025/09/11/apple-watch-hypertension-alerts-feature-receives-fda-clearance/); [AAFP blog](https://www.aafp.org/pubs/afp/afp-community-blog/entry/smartwatch-screening-for-hypertension.html), accessed 2026-10-31 | | **Continuous glucose monitors (CGMs)** | NICE (National Institute for Health and Care Excellence) guideline NG28 was amended in June 2022. It offers intermittently scanned CGM to adults with type 2 diabetes on multiple daily insulin injections who meet set criteria, such as recurrent or severe hypoglycaemia (low blood sugar). Real-time CGM is allowed if it costs the same or less. | AGREEMENT (via secondary summaries) | [NICE resource impact report](https://nice.org.uk/guidance/ng18/resources/resource-impact-report-type-1-and-type-2-diabetes-and-continuous-glucose-monitoring-pdf-11020390813); [NHS London implementation document](https://www.england.nhs.uk/london/wp-content/uploads/sites/8/2025/10/1.London-T2DM-CGM-Implementation-document-1.0-1.pdf), accessed 2026-10-31 | | **Smart rings** | Samsung announced FDA-cleared sleep apnoea risk screening for the Galaxy Ring (July–August 2026). It flags risk and does not diagnose. Oura's ring does not have that clearance. Apple Watch sleep apnoea notification was cleared in September 2024. | CLAIM | [MobiHealthNews](https://www.mobihealthnews.com/news/samsung-announces-fda-cleared-smart-ring-sleep-apnea-risk); [Tom's Guide](https://www.tomsguide.com/wellness/smart-rings/the-samsung-galaxy-ring-is-about-to-be-the-worlds-first-fda-cleared-smart-ring-for-sleep-apnea), accessed 2026-10-31 | | **Home blood pressure cuffs (NHS BP@Home)** | More than 220,000 monitors distributed in England since October 2020. A DECIDE rapid evaluation (March 2024 to May 2025) reported better blood pressure control, case-finding and spread of workload as perceived impacts. | CLAIM | [Health Innovation Network](https://healthinnovationnetwork.com/projects/bp-home-evaluation/); [RAND](https://www.rand.org/pubs/external_publications/EP71121.html), accessed 2026-10-31 | Note: FDA clearance is a US decision. UK use needs UKCA or CE marking (section 6). ###### 4.2 NHS plans and pilots | Fact | Label | Source | |---|---|---| | The 10 Year Health Plan (July 2025) says the NHS App will read real-time data from wearables and connected devices. Commentators count 24 mentions of wearables in the plan, and it builds on the Wolverhampton Prevention Pilot. | CLAIM; PREDICTION (plan) | [HTN](https://htn.co.uk/2025/07/04/government-outlines-role-and-future-functionality-of-nhs-app-in-10-year-health-plan/); [Digital Health](https://www.digitalhealth.net/2025/07/what-does-the-10-year-plan-mean-for-digital/), accessed 2026-10-31 | | About 10 million people in the UK have smartwatches (2025). | CLAIM | [Digital Health](https://www.digitalhealth.net/2025/07/what-does-the-10-year-plan-mean-for-digital/), accessed 2026-10-31 | | No formal NHS wearables standard, device list or funded national rollout was found. | UNKNOWN | as above | ###### 4.3 Portable and smartphone fundus (retina) imaging | Fact | Label | Source | |---|---|---| | AEYE-DS, an autonomous AI for diabetic retinopathy, was FDA-cleared with the handheld Optomed Aurora camera (announced around April 2024). Sensitivity 92–93%, specificity 89–94%, and more than 99% of patients got a result. | CLAIM | [Ophthalmology Times](https://www.ophthalmologytimes.com/view/aeye-health-receives-fda-clearance-for-first-ever-fully-autonomous-ai-for-portable-dr-screening); [Retinal Physician](https://retinalphysician.com/issues/2024/julyaugust/fda-clears-ai-device-for-portable-diabetic-retinopathy-screening/), accessed 2026-10-31 | | Peek Retina smartphone adapter in Uganda: 286 patients against a tabletop camera, with "high sensitivity and specificity" for diabetic retinopathy (2022). | CLAIM | [PMC9447889](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9447889/), accessed 2026-10-31 | | Eyer handheld camera, multicentre study: 0.906 sensitivity and 0.808 specificity for referable diabetic retinopathy. | CLAIM | [PMC10289975](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10289975/), accessed 2026-10-31 | | A 2024 Indian study of a fundus-on-phone camera without pupil dilation said it was "not a feasible method", because of low sensitivity and many poor-quality images. | CLAIM | [Kerala Journal of Ophthalmology via DOAJ](https://doaj.org/article/e8f9a982281a4563abb621bc62a23d30), accessed 2026-10-31 | | Deep learning on smartphone retinal images (2020): accuracy 61% (iExaminer), 62% (D-Eye), 69% (Peek Retina) and 75% (iNview). | CLAIM | [PMC7336606](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7336606/), accessed 2026-10-31 | **What this means for EDGE Clinic (PREDICTION).** Image quality is what limits portable imaging. Any EDGE retinal unit needs an "ungradable image" route, such as a retake, dilation or referral, and must never report an ungradable image as "normal". ###### 4.4 The false-positive problem | Fact | Label | Source | |---|---|---| | A Mayo Clinic review of notes mentioning "Apple Watch" (published in JAMIA, the Journal of the American Medical Informatics Association): only 11.4% of patients got a clinically actionable diagnosis. Among people without symptoms who came in after an alert, 1 in 15 got an actionable heart diagnosis. The authors warn of overuse of care and anxiety among the "worried well". | CLAIM | [MobiHealthNews](https://www.mobihealthnews.com/news/apple-watchs-abnormal-pulse-feature-driving-many-unnecessary-healthcare-visits-mayo-clinic), accessed 2026-10-31 | | In the Apple Heart Study, only 34% of returned ECG patches confirmed AF after a notification. | CLAIM | [Stanford Medicine](https://med.stanford.edu/news/all-news/2019/11/through-apple-heart-study--stanford-medicine-researchers-show-we.html), accessed 2026-10-31 | | AI retinal graders are weaker at clearing people with no disease, which means more false positives. | OBSERVED | [UK NSC blog](https://nationalscreening.blog.gov.uk/2026/05/11/uk-nsc-consults-on-use-of-artificial-intelligence-in-the-diabetic-eye-screening-programme/), accessed 2026-10-31 | | Fitness-tracker data can become health data (special category data under UK GDPR) when it reveals or is used to infer health status. | OBSERVED | [ICO special category data](https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/lawful-basis/special-category-data/what-is-special-category-data/), accessed 2026-10-31 | **Design rule (PREDICTION).** Treat a wearable alert as a reason to measure properly, not as a diagnosis. For example, confirm a pulse alert with a 12-lead or single-lead ECG, or a blood pressure alert with a validated cuff or ambulatory monitoring. Track the confirmed-yield rate for each device so EDGE Clinic does not pass a flood of false alarms on to GPs. --- ##### 5. Physical presence: places and devices that come to the person | Service | Facts | Label | Source | |---|---|---|---| | **Community diagnostic centres (CDCs)** | About 170 operating in England. 10.93 million tests and scans since July 2024. 108 open 12 hours a day, 7 days a week. £237m for 36 centres, with 4 new and 32 expanded during 2026/27. | CLAIM | [NHS England CDC page](https://www.england.nhs.uk/long-read/community-diagnostic-centres/); [GOV.UK](https://gov.uk/government/news/nhs-patients-to-get-quicker-tests-and-scans-closer-to-home), accessed 2026-10-31 | | **Pharmacy First** (launched early 2024; seven common conditions treated by pharmacists) | 4,941,308 consultations claimed from February 2024 to January 2025. Sore throat was most common (835,679 in the first 13 months), then uncomplicated urinary tract infections (665,409). | CLAIM | [The Pharmacist](https://www.thepharmacist.co.uk/pharmacy-first/almost-five-million-pharmacy-first-consultations-delivered-in-first-year/); [Pharmaceutical Journal](https://pharmaceutical-journal.com/article/news/pharmacy-first-passes-5m-consultations-landmark), accessed 2026-10-31 | | Pharmacy blood pressure checks | Almost 7 million since October 2021 (see 3.3). | CLAIM | [Pharmaceutical Journal](https://pharmaceutical-journal.com/article/news/almost-7-million-blood-pressure-checks-delivered-in-community-pharmacies-since-service-began), accessed 2026-10-31 | | **Mobile units: Targeted Lung Health Checks** | About 70% of screening in the opening phase took place in mobile units. 76% of cancers found were at stage 1 or 2. By March 2025, 7,193 cancers had been found, 63.1% at stage 1 and 12.6% at stage 2. Under half (49.0%) completed the assessment. | CLAIM | [GOV.UK](https://gov.uk/government/news/new-lung-cancer-screening-roll-out-to-detect-cancer-sooner); [Roy Castle Lung Cancer Foundation](https://roycastle.org/over-three-quarters-of-lung-cancers-caught-early-through-lung-health-checks/), accessed 2026-10-31 | | **Neighbourhood health centres** | The first 27 were named on 26 March 2026 with up to £50m, to open by 2027. The plan is 250 in total, 120 of them by 2030. Further new-build sites to be announced. | PREDICTION (plan) | [Pharmaceutical Journal](https://pharmaceutical-journal.com/article/news/first-neighbourhood-health-centres-to-open-by-2027); [NHS England North East and Yorkshire](https://www.england.nhs.uk/north-east-yorkshire/2026/03/26/first-wave-of-neighbourhood-health-centres-in-north-east-and-yorkshire/), accessed 2026-10-31 | | **NHS Health Check** (community health check for ages 40–74) | Uptake has never been above 50%. Digital version piloted from 2025 (see 3.4). | CLAIM | [Digital Health](https://digitalhealth.net/2024/09/digital-nhs-health-checks-to-be-piloted-at-three-councils-in-2025), accessed 2026-10-31 | **What this means for EDGE Clinic (PREDICTION).** The NHS is already building physical front doors: pharmacies, CDCs and neighbourhood health centres. EDGE Clinic's physical presence is most useful where those doors do not reach, such as workplaces, faith and community venues, housing estates, rural areas, and people who never book. It should also hand people over into these NHS sites, not compete with them. --- ##### 6. Regulation and safety conventions ###### 6.1 Medical device regulation | Fact | Label | Source | |---|---|---| | The MHRA's Software and AI as a Medical Device Change Programme has 11 work packages. They cover qualification, classification (including an "airlock" sandbox), premarket rules, post-market monitoring, cyber security, AI rigour, interpretability and adaptive AI. The roadmap page was last updated 14 June 2023. | OBSERVED | [MHRA roadmap](https://www.gov.uk/government/publications/software-and-ai-as-a-medical-device-change-programme/software-and-ai-as-a-medical-device-change-programme-roadmap), accessed 2026-10-31 | | UK guidance on symptom checkers: they are Class I unless they "allow direct diagnosis", in which case they are Class IIa. For lay users, an indicative diagnosis may be enough to count as direct diagnosis. | AGREEMENT (guidance; read via search summary) | [MHRA software guidance, Appendix 1: symptom checkers](https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/1103566/Appendix_1_-_symptom_checkers.pdf), accessed 2026-10-31 | | **AI Airlock** (MHRA regulatory sandbox for AI as a medical device). Phase 2 cohort (published 16 October 2025): TORTUS, Numan, Panakeia, Octopath, NHS England Federated Data Platform Safe Summarisation, Eye2Gene and DeepX Health. The page says testing runs to March 2026 in one place and April 2026 in another. | OBSERVED | [GOV.UK AI Airlock phase 2 cohort](https://www.gov.uk/government/publications/ai-airlock-phase-2-cohort), accessed 2026-10-31 | | £3.6m over three years for the AI Airlock, announced April 2026. | CLAIM | [Fieldfisher](https://www.fieldfisher.com/en/insights/mhra-ai-airlock-from-pilot-sandbox-to-scaling-regulatory-pathway-for-ai-medical-devices), accessed 2026-10-31 | | **UKCA and CE.** On 21 October 2026 the government confirmed it will accept CE-marked devices long-term. The UKCA mark remains the domestic route. A statutory instrument will update UKCA requirements, and primary legislation will seek powers for a future licensing approach for software and AI. More than 90% of devices registered in Great Britain carry a CE mark. | OBSERVED | [GOV.UK, 21 October 2026](https://www.gov.uk/government/news/government-confirms-indefinite-recognition-of-ce-marked-medical-devices), accessed 2026-10-31 | | Draft UK medical device amending regulations (May 2026) and the MHRA's July 2026 statement on what counts as a device are in the companion file. | see companion | `research/clinic/COMPETITORS-REGULATORS-REUSE.md` | ###### 6.2 NHS assurance | Fact | Label | Source | |---|---|---| | **NICE Evidence Standards Framework (ESF)** for digital health technologies. Tier A has the lowest risk (system services), Tier B helps people manage their own health, and Tier C is the highest risk (treat, diagnose, guide care). There are 21 standards, 17 of which apply to every tier. The refresh renamed the tiers to avoid confusion with MHRA device classes and added requirements for AI and adaptive algorithms. A symptom checker or triage tool would likely sit in Tier C. | CLAIM (NICE pages blocked automated reading); PREDICTION (tier placement) | [NICE ESF user guide](https://www.nice.org.uk/corporate/ecd7/resources/evidence-standards-framework-for-digital-health-technologies-user-guide-9024813901/chapter/how-to-use-the-framework); [Fieldfisher](https://www.fieldfisher.com/en/insights/new-changes-to-evidence-standards-framework-and-digital-health-technologies), accessed 2026-10-31 | | **DTAC (Digital Technology Assessment Criteria)**: version 2 has been live since 24 February 2026, with about 25% fewer questions. The old form was retired by 6 April 2026. The guidance moved to digital.nhs.uk from 31 March 2026. | CLAIM | [NHS Innovation Service](https://innovation.nhs.uk/news/updated-nhs-england-digital-technology-assessment-criteria-dtac-form-and-guidance/); [Burges Salmon](https://www.burges-salmon.com/articles/102mnjh/new-nhs-digital-technology-assessment-criteria-what-health-tech-suppliers-need-t), accessed 2026-10-31 | | **DCB0129** (for manufacturers) and **DCB0160** (for organisations that deploy health IT) are "mandatory standards" for clinical risk management. They are under review (29 June to 11 September 2026), with AI named as a reason. | OBSERVED (AGREEMENT for the current versions) | [NHS England consultation](https://www.engage.england.nhs.uk/patient-safety/national-review-of-clinical-risk-management-standa/), accessed 2026-10-31 | ###### 6.3 Health data protection | Fact | Label | Source | |---|---|---| | Health data is special category data under UK GDPR (the UK General Data Protection Regulation). Processing it needs a lawful basis plus an Article 9 condition. Inferring health status through profiling also counts. | OBSERVED (AGREEMENT) | [ICO](https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/lawful-basis/special-category-data/what-is-special-category-data/), accessed 2026-10-31 | | Fitness-tracker data is not automatically health data, but becomes it when it reveals health status. The ICO says to "take more care with this type of information". | OBSERVED | [ICO](https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/lawful-basis/special-category-data/what-is-special-category-data/), accessed 2026-10-31 | | Whether a DPIA (data protection impact assessment) is required for EDGE Clinic's processing was not confirmed on the page read. Large-scale special category processing is usually treated as high risk. | UNKNOWN (verify with ICO DPIA guidance) | as above | ###### 6.4 Emergency signposting conventions (NHS wording) | Route | NHS wording | Label | Source | |---|---|---|---| | **999** | For "life-threatening emergencies like serious road traffic accidents, strokes and heart attacks". On the 111 page: "call 999 now" if someone has signs of a life-threatening illness or has tried to end their life. | OBSERVED (AGREEMENT as NHS guidance) | [NHS: when to call 999](https://www.nhs.uk/nhs-services/urgent-and-emergency-care-services/when-to-call-999/); [NHS: when to use 111](https://www.nhs.uk/nhs-services/urgent-and-emergency-care-services/when-to-use-111/), accessed 2026-10-31 | | **NHS 111** | "NHS 111 can help if you think you need medical help right now." Use it if you are unsure whether you need 999. 111 online is for ages 5+. Phone 111 for a child under 5, complex problems from an existing condition, an existing care plan, or end-of-life care. | OBSERVED | same two pages, accessed 2026-10-31 | | **A&E** | For "serious injuries and life-threatening emergencies". If you do not need A&E you "might be sent to a nearby urgent treatment centre". | OBSERVED | [NHS: when to go to A&E](https://www.nhs.uk/nhs-services/urgent-and-emergency-care-services/when-to-go-to-ae/), accessed 2026-10-31 | | **111 press 2** (mental health crisis, England) | Rolled out nationally in 2024. | CLAIM | [North London NHS FT](https://www.northlondonmentalhealth.nhs.uk/news/nhs-111-available-for-mental-health-across-north-london-5150), accessed 2026-10-31 | **Signposting rules for EDGE Clinic (PREDICTION).** 1. Show the emergency check before any login or questionnaire. 2. Use the NHS wording above word for word. 3. Never let AI downgrade a red-flag answer. 4. Offer 111 when unsure, and the phone route for children under 5. 5. Treat any positive answer to PHQ-9 item 9 as an immediate safety step (999, 111 press 2, or same-day clinical contact), not as a score. 6. Record every signposting event for the DCB0129 hazard log. --- ##### 7. Competitors and the gap The full competitor table is in the companion file. In summary: | Who | What they cover | What they leave out (PREDICTION, my reading) | |---|---|---| | NHS 111 online and NHS App | Free, national triage to the right urgency (section 2.1) | Social needs, early detection, follow-through after advice | | Online consultation tools (Accurx, eConsult/Huma, Patchs, Klinik) | Routing requests inside one GP practice | Anyone who is not registered or not online; non-clinical needs | | Consumer symptom checkers (Ada, Healthily) | Condition suggestions and triage | Validated questionnaires and detection hardware; NHS integration | | Social prescribing platforms (Joy, Access Elemental) | Link worker case management | Clinical triage and detection; mostly reached only after a GP referral (87% of referrals come from GPs; 26% are self-referrals, per the [NASP survey](https://socialprescribingacademy.org.uk/media/eznh1x3a/splw-survey-report-2025.pdf), accessed 2026-10-31) | | Pharmacies, CDCs, Lung Health Check trucks, NHS Health Checks | Physical tests and treatment | One condition or programme at a time; no single "anything" door | | Consumer wearables (Apple, Fitbit/Google, Samsung, Oura) | Continuous signals | Confirmation, clinical follow-up, NHS hand-off; false-positive burden | **The gap (PREDICTION).** No one service found in this research combines all of these: (a) open to anyone about anything, without a GP referral; (b) emergency-safe triage using NHS wording; (c) validated questionnaires; (d) early detection with confirmation steps; (e) social prescribing with outcome recording; (f) physical presence that comes to the person; and (g) closing the loop back to the GP record and NHS pathways. ###### Bare minimum vs ambitious (PREDICTION) | Area | Bare minimum (safe to pilot) | Ambitious | |---|---|---| | Emergency safety | Red-flag screen first; NHS 999/111/A&E wording; human escalation; audit log | Real-time clinical oversight; post-market safety monitoring under DCB0129/0160 | | Triage | Licensed, already-regulated triage engine or rule-based pathways; no free-text AI diagnosis | Own UKCA/CE Class IIa triage, proven on real patients and in an equity audit | | Questionnaires | PHQ-9, GAD-7 (with the item 9 safety route) and ONS4, free to use | Wider validated set, repeat measurement, outcome dashboards for link workers | | Social prescribing | Self-referral plus warm hand-off to the local link worker or service directory; outcomes recorded with SNOMED codes | Live service capacity, closed-loop referrals, outcome evidence that can be published (fills the evidence gap in 1.4) | | Detection | Blood pressure with a validated cuff; pulse check; referral into NHS Health Check and pharmacy pathways | Single-lead ECG for AF case-finding; HbA1c finger-prick test (a blood test for diabetes); portable retinal imaging evaluated in partnership with an NHS DESP | | Wearables | Accept readings as prompts only; confirm with a clinical-grade device | Device-specific yield tracking; NHS App wearable integration as the plan matures | | Physical presence | Pop-ups in community venues, linked to pharmacies and CDCs | Mobile units in deprived areas, matching the Lung Health Check model; co-located in neighbourhood health centres | | Assurance | DTAC v2, DCB0129, DPIA, NICE ESF evidence plan | AI Airlock or similar sandbox; independent evaluation published | --- ##### 8. The costliest burdens EDGE Clinic could address | Burden | Figure | Label | Source | |---|---|---|---| | **Being seen** | A record 388 million general practice appointments in the year to February 2026. | CLAIM | [NHS England](https://www.england.nhs.uk/2026/03/record-gp-access-figures), accessed 2026-10-31 | | Being seen: ease of contact | GP Patient Survey 2025: 52.9% found it easy to contact their practice by phone and 49.0% by the NHS App. 69.6% had a good overall experience of contacting their practice. | CLAIM (Ipsos; definitions differ from Nuffield Trust's figures) | [Ipsos](https://www.ipsos.com/en-uk/2025-gp-patient-survey-results-released); [Nuffield Trust](https://www.nuffieldtrust.org.uk/resource/access-to-gp-services), accessed 2026-10-31 | | **Waiting** | Hospital waiting list (RTT, referral to treatment) at the end of August 2026: 7.4 million pathways, about 6.2 million unique patients. 65.2% waiting up to 18 weeks, against a 92% standard. Median wait 12.2 weeks. 113,339 pathways waiting more than 52 weeks. Four trusts did not submit data, and the national figures include estimates for them. | OBSERVED | [NHS England RTT press notice, 8 October 2026](https://www.england.nhs.uk/statistics/wp-content/uploads/sites/2/2026/10/Aug26-RTT-statistical-press-notice-PDF-389K-vjgRdG.pdf), accessed 2026-10-31 | | Waiting: the month before | July 2026: about 7.33 million, with 65.3% within 18 weeks. | CLAIM | [COVER magazine](https://www.covermagazine.co.uk/news/4535557/nhs-waiting-list-increases-733m), accessed 2026-10-31 | | Waiting for cancer diagnosis | 79.3% within 28 days (July 2026), against an 80% target. | CLAIM | [Cancer Research UK](https://news.cancerresearchuk.org/2026/09/13/cancer-waiting-times-latest-updates-and-analysis/), accessed 2026-10-31 | | Waiting for social prescribing | The top barrier link workers report is "lack of provision/waiting lists". | OBSERVED | [NASP survey 2025](https://socialprescribingacademy.org.uk/media/eznh1x3a/splw-survey-report-2025.pdf), accessed 2026-10-31 | | **Missed appointments** | About 15.4 million general practice slots missed a year (5% of about 307 million), costing more than £216m at £30 each. 7.2 million of them were GP slots (1.2 million GP hours). This is an NHS England figure from January 2019 and is now old. | CLAIM | [NHS England, January 2019](https://england.nhs.uk/2019/01/missed-gp-appointments-costing-nhs-millions/); [PharmaTimes](https://pharmatimes.com/news/1_in_20_gp_appointments_missed_costing_216m_1273651/), accessed 2026-10-31 | | **Preventable illness: blood pressure** | Up to 5 million undiagnosed; more than 125,000 heart attacks and strokes preventable in a decade. | OBSERVED | [BHF, July 2025](https://www.bhf.org.uk/what-we-do/news-from-the-bhf/news-archive/2025/july/slash-heart-attack-strokes-find-millions-undiagnosed-high-blood-pressure), accessed 2026-10-31 | | Preventable illness: diabetes | About 1.3 million with undiagnosed type 2 diabetes. | CLAIM | [Diabetes UK](https://www.diabetes.org.uk/about-us/about-the-charity/our-strategy/statistics), accessed 2026-10-31 | | Preventable illness: late cancer | Late-stage treatment costs about 2.5 times as much (2014 prices). | CLAIM | [CRUK 2014](https://news.cancerresearchuk.org/2014/09/22/saving-lives-and-averting-costs-the-case-for-earlier-diagnosis-just-got-stronger/), accessed 2026-10-31 | | Preventable illness: obesity | A news report says the government's estimate of obesity's cost to the NHS rose from about £6.5bn a year (2023) to about £11.4bn (2025). Another submission cites about £11bn of direct NHS spending within £126bn of total UK costs. The government document itself was not found. | CLAIM | [AOL news](https://www.aol.com/news/true-cost-uk-obesity-crisis-182244721.html); [Association of Stroke Organisations submission](https://aso.org.uk/?p=2096), accessed 2026-10-31 | | Ill health and work | About 2.81 million people economically inactive because of long-term sickness. | CLAIM (secondary report of ONS data) | [COVER magazine](https://www.covermagazine.co.uk/news/4337255/81m-economically-inactive-sickness), accessed 2026-10-31 | | Wasted capacity in online consultation | 70% of online consultations still needed a GP phone or face-to-face contact, and staff cost per online consultation was higher. | CLAIM | [University of Bristol](https://research-information.bris.ac.uk/en/projects/evaluating-an-online-consultation-system-in-gp-practices/), accessed 2026-10-31 | **Where EDGE Clinic could plausibly move the numbers (PREDICTION).** - **Being seen:** sending social needs (money, housing, loneliness) to link workers and community services instead of GP slots. - **Preventable illness:** blood pressure, diabetes and AF case-finding in people who do not come forward, given that NHS Health Check uptake is below 50%. - **Missed appointments:** reminders and help with getting there for people referred onward. - **Waiting:** better-prepared referrals, with questionnaire scores and measurements attached. Each of these is a hypothesis to test, not a proven saving. --- ##### 9. Gaps and unknowns (what I could not confirm) - The content of the 2026/27 SPLW minimum-role changes (Annex B.3). UNKNOWN. - Any UK NSC decision on AI diabetic eye screening after the 3 August 2026 consultation closed. UNKNOWN. - Whether the SAFER trial has reported, and whether the UK NSC has updated its AF position. UNKNOWN. - Healthily's current trading status and device class, and any independent accuracy evaluation of Klinik. UNKNOWN. - Ada-specific figures in Gilbert et al. 2020: BMJ Open blocked automated reading. UNKNOWN. - A full-year 2025 Pharmacy First total, and an up-to-date cost of missed appointments (the £216m figure is from 2019). UNKNOWN. - A primary government source for the cost of obesity (only news and submissions found), and a current UK-wide cost of preventable illness overall. UNKNOWN. - The exact current NICE ESF tier names and DTAC v2 question set. NICE and NHS England pages blocked automated reading. CLAIM only. - Per-product shares of link workers' referral tracking tools: the PDF chart labels could not be matched reliably to their percentages. UNKNOWN. ### research/design/BASE44-V7BRO-STUDY.md `research/design/BASE44-V7BRO-STUDY.md` · sha256 `adf5b2032ff5585291c7b967d43e1a473be401f40a9c6dc23fe78ec09ac12e05` #### Base44 template "V7BRO": what it is made of Observed 1 November 2026, 04:05 to 04:12 UTC, by Claude-Jegudiel 00:5531102026-1. Observed facts only; every line names its source. ##### 1. What could and could not be read - Our copy of the template is Base44 app `6ae681783ae584c8c8890bf7`, named "V7BRO (Copy)" (`list_user_apps`). - The Base44 MCP file tools (`list_directory`, `read_file`, `grep`) refused: "External coding agents require the Builder plan or above … sandbox-bridge tools (direct file and shell access) require the Builder plan". The source tree of the copy was therefore **not** read. - `list_entity_schemas` on the copy returned four entities: `FinanceApplication`, `EvidenceEntry`, `EvidenceAnchor`, `User`. These are estate entities written into the copy by an earlier lane, not the template's own data model; the template keeps its grid data as static arrays inside its JavaScript (section 4). - Fallback used: the public template page `https://base44.com/templates/apps/view/v7bro` embeds the running app in an iframe at `https://v7bro.base44.app/?hide_badge=true` (page DOM, `iframe[src]`). That app's two build assets were downloaded and searched: - `https://v7bro.base44.app/assets/index-DLgXL3ED.js` (546,823 bytes, one Vite bundle) - `https://v7bro.base44.app/assets/index-DKJONCwP.css` (85,239 bytes) - The app presents itself as "V7 Go | Generative Business Process Automation" (document title); the template is a clone of V7 Labs' "Go" product UI. ##### 2. Libraries (string counts in the bundle, case-insensitive) | Candidate | Count | Reading | |---|---|---| | `react-dom.production.min.js`, `"18.3.1"` | 1, 2 | React 18.3.1 + ReactDOM | | `useNavigate` | 1 | React Router (data router API) | | `radix`, `data-radix-collection-item`, `--radix-toast-swipe-move-x`, `Symbol.for("radix-ui")` | 24 | Radix UI primitives (toast, dismissable layer, collection) as used by shadcn/ui | | `lucide-*` class names (`lucide-chevron-down`, `lucide-folder`, `lucide-search`, …) | 73 | Lucide icons | | `tw-`, `--radius`, `--primary`, `--sidebar-*` in the CSS | 871, 28, 35, 48 | Tailwind with shadcn/ui CSS variables | | `react-flow`, `@xyflow`, `xyflow`, `reactflow` | 0 | **No React Flow / xyflow** | | `tanstack`, `react-table`, `ag-grid`, `glide` | 0 | **No table or data-grid library** | | `zustand`, `d3-zoom`, `d3-drag`, `dagre`, `elkjs`, `framer-motion`, `dnd-kit`, `react-pdf`, `pdfjs`, `recharts`, `tiptap` | 0 | none of these | | `base44`, `base44_access_token`, `base44SharedInstances`, `x-base44-connector-error`, `serverUrl:"https://base44.app"`, `window.open(t,"base44_auth",…)` | 34 | the Base44 JS SDK: auth popup, access-token storage, analytics session, backend-function fetch | Conclusion: the grid and the node canvas are **hand-written React components**, not libraries. The only platform-specific code is the Base44 SDK (auth, analytics, `functions.fetch`). The grid rows, columns, stages, nodes and their positions are literal arrays in the bundle. ##### 3. Layout measurements (DOM of the running app at 1440×900) - Left sidebar 240 px wide, `bg: hsl(0 0% 98%)` (`--sidebar-background`), items 14 px; tree = workspace switcher → Home / Knowledge / Admin → "Agents" with search, new-folder and plus → folders (Real Estate & PE, Insurance & Risk, …) → projects (CIM Triage & Valuation, FNOL Claims Processing, …) → under the open project: Build / Review / Visualize. - Main column `x=240, w=1200`: a 56 px header row (`.v7-section-header`, min-height 3.5rem, padding .75rem 1.25rem; breadcrumb "All projects / FNOL Claims Processing · Live", segmented Build | Review | Visualize, "Export data" primary control), then a stage-tab row (Main, Human Review, Outdated Information, + New stage), then a toolbar row (sort, filter, "⚠ 2", "Look up row, column…" search, Add Row, Add Column), then the grid. - Grid: no ``; ARIA `role=row` (4), `role=columnheader` (10), `role=gridcell` (30). Row height **36 px**. Column widths are declared per column in the data (`width:"190px"`, `"140px"`, …). Row 0 is a check-box column, then a row number. - Record detail (click a row): an overlay `absolute inset-0 z-30 flex bg-background` over the grid; left pane `w-[42%] border-r bg-muted` (504 px) with a header (close ✕, previous ▲, next ▼, "…" menu) and the file name, body `p-5` holding a document preview card; right pane `min-w-0 flex-1 overflow-y-auto` (696 px) with the fields in two columns. - Pills (computed styles): `font-size 11px; padding 2px 6px; border-radius 4px`. Observed pairs: rose `bg rgb(253,236,235) / fg rgb(202,58,49)` (Vehicle Collision, Special Investigations (SIU), Yes - Urgent); green `bg rgb(229,246,236) / fg rgb(22,163,74)` (Verified, Auto-Approved); cyan `bg rgb(224,244,248) / fg rgb(8,145,178)` (Water Damage); orange `bg rgb(255,239,224) / fg rgb(229,107,0)` (Kitchen Fire). - Controls: `.v7-control` height 2rem, 13 px/500, 1 px border `hsl(var(--border))`, radius `calc(var(--radius) - 2px)`; `.v7-primary-control` same with foreground background; `.v7-icon-tile` 2rem square; `.v7-surface-card` 1 px border, radius `var(--radius)`; `.v7-menu` popover with two shadows. - Fonts: `Inter, -apple-system, system-ui, sans-serif` from Google Fonts (Inter, Newsreader, JetBrains Mono). Tokens: `--radius .5rem; --primary 262 83% 58% (violet); --brand 24 100% 50% (orange); --border 240 6% 90%; --muted 0 0% 96%`. ##### 4. Data model inside the bundle - Columns: `{key, label, type, tool, icon, width, format?, groundable?, citationId?}` with `type ∈ {file, text, select, number, python, collection, multiselect, json}` and `tool ∈ {Input, AI Model, Python Tool, Workflow}`. Examples (FNOL): `file "Claim File / FNOL" Input`, `claimId text AI Model`, `incidentType "Peril Type" select AI Model`, `lossEstimate number AI Model format currency groundable citationId "11"`, `fraudRiskScore "Fraud Risk (0-100)" python "Python Tool"`, `policyActive "Policy Verified" select`, `routingStage "Routing Stage" select Workflow`, `humanReview "Review Required" select`. CIM Triage: `tenants "Major Tenants" collection`, `condition multiselect`, `roiPotential select`, `extractJson json`. - Rows: literal objects, e.g. `condition:["Excellent Condition","Deferred Maintenance"], roiPotential:"Excellent", recommendation:"Strong Buy", state:"Under Review", tenants:[{name:"FedEx Ground Distribution", sqft:180000, expiration:"2036-04", annualRent:1620000}, …]`. - Cell rendering: `h == null || h === "" ? "—" : h === "calculating" ? "Calculating..." : number → format`; the calculating cell is a 13 px muted span "Calculating..." with a 6 px pulsing primary dot. - Stages: `[{id:"main",name:"Main",color:"#1a1a1a"},{id:"human-review",name:"Human Review",color:"#e56b00"},{id:"outdated",name:"Outdated Information",color:"#743ee4"}]`. - Visualize nodes: `{id, label, caption ("Source" | "Condition" | "Stage"), count, color, kind ("stage" | "condition"), x, y, width, prompt?}`, e.g. `main "Main" Source x:22 y:235 width:190`, `human-review Stage #e56b00 x:625 y:88 width:210`, `outdated Stage #743ee4 x:625 y:272`, `auto-approve Stage #16a34a x:625 y:432`, a condition node `"Strong Buy" caption Condition x:305 y:422 width:225`. Positions are hard-coded. - Edges are cubic Béziers drawn in an SVG: `` `M ${x1} ${y1} C ${x1+R} ${y1}, ${x2-R} ${y2}, ${x2} ${y2}` ``; new edges get ids `` `edge-${Date.now()}` ``. - Node chrome: `relative flex h-10 cursor-grab items-center gap-2 rounded-lg border bg-card px-3` with a 9 px uppercase caption above; the input dot is a `button[data-flow-target]` `absolute -left-1.5 top-1/2 h-3 w-3 rounded-full border-2` ("Connect into …") and node drag starts on `onPointerDown` of the node body; `onPointerDown` occurs 10 times in the bundle, `wheel` 10, `zoom` 7, `getBoundingClientRect` 3, `requestAnimationFrame` 2. No minimap, no fit-view (0 occurrences of `minimap`, `fitView`). - Inspector labels present once each: "Input", "Model" (value shown in mono: "Concierge"), "Reasoning" (a control reading "Low"), "Tool" (twice). ##### 5. What this means for Estate Studio - Nothing needs to be vendored to match the original: it uses no grid or graph library. Estate Studio's grid, detail view and canvas are therefore plain JavaScript served from our own origin (the site's CSP is `default-src 'self'`, so no inline scripts and no CDN), with the same interaction grammar: 36 px rows, ARIA grid roles, pills 11 px / 2×6 px / 4 px radius, a 42 % document pane, Bézier edges from an output dot to an input dot, node drag on pointer down, pan and zoom on the canvas, an inspector with Source / Tool / Reasoning / Model / Input. - Where the original hard-codes rows, columns, stages and node positions, ours read the capability's unit (`spec/units`), its records in the estate record store (`objects`), its journeys and actions, and the TOK dependency index; computed columns call the estate's reasoning plug and never show a value the model did not return. ### research/edge-consortium/GOVERNED-INTELLIGENCE-DEGREE.md `research/edge-consortium/GOVERNED-INTELLIGENCE-DEGREE.md` · sha256 `c93c23ededa1d05e46aa6285fed9d0d4528720b52002617290e2d6ff2b1d8380` #### Governed intelligence: the degree Generated by `tools/degree.py`. It is awarded by **The Edge Consortium** and offered with leading institutions. It runs as an integrated BSc (Hons) of 360 credits over FHEQ levels 4 to 6, and an MSc of 180 credits at level 7: 30 modules. The program checks that every level sums exactly, that every module is assessed by real work, and that all 47 topics the owner asked for are taught. **Governed intelligence** is the discipline of taking an intention to a verified outcome by the shortest path, with people and agents. It is deterministic where it can be, measured where it cannot be, reversible by design, and glass box throughout. Maintaining and reinforcing living systems built this way is the profession it prepares for. ##### Assessment: real work, never recall - **C**: Coursework: a working specification with proof. It must pass the estate's own checks (no dead controls, every journey linked, every outcome measured) and be reproduced by a second examiner from its evidence alone. - **L**: Live challenge under pressure: a real, unseen problem solved in a fixed time, observed and recorded. - **S**: Live spec-driven development: intention to a running, checked capability in the session, with every prompt and decision recorded. - **A**: Live audio and speech: viva, listening and transcription. The student explains, defends and answers; listens to a recorded case and transcribes and summarises it; marked from the recording. Marking: - **Two examiners and a recording** for every live and spoken assessment. - **Deterministic checks** for every specification. - **Reproduction from evidence by a second examiner** for every mark. A student passes a module by producing work that works, with proof. ##### Level 4 (BSc year 1) | Code | Module | Credits | What it teaches | What the student builds or does | Domain of the case | Assessed by | |---|---|---|---|---|---|---| | GI101 | Intelligence, philosophy and responsibility | 15 | What intelligence is, human and machine: goals, intentions, outcomes, knowledge and judgement; epistemology and philosophy of mind; ethics of autonomy, accountability and consent; who answers when an agent acts. | A map of one real decision from intention to outcome, naming where intelligence was used or missing and who is accountable at each step. | health and medicine | coursework with proof, live audio and speech | | GI102 | Logic, sets and proof | 15 | Propositional and predicate logic, sets, relations, induction, proof by construction; specifications as logical statements. | Proofs that a small specification is complete (every state reachable, every control leads somewhere) and consistent. | finance and money | coursework with proof, live challenge | | GI103 | Probability and Bayesian reasoning | 15 | Probability, Bayes' theorem, priors and evidence, likelihood, base rates, calibration, decision under uncertainty. | A calibrated forecast record and a Bayesian update of a real screening decision from published sensitivity and prevalence. | society and people | coursework with proof, live challenge | | GI104 | Laws, agreements, decisions, facts and truth | 15 | The difference between a law, an agreement, a decision, an observation, a claim and the truth; who may make each, how each is evidenced and how each changes. | A case file sorting every statement in a real dispute into its class, with its source and authority. | government and public services | coursework with proof, live audio and speech | | GI105 | Writing, reading, listening and speaking for people and agents | 15 | Plain language, instructions an agent can execute without guessing, reading for exactness, active listening, transcription. | One procedure written so that a person and an agent both carry it out identically, tested on both. | education and research | coursework with proof, live spec-driven development, live audio and speech | | GI106 | Smart internet technologies: networks and connected systems (taught by LSBU SITHub) | 15 | Taught by London South Bank University's Smart Internet Technologies Hub (SITHub, Prof. Tasos Dagiuklas) from its own research and LSBU's 5G testbed: 5G and 6G, non-terrestrial networks and UAV coverage, the Internet of Things, knowledge-defined networking to cloud-native AI, software-defined networking and self-organising Routing as a Service (self-optimising, self-configuring, self-healing), multi-access edge computing with containerised services, predictive migration and caching, intent-based networking orchestrated by agents (intent agent, junior agents, a senior agent that validates and generates infrastructure as code), O-RAN analytics, positioning as a service, and quality of experience. | A network decision made by an agent in SITHub's instrumented world, simulation and emulation first (VFCSIM, Mininet-WiFi, GNS3, Docker Open vSwitch, OpenDaylight): the environment, not the student, measures the outcome (latency, throughput, service-level breaches, utilisation, energy), and the agent must beat static and heuristic baselines on unseen scenarios. | defence and security | coursework with proof, live challenge | | GI107 | Digital societies | 15 | Sociology of digital life: power, inclusion, trust, platforms, work, exclusion by design and frictionless public systems. | Field study of one public service journey: where people are blocked, who is excluded, and why. | environment and energy | coursework with proof, live audio and speech | | GI108 | Prototyping studio: intention to outcome | 15 | Prototyping from intention: the shortest path from what a person wants to a verified outcome. | A working prototype of one capability, measured from intention to outcome, with its evidence. | law and justice | live spec-driven development, live challenge | ##### Level 5 (BSc year 2) | Code | Module | Credits | What it teaches | What the student builds or does | Domain of the case | Assessed by | |---|---|---|---|---|---|---| | GI201 | Error-rate control and failure mining | 15 | Type I and II errors, family-wise error and false discovery rate (Benjamini-Hochberg), online error-rate control as data arrives (alpha-investing, LORD, LORD++, SAFFRON, ADDIS), guaranteed miss rates, failure mining from records. | An online evaluation of a live signal stream with its false discovery rate pre-registered (for example 0.05) and held, reproduced exactly by an independent re-run from sealed data and code (white paper 1); and a failure-mining report from a real incident log. | health and medicine | coursework with proof, live challenge | | GI202 | Systems engineering, bottleneck and blocker science | 15 | Systems engineering, reliability, FMEA, fault trees, queueing (Little's law), theory of constraints, bottlenecks and blockers removed at the root. | A journey measured step by step; its bottleneck and blockers traced to root causes and removed by a system change, then re-measured. | finance and money | coursework with proof, live challenge | | GI203 | Secure by design | 15 | Cybersecurity mapped to the Cyber Security Body of Knowledge: threat modelling, identity, least privilege, provenance, supply chain. | A threat model and the controls that close it, each proven by a test that fails without the control. | society and people | coursework with proof, live challenge | | GI204 | Cloud architecture, efficient by design and zero waste | 15 | Cloud architecture, cost, energy and carbon; deterministic computation first; measuring words, prompts, tokens, cost and carbon per verified outcome. | An architecture with its waste measured and driven to zero, with a blocking gate that proves it. | government and public services | coursework with proof, live spec-driven development | | GI205 | Psychology, counselling and human domain expertise | 15 | Cognition, behaviour, motivation, counselling skills (active listening, motivational interviewing), safeguarding; how to learn an expert's domain. | A recorded counselling-skills session and a domain-expert interview turned into a specification the expert signs off. | education and research | live audio and speech, coursework with proof | | GI206 | User journeys and accessibility | 15 | User research, journeys, click-by-click design with no dead controls, WCAG 2.2, inclusive design for every ability. | A full journey specification with every control linked and an accessibility audit passed with assistive technology users. | defence and security | coursework with proof, live spec-driven development | | GI207 | Communicating with autonomous agents; prompt engineering with proof | 15 | Mandates, tools, limits, expiry; authority envelopes the agent cannot widen, hold and decide, signed receipts, kill switch and rollback (white paper 2); the twelve universal agentic capabilities from intent capture to verification; instructions an agent cannot misread; frontier benchmarks and their traps (time horizon, pass^k, contamination). | A private evaluation suite (UK AI Security Institute Inspect) for one real task: pass^k over repeated trials, cost and latency per successful task, the harness and version reported, and a proof that the agent never games the verifier. | environment and energy | live spec-driven development, live challenge | | GI208 | Scope discipline: scope creep as raw material | 15 | Scope freezing, change requests, triage of new requests, the whole map, increments; scope creep captured as raw material, never lost and never silently admitted. | A scope register run through a live stream of changing requests: every request captured, classified, costed and admitted or deferred. | law and justice | live challenge, coursework with proof | ##### Level 6 (BSc year 3) | Code | Module | Credits | What it teaches | What the student builds or does | Domain of the case | Assessed by | |---|---|---|---|---|---|---| | GI300 | The whole map: every domain and how they connect | 15 | Health and disease, money, society, government, education and research, defence and security, environment and energy, law: each domain's processes, failure modes, measures and impact, and how a change in one moves the others. | A whole-map model tracing one real event through every domain, with its impact measured in each. Passing it is the gate to specialising. | the specialism | coursework with proof, live challenge, live audio and speech | | GI301 | Glass box: observability, traceability and reproducibility | 15 | Evidence records, digests, provenance, observability, traceability from intention to effect, reproducing any decision from its evidence. | A system whose every decision a second examiner reproduces exactly from its evidence. | the specialism | coursework with proof, live spec-driven development | | GI302 | Assurance by design, capability manufacturing and continual improvement | 15 | Specifications that compile, checks proven to catch their faults, generators that turn one declaration into a complete capability, parity as a minimum then exceeding it, measured improvement. | A generator and its checks, each check proven to fail on a planted fault, and one measured improvement cycle. | the specialism | live spec-driven development, coursework with proof | | GI303 | Reversibility, proximity settlement and the analog human switch | 15 | Reversible by design, provisional value, grace windows, in-person sign-off on registered devices, and the analog human act for irreversible decisions, including defence. | A replay of a real theft or misuse through a design that reverses it, and the specification of an analog switch with identical backups. | the specialism | coursework with proof, live challenge | | GI305 | Records, breach and incident response | 15 | Record keeping, retention, change control, breach notification, incident command, post-incident learning. | A live incident exercise: detect, contain, notify, recover and report within the statutory clocks, from records alone. | the specialism | live challenge, live audio and speech | | GI310 | Specialist pathway I (15 credits, from the chosen pathway) | 15 | The first specialist module of the student's pathway, taken only after passing GI300 (see Specialising). | As set by the pathway module. | the specialism | coursework with proof, live challenge | | GI307 | Capstone in the specialism: a governed capability, live (30 credits) | 30 | Everything above, applied in the chosen domain (or across domains for the generalist award): a real organisation's intention taken to a verified outcome under scope change and pressure. | A governed capability for a real client in the specialism, built live, defended in a viva, and reproduced by a second examiner. | the specialism | live spec-driven development, live challenge, live audio and speech, coursework with proof | ##### Level 7 (MSc) | Code | Module | Credits | What it teaches | What the student builds or does | Domain of the case | Assessed by | |---|---|---|---|---|---|---| | GI401 | Frictionless, gapless public systems | 15 | Designing whole societies' services so bottlenecks and blockers are removed by design; the shortest intention to outcome at national scale. | A redesign of one national service, measured before and after on a simulation from public data. | the specialism | coursework with proof, live challenge | | GI402 | Research methods and the science of reproducibility | 15 | Experimental design, pre-registration, replication, selective inference, measurement validity. | A pre-registered study with its analysis code reproduced by another student. | the specialism | coursework with proof, live spec-driven development | | GI403 | Governance of autonomous agents: law, regulation and standards | 15 | Agent identity, mandates, liability, regulation of AI and finance, management-system standards, and approval by regulators. | A regulator-ready evidence pack for an agent deployment, reviewed live by a practising reviewer. | the specialism | coursework with proof, live audio and speech | | GI405 | Leading governed intelligence | 15 | Leading teams of people and agents, scope discipline at organisation scale, maintenance and innovation as a service. | A live leadership challenge: a team of people and agents brought through a changing scope to a delivered outcome. | the specialism | live challenge, live audio and speech | | GI406 | Advanced Bayesian decision and failure science | 15 | Hierarchical models, decision theory, value of information, extreme events, near-miss mining. | A decision model for a real high-stakes choice with its value of information computed. | the specialism | coursework with proof, live challenge | | GI410 | Specialist pathway II (15 credits, from the chosen pathway) | 15 | The advanced specialist module of the student's pathway. | As set by the pathway module. | the specialism | coursework with proof, live spec-driven development | | GI407 | Dissertation in the specialism: a governed intelligence system with proof (90 credits) | 90 | Original work in the chosen domain: a system, its specification, its evidence and its measured outcome. | A dissertation whose every claim is reproduced from its evidence, defended in a recorded viva. | the specialism | live spec-driven development, coursework with proof, live audio and speech | ##### Specialising: only after the whole map Everyone learns every domain and how they connect before going deep in one. 1. **Levels 4 and 5 (every student, every domain twice).** Each of the 16 modules sets its coursework case in one of the eight domains by a fixed rotation (the table above). By the end of Level 5 every student has worked in all eight domains, twice each, using every core method. 2. **GI300 The whole map (the gate).** At the start of Level 6, the student traces one real event through every domain and measures its impact in each. Specialising is unlocked only when all Level 4 and 5 modules and GI300 are passed (`runtime/programme.py`, `may_specialise`). 3. **Going deep.** The student chooses one pathway: its Level 6 module (GI310 slot), a capstone in that domain, and for the MSc its advanced module (GI410 slot) and a dissertation in that domain. A student who chooses no pathway takes two pathway modules from different domains and a cross-domain capstone, for the generalist award. 4. **The award names the specialism**, for example *BSc (Hons) Governed Intelligence (Medical Intelligence)*. The specialism is always built on the whole map, never instead of it. | Pathway | Award | Builds on (core modules) | Level 6 module | Level 7 module | Where graduates work in the estate | |---|---|---|---|---|---| | Health and medicine | Governed Intelligence (Medical Intelligence) | GI103, GI201, GI205, GI300 | GI311 Medical intelligence: disease, detection and reversibility: Disease processes and counter-processes, screening and detection with stated error rates, triage, prescribing safety, the reversibility register, impact reversal for conditions that are part of who a person is; clinical decisions stay with an accountable clinician. | GI411 Advanced medical intelligence: trials, devices and safety cases: Trial design and reading, medical device and clinical safety obligations, safety cases, post-market surveillance. | clinic | | Finance and money | Governed Intelligence (Financial Intelligence) | GI103, GI201, GI203, GI303 | GI312 Financial intelligence: money, risk and agents as customers: Payments, lending, fraud, sanctions with guaranteed miss rates, agents as lenders and borrowers, provisional value and reversal. | GI412 Advanced financial intelligence: markets, prudential risk and supervision: Market structure, prudential and conduct risk, supervision and regulator evidence. | finance | | Society and people | Governed Intelligence (Social Intelligence) | GI107, GI205, GI101, GI206 | GI313 Social intelligence: sociology, psychology and counselling applied: Group behaviour, wellbeing, counselling practice, community systems, inclusion and exclusion by design, social prescribing; built on GI107 and GI205. | GI413 Advanced social intelligence: societies without blockers: Designing services and communities around people's lives, measuring social outcomes, safeguarding at scale. | judiverse | | Government and public services | Governed Intelligence (Public Intelligence) | GI107, GI202, GI208, GI104 | GI314 Public intelligence: government, councils and public services: Statutory duties, decision-making, procurement, public records, service journeys and their bottlenecks. | GI414 Advanced public intelligence: policy, regulation and national systems: Policy design with evidence, regulation, national-scale service redesign. | public-edge | | Education and research | Governed Intelligence (Research and Educational Intelligence) | GI102, GI105, GI201, GI301 | GI315 Research and educational intelligence: learning, evidence and peer review: How people learn, assessment by real work, research integrity, peer review, open knowledge. | GI415 Advanced research intelligence: reproducible science at scale: Research programmes, replication, research infrastructure and funding. | edge-consortium | | Defence and security | Governed Intelligence (Defence and Security Intelligence) | GI203, GI106, GI303, GI305 | GI316 Defence and security intelligence: threats, autonomy and human control: Threat assessment, autonomy and human control, two-person rules, analog interlocks in the actuation path, cyber defence. | GI416 Advanced defence intelligence: mass harm, escalation and verification: Weapons, mass harm, escalation control, arms control verification, irreversible decisions. | public-edge | | Environment and energy | Governed Intelligence (Environmental and Energy Intelligence) | GI204, GI202, GI103, GI300 | GI317 Environmental and energy intelligence: systems, carbon and supply: Energy systems, grids and prices, carbon accounting, resource flows, climate risk. | GI417 Advanced environmental intelligence: resilience and transition: Energy crises, transition planning, resilience of supply. | public-edge | | Law and justice | Governed Intelligence (Legal Intelligence) | GI104, GI101, GI305, GI403 | GI318 Legal intelligence: laws, agreements, decisions and evidence: Statute, contract, decisions, evidence and proof, rights and remedies; agreements kept separate from facts. | GI418 Advanced legal intelligence: liability, agents and dispute resolution: Liability for agents, dispute resolution with evidence, licensing and enforcement. | foundation | | Networks and connected systems | Governed Intelligence (Network and Edge Intelligence) | GI106, GI204, GI203, GI202 | GI319 Network and edge intelligence (taught by LSBU SITHub): Software-defined and intent-based networking with agents, multi-access edge orchestration with predictive migration and caching, O-RAN analytics, 5G in a box, positioning as a service, AI video analytics on the edge. | GI419 Advanced network intelligence: 6G, non-terrestrial networks and federated learning under attack (taught by LSBU SITHub): 6G sustainability, UAV coverage and energy, integrated terrestrial and non-terrestrial networks, federated learning across the edge-cloud continuum, data poisoning and adaptive zero-trust trust management. | judiverse | | Every domain | Governed Intelligence | all | GI310 and GI410 chosen from any two pathways; capstone and dissertation across domains | | all | For example, Social Intelligence is built on sociology (GI107), psychology and counselling (GI205), philosophy and ethics (GI101) and user journeys (GI206). Every specialist reaches their domain already knowing how it connects to all the others. ##### Scope discipline, and bottleneck and blocker science (GI208, GI202, GI401, GI405) Scope creep is inevitable, so the programme teaches a system that uses it as raw material. - **Capture and triage:** every new request is captured in the person's words and triaged (`tools/intent_triage.py`). - **Classify:** each request is marked as covered, extends or new. - **Place and cost:** it is placed on the whole map against the frozen scope and costed in time to outcome. - **Decide:** it is admitted to the next increment or the backlog by the accountable owner. Nothing is lost, and nothing silently enters a frozen scope. The estate runs this as `foundation/scope-change-intake`. Bottlenecks and blockers are handled by measurement and design, not effort. - **Bottleneck:** the slowest step of a journey measured from intention to outcome. - **Blocker:** a step that cannot proceed. - **Root cause:** each one is traced to fragmentation, weak controls, weak risk assessment, incomplete information or no whole map. - **Fix:** the system is changed at that root, then the journey is re-measured (`foundation/bottleneck-and-blocker-removal`). The aim is frictionless, gapless systems: the shortest intention to outcome, by design. ##### Set against leading programmes in neighbouring fields | Programme | What it offers | Source | |---|---|---| | King's College London, MSc Applied Artificial Intelligence for Security (online, 2026/27) | Responsible AI core; AI governance and AI and public policy options | [link](https://onlinecourses.kcl.ac.uk/programme-delivery-guide/applied-artificial-intelligence-security-programme-delivery-guide) | | Cranfield University, Systems Engineering MSc: Dependability and Resilience | Failure, fault, FMECA, fault trees, prognostics, design for resilience | [link](https://www.cranfield.ac.uk/courses/short/defence-and-security/module-only/dependability-and-resilience) | | UCLA Garrick Institute, MS in Risk Sciences | Reliability engineering, design for reliability and resilience, model-based systems engineering | [link](https://risksciences.ucla.edu/ms-online) | | Vanderbilt, MEng Risk, Reliability and Resilience | Foundations, policy and an applied capstone, 30 credit hours | [link](https://go.grad.engineering.vanderbilt.edu/risk) | | University of Maryland, Reliability Engineering (ENME427, M.Eng.) | Failure analysis, root cause analysis, physics of failure | [link](https://academiccatalog.umd.edu/graduate/programs/reliability-engineering-pmre/) | | UCL, Human-Computer Interaction MSc | Interaction science and design; 60-credit research project | [link](https://www.ucl.ac.uk/study/prospective-students/graduate/courses/human-computer-interaction-msc) | | NCSC-certified degrees, mapped to CyBOK's 21 knowledge areas | The UK reference for cyber security degree content | [link](https://www.ncsc.gov.uk/section/education-skills/cybok) | | UC Berkeley, Stat 212A selective inference | False discovery rate, Benjamini-Hochberg, empirical Bayes | [link](https://www.stat.berkeley.edu/~wfithian/courses/stat212a/syllabus.pdf) | | International Programme on AI Evaluations (pilot, 2026) | Alignment and control evaluations; governance, policy and regulation of AI | [link](https://www.lesswrong.com/posts/c5dSmRk4HfQqGH6ST/international-programme-on-ai-evaluations) | | MIT AI Safety Fundamentals (reading group) | Alignment, control, interpretability, governance and liability | [link](https://aialignment.mit.edu/aisf/) | | University of Groningen, Computing Science assessment plan 2026-27 | Oral exams with two examiners or recorded and kept two years | [link](https://www.rug.nl/fse/education/ter/assessment-plans/bachelors/2026-2027/bsc-cs-assessment-plan-2026-2027.pdf) | What this programme adds: - No neighbour teaches the whole chain from intention to verified outcome; each covers one part (security, reliability, interaction, statistics or AI safety). - Every assessment is real work checked by a program or reproduced from its evidence. None tests recall. - Every live and spoken assessment has two examiners and a recording, and the recording is kept for the life of the award. The best published rule found requires two examiners or a recording, kept for two years. - Marks are reproducible: a second examiner given only the evidence reaches the same mark, and any difference is recorded and resolved. - Prompt engineering is assessed by repeated-run test suites with stated success and error rates, not by single demonstrations. - Scope discipline, bottleneck and blocker science, and the analog human switch for irreversible decisions are taught as subjects in their own right. - The whole working platform (Judiverse) is included in tuition, and every assignment is a real, sandboxed project. Work a client uses is paid, and the student can offset it against tuition: earning while studying, with a route to graduate debt-free and into long-term contracts. ##### The research it is built on | Source | What it gives the degree | Modules | |---|---|---| | White paper 1: Online error rate control for faster NHS service evaluation (X-X.TECH, 19 October 2026) | Testing as data arrives with the false discovery rate held below a pre-registered limit; independent re-run | GI201, GI402, GI311 | | White paper 2: Governed AI agents: authority, receipts and recovery (X-X.TECH, 19 October 2026) | Authority envelopes, hold and decide, signed receipts, kill switch and rollback, attack testing | GI207, GI303, GI403 | | White paper 3: Cost and verified outcomes (X-X.TECH, 19 October 2026) | Cost per verified outcome; a saving claimed only after it is measured against a control and re-run | GI204, GI302 | | Frontier Agent Capability Reference, UK (7 October 2026) | Frontier benchmarks and their traps; twelve universal agentic capabilities; UK domain controls; verification gates | GI207, GI312, GI316, GI318 | | LSBU SITHub research deck (Prof. Tasos Dagiuklas, 30 June 2026) | The networking world: 5G/6G, SDN, edge, intent-based networking with agents, O-RAN, federated learning, UAV and positioning | GI106, GI319, GI419 | Digests and summaries: `research/edge-consortium/SOURCES-2026-10-31.md`. ##### Partner institutions Every university should offer the degree. London South Bank University is the first, alongside The Edge Consortium. | Institution | Role | What it does | Status | |---|---|---|---| | London South Bank University (LSBU) | First partner institution and flagship (decision tok:decision:01m69m36psxrwhsa148z746ppv) | Offers the governed intelligence degree with The Edge Consortium. SITHub, LSBU's Smart Internet Technologies Hub, teaches GI106 from its own research and testbeds: 5G and 6G, software-defined networking, edge orchestration, intent-based and agentic networking, O-RAN analytics and federated learning under attack. The campus is shared, and The Edge Consortium is LSBU's partner institution for research and development (first project: agents that learn from measured failure in SITHub's networks, simulation first) and for international technical students. | SITHub has been the consistent research partner since October 2026 (collaboration brief of 3 October, partner meeting of 6 October, consortium plan of 10 October, partner pack of 20 October; research/legacy/STRATEGY-FUNDING-OTHER.md). Degree, shared-campus and R&D terms are still to be signed. | ##### Studying it: Judiverse included, real work, paid (decision tok:decision:01m69k8c61zde0er6j94yb4ekz) Rules in `spec/offers.json`, computed by `runtime/entitlements.py`, proved by `tests/offers.py`. - **Judiverse is included in tuition.** Every enrolled student gets every Judiverse capability and every course material, with no other cost for materials. - **Assignments are real projects in a sandbox.** Partner institutions and clients post real problems, and students solve them with synthetic or consented data. Nothing touches a live system until the client accepts. - **No work is ever taken free.** A client can only take a student's work into use by paying for it, at least the National Living Wage per hour, whatever the student's age. - **The student decides where the pay goes.** For each payment they choose the split between cash and tuition offset. The offset never exceeds the tuition owed, so a student can finish debt-free. - **Academic credit and pay are independent.** Credit follows the evidence standard and pay follows the client's acceptance; neither decides the other. - **Visa work-hour limits are enforced by the scheduler.** - **Students can become long-term contractors.** A client can offer a contract, and the student accepts or declines freely. - **Outcomes are counted.** The programme counts and publishes, every term: - students in paid work; - hours and earnings paid; - tuition offset; - graduates debt-free; - contracts after graduation; - jobs created at clients. **Student finance:** The Edge Consortium petitions for its course to be funded by student finance (`edge-consortium/petition-for-student-finance`). In England, a tuition fee loan needs the provider to be registered with the Office for Students; this needs the owner. ##### Awarding it Two routes to award a degree in England: - **Degree awarding powers:** granted by the Office for Students. - **Validation:** by a university that holds those powers. The Edge Consortium offers the programme through validation and teaching partnerships with leading institutions while it applies for its own powers. This is recorded in the scope map as needing the owner. Until then, the modules run as credit-bearing short courses and as continuing professional development. ### research/edge-consortium/REFERENCE-STUDIO.md `research/edge-consortium/REFERENCE-STUDIO.md` · sha256 `e6516dd59e8132d83c06617c1dadd18813ec47e95b6da2a7a6f6b7e83db75ea9` #### Reference for /studio: the Base44 template V7BRO Owner order, 1 November 2026 04:14Z: every layout maps to one named reference; the only differences are added controls or journeys. The reference for Estate Studio (`/studio`, runtime/apps/studio.py, runtime/studio_assets.py) is the Base44 template **V7BRO** (our copy: Base44 app `6ae681783ae584c8c8890bf7`, "V7BRO (Copy)"; the running template at `https://v7bro.base44.app/`, a clone of V7 Go). Measurements and library facts: research/design/BASE44-V7BRO-STUDY.md. Written by Claude-Jegudiel 00:5531102026-1; the same file stands in research/edge-consortium and research/public-edge. ##### 0. Entry | V7BRO element | /studio renders | Difference | |---|---|---| | The workspace opens on the tree with the first project's Build grid; there is no landing page | `GET /studio` → 302 → the first application's Build (`/studio/finance/send-money/build?tree=1`); `/studio/` → 302 → that application's worked example (`/studio/consortium/handle-a-contract-dispute/build`, `/studio/public/employment/build`); `?tree=1` opens the tree on a phone, the grid shows on desktop | none; no hero, no cards | | Base44 sign-in before the workspace | a visitor sees the same shell and tree with, in place of the grid, the capability's name, its requirement and "Continue with the demo" (POST /signin/demo, next = that Build) or "Sign in with EDGE ID" | the demo entry is the estate's | ##### 1. Tree (left) | V7BRO element | /studio renders | Difference | |---|---|---| | 240 px sidebar, `bg hsl(0 0% 98%)` | 240 px `.st-tree`, `--bg-2 #f8fafc`, no margins, border-right 1 px | none | | Workspace switcher (avatar "JW", "Jason's Workspace", chevron, collapse icon) | `.st-ws`: the approved three-dot mark, "Estate Studio", the signed-in identity ("Demo workspace" or the EDGE ID) | switcher replaced by identity; no collapse icon | | Home / Knowledge / Admin | not rendered | removed: the estate's home is `/`, Help is in the page header | | "Agents" eyebrow with count, search box, new-folder and plus icons | "Search capabilities" search box filtering the whole tree | folder-create and plus removed (capabilities come from spec/units, not from the tree) | | Folders (Real Estate & PE, Insurance & Risk, …) with counts | Applications from the contract (EDGE Bank, EDGE Clinic, EDGE Consortium, EDGE Public, Judiverse Suite) with the approved app mark and the live capability count; other applications load on expand | folders are the contract's applications | | Projects under a folder with a green status dot | Capabilities under an application (research units excluded) with a green dot | same | | Build / Review / Visualize under the open project | Build / Review / Visualize under the open capability | same | | Footer: user, plan, tokens, settings | none | removed: identity and sign-out are in the page header | ##### 2. Header row | V7BRO | /studio | Difference | |---|---|---| | 56 px row: breadcrumb "All projects / FNOL Claims Processing · Live" | 56 px `.st-head`: "Estate Studio / EDGE Consortium / Handle a contract dispute · Live" | same shape; crumbs link to `/studio` and `/studio/` | | Segmented Build · Review · Visualize | same, `aria-current` on the open view | same | | "Export data" primary control (black) | "Export data" primary control (teal #005f8a) downloading `export.csv` | brand colour; export is a real CSV of the records and agent columns | | "Ask Go" composer, "…" menu | not rendered | removed | ##### 3. Build grid | V7BRO | /studio | Difference | |---|---|---| | Stage tabs: Main, Human Review, Outdated Information, "+ New stage" with colour squares | "Main n", one tab per value of the entity's status/state/stage field with its count, saved views, "+ New stage" | stages come from the record's state field; New stage creates a saved view (field = value) in Estate Studio's store, the record store is unchanged | | Toolbar: sort, filter, "⚠ 2", "Look up row, column…", Add Row, Add Column | sort, filter (active filter shown as a pill), "Look up row, column…", entity switcher, Add row, Add column | warning count removed (the Review view holds it); entity switcher added | | `role=grid`, checkbox column, row number, 36 px rows, header icon + name | same roles, same columns, 36 px rows, header icon by field kind + label + "Input" or "Agent" | same | | Column types file, text, select, number, python, collection, multiselect, json, with tool Input / AI Model / Python Tool / Workflow | declared field types of the unit (text, textarea, select, enum[], set[], list[], file, number, money, date, boolean …) rendered as text, pills, file chip, "n rows"; agent columns computed by the reasoning plug | columns come from spec/units data, not a literal array | | "Calculating..." cell with pulsing dot | "Calculating…" with pulsing dot, resolving by polling; "Model not connected" when the plug refuses; "— compute" before the first run | never a fake value | | Pills 11 px / 2 px 6 px / 4 px, rose / green / cyan / orange | same metrics; rose, green, orange by meaning, cyan / violet / teal by hash | same | | "+ Add Row" last row; footer "3 rows · Live" | "+ Add row" last row (disabled with the reason when no operation writes the entity); footer "n rows · entity · store · Live" | add row runs the capability's own create / change / send / decide / plan operation | | Row click opens the record detail | same, URL `/studio///build/` | same | ##### 4. Record detail (row click) | V7BRO | /studio | Difference | |---|---|---| | Overlay over the grid; left pane 42 % `bg-muted`, header ✕ ▲ ▼ …, file name; document preview card | same: `.st-doc` 42 %, ✕ ▲ ▼ … , name, "n / m"; PDF or image inline when the record's file field is one this site serves, else a skeleton with the reason | the "…" menu holds copy id, download JSON, inspect, open the journey | | Right pane: two-column field cells (icon + name + chevron), select fields open a Search box with radio rows and the chosen value as a pill, numbers plain, text plain | same: `.st-fgrid` two columns; enum/select → Search + radio pills; set/list[enum]/multiselect → Search + checkbox pills + Save; list of objects → nested table; file → chip; number / date / text → inline input; textarea → textarea | edits save through the capability's change (or send / decide / plan) operation with `expected_version`; "Saved · operation · verified · vN" or the refusal code | | Agent fields show the number (Fraud Risk 84) | agent-computed section: value, pill or number; click opens how it was computed in the inspector | same | | — | "Other data on this record": keys the unit does not declare | added | ##### 5. Visualize canvas | V7BRO | /studio | Difference | |---|---|---| | Nodes Source / Condition / Stage: 9 px uppercase caption, 40 px rounded box with colour dot, label, count; input dot left, output dot right | Screen / Condition / Stage nodes with the same chrome; screens and conditions from the unit, stages from action outcomes (success, error, journey:end, other capabilities) | node kinds come from the unit | | Cubic Bézier edges `M x1 y1 C …`; drag from an output dot onto an input dot creates an edge | same path; failure edges dashed red, condition edges amber; dot-to-dot drag saves a **draft** edge (badge "draft ×n") in Estate Studio's store; the unit is never changed | drafts instead of edits | | Node drag on pointer down; hard-coded x/y | node drag; positions saved per tenant; first layout three screens per row, conditions beneath, stages at the right | positions persist | | Pan and wheel zoom | pan, wheel zoom 0.2–2.5 around the cursor, + / − / fit buttons, minimap with viewport (click to jump) | minimap and fit added (the original has none) | | — | second tab **Knowledge graph**: the capability's TOK neighbourhood from `.cache/tok-dependencies.sqlite` (read-only, indexed lookups, 8 s deadline), draggable; says so when the index is absent, empty or slow | added | ##### 6. Right inspector | V7BRO | /studio | Difference | |---|---|---| | Sections Source, Tool, Reasoning (Low), Model (Concierge), Input | Source (store, tenant, capability, entity, id, version), Tool (operation, endpoint, MCP tool), Reasoning (error bound, evidence, guards), Model (plug, provider, model name, connected; never a key), Input (prompt or raw value) for the selected row, field, cell, node or edge | same sections, real provenance | ##### 7. Added by the estate (not in V7BRO) Applications and capabilities from the contract and spec/units; records from the estate record store per tenant; every write through the capability's own operations with receipts and versions; agent columns through the reasoning plug with the error bound shown; saved views; entity switcher; CSV export of real data; Review view (what the store confirmed, events, what the model produced); Knowledge graph tab; minimap and fit; positions and drafts per tenant; phone layout (tree behind a button, opened on entry at /studio; grid scrolls, canvas full-bleed, inspector as a bottom sheet); demo entry for visitors; the approved marks and Inter; teal #005f8a instead of violet/orange; no external script, font or CDN. ##### 8. Not yet built Document preview for files the site does not serve; file upload from the detail view; Review as a workflow queue (it is a ledger today); editing agent columns' prompts in place (delete and re-add); knowledge-graph depth beyond one hop; the "Ask" composer. ### research/edge-consortium/SOURCES-2026-10-31.md `research/edge-consortium/SOURCES-2026-10-31.md` · sha256 `a71ab009f0ee5b5a4655dfae038a1c4f3f2852bf3e83994182e3383f7c1ae98e` #### Sources given by the owner, 31 October 2026 Each source is held outside the repository and referenced by its sha256. Statement class: each is a CLAIM by its author, used as research and never as authority. | Source | Author and date | sha256 | What the estate takes from it | |---|---|---|---| | White paper 1: Online error rate control for faster NHS service evaluation | X-X.TECH, 19 October 2026 | `8c56e612743546fdc27fe7502ab41e3db6f91d0a1939a1c99e564ead382d5eaf` | Online FDR control as data arrives (alpha-investing, LORD, LORD++, SAFFRON, ADDIS). Pre-registered targets: time to evidence at least 20% lower, FDR below 0.05, independent re-run matches. A clinician decides every action. Taught in GI201, GI402 and GI311. | | White paper 2: Governed AI agents: authority, receipts and recovery | X-X.TECH, 19 October 2026 | `3ad034c715ff1535702966bd84824872a4ba8b99ed9bdba51912ec12a90578d7` | Authority envelope (0 actions without authority), hold and decide by a named person, signed chained receipts, kill switch and rollback, 70 attack scenarios, a risk engine. Reported: 17 runtime tests and 115 bank tests pass. Taught in GI207, GI303 and GI403. | | White paper 3: Cost and verified outcomes | X-X.TECH, 19 October 2026 | `900d123f1162fbecf0b947502d5c746a3dfae1a4d2adcfcb4e8161f674e6edf6` | Two challenges on one platform: 24 FTE and £5,225,000, against 40 FTE and £9,960,000 the usual way (48% saved). Cost of the problems addressed: £7,459m a year. A saving is claimed only after it is measured against a control and re-run. Taught in GI204 and GI302. | | Frontier Agent Capability Reference (UK) | @AX, 7 October 2026 | `b2aacac5dddf4bb0ff525c1e85bca6c80868918da2e21a4a037867d9f23896ed` | Frontier benchmarks and their traps (METR time horizon, pass^k, contamination); twelve universal agentic capabilities written as trigger, action and evidence; UK financial crime, cyber, resilience and legal controls; master verification gates. Taught in GI207, GI312, GI316 and GI318. | | SITHub research deck | Prof. Tasos Dagiuklas, LSBU Smart Internet Technologies Hub, 30 June 2026 (47 slides) | `c4c41bbfeed6ce9d6dbb51e167cf52855a1b9ae71eafa43c72b3cb956f477417` | 5G/6G, non-terrestrial networks, IoT, immersive media. LSBU 5G testbed (v1, v2). Cognitive self-organising Routing as a Service over SDN. Intent-based networking orchestrated by agents (intent agent, junior agents, senior agent that validates and generates infrastructure as code; LangGraph; A2A; dual modular redundancy). Wireless-SDN experimentation (Matlab, Mininet-WiFi, GNS3, Docker OVS, OpenDaylight). Edge computing: containerised services, LSTM predictive migration, PCR caching. Federated learning across the edge-cloud continuum: data poisoning and adaptive zero-trust (TOPSIS and EMA). 6G sustainability, 5G in a box, O-RAN analytics, UAV coverage and energy, positioning as a service (DELTA), real-time AI video analytics on the edge (VFCSIM), QoE modelling, smart farming. Taught in GI106, GI319 and GI419. | ### research/experience/REFERO-FLOWS-1.md `research/experience/REFERO-FLOWS-1.md` · sha256 `b9e6b42ca8672cb349e6d9ef87ddc1f8cbfb022c777db019088fb6a8b8b0d08d` #### Customer experience (all applications) research: real product flows from Refero, part 1 of 2 **What this is.** Real screens and user flows from existing products, recorded by Refero (a library of real app and website screenshots, each journey described step by step). They show the floor (what users already expect) and good customer experience from any industry worth borrowing. **Statement class.** Every flow is an OBSERVATION by Refero of a real product, searched 2026-10-31. Descriptions are Refero's wording. Product names belong to their owners. **How it was gathered.** Semantic searches for this application's tasks on iOS and web. Results from other industries are kept on purpose: they are often the best customer experience to learn from. Refero does not hold many incumbents (and few government, enterprise or NHS systems), so this is a sample, not a census. **Products in this sample:** Airbnb (34), WhatsApp (33), District (29), Dropbox (26), PayPal (22), Instagram (20), Fresha (19), Lovable (18), Mercury (15), komoot (15), TikTok (15), Meta AI (15), ElevenLabs (14), Linear (14), ElevenMusic (14), X (13), ElevenReader (13), On (13), Nextdoor (13), Omio (13), Runey (11), Riverside (11), Willow (10), Faire (10), Cushion (10), Tinder (10), Artsy (10), Fernand (10), Pinterest (10), Substack (10), Structured (10), Intercom (9), GoFundMe (9), YouTube Music (9), IKEA (9), Napper (9), Tripsy (9), Rork (9), XChat (9), Missive (9), Ada (8), Wabi (8), DailyArt (8), HYPE (8), Sora (8), Anam (8), Foodllama (8), Zendesk (8), Voicenotes (8), Podimo (8), OpenAI (8), Trip (8), ChatGPT (8), Blackbird (7), Phantom (7), Revolut (7), Comet (7), Train Fitness (7), Google Gemini (7), Apple Books (7), Twitch (7), Family (7), Binge (7), Wealthsimple (7), Raycast (7), Doctronic AI (7), Longevity Deck (7), Leonardo.Ai (7), Melius (7), Manus (7), Asos (7), Brilliant (7), Maze (7), YouTube (7), Aboard (7), SSENSE (7), Threads (7), Grok (6), Givingli (6), Fingerprint (6), Adobe Photoshop (6), Shopify (6), Mapbox (6), Instacart (6), Duolingo (6), Chance AI (6), Vinyls (6), Handshake (6), Hello Ivy (6), Vimeo (6), Synthesia.io (6), Canva (6), DailyArt Magazine (6), Сontra (6), Medium (6), Tango (6), HBO Max (6), Steps (6), Acctual (5), Google Translate (5), TIDAL (5), Granola (5), Grammarly (5), Shop (5), Loowner (5), Teal (5), GlossGenius (5), Reown (5), Ground News (5), Cohere (5), Poppy (5), Mozi (5), Apple Maps (5), Exoplan (5), Care (5), Steel (5), Roots (5), Rox (5), Bond (5), Jace AI (5), Skiff (5), OKURI (5), Figma (5), Trip Way (5), Bezi (5), Dropset (5), Genie (5), LEGO® Builder (5), BeReal. (5), Reeder. (5), ABY Journal (5), Daylish (5), Jasper AI (5), Verse (5), RAD Weather (5), Mela (5), UGLYCASH (4), Fourmula.ai (4), Fuse (4), Introspect (4), Waterllama (4), The Outsiders (4), Cake Equity (4), Pi (4), Pinkllama (4), Linktree (4), Apple TV (4), Product Hunt (4), Chargetrip (4), fal (4), ALO (4), Nike (4), Factory (4), Imprint (4), Mindllama (4), Matter (4), Netflix (4), Wrike (4), Microsoft Teams (4), Cursor (4), Whereby (4), TIDE (4), Glorify (4), Kickstarter (4), Microsoft Copilot (4), 222 (4), Replo (4), Airtable (4), Spotify (4), LinkedIn (4), Delphi (4), Wispr Flow (4), Stellate (4), Parallel (4), DoorDash (4), Square (3), Nooka (3), Turf (3), TravelPerk (3), Spoil me (3), Literal.club (3), AVNIER (3), Copperx (3), Gumroad (3), UNHINGED (3), Carbon (3), Understory (3), Claude (3), Kin (3), Appwrite (3), n8n (3), Kit (3), Dribbble (3), Digg (3), Saint Laurent (3), Telegram (3), Remy (3), Preply (3), The Leap (3), MonoDesk (3), Miles (3), TidyCal (3), Programa (3), Polar (3), Exa (3), Acne Studios (3), West Elm (3), Microsoft (3), Reclaim AI (3), Apple Invites (3), Daze (3), Calendly (3), PocketShelf (3), Joi (3), Canopi (3), Runway (3), Aesop (3), Tripadvisor (3), ManyChat (3), Hashnode (3), Typeform (3), Peak Performance (3), Stardust (3), Visitors (3), Mocha (3), Spline (3), ZARA (3), The Washington Post (3), Julienne (3), Navan (3), adidas (3), Box Box Club (3), Luminar (3), pushr (3), Pop Site (3), MWM (3), pillowtalk (3), Abode (3), Wise (2), Cashcaded (2), Pool (2), Tidal (2), Nudea (2), Kann (2), Spyglass (2), Haptic (2), Weekrise (2), Shazam (2), Cofounder (2), Sunsama (2), Prose (2), time2book (2), Zara Home (2), ROVR (2), Dopper (2), Loom (2), Days (2), Lemni (2), Memotron (2), Epidemic Sound (2), Volv (2), SoundCloud (2), FocusPomo (2), Agora 2 (2), Tella (2), Modyfi (2), Descript (2), Cal (2), Luminary (2), Pattern Brands (2), Acuity Scheduling (2), Fourthwall (2), Onefinestay (2), Breathwrk (2), raum. (2), Showcase (2), PamPam (2), Ozone (2), Polaroid (2), Picnic (2), FocusFlight (2), Patreon (2), Playground (2), Enode (2), Plane Finder (2), Dub (2), Ballpark (2), Drinkit (2), Peerlist (2), Navia (2), Rive (2), [untitled] (2), Stripe (2), Moises (2), Expedia (2), Telescope (2), voilà (2), Sky Guide (2), Endless Tools (2), Urban Outfitters (2), Brink (2), Clipchamp (2), Exactly.ai (1), Apple Stocks (1), Denim (1), Allbirds (1), Butter (1), Farfetch (1), Make (1), Column (1), Yelp (1), Stemz (1), Clime (1), SavvyCal (1), Homerun (1), Shuttle (1), Play (1), Gizmo (1), MD Vinyl (1), Savee (1), Vocabulary (1), Moonlitt (1), Endless (1), Soundtrap (1), TwelveLabs (1), WURRD (1), Lummi (1), ARKET (1), Douze (1), ten ten (1), MyMind (1), Portrait (1), Womp (1), Rarible (1), VSCO (1), Skyscanner (1), Macaw (1), Touch Time (1), (Not Boring) Weather (1), Unwind (1), Moonly (1), (Not Boring) Habits (1), Miro (1), Prime Video (1), Pitch (1), Doji (1), Athleats (1), Scape (1), Apple Podcasts (1), B150 (1), Stocktwits (1), Project Indigo (1), StrideGate (1), tone (1), Queue (1), WishWish (1), Craft (1), Open (1), Marathon (1), Oppie (1), Weather mini (1), pico cam (1), Pryzm (1), one year (1), (Not Boring) Camera (1), ShareWillow (1), Uber (1), Resend (1), LEGO (1), Volkshotel (1), CapWords (1), Orbit (1), Variant (1), Apple Games (1), Excalidraw (1), Dona (1), GO Club (1), Neuecast (1), PlayStation (1), Wittl (1), Hello Klean (1), Xbox (1), Checker (1), Tilt (1), Slite (1), Subo (1), CocoonWeaver (1), Uber Eats (1), New Balance (1) Tasks in this part: AI agent setup permissions to loading progress state. Part 2 is REFERO-FLOWS-2.md. ##### Task: AI agent setup permissions **Comet: Agent card gallery** (ios, 3 screens, AI Tool, Utility) [Refero](https://refero.design/flows/10948) > Happy-path summary: The user starts on a mobile "card gallery" that shows two active AI agent cards. From this gallery the user inspects either card, brings one into focus (front), optionally drags to reorder the two cards, and returns to a side-by-side overview. Steps: Card gallery overview → Bring card to front / partial overlap → Updated overview / reordered cards **Tripsy: Onboarding and First Launch** (ios, 8 screens, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/2326) > First-launch onboarding for a travel organizer (iOS). The user is introduced to the product via a multi-slide carousel that highlights key features (flight updates, collaboration, documents). During or immediately after onboarding the app requests runtime permissions (notifications and location). An upsell (Tripsy Pro) promotional modal can be shown before the user lands on the main 'My Trips' screen. Steps: Notification Permission Prompt → Welcome / Intro Slide → Feature Slide — Flight Details → Feature Slide — Collaboration → Feature Slide — Documents → Location Permission Prompt → Promotional Upsell (Tripsy Pro) → My Trips — Empty State Home **Dropbox: Folder access management** (web, 4 screens, Collaboration Tools, Business, Productivity) [Refero](https://refero.design/flows/13031) > A logged-in user manages sharing and permissions for a team folder in a cloud file manager. The flow starts from the folder’s action menu, opens an access management dialog, changes a collaborator’s permission level, and confirms the update with a success message. **Leonardo.Ai: Profile Settings Management** (web, 2 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/12914) > This is a settings screen in Leonardo.ai focused on personal profile management. The entry point is the user already being logged in and opening Settings, with the left sidebar showing Personal Settings and Workspace Settings. The user can review profile information, choose interest tags, confirm adult content access, and optionally delete the account. The exit point is after changes are saved and reflected in the profile section with confirmation feedback, or after the user leaves settings. Steps: View Profile Settings → Update Preferences **WhatsApp: Create Community Setup** (ios, 6 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12677) > Happy path for creating a new community on mobile. The user enters a community name, customizes the community image from the Edit control, selects either an emoji or sticker with a background color, then returns to the main creation screen with the chosen avatar and name ready to submit. **Rork: Collaborator Invitation** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12168) > Happy-path project sharing in a collaboration modal. The user opens an invite dialog from a project workspace, enters an email, chooses a permission level, and either sends an invitation or creates a shareable invite link. Steps: Invite Collaborators → Permission Selection → Invite Link Created **Pool: Refero MCP item review** (ios, 4 screens, Utility) [Refero](https://refero.design/flows/11989) > This is a mobile screenshot sequence showing a Refero content item and the user opening its overflow menu, then opening an agent selection bottom sheet and selecting one agent. Entry appears to be a detail viewer for a Refero MCP AI Agents item, with a close button, overflow menu, and embedded preview content. The user can either dismiss overlays, open the action menu, or proceed into the MCP exploration/pooling step. Assumption: the user is already inside the viewer on a mobile device and has permission to interact with the item. The flow ends with one agent tile selected in the bottom sheet. Steps: Item Detail Viewer → Overflow Actions Menu → Pool Selection Sheet → Agent Selected **WhatsApp: Meta AI Voice Selection** (ios, 3 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12674) > Happy path for choosing a voice persona for Meta AI and optionally enabling automatic voice start. The user enters a Meta AI voice/settings view from a chat context, browses available voice personas, selects one, and can turn on the 'Start with voice' toggle so Meta AI opens directly in voice mode from the Chats tab. **Meta AI: Meta AI Account Sign-In** (ios, 10 screens, AI Tool) [Refero](https://refero.design/flows/13342) > The user opens Meta AI and starts from the welcome screen. They choose to get started, select a suggested account, and authenticate using a six-digit code sent to the masked email address. After the code is submitted and authentication completes, Meta AI opens and requests permission to send notifications. The user proceeds to a short feature introduction and taps Get started to reach the main assistant screen. The flow assumes the user is on an iPhone, has a previously recognized Meta account, can access its email, and is not setting up a device. Steps: Welcome → Choose Saved Account → Account Login Loading → Email Code Entry → Code Partially Entered → Code Completed → Authentication Loading → Notification Permission → Meta AI Introduction → Assistant Home **Structured: Structured AI Onboarding** (ios, 9 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3284) > Happy path: From the New Task composer the user selects the "Plan your day with AI" suggestion which launches a multi-step Structured AI onboarding carousel. The user advances through three marketing/feature slides (Speed up planning, Auto-generate schedule, Digitize planner), then reviews privacy & data details and data categories sent to the AI, optionally reads FAQ items, and accepts the Terms. After acceptance the user lands on the Structured AI input screen where they can type, speak or scan a handwritten plan to have the AI generate schedule suggestions. Steps: New Task — Suggestions → Onboarding Slide — Speed Up Planning → Onboarding Slide — Auto-Generate Schedule → Onboarding Slide — Digitize Analog Planner → Privacy & Data — Consent Intro → Privacy — Data Used For Structured AI → Privacy — FAQ Expansion → Accept Terms & Finish Consent → Structured AI Input — Prompt Entry **Cursor: Agent prompt composer** (web, 2 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9941) > Desktop web interface for an 'Agents' workspace. Steps: Workspace with left navigation → Collapsed navigation, focused canvas ##### Task: AI assistant chat **Melius: Image Upload Assistance** (web, 6 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/14073) > The user starts on a visual generation canvas containing a reference image node, a descriptive prompt node, and an image-generation node. They open the prompt node’s context menu, choose the assistant option, ask how to upload an image, and receive written instructions in a side conversation panel. The flow begins on an existing canvas and ends when the assistant displays the upload guidance. The screenshots do not establish whether the user is logged in, but the canvas is already open and interactive. The user appears to be on a desktop device with permission to edit the canvas. Steps: Canvas Workspace → Node Context Menu → Assistant Prompt → Question Entry → Assistant Processing → Upload Instructions **Jace AI: Onboarding assistant conversation** (web, 4 screens, Communication Tools, AI Tool) [Refero](https://refero.design/flows/10221) > Happy-path: A signed-in desktop user lands on a conversational assistant home view that surfaces three template cards (suggested prompts). The user clicks a template, which creates a new conversation and pre-fills a question. The assistant processes the prompt (typing indicator shown) and returns a structured reply (table + bullets) in the conversation. Steps: Welcome + Template Suggestions → Template Selection → Prompt Sent — Waiting State → Structured Assistant Reply **Jasper AI: Slogan Creation with Conversational Assistant** (web, 6 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/5297) > Happy path: A signed-in user working in a two-column workspace (document on right, conversational assistant on left) opens the assistant, asks it to generate a catchy slogan, answers a clarifying question, supplies the brand name, and receives a generated slogan with options to iterate or request related assets. Entry: user has the editor open and assistant panel visible. Exit: assistant returns one or more slogan options and presents follow-up actions (generate more, create logo). Assumptions: desktop layout, user is authenticated and has access to the assistant ('Jasper'), the assistant panel remains anchored to the left while the document remains visible on the right. Steps: Open assistant and choose task → Assistant asks for clarification → User provides context via quick replies → User supplies brand name → Assistant generating content → Generated slogan and follow-ups **Doctronic AI: Response Style Preference** (web, 3 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13703) > A logged-in user updates the assistant’s response style from the account settings screen. The user opens the Response style dropdown, reviews the available tone options, and selects “Just the facts.” The setting is applied immediately, the dropdown closes, and a confirmation banner appears. The flow starts on the user settings screen and ends with the updated response style visible. **Fourmula.ai: Background Generation Support** (web, 6 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/11844) > A user is viewing a fashion image project dashboard and opens the Formula Assistant for help with background generation. They enter a complaint about prompts not working, send it, and wait while the assistant provides troubleshooting guidance. The flow starts on the project gallery and ends with suggested prompt-writing actions. The user appears to be signed in and working in a desktop browser. A persistent notice indicates that no credits remain and recommends upgrading or purchasing credits. Steps: Project Gallery → Open Formula Assistant → Enter Support Request → Send Message → Assistant Processing → Review Troubleshooting Guidance ##### Task: AI chat assistant **PayPal: PayPal Assistant Conversation** (ios, 16 screens, Finance) [Refero](https://refero.design/flows/7266) > Happy-path walkthrough of a logged-in mobile user opening the Message Centre, launching the PayPal Assistant, asking for transaction help, receiving an automated answer with detailed steps and links, giving feedback, and ending the conversation. Steps: Message Centre (empty) → Assistant greeting and quick options → User quick reply sent → Assistant 'no transactions found' → User declines search change → User enters free-text question → Assistant processing response → Assistant provides step-by-step answer → Related links and feedback prompt → User taps thumbs up (positive feedback) → Feedback input (user typing) → Send feedback → Open End Conversation action → Confirm end conversation → Message Centre updated → End state and return to idle **Granola: Review Generated Email** (ios, 2 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14157) > The user is viewing an AI-generated follow-up email within a mobile assistant conversation. The flow starts with an existing prompt and generated email response, then continues when the user selects the positive feedback control. It ends with the response marked as helpful. The user is assumed to be in an active conversation on a mobile device; authentication and permission state are not visible. Steps: Generated Email Review → Positive Feedback Submitted **GlossGenius: Analytics with AI Growth Analyst** (web, 4 screens, Business) [Refero](https://refero.design/flows/9979) > Happy path: A signed-in user opens the Analytics section from the left navigation and sees key performance indicators (net revenue, projected revenue, service/retail revenue) for the selected timeframe and staff filter. A persistent right-hand panel titled 'AI Growth Analyst' offers suggested prompts and an input area. The user selects (or types) a suggested prompt to set a revenue goal; the assistant shows brief processing states and then returns a written recommendation with numeric targets and tactical suggestions. Steps: Analytics overview with AI suggestions → User submits suggested prompt → Assistant processing state → Assistant provides recommendation **Granola: Copy Generated Response** (ios, 2 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14156) > A user views a generated follow-up email in a mobile AI conversation and copies the response to the clipboard. The flow starts on an open conversation with the generated message visible and ends with confirmation that the message was copied. The user is assumed to be on a mobile device with access to the conversation; no additional authentication or permission state is visible. Steps: Generated Email Response → Copy Confirmation **Cohere: Tool-assisted Charting with Python Interpreter** (web, 6 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3837) > Happy path: A logged-in user opens the Chat workspace in Tool Use mode, selects helper tools (Python Interpreter, Calculator, Web Search), types a single prompt requesting a plot of average 1-bedroom rental prices for the top 5 most expensive North American cities in Jan 2024, and sends it. The system invokes tools (web search then Python interpreter), generates code, executes it, and returns a matplotlib bar chart image embedded in the conversation along with code and references. Steps: Tool selection & prompt composer → Prompt submitted → Assistant begins reasoning & tool invocation → Python code produced and executed → Plot rendered inline → Tool result preview & continued interaction **Doctronic AI: Passwordless AI Doctor Sign-In** (web, 8 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13706) > The flow begins on Doctronic’s public landing experience, where an unauthenticated visitor can start a chat or access account options. The user chooses email-based login, enters an email address, receives a six-digit verification code by email, and submits the code. After successful verification, the user is taken to the authenticated AI health assistant screen, ready to ask a health question. The user is assumed to be logged out and able to access the entered email inbox. **Google Gemini: Voice Query to Suggestions** (ios, 6 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/3207) > Happy-path: A signed-in user interacts with the Gemini conversational assistant on a mobile device. Steps: Conversation (idle with assistant intro) → Start Voice Input (listening) → Live Transcription (partial) → Full Transcription / Submit → Assistant Responding (audio playing) → Answer Rendered (suggestions with media) **WhatsApp: Meta AI Voice Chat Setup** (ios, 6 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12673) > Happy path for starting a voice conversation with Meta AI. The user enters the Meta AI profile/settings area, opens the voice feature, reviews the informational consent-style modal, and continues into a live voice session. **Meta AI: Save Recipe Instructions** (ios, 2 screens, AI Tool) [Refero](https://refero.design/flows/13345) > A user views an assistant-generated response containing ingredients and a 20-minute pasta cooking method. The user copies the response using the copy action beneath the message. A confirmation toast appears stating that the content was copied. The flow starts with an existing assistant response and ends with the response available on the device clipboard. Steps: Review Recipe Response → Copy Recipe **Raycast: Save AI Message to Note** (ios, 4 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/10166) > Happy path: A user is reading an AI assistant conversation on a mobile device and wants to save a specific assistant message as a persistent note. From the conversation view the user opens a contextual '+' menu, selects "New Note," receives a transient confirmation, and then navigates to the newly created note where the assistant's long prompt text is displayed. Steps: Conversation — locate message → Open action menu → Show creation confirmation → Note viewer — review saved content **Grok: Private Chat — Compose and Receive Response** (ios, 6 screens, AI Tool) [Refero](https://refero.design/flows/6093) > Happy path: A signed-in mobile user opens the assistant, taps the private (ephemeral) chat mode, composes a long prompt, submits it to the selected model (Grok 3), waits while the model processes, and then reads the generated response. The user can stop generation mid-way, scroll the resulting content, and use inline actions (copy, share, thumbs up/down). Steps: Chat Home / Composer → Private Chat — Empty State → Compose Prompt → Submit & Processing → Assistant Response Appears → Read Full Reply & Actions **Genie: Ask Genie With Photos** (ios, 9 screens, Social Networking, AI Tool) [Refero](https://refero.design/flows/12422) > A mobile user opens a Genie assistant conversation, enters a question about the assistant’s capabilities, attaches three photos from the device photo library, and sends the message. Genie shows a loading state, then responds with an overview of its capabilities and references the uploaded photos. The flow begins in an empty assistant conversation with the user already signed in and ends with the assistant’s response visible; the user remains on the conversation screen and can continue using the Follow up field. Steps: Assistant Home → Enter Question → Open Attachments → Choose Photos → Review Attachments → Message Submitted → Assistant Processing → Capabilities Response → Continue Conversation **WhatsApp: Meta AI Message Feedback** (ios, 2 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12672) > A mobile chat conversation with Meta AI. The screen opens inside an existing conversation, showing an AI-generated reply, a disclaimer about AI accuracy, message actions, and the input bar. The user can react to the AI message using the feedback controls, and a confirmation toast appears after feedback is submitted. Assumptions: the user is already inside the chat and has permission to interact with the conversation on a phone device. **Meta AI: Incognito AI Conversation** (ios, 7 screens, AI Tool) [Refero](https://refero.design/flows/13350) > A logged-in user opens the navigation drawer, selects the private/incognito Meta AI chat option, reviews the privacy introduction, and continues into a temporary conversation. The user enters a prompt requesting two explanation levels, submits it, and waits while Meta AI generates the response. The completed answer presents a child-friendly explanation followed by a physics-professor-level explanation, which the user can scroll through. The flow begins in the main chat area with the navigation drawer open and ends on the completed incognito response. Steps: Open Incognito Chat → Review Privacy Introduction → Enter Private Prompt → Complete Prompt → Submit Prompt → Review Generated Response → Read Full Explanation **Meta AI: Mythical Creature Field Guide** (ios, 7 screens, AI Tool) [Refero](https://refero.design/flows/13351) > A user submits a request to Meta AI on a mobile device asking for a detailed walkthrough of dragons, phoenixes, krakens, unicorns, and mermaids, using multiple sub-agents. Meta AI first plans the work, then runs five specialist biology agents, regenerates a corrected poster asset, and presents a consolidated field-guide response with supporting infographics. The user can scroll through the answer, provide feedback, copy or share it, and choose from suggested follow-up requests. The user appears to be using the assistant without any visible authentication or permission step. Steps: Prompt Entry → Request Planning → Sub-Agent Execution → Asset Regeneration → Field Guide Response → Answer Review → Completion and Follow-Up **Voicenotes: Ask AI from Filtered Notes** (ios, 11 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14249) > A signed-in user starts from the Voicenotes note library and opens Ask AI to query their recorded notes. They choose Web Search and restrict the note context to the last 30 days, then dictate a request by voice. The system converts the recording into editable text, submits it, and displays an AI-generated response with source and copy controls. The flow begins in the note library and ends on the completed response view. The screenshots show an iPhone-sized interface with keyboard and microphone permissions already available; no authentication or permission setup is shown. **Meta AI: AI Image Creation** (ios, 4 screens, AI Tool) [Refero](https://refero.design/flows/13358) > A user enters a detailed text prompt asking Meta AI to create a hyper-realistic image. After submitting the request, the interface shows a prompt-processing state, then presents the generated image with an explanatory response. The user can review the result, provide feedback, copy or share the response, and choose suggested refinements. The flow begins on the Meta AI prompt composer and ends with the generated image available for review or further editing. The screenshots imply a mobile device and an available Meta AI session. **Google Gemini: Request Java Ping Pong Example** (ios, 4 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/3222) > Happy path: A signed-in user opens the Gemini chat assistant on a mobile device, types (or speaks) a request for a simple Ping Pong game written in Java, and receives a multi-part response: a labeled code block (scrollable), a caution/copy affordance, a structured explanation, and a source link with feedback/share controls. Steps: Welcome / Composer → Request Sent / Code Response → Code Continuation & Caution → Explanation, Sources & Feedback **Meta AI: Recipe Response Rating** (ios, 3 screens, AI Tool) [Refero](https://refero.design/flows/13344) > A user reviews a conversational cooking response containing ingredients and a 20-minute pasta preparation plan, then marks the response as helpful using the thumbs-up control. The flow starts on an open assistant conversation and ends with the positive rating selected. No login, permissions, or additional setup are visible. Steps: Review Recipe Response → Submit Positive Feedback → Confirm Selected Rating **Voicenotes: Configure Ask AI Search** (ios, 4 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14250) > The user starts in an Ask AI note-assistant interface with the message composer focused and the keyboard open. They open the composer options, choose Web Search as the assistant mode, and select Last 30 days as the note date filter. The options panel closes, and both selections appear as removable chips above the message field. The user remains ready to enter a question; no message has been submitted. The flow assumes the user has access to at least one recorded note and is using a mobile device with the keyboard available. **Genie: Dinner Recommendation Chat** (ios, 6 screens, Social Networking, AI Tool) [Refero](https://refero.design/flows/12427) > Happy-path conversational assistant experience on a mobile device. The user starts on a home/chat screen with a Genie profile card and prompt chips, enters a dinner-related question in the message composer, sends it, and receives suggested nearby restaurant options with image cards and explanatory text. Steps: Assistant Home → Message Entry → Send Message → Assistant Loading → Recommendation Preview → Detailed Suggestions **Meta AI: Trending Recipe Search** (ios, 8 screens, AI Tool) [Refero](https://refero.design/flows/13349) > The user is viewing an existing dinner-related assistant response and starts a new recipe search from the reply composer. They enter “Find recipes with chicken and avocado,” open the response-mode selector, switch from Instant to Thinking, and submit the request. The interface shows a loading state while gathering trending recipes, then displays a multi-item recipe response that can be scrolled for additional recommendations. The flow assumes the user is already inside a conversation named “Dinners.”. Steps: Existing Dinner Response → Activate Reply Composer → Enter Recipe Query → Open Response Mode Menu → Choose Thinking Mode → Submit Recipe Search → View Initial Recommendations → Review Full Recipe List **Synthesia.io: AI Presentation Review and Summary** (web, 8 screens, AI Tool) [Refero](https://refero.design/flows/13744) > The user is working in a presentation and video editor with a selected scene and an AI assistant available in the right-hand panel. They first ask the assistant for an opinion about the presentation, receive an explanation of the assistant's capabilities, and then request a summary of the video. The assistant returns a scene-based summary and identifies that scene 8 is currently selected, offering to provide a more detailed scene-by-scene explanation. The flow starts with the editor open and ends with the summary expanded in the chat. **ChatGPT: Voice prompt to generated email** (web, 4 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/6893) > Happy-path flow for creating a new conversation by dictating a prompt and receiving an AI-generated response. Steps: Landing / Input ready → Recording voice input → Transcription visible for review → Assistant generates response **Canva: Revise cover title via voice or text** (web, 4 screens, Design & Media Editing, Collaboration Tools) [Refero](https://refero.design/flows/8061) > User is in a desktop web design assistant screen showing a project titled "Tokyo Travel Vlog Cover Design." The interface displays four AI-generated design thumbnails, a row of suggestion chips (e.g., "More Instagram Story designs", "Add a cherry blossom illustration"), and a chat-style input area at the bottom that supports typing and voice input. Steps: Design suggestions preview → Start voice input → Transcription inserted → Edit and submit instruction ##### Task: API keys developer settings **Willow: Willow onboarding intro** (ios, 3 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/13227) > A short onboarding carousel for Willow shown to a new or returning user before entering the main experience. The user can swipe through several promotional intro screens and tap the primary Get started button to continue. **Fuse: New wallet signup and secure key setup** (ios, 22 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8729) > Happy-path onboarding for a crypto wallet: user is introduced to product pillars (Spend / Save / Invest), chooses to create an account using Sign in with Apple, and runs through a 3-part key setup (Device Key — biometric, 2FA Key — iCloud, Recovery Key — email OTP). After key creation the wallet is provisioned, a short promo for a paid membership is shown, and the user arrives at the empty wallet home. Steps: Onboarding carousel — Spend → Onboarding carousel — Save → Onboarding carousel — Invest → Welcome / Sign-up choices → Sign in with Apple sheet → Signing-in progress → Secure your wallet — Device Key intro → Face ID permission prompt → Creating Device Key (progress) → Create 2FA Key (iCloud) prompt → Recovery Key — email entry → OTP entry — keyboard & autofill → OTP verification and provisioning → Locked state — Unlock CTA → Promo modal — free transactions → Fuse Plus membership detail → Wallet provisioning complete — home (empty) → Empty wallet CTA — Receive → Default bottom navigation & quick actions → End state — ready wallet **Poppy: Recommended Connection Setup** (ios, 4 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/12177) > Happy-path onboarding/settings sequence for a logged-in user named Connie. The user opens the profile/settings area, chooses a recommended connection type (Fitness), is shown a system Health permission request, and returns to the recommendation screen after granting access. Entry is from the profile/settings menu; exit is back to the recommendation step with the selected connection highlighted and a Next button available. Steps: Profile and Settings → Connection Recommendations → Health Permission Request → Connection Selected **komoot: New user onboarding and consent setup** (ios, 39 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8281) > Happy-path onboarding for a first-time mobile (iOS) user. Entry: app launch -> privacy/tracking consent banner. User can Accept All, Refuse, or Customize preferences. If user customizes they can toggle categories and open vendor details. After consenting, the flow presents sign-in options (social or email). Choosing email leads to email validation then account creation (password). Post-signup the user progresses through product onboarding cards that request push notifications and access to device integrations (Apple Health). The flow ends with device connections (optional) and a completion screen, then the main map/routes home screen. Steps: Privacy notice (top-level) → Customize preferences (categories visible) → Category list (scroll) → Analytics expanded (vendor row visible) → Vendor info modal (Firebase) → Sign-in options → Enter email (initial) → Email validation error → Email accepted (proceed) → Create account (password) → Onboarding hero / intro → Onboarding card: Navigation → Onboarding card: Notifications → iOS push permission dialog → Device integrations list → Health Access (iOS modal) → Integration connected state → Onboarding completion → Main map / routes home → Optional: Connect additional devices (post-onboarding) **Exa: Developer Workspace Home** (web, 2 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/12010) > This is a dashboard-style entry screen for an Exa developer workspace. The user is logged in and can either begin onboarding/integration from the main hero card, inspect documentation, manage API credentials, add credits, or explore product capabilities and templates. The flow enters from the workspace home dashboard and exits when the user clicks a primary action such as Get Started, Docs, a product card, a template card, or API key-related controls. Steps: Workspace Dashboard → Compact Dashboard View **Google Gemini: Gemini Welcome & Consent** (ios, 3 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/3201) > This is a mobile-first, modal-style welcome and consent sequence presented to a logged-in Google user before they access Gemini. Steps: Welcome modal with summary → Expanded detail card → Legal links, opt-in, and final CTAs **ElevenMusic: Account Settings Feedback** (ios, 4 screens, AI Tool, Entertainment, Music Streaming) [Refero](https://refero.design/flows/12446) > This is a mobile account settings area with a feedback subflow. The user starts on Account Settings, where they can review subscription status and access support/help links. Selecting Give feedback opens a Send Feedback modal with a text field and submit action. After entering text and submitting, a confirmation toast appears and the user returns to Account Settings. Steps: Account Settings → Feedback Form → Feedback Entry → Submission Confirmation **Willow: Willow Home Dashboard** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/13235) > Logged-in Willow home screen on iPhone. The user sees a stats header, a dismissible promo banner, and a rotating onboarding/promotional card carousel. From this screen they can start dictation, set up an Action Button shortcut, watch a tutorial, add a personal term in Dictionary, or invite friends for a referral reward. The screen also shows a transcripts section with search and filter controls, and a bottom tab bar for navigating to Home, Dictionary, Style Matching, and Profile. Steps: Home Overview → Action Button Setup Card → Dictation Tutorial Card → Personal Term Card → Referral Reward Card **ElevenMusic: Music Creation Workspace** (ios, 4 screens, AI Tool, Entertainment, Music Streaming) [Refero](https://refero.design/flows/12459) > This is a mobile music-creation workspace. The user enters the Create section, where they can browse draft or published generations, open a prompt/composer panel, and adjust generation settings in a bottom sheet. The main entry point is the Create tab; the exit point is either returning to the list or submitting a creation prompt. Steps: Create Dashboard → Settings Overview → Settings Adjustment → Prompt Composition **Willow: Offline Mode Settings** (ios, 3 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/13251) > Mobile settings screen for managing offline dictation support. The flow starts on an Offline mode settings screen and can end either by deleting the downloaded offline model or by returning to the same screen after viewing the delete confirmation. Steps: Offline Mode Settings → Delete Confirmation → Offline Model Removed ##### Task: CRM contacts pipeline **Lovable: Missing Connector Request** (web, 4 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12342) > Happy-path for requesting a missing connector from the Connectors directory. The user starts in the Connectors section, filtered to the Messaging category, and uses the ‘Missing a connector?’ form in the lower-left panel to request an integration. The visible state assumes the user is already signed in and has access to the connectors catalog. After entering a connector name and submitting, the form confirms the request was received and disables further submission. Steps: Browse Messaging Connectors → Enter Connector Request → Submit Request → Request Confirmed **WhatsApp: Add WhatsApp Contact** (ios, 7 screens, Social Networking, Collaboration Tools) [Refero](https://refero.design/flows/12606) > Happy path for adding a WhatsApp contact from the Contacts screen using either the add menu or QR scan. Steps: Contacts Overview → Add Options Menu → QR Scanner Prompt → Scanner Ready → QR Code Detected → Account Found → Contacts Updated **WhatsApp: Calls Onboarding** (ios, 3 screens, Social Networking, Collaboration Tools) [Refero](https://refero.design/flows/12600) > A Calls tab onboarding screen shown in a logged-in mobile messaging app. The user can swipe through three informational cards and use the primary action to start a call or secondary action to invite a friend. The entry point is the Calls tab from the bottom navigation, and the exit point is likely the next call setup or invite screen after tapping an action. Steps: Private Calls Intro → Group Call Intro → Effects Intro **Cursor: Enterprise sales contact form** (web, 5 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9915) > Multi-step enterprise sales contact form presented alongside marketing content. Steps: Initial contact prompt → Qualification completed (pre-fill) → Detailed contact information → Filled form review → Submission confirmation ##### Task: QR code scan **Mozi: Invite & Connect People** (ios, 5 screens, Social Networking, Travel & Booking) [Refero](https://refero.design/flows/4144) > Happy path: A signed-in user opens the 'My People' section, sees either an empty-state prompting them to invite or check contacts, opens the in-app 'Connect' screen to show/share their personalized QR code and invite link, optionally views a short QR onboarding tip, and returns to My People where newly connected contacts appear. Entry: My People tab. Exit: My People populated with one or more connections (or QR Connect screen is visible). **komoot: Invite participants** (ios, 3 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8356) > Happy path: A signed-in user opens the Invite people screen for a group/activity on an iOS device, then either shares the QR/link or searches their contacts to invite individuals. Steps: Invite overview → Search activated → Filtered results and invite **PayPal: Save personal QR code** (ios, 3 screens, Finance) [Refero](https://refero.design/flows/7229) > Happy path: a logged-in user opens the QR Code area and views their Personal QR code under the 'Your code' tab. They tap the download icon to export the QR image to the device Photos app. Because the app has not yet been granted photo access, the OS photo-permissions dialog appears. After the user grants access (or allows saving), the image is written to Photos and a transient confirmation toast appears. Entry: QR Code screen (Your code). Exit: QR image saved to device photo library with a confirmation message. Assumptions: user is authenticated, running on an iOS device, and the app does not currently have required photo-library permission. Steps: Personal QR code display → Photo library permission prompt → Saved confirmation **Instagram: Profile Nametag Card** (ios, 3 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/3592) > Happy path: A logged-in user opens a nametag/QR card for their profile to let others follow them. The UI presents a large scannable card with the username, a style selector ("COLOR") and brief step hints at the top. The user can toggle style options, tap the card background to customize visuals, then share the profile using the system share sheet or copy a profile link. Entry: user opens the nametag/QR screen from their profile. Exit: user closes the screen (X) or uses Share / Copy link. Assumptions: user is authenticated and viewing their own profile's nametag; device has standard share and clipboard capabilities. Steps: Show scannable nametag → Style selection hint → Share or copy link **Instagram: Reel QR share & color selector** (ios, 3 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/3591) > Happy path: A logged-in user viewing a Reel opens the Reel's QR-code sharing modal, previews the QR code, chooses a color theme from a horizontal swatch list, and optionally saves the QR image to the device or dismisses the modal. Steps: Open QR share modal → Select second color swatch → Select fourth color swatch **komoot: Share Profile (QR + Copy Link)** (ios, 3 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8368) > Happy path: A logged-in user views another user's profile and uses the top-right share control to open a QR-code modal that exposes a scanable profile link plus two actions: Copy link and Share profile. The user taps Copy link and receives a transient confirmation banner. Steps: Profile overview → Profile QR and actions → Copy confirmation **PayPal: Seller QR checkout settings** (ios, 10 screens, Finance) [Refero](https://refero.design/flows/7231) > Happy path: a seller views their on-screen QR code and opens an “Optional checkout features” sheet to add a fixed amount and enable tipping presets. The seller enables "Add an amount", types a value using the numeric keypad, enables "Tipping option", chooses whether tips are percentage-based or fixed-amount and edits the three tip presets, then applies the settings. The system shows a loading state, confirms via a toast, and updates the in-app QR to show the selected fixed amount. Steps: View Seller QR → Open Optional features → Enable Add an amount → Focus amount input → Enable Tipping option → Choose tipping mode → Edit tip presets → Apply Settings → Applying settings (loading) → Confirmation and updated QR **PayPal: Tip Jar QR Setup** (ios, 3 screens, Finance) [Refero](https://refero.design/flows/7232) > Happy path: A logged-in seller (Marcus Bennett) creates/activates a standalone Tip Jar QR code, configures preset tip amounts, then views the generated QR code which can be edited or downloaded. Steps: Tip Jar Activation → Configure Tip Options → Tip Jar QR Display **komoot: Invite participants and manage QR invite** (ios, 5 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8353) > Happy path: A user (assumed organizer/owner and signed in on a mobile device) opens the Invite people screen to add participants to a group activity. From this screen they can: search for people or email, share or show a QR code, open QR/link options, revoke/regenerate the QR invite, and invite individuals from a suggested contacts list. Entry: Invite people screen. Exit: User either returns (back) or continues after generating a new QR (toast confirms success). Assumptions: user is authenticated and has permission to manage invites; device is a mobile OS (iOS-like chrome/status bar visible). Steps: Invite people screen → QR options revealed → Invite access info → Revoke confirmation → Regenerate feedback **Linktree: Linktree QR Code Customization** (web, 6 screens, Business, Social Networking) [Refero](https://refero.design/flows/12396) > Logged-in Linktree creator on the Links dashboard opens the share/QR tools from the top-right controls. A share panel appears with the public Linktree URL, a copy action, a scannable QR code, and platform shortcuts. From there, the user opens QR Code details, then Customize QR code to change QR color, hide/show the Linktree logo, and add a custom logo. The main exit points shown are returning to the QR Code screen or using download options for PNG/SVG. Assumption: the user is authenticated and has access to their Linktree profile settings on desktop. Steps: Open Share Panel → View QR Code → Customize QR Color → Toggle Branding → Add Custom Logo → Return to QR Export **Reown: MetaMask QR Connect** (web, 2 screens, Development Tools, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/5703) > Happy path: From the AppKit Demo left panel the user opens a wallet connection modal for MetaMask. The modal presents a large QR code (with Mobile/Browser toggle), instructions to scan, and a 'Copy link' affordance. The user either scans the QR with their phone or taps Copy link. When Copy link is used the UI shows an immediate confirmation ('Link copied') while keeping the modal visible and leaving sharing/reset controls available. Entry: user triggers the wallet/connect action from the demo panel. Exit: user either completes connection externally (by scanning) or copies/shares the connection link and closes the modal. Assumptions: user is on a desktop/browser viewing the demo, has clipboard permissions, and may have a phone available to scan the QR. Steps: Open MetaMask QR Modal → Link Copied Confirmation **Revolut: Phone number sign-in** (web, 4 screens, Finance, Business) [Refero](https://refero.design/flows/1322) > Desktop welcome / sign-in screen for returning users. Entry: a user lands on the "Welcome back" screen (not signed in) and can authenticate using their phone number, a QR code, or alternative social/email providers. The main happy path shown: user opens the country selector, searches or selects their country code, enters their phone number, then the Continue button becomes enabled so they can proceed. Assumptions: user is on a desktop/laptop, not currently authenticated, and has or remembers the phone number associated with their account. Exit: after tapping Continue the UI will submit the phone number and progress to the next authentication step (e.g., verification or sign-in handoff). **Instagram: Post Share — QR Generator** (web, 3 screens, Social Networking, Entertainment) [Refero](https://refero.design/flows/889) > Happy-path: A signed-in desktop user views a post in a lightbox/modal and uses the post's share control to open the share menu, then selects the QR code option to generate and download a scannable QR for that post. Steps: Post Lightbox → Share Menu → QR Code Generator **Telegram: Telegram web sign-in** (web, 6 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/564) > Happy path: A user arrives at Telegram's public landing page, chooses to sign in to the web/desktop client, and authenticates either by scanning a QR code with their phone or by entering their phone number to receive a confirmation code. Steps: Landing / Entry → QR Code Login Screen → Phone Number Entry → Confirmation Code Entry → Country Selector → Authenticated Chat View **Riverside: Snapshot Sharing** (web, 3 screens, Productivity, Entertainment, Collaboration Tools) [Refero](https://refero.design/flows/4675) > Happy-path: A signed-in desktop user opens a project’s snapshot gallery, selects a snapshot from a carousel, and shares it. The flow is entirely modal-driven: first the user picks a snapshot (carousel view) then chooses Share which opens a social-share modal with multiple destinations and actions. If the user selects "Share via mobile," a QR-code modal is shown for scanning with a phone. Steps: Pick a snapshot → Share options → Share via mobile (QR) **Linktree: Digital Business Card Setup** (web, 5 screens, Business, Social Networking) [Refero](https://refero.design/flows/12395) > The user starts in the Linktree Links dashboard and opens the sharing panel. They select the Cards option under My platforms, review an introduction to digital business cards, and continue to the card editor. The user updates the displayed name, chooses a background color, and proceeds to the QR-code view for saving the card to Google Wallet. The flow assumes the user is signed in to Linktree and has access to the Links dashboard. The visible flow ends on the QR-code screen. Steps: Open Sharing Options → Review Card Introduction → Customize Card Details → Confirm Card Appearance → Save Card with QR Code **TikTok: Sign up and verify via mobile** (web, 13 screens, Video Streaming, Social Networking, Entertainment) [Refero](https://refero.design/flows/1345) > Happy-path sign-up and authentication flow for an unauthenticated desktop user. Entry: user clicks Log in / Sign up from the top-right and is shown a modal. The user selects Sign up → chooses 'Use phone or email' → switches to email sign-up → enters birthday, email, password, and a 6-digit verification code → presses Next → is asked to continue authentication on the mobile TikTok app via a QR code → after scanning on their mobile, the desktop session becomes authenticated and the user lands on the main For You feed. Assumptions: user is on a desktop browser, not signed in, can receive verification codes by email (or phone) and has the TikTok mobile app to scan the QR for final confirmation. **Revolut: Download Revolut App via SMS or QR** (web, 3 screens, Finance, Business) [Refero](https://refero.design/flows/79) > User initiates the process to download the Revolut app by clicking 'Get the app' on the main page. A modal opens presenting two options: scan a QR code or enter a mobile phone number to receive a download link via SMS. The user inputs their phone number and submits it. Upon submission, the system validates the phone number and, if successful, shows a success confirmation message. The flow starts with the user on the homepage and ends with confirmation that the download link has been sent. Steps: App Download Modal Opened → Phone Number Entered → Download Link Sent Confirmation **IKEA: Share design via QR** (web, 4 screens, Marketplace, Home & Decor) [Refero](https://refero.design/flows/8245) > Happy path: A signed-in user works in the 3D room canvas, opens the main navigation to access save/share and other project actions, selects Share, and generates a shareable QR code so the design can be opened on another device. Steps: 3D Room Canvas → Main Navigation → Share Design Options → QR Code Share **Uber: Sign-in to access rentals** (web, 6 screens, Business, Real Estate) [Refero](https://refero.design/flows/442) > Happy-path summary: The user starts on the Uber marketing/entry screen, navigates to the sign-in prompt, and is offered multiple login methods (phone/email, Google, Apple, or QR code). The QR modal appears with scanning instructions when chosen. An email/phone submission can trigger an account-locked modal with options to restart or contact support. The final screen shows the Rentals landing (search controls and partner brands) where the header indicates a logged-in user. **Ballpark: Share & Protect Test Link** (web, 8 screens, Business, Productivity, Communication Tools) [Refero](https://refero.design/flows/2658) > Happy path: From the survey/build editor the user opens the "Share test" modal to distribute a test by link or QR and optionally protect it with a password. Steps: Survey editor - open share → Share modal - link visible → Show QR code → Enable password protection → Enter password (initial) → Password validation error → Set valid password → Post-save state **Resend: MFA Enrollment** (web, 4 screens, Development Tools, Communication Tools) [Refero](https://refero.design/flows/10641) > Logged-in user (admin) is on the account Settings > Authentication area and chooses to add multi-factor authentication. The flow opens a modal that displays a QR code and six-digit input fields for an authenticator app. After entering the code and submitting, the modal closes and the Authentication section updates to show MFA as Enabled with an option to Revoke Access. Steps: Settings — Authentication overview → Setup modal — QR & instructions → Verify code — enter token → Settings — MFA enabled confirmation **IKEA: Add co-registrant to gift registry** (web, 4 screens, Marketplace, Home & Decor) [Refero](https://refero.design/flows/8267) > Happy path: A logged-in registry owner viewing their gift registry opens the "Invite to collaborate" drawer, enters a co-registrant name, and creates an invitation. The UI generates a unique invitation URL and displays a QR code and a copy-button for sharing. Steps: Registry overview → Open invitation drawer (empty state) → Enter co-registrant name → Invitation link generated ##### Task: access review permissions audit log **TIDAL: Track Report Reason Selection** (ios, 3 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/13140) > A mobile modal for reporting a track, starting with a searchable list of report reasons and ending with a selected reason shown after search. The user opens the report dialog, searches or browses categories, and selects the matching reason. Steps: Browse Report Reasons → No Match Search State → Filtered Result **Denim: Gallery permissions onboarding** (ios, 5 screens, Design & Media Editing) [Refero](https://refero.design/flows/5929) > Happy-path discovery and permission gating inside the Gallery tab on an iPhone. Entry: user opens the app and lands on the Gallery tab (bottom navigation). The screen shows seasonal hero cards, a “Get the Best Experience” section with action cards (Connect Account — Apple Music, Enable Alerts), and a Free Artworks carousel below. Two system permission dialogs can be triggered from the Gallery: (1) Apple Music / media library access and (2) push notifications. Exit: user remains in the Gallery; if permissions are granted the app can personalize covers from playlists and schedule alerts. Steps: Gallery Home → Apple Music / Media Permission Prompt → Gallery (after music permission) → Push Notifications Permission Prompt → Gallery with Free Artworks **TikTok: Activity Status Visibility** (ios, 2 screens, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/2798) > Happy path: A logged‑in mobile user opens the Inbox to review notifications and then opens the activity status information modal to review or change who can see their active/last‑active presence. Steps: Inbox / Notifications → Activity Status Modal **Product Hunt: Product Team Access Request** (web, 3 screens, Business, Social Networking) [Refero](https://refero.design/flows/11461) > A logged-in user views Notion’s Product Hunt profile on the Team tab and requests access to be added to the product team. Because the product page already has an owner, the system opens an access-request modal. After the user submits the request, the modal confirms that the request was sent and the team section changes to indicate a pending request. The flow begins on the product team profile and ends with the request awaiting owner approval. Steps: Notion Team Profile → Access Request Prompt → Request Confirmation **fal: Recent History Preview Visibility** (web, 2 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/13542) > The user opens the Recent History area while authenticated and reviews recent API requests across endpoints. Results are displayed in a four-column card grid with generated previews, request metadata, timestamps, status codes, prompts, and actions such as Playground, Copy prompt, and Model Analytics. The user can use the Show preview toggle to switch between visible generated images and hidden preview placeholders. The flow ends with the history list displayed in the selected preview state; pagination controls and the site footer remain available below the results. Steps: Review History With Previews → Hide History Previews **District: Event Feedback Submission** (ios, 5 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12916) > Happy-path mobile feedback journey after attending an event. The user enters a rating screen for a specific event ('Sip and Act at 3 Art House'), selects an overall sentiment, optionally rates individual aspects, adds a text comment, and submits. Steps: Overall Rating → Detailed Ratings → Comment Entry → Submission Confirmation → Completion State **Podimo: Podcast Comment Moderation** (ios, 3 screens, Entertainment) [Refero](https://refero.design/flows/12274) > Happy path in a podcast comment thread on mobile. The user is already viewing a specific episode’s comments, can read existing replies, post a new comment, and open the report action for a comment. After submitting the report, a confirmation sheet appears with a success message, a link to community guidelines, and a Done button to dismiss it. Steps: Comment Thread → New Comment Posted → Report Confirmation **Dropbox: Activity Log Filtering** (web, 4 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/13064) > An admin reviews the Activity audit log in Dropbox Admin Console and narrows results using filter chips. The entry point is the Activity section from the left navigation. The user can inspect event rows, open the People filter, select a person, and see the table update to show only matching activity. The exit point is either continuing to refine filters or creating a report from the filtered results. Steps: Activity Log Overview → People Filter Picker → Person Filter Applied → Filtered Activity Results **Stemz: Record Audio Import** (ios, 5 screens, AI Tool, Music Streaming) [Refero](https://refero.design/flows/12977) > Happy-path flow for importing a track by recording audio on a mobile device. The user opens an Import Track chooser, selects Record Audio, grants microphone permission, records audio, stops the recording, and reaches a review state where they can delete, play back, or continue. Steps: Import Source Selection → Microphone Permission → Recording Start → Active Recording → Recording Review **komoot: Revoke Pending Invitation** (ios, 3 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8354) > Happy path: a logged-in organizer opens the Activity participants screen on a mobile device to review the organizer and pending invites. The screen lists the Organizer section and an "Invite sent" list with one or more pending email invites. The organizer taps the remove icon (X) next to a pending invite, which opens a modal asking to confirm revocation. If the organizer confirms, the pending invite is removed from the list and the view is updated. Steps: View participants list → Confirm revocation → Updated participants list **Roots: Developer Tools Logs** (ios, 3 screens, Productivity) [Refero](https://refero.design/flows/5765) > Happy path: a developer opens the in-app Developer Tools to inspect chronological log events, narrows the list by entering a short identifier (partial UUID) in the Filter field, and refines results by selecting a category chip (e.g., Daily Limit, Games). Steps: Developer Tools — initial log list → Filter by identifier — Daily Limit selected → Switch category to Games **Meta AI: Gmail Connection Authorization** (ios, 6 screens, AI Tool) [Refero](https://refero.design/flows/13355) > A logged-in user asks an AI assistant to identify active subscriptions and assess which are useful. Because Gmail is not connected, the assistant explains that email access is required and presents Gmail and Outlook connection options. The user selects Gmail, reviews Meta’s connection disclosure, continues to Google sign-in, reviews the requested Gmail permissions, and returns to the assistant with Gmail shown as connected. The depicted flow ends before the subscription audit results are displayed. **Dropbox: Activity Log Filtering** (web, 4 screens, Collaboration Tools, Business, Productivity) [Refero](https://refero.design/flows/13036) > A logged-in user is viewing a team file storage area with an activity sidebar open on the right. The user opens the Activity filter menu, selects an event type filter, and the activity feed updates to show matching events. ##### Task: accessibility settings text size **XChat: Appearance Settings** (ios, 2 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/11675) > Compact mobile settings screen for configuring app appearance. Entry: user navigates into Appearance from a higher-level Settings area. The screen shows three theme options (System, Day, Night), a Text Size control with a horizontal slider, and a live preview of chat bubbles. Exit: user returns via the top-left back control. Assumptions: user is on a modern mobile device (iPhone-style status bar visible) and is in the app's settings context with permission to change display preferences. Steps: Theme selection + text size preview → Adjust text size and show reset **Reeder.: Reader typography settings** (ios, 2 screens, News) [Refero](https://refero.design/flows/3576) > Settings screen that lets the user customize typography for the in-app article viewer (Reader View, Feeds, and in-app content). Steps: Reader Typography Settings — Default state → Reader Typography Settings — Adjusted values **XChat: Appearance Settings** (ios, 2 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/11676) > User is on an Appearance settings screen where they can choose between three visual modes (System, Day, Night) and adjust text size with a slider. A message-preview area updates to reflect the current selection and text-size setting. Steps: Choose Mode — System (light preview) → Choose Mode — Night (dark preview) **Substack: Article typography customization** (ios, 9 screens, News, Social Networking) [Refero](https://refero.design/flows/4469) > Happy path: user opens an article, opens the article Options sheet, selects Display Settings, chooses a different article font and increases text size, sees an immediate live preview of changes in the article, then closes the settings. Steps: Article view with action bar → Open Options sheet → Open Display Settings → Select a different font (Bitter) → Select another font (Söhne) → Increase text size (first increment) → Increase text size (further increments) → Preview final typography → Close settings and view updated article **Apple Books: Reading Appearance Settings** (ios, 3 screens, Books) [Refero](https://refero.design/flows/11929) > A reader is viewing a book and opens a Themes & Settings bottom sheet to adjust reading appearance. The user remains in the reading view while changing typography, theme, brightness, and other display options. The sheet can be dismissed with the close button, returning to the current page with the selected reading settings applied. Steps: Open Themes Panel → Adjust Reading Preferences → Preview Updated Theme **Google Translate: Translation Home Settings** (ios, 4 screens, Reference, Utility) [Refero](https://refero.design/flows/12137) > This is a translation/conversation home screen showing stacked translation cards and quick controls. The user enters from the Home screen and can review translated text in multiple language directions, open settings, adjust text size, toggle auto playback, then dismiss the settings sheet and return to the main translation view. Assumptions: user is already on the translation home screen and has access to the settings icon; device is mobile and logged-in state is not relevant. Steps: Translation Home → Settings Sheet → Text Size Adjustment → Playback Enabled **Apple Books: Reader Theme Customization** (ios, 8 screens, Books) [Refero](https://refero.design/flows/11931) > A reader is viewing a book chapter and opens a Themes & Settings sheet to change reading appearance. The entry point is the reading screen; the exit point is returning to the reader with the selected theme and text settings applied. Steps: Themes & Settings Sheet → Customize Theme Home → Font Options → Font Selection List → Updated Theme Preview → Layout Controls → Justify Text Option → Apply Theme to Reader **Verse: Text overlay editor** (ios, 2 screens, Social Networking, Design & Media Editing) [Refero](https://refero.design/flows/3987) > Happy path: A user composes or edits a block of text that will appear as a styled overlay on a black canvas (likely a story/graphic editor). **Grok: Appearance Settings** (ios, 3 screens, AI Tool) [Refero](https://refero.design/flows/6097) > Happy path: A logged-in user opens the Settings modal from the main app and navigates to Appearance to change theme and font size. Steps: Settings (Profile & Sections) → Appearance — Light Preview → Appearance — Night Mode Preview **Memotron: Appearance Customization** (web, 3 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/8344) > Happy-path: A signed-in desktop user opens the Settings modal and customizes visual appearance options. The modal contains a left-side settings navigation and a right-side Appearance panel where the user can choose font, toggle "Sync theme with system", pick light and dark color schemes (via swatches), and enable a blurred background for popups. Steps: Appearance panel open (initial state) → Change light color swatch → Enable blurred background **Excalidraw: Text properties editing** (web, 2 screens, Collaboration Tools, Productivity) [Refero](https://refero.design/flows/1968) > Happy path: A user is editing an existing diagram/whiteboard and wants to change the appearance of a selected text element (a quote). Steps: Select text object → Change font family **Canva: Brand Typography Editor** (web, 6 screens, Design & Media Editing, Collaboration Tools) [Refero](https://refero.design/flows/8031) > Happy path for adjusting a brand kit's predefined typography role (Title) from the Brand > Fonts area. Entry: user is signed in, has a brand kit selected (kit) and navigates to Brand > Fonts on a desktop. The user expands a preset (Title), uses the inline controls to pick a font family, heading role, size, and toggles weight/italic, then confirms. Exit: the card collapses and the Fonts list shows the updated preview for the Title entry. Steps: Fonts list (overview) → Open inline editor for preset → Open font family dropdown → Set heading type mapping → Pick font size → Confirm edits and update preview **Figma: About section text edit** (web, 9 screens, Design & Media Editing, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/5037) > Happy path: A designer opens a desktop design file, selects the 'About' text block on the canvas, replaces or edits the paragraph copy, and updates typography (font family, weight, size) using the right-hand properties panel. Steps: Select About paragraph → Enter inline edit mode → Clear existing copy → Paste or type new paragraph → Open font family menu → Browse and pick font → Set font weight → Adjust font size → Review and finalize **Shopify: Change theme typography** (web, 5 screens, Business) [Refero](https://refero.design/flows/5416) > Happy path: A logged-in merchant opens the theme editor, navigates to Theme settings → Typography, opens the font picker, selects a new font for headings (and/or body), confirms the selection, and returns to the Theme settings where the new font is applied in the live preview. **Medium: Image Alt Text Editor** (web, 3 screens, Social Networking) [Refero](https://refero.design/flows/11125) > Happy path: A logged-in author working on a draft selects an inserted image in the article editor, opens the Alternative text dialog, types a brief description, and saves it. Steps: Image selected in article → Open Alternative text dialog → Enter and save alt text **ElevenLabs: Caption Styling and Font Selection** (web, 3 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/8686) > Happy path: A user opens the Captions panel to change the visual style of subtitles while reviewing a video. They inspect the caption controls, open the font-family dropdown, select a new font, and increase the caption size. The preview on the right updates in real time to reflect style changes; the caption segments remain visible in the timeline at the bottom. Steps: Open Captions Panel and Inspect Controls → Open Font Family Dropdown → Select Font and Increase Size **Skiff: Appearance Settings** (web, 2 screens, Communication Tools, Paperwork, Scheduling Tools) [Refero](https://refero.design/flows/1910) > User opens the Appearance section in the account settings to choose how the email interface looks. The modal/panel shows three theme-preview cards (Light, Dark, System) plus controls for Time format and Date format. In the happy path the user clicks a theme preview (e.g., Dark) and the interface appearance updates to reflect that selection. Steps: Appearance panel — Light selected → Select Dark theme **Threads: Change appearance (theme) settings** (web, 4 screens, Social Networking, Entertainment) [Refero](https://refero.design/flows/10671) > Happy-path: A logged-in user on a desktop browser is viewing their "Following" feed and wants to change the site's visual theme. They open the main/left menu, select "Appearance," choose a theme option (Light / Dark / Auto), and the page updates immediately to reflect the chosen appearance. Entry: user is on the feed (Following tab). Exit: the feed UI updates to the selected theme (dark in this sequence). Assumptions: user is authenticated, using a wide/desktop viewport, and has access to the left navigation/hamburger menu. Steps: Feed (Following) - Default appearance → Main menu opened → Appearance picker shown → Dark theme applied **Linear: Account Preferences and Theme Settings** (web, 3 screens, Business, Productivity, Project Management, Collaboration Tools) [Refero](https://refero.design/flows/6688) > Happy path: A signed-in desktop user opens the Preferences area from the left navigation, reviews and updates General settings (default home view, display names, first day of week, emoticon conversion), and then configures Interface and theme options (sidebar behavior, font size, pointer cursor, and interface color theme). The flow shows the theme dropdown expanded (light-themed UI) and then the interface with a dark theme applied. Steps: Preferences — General & Theme (Light) → Preferences — Theme Dropdown Expanded (Light) → Preferences — Dark Theme Applied **Canva: Text font selection for vertical video** (web, 4 screens, Design & Media Editing, Collaboration Tools) [Refero](https://refero.design/flows/8043) > Happy path: User is editing a vertical video in a visual editor, opens the Text panel, then searches for a specific font style ("neon") and inspects the available fonts and font-combination tiles. Entry: user already in the editor with a media clip on the canvas and timeline selected. Exit: the editor displays filtered font results and font-combination previews (ready for the user to select and apply to a text element). Assumptions: user is signed into the editor, left-side Text panel is visible, a video clip is on the canvas and selected in the timeline, and the editor provides live previews for text and fonts. Steps: Open Text panel / default view → Open recent search / search field focused → Enter search term → View filtered fonts and combinations **Mapbox: Map typography font selector** (web, 2 screens, Development Tools, Map & Navigation) [Refero](https://refero.design/flows/9665) > Happy-path: A signed-in user, working in the map style editor, opens the Fonts chooser from the top bar to change the map’s typography. The user scans the available font families, uses the search field to find a specific family (‘Rubik’), then reviews available font weights/variants and can choose one. Steps: Open Fonts chooser → Search and inspect family variants **Linear: User Preferences Center** (web, 3 screens, Business, Productivity, Project Management, Collaboration Tools) [Refero](https://refero.design/flows/6687) > Logged-in user opens the Preferences area from the left navigation to adjust personal settings. The screenshots show a single, scrollable Preferences screen divided into sections: General (default view, display names, date start, emoticons), Interface and theme (sidebar customization, font size, pointer cursor, light/dark/system theme), and Desktop application (open in desktop, notification badge, etc.). Steps: General preferences → Interface and theme adjustments → Desktop application options **Replo: Responsive preview control** (web, 4 screens, Design & Media Editing, Business) [Refero](https://refero.design/flows/3893) > Happy path: a user working inside a visual site editor opens the responsive preview control, inspects available device presets, and switches the editor's preview to a chosen device/pixel width. Steps: Preview control visible → Open preview dimensions → View device presets → Select device / apply width **Airtable: Adjust Grid Row Height** (web, 2 screens, Project Management, Collaboration Tools, Scheduling Tools, Paperwork, Productivity) [Refero](https://refero.design/flows/1562) > Happy path: a user viewing a Records grid (Creative Asset Development > Concepts) opens the view controls and changes the row height to improve readability of multi-line descriptions and image thumbnails. Steps: Open Row Height Menu → Select Row Height **The Washington Post: Change display theme** (web, 4 screens, Magazines & Newspapers, News) [Refero](https://refero.design/flows/11183) > Happy path: A signed-in reader on a desktop/laptop opens the account/person menu from the header, changes the site display theme from Light/Match Device to Dark, and returns to the homepage rendered in the Dark theme. Steps: Home (Light) - initial state → Account panel open → Select Dark theme → Home (Dark) - result **HBO Max: Subtitle style customization** (web, 4 screens, Entertainment, Video Streaming) [Refero](https://refero.design/flows/10261) > Happy path: a signed-in user opens Settings, selects "Subtitle Style," adjusts caption appearance using dropdown controls with a live preview, saves changes, and is returned to the Settings list with a confirmation toast. Steps: Settings → Subtitle Style (default view) → Subtitle Style (edit preview) → Settings (confirmation) **Dona: Task dashboard theme switch** (web, 4 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/5664) > Happy-path: A signed-in user views their task dashboard, opens the global settings menu (kebab / three-dot), and changes the app appearance theme from Light → Dark → Black. Steps: Dashboard — Light theme (default) → Open settings — Theme options visible → Select Dark theme → Select Black theme (high-contrast) ##### Task: account closure **Voicenotes: Account Deletion Request** (ios, 6 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14276) > The flow starts with a signed-in user on the Settings screen. The user opens Account, selects Delete account, and confirms the intent in a native confirmation dialog. The system then opens an external Typeform deletion request in an in-app browser. The user reviews the permanent-deletion warning, completes the required account email, full name, and deletion reason fields, and submits the form. The flow ends on the external form’s completion screen. The screenshots do not show the account being immediately deleted or the user being signed out; they only confirm submission of the deletion request. **HYPE: Account Deletion** (ios, 8 screens, Lifestyle, News) [Refero](https://refero.design/flows/12254) > An authenticated user starts on the Account section, opens Profile, and scrolls to the account-management actions. They select Delete Account, review the permanent-deletion warning, choose a reason, and enter additional feedback when selecting Other. After continuing, a final confirmation modal asks whether to permanently delete the account. Confirming returns the user to the Hype welcome screen, where login, sign-up, or guest access options are available. The screenshots imply that the account deletion succeeds, although no explicit success message is shown. Steps: Account Overview → Profile Details → Profile Account Actions → Deletion Warning → Deletion Reason → Custom Feedback → Final Deletion Confirmation → Signed-Out Welcome **PayPal: Account deletion and data removal** (ios, 13 screens, Finance) [Refero](https://refero.design/flows/7271) > A logged-in user navigates from Privacy/Manage your data to request deletion of their PayPal data and close their account. The flow: user selects "Delete your data / close account", reads eligibility and pre-steps, gives explicit consent, reviews a notice about linked services, chooses an email for updates, provides exit reasons (optional), completes a quick security check (SMS code), and receives a final confirmation that the request was received. Entry: Manage your data settings. Exit: confirmation screen stating request received and that an email will follow. Assumptions: user is authenticated, has a verified mobile number for SMS, has an email on file, and can complete any outstanding account requirements (balances, limitations) outside this flow. Steps: Manage your data list → Deletion info & prerequisites → Consent acknowledgement (unchecked) → Consent acknowledgement (checked) → Linked services notice → Choose email for updates → Exit feedback (reason selection) → Quick security check → Enter security code (empty) → Submit security code → Processing / success animation → Request received confirmation → Final status / completion **DailyArt: Account Deletion** (ios, 6 screens, Education, Magazines & Newspapers) [Refero](https://refero.design/flows/3344) > Happy path: A signed-in user opens Settings, navigates to Account Details, chooses Delete Account, selects a reason, confirms intent on a warning screen (with subscription notice) and completes deletion which signs them out. Steps: Settings - Signed In → Account Details → Delete Account — Reason (initial) → Delete Account — Reason Selected → Final Confirmation & Subscription Notice → Settings - Signed Out **ElevenReader: Account deletion and confirmation** (ios, 5 screens, AI Tool, Books) [Refero](https://refero.design/flows/8988) > Happy path: A logged-in user navigates from their account overview into the settings list, selects the 'Delete account' option, confirms the deletion by typing their email address exactly, and taps the destructive CTA. The UI requires an exact email match before enabling the red Delete action. After tapping Delete, the user receives an inline success confirmation banner and is effectively signed out. Steps: Account overview → Settings list (account actions) → Delete confirmation (initial) → Delete confirmation (email entry) → Deletion success feedback **Pinterest: Account Deletion** (ios, 4 screens, Reference, Inspiration Board, Social Networking, Lifestyle) [Refero](https://refero.design/flows/2470) > Happy path: A signed-in mobile user navigates from Account management to permanently delete their account. From Account management they choose "Delete your data and account", read a final confirmation screen, tap Continue, and a bottom-sheet modal asks for a reason. After selecting a reason the primary action (Send email) becomes enabled; tapping it sends an email to the account address with the final step to complete deletion. Steps: Account management → Delete confirmation screen → Reason selection (initial) → Reason selected and confirm **Tinder: Account Pause and Deletion Options** (ios, 3 screens, Dating & Relationships) [Refero](https://refero.design/flows/3476) > Happy path: A logged-in user navigates to Settings -> Delete Account to either pause or permanently delete their account. The Settings list contains a 'Delete Account' entry. Tapping it opens a screen that offers 'Pause Account' (prominent gradient CTA) and a secondary 'Delete Account' option. When the user taps 'Pause Account', the app hides the user's profile, returns them to the main discovery view, and shows a clear paused-state screen with a 'Jump back in' CTA to unpause. Steps: Settings list → Pause vs Delete decision → Paused state confirmation **Kit: Account deletion via in-product support** (web, 4 screens, Business, Communication Tools) [Refero](https://refero.design/flows/2302) > Happy path: A signed-in user on desktop is in their Settings > Login Details screen, opens the on-screen support chat, and requests permanent account deletion. The support agent replies asking for confirmation and additional details. After no user response, the agent cancels the account and closes the ticket. When the user later attempts to access the product they land on an account-closed screen that offers a reinstatement CTA and contact link for Customer Success. Steps: Open support chat and request deletion → Support asks for confirmation and details → Agent follows up and closes ticket → Closed-account landing with restore option **Grammarly: Account Deletion Confirmation** (web, 2 screens, Paperwork) [Refero](https://refero.design/flows/1883) > Happy-path: A logged-in desktop user navigates to Account Settings (left navigation). From the Profile/Account Settings area they select the Delete Account action. A centered confirmation modal overlays the screen, shows the account email for context, and offers two choices: cancel and return to saved documents, or continue with account deletion. Steps: Account Settings (Profile) → Delete Account Confirmation Modal **Dribbble: Account deletion and redirect** (web, 3 screens, Communication Tools, Job Search) [Refero](https://refero.design/flows/783) > Happy path: A signed-in user goes to Settings → General, clicks the 'Delete Account' entry in the left navigation, is shown a centered confirmation card with a strong warning and two actions (Nevermind / Delete my account), and after confirming deletion they are redirected to the public homepage where a banner or top message indicates the account is being deleted and the user is logged out. Steps: Settings — General → Delete confirmation card → Homepage — post-deletion **Digg: Account Deletion** (web, 4 screens, News) [Refero](https://refero.design/flows/14353) > The user starts logged in on a desktop Digg experience with the Account settings modal open on the Security section. They select the account deletion option, review the warning, type the required confirmation phrase, and submit the deletion request. The flow then returns to the public home feed, indicating the account action completed or the user was redirected after submission. The screenshots do not show an explicit success message, error state, or sign-out confirmation. Steps: Security Settings → Deletion Warning → Confirm Deletion → Home Feed Redirect **Product Hunt: Permanent Account Deletion** (web, 5 screens, Business, Social Networking) [Refero](https://refero.design/flows/11470) > A signed-in user opens their account Settings and chooses to permanently delete their account and associated data. The system presents a confirmation modal requiring the user to enter the displayed username, then processes the deletion. After completion, the user is returned to the public Product Hunt home experience in a signed-out state. The screenshots do not show an error state or an explicit completion message. Steps: Account Settings → Deletion Confirmation → Username Verification → Deletion Processing → Public Home ##### Task: activity history timeline **Vinyls: Monthly Listening History** (ios, 2 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/8935) > Summary: A mobile "History" screen that shows the user's past items grouped by month (example: December, 2025). Entry: user arrives at the History section (from main navigation — not shown). Exit: user can either tap an item (presumably to view details) or interact with the item overflow menu. Assumptions: user has historical items present; the device is a mobile phone in portrait orientation; the screenshots show two layout modes (grid and list). Happy path: user lands on History → scans items for a month in grid view → uses the top-right toggle to switch to list/timeline view → optionally uses the search box to filter items or taps an item or its overflow menu for more actions. Steps: Monthly History — Grid → Monthly History — List / Timeline **Linear: Edit Issue Priority** (ios, 3 screens, Collaboration Tools, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/6610) > Happy path: A signed-in user views an issue's detail screen, opens the Priority field, selects a new priority from a modal list, and returns to the issue where the activity timeline reflects the change. Entry: user is on the issue detail screen. Exit: the issue detail screen shows the updated activity entry and the Priority value has been changed. Assumptions: the user can edit issue attributes from this screen and the device is a mobile phone with a bottom navigation bar. Steps: Issue Detail — View → Priority Picker — Select → Issue Detail — Confirm Change **Givingli: Send and view a digital card** (ios, 16 screens, Lifestyle, Scheduling Tools) [Refero](https://refero.design/flows/8217) > Happy-path for sending a digital card (Givingli) inside a messaging conversation and then viewing/replying. Steps: Inbox list → Conversation preview → Add a Card - picker → Browse cards grid → Category gallery → Card detail → Personalize - preview → Message editor → Send confirmation → Thread updated with card → Envelope interaction → Swipe to reveal → Card viewer (paged) → Reply composer open → Send reply → Conversation continues **Haptic: Activity timeline** (ios, 3 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/2122) > Vertical chronological activity/timeline list showing individual content cards grouped by date. Each card contains an icon, a small status label (e.g., LISTENED, ADDED, TRAVELED), a title, a secondary line (artist/location), a star-count badge, and a short text preview. The screen is on a mobile device with a persistent bottom bar containing a left menu icon, a prominent central circular action, and a right settings/controls icon. **Telegram: Manage conversation: record, call, select & delete message** (web, 6 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/565) > Happy path: a signed-in user on a desktop/tablet Telegram client opens the Conversations list, selects a contact (Henry) to view the chat, records and sends a voice message, makes an outgoing call, selects one or more messages (including a photo) and invokes Delete. The flow entry point is the main chat list (left column). Exit point is the Delete confirmation modal (final step) — after confirming, the message(s) will be removed from the conversation. Assumptions: user is authenticated, has existing conversations, and is using a wide (two-column) layout where the chat list is always visible on the left and the conversation is shown on the right. Steps: Conversation list (empty state) → Open contact conversation → Record voice message → Make an outgoing call → Select message(s) or media → Confirm deletion **ElevenLabs: Narration editing and export setup** (web, 54 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/8677) > Happy path: A user is editing an imported text document to produce narrated audio, using both generated voices and a recorded actor take, managing per-segment voice settings, collaborating with comments, and preparing export metadata. Steps: Document selection and playback → Switch text type to plain Text → Choose heading level → Open voice override controls → View generation history → Open Actor mode modal → Start recording (countdown) → Recording in progress → Recording preview and generate → Actor override settings → Open Comments panel → Expand a comment thread → Compose a reply → Reply submitted → Comment status update → Open global menu → Settings - General tab → Settings - Export tab → Fill metadata and toggle normalization → Add cover image and description **Zendesk: Handle incoming support ticket** (web, 4 screens, Business) [Refero](https://refero.design/flows/1351) > Happy-path: An authenticated support agent starts at the dashboard, selects an open ticket from the 'Tickets requiring your attention' list, reviews the requester and conversation, composes a public reply, optionally consults the Knowledge sidebar, inspects or opens the Apps panel, then submits the ticket as open. **Steel: Create and Monitor Browser Session** (web, 7 screens, AI Tool, Development Tools) [Refero](https://refero.design/flows/14282) > The user starts from the Sessions area of an authenticated browser automation workspace, configures a new browser session, optionally expands advanced settings, and launches it. The system creates a live session, adds it to the sessions list, and opens a session detail view where the user can monitor browser activity and loaded content. The flow begins on an empty Sessions list and ends with an active session displaying navigation traces and the rendered target content. The screenshots imply a desktop environment and a user with permission to create and manage sessions. **Leonardo.Ai: AI Music Generation** (web, 7 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/12886) > Logged-in user in Leonardo AI’s AI Creation area starts on the Audio tab, selects a music generation model, enters a text prompt, and generates audio. The interface shows generation controls in a left sidebar, a prompt field across the top, and a results feed in the main area. After generation, the user can preview the waveform, scroll through past outputs, and open a detailed player view with playback controls and item metadata. Steps: Audio Creation Setup → Model Selection → Prompt Entry → Generation in Progress → Results Preview → Output History → Full Player View **Modyfi: In-context Commenting & Attachment Thread** (web, 11 screens, Design & Media Editing) [Refero](https://refero.design/flows/488) > Happy path: A signed-in user adds an in-context comment on a design/artboard, enriches that comment with emoji and an attached image, and then opens/continues the threaded conversation in a larger panel. **DoorDash: Cancel Scheduled Order with Refund and Feedback** (web, 13 screens, Food & Drink, Food Delivery) [Refero](https://refero.design/flows/10316) > Happy-path: A logged-in customer on desktop navigates from Customer Support to the in-progress order, chooses to cancel that scheduled/pickup order, selects a cancellation reason, picks a refund method (credits or original payment), confirms cancellation and sees a cancellation confirmation. After cancellation the product surfaces an in-context feedback modal (2-step) and then the Orders list and order details reflect the refund issuance. Steps: Customer Support Hub → Select In‑Progress Order → Choose Support Action → Select Cancellation Reason → Choose Refund Method (Credits) → Choose Refund Method (Card) → Feedback Prompt — Step 1 → Feedback Prompt — Step 2 (Rating) → Feedback Details → Feedback Submitted (Selected) → Cancellation Confirmation → Orders List Updated → Order Details / Refund Receipt **Zendesk: Guided demo walkthrough** (web, 21 screens, Business) [Refero](https://refero.design/flows/1384) > Happy-path guided product demo that moves a prospective user from a marketing landing area into an in-context interactive demo of Zendesk features, then into agent and sales (Sell) interfaces showing onboarding tooltips and task workflows. **Visitors: People list - period filter** (web, 4 screens, Business) [Refero](https://refero.design/flows/9823) > This flow shows a desktop "People" screen where the user inspects recent contacts and filters them by time period. Steps: People overview (Today) → Open period selector → Empty state for Yesterday → All time results **Reclaim AI: Personal time stats review** (web, 5 screens, Scheduling Tools, Collaboration Tools) [Refero](https://refero.design/flows/4248) > Happy-path: A signed-in user opens the Stats section to review personal time analytics for a selectable date range. **Exa: Usage Date Range Selection** (web, 3 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/12023) > A logged-in user is viewing the Usage section of a management dashboard and adjusting the reporting date range. The entry point is the Usage screen with activity analytics visible; the exit point is the same screen updated to a new date range, with charts and totals reflecting the selected period. Assumption: the user has access to analytics for the current workspace/account and can use the date picker and API key filter. Steps: Usage Dashboard → Date Range Picker → Updated Usage Metrics ##### Task: apply for loan credit **District: Add Card** (ios, 4 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12836) > Happy path for adding a payment card. The user enters from an Add Card screen, optionally chooses to scan the card to autofill details, grants camera access when prompted, confirms or reviews the scanned card number and expiry date, selects a card nickname category, and submits the card. Steps: Add Card Entry → Camera Permission → Card Scan → Review and Save **District: Wallet Balance and History** (ios, 3 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12877) > A wallet/balance screen showing the current District Money balance, an Add money action, and a transaction history list. The user can filter history by All, Debit, Credit, or Refunds. Entry is from the back navigation into the balance/history view; exit is via the back button or by selecting Add money. Steps: Balance Overview → Debit Filter Applied → Credit Filter Applied **Introspect: Friend decision response** (ios, 3 screens, Finance) [Refero](https://refero.design/flows/11952) > Happy path for a recipient responding to a friend’s product decision request on mobile. Steps: Decision Request → Sending State → Completed State **Airbnb: Artist experience request** (ios, 11 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6413) > Happy path for requesting entry to a limited/lottery artist event. Steps: Event detail / Request CTA → Guests selection → Application question prompt → Text entry (keyboard active) → Answer review → Request details summary → Before you submit (rules & process) → Submission (start animation) → Submission animation sequence → Confirmation screen → Exit / Return **DailyArt: Checkout Payment Entry** (ios, 2 screens, Education, Magazines & Newspapers) [Refero](https://refero.design/flows/3332) > This is the final payment step of a mobile checkout flow. The screen lets the user choose a payment method (credit/debit card or PayPal), enter card details (card number, expiration date, security code), read a short privacy notice, check a required terms-and-conditions box, and submit the order via the PLACE ORDER button. Steps: Payment method & card input (initial) → Inline validation error (invalid card) **Lovable: Plans and credits management** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12301) > Happy path for managing a workspace subscription from the Plans & credits area. Entry is the workspace settings sidebar on the Plans & credits screen. The user can review current plan status, compare Pro/Business/Enterprise options, and initiate an upgrade or credit purchase. If upgrading, the flow continues to a Stripe-style checkout where the user enters payment details and confirms the subscription. Exit is the updated Plans & credits screen showing the new active plan and refreshed credit balance. Steps: Current Plan Overview → Checkout and Payment → Updated Subscription Status **On: Apply gift card at checkout** (ios, 5 screens, Marketplace) [Refero](https://refero.design/flows/9603) > A single-screen modal-style mobile checkout where the user selects a payment method, optionally applies a gift card, and completes payment. Steps: Payment selection and gift card collapsed → Enable gift card entry → Enter gift card details → Apply submission (loading) → Validation error shown **Faire: Request Unlisted Product** (web, 5 screens, Business, Marketplace) [Refero](https://refero.design/flows/13083) > Happy path on a brand storefront where a buyer requests products that are not currently listed. Entry is the brand shop/product listing page. The user opens the brand actions menu, chooses Request unlisted product, fills in one product name, URL, and quantity in the modal, then submits. The modal closes, a confirmation toast appears, and the storefront remains visible. Steps: Brand Storefront → Open Request Form → Enter Product Details → Submit Request → Request Confirmed **Lovable: Start Project Prompt** (web, 6 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12303) > A logged-in user lands on the Lovable home screen with a left sidebar, a central prompt composer, and a template gallery below. The user can start by typing or using voice input in the composer, optionally connect tools, then submit the prompt to generate a project. The flow ends when the prompt is fully entered and ready to build/submit, with the system showing the composed request and available build controls. Steps: Home Prompt Start → Template Inspiration → Prompt Refine → Voice Capture → Transcribing State → Prompt Ready to Build **Lovable: Help and Resources Center** (web, 5 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12206) > This is a help and support entry experience for Lovable. The user starts on a public support/resources landing page with navigation, a central help search prompt, resource cards, FAQs, and a build prompt at the bottom. The user can also open a centered search modal, enter a query or numeric text, and review an AI-generated response with suggested follow-up chips. Assumption: the user is on desktop and not required to be logged in to access the visible help content. Steps: Help Landing Page → Search Modal Opened → Query Entered → Search Suggestions → Assistant Response **ElevenLabs: Request License for Iconic Voice** (web, 6 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/8622) > Happy-path: A visitor on desktop browses the ElevenLabs "License Legendary Voices" marketplace, opens a specific voice detail (John Wayne™), opens the in-page modal to request a license, completes the request form (contact + project details + estimated start date), agrees to terms and submits. The flow ends with an in-page confirmation toast. Steps: Marketplace listing / Voice discovery → Open Request License modal → Choose estimated start date → Complete contact and project fields → Acknowledge terms and submit → Show submission confirmation ##### Task: approval request decision **222: Decline invite with feedback** (ios, 7 screens, Lifestyle, Dating & Relationships) [Refero](https://refero.design/flows/9316) > Happy path: A signed-in user views an invite (event title, date/time, location, party composition, payment status) and chooses to decline (No). Tapping No opens a modal asking "why wasn't this experience for you?" with multiple selectable reasons and an "other..." text field. The user can check items, open the keyboard to type a custom reason, then submit by tapping RSVP NO. After submitting, the interface offers to "request another invite" via a compact date-picker overlay so the user can ask for an alternative day. Steps: Invite summary → Decline reason modal → Select reasons → Open custom reason input → Enter custom text → Submit decline (RSVP NO) → Request another invite (date picker) **BeReal.: Manage Friend Requests** (ios, 3 screens, Social Networking) [Refero](https://refero.design/flows/4925) > Happy-path: A logged-in user opens the Friends area and handles an incoming friend request. The user sees a pending request in the Requests tab, taps Accept, then the Requests tab shows an empty state. The user can switch to the Connections tab to confirm the new friend appears under My Friends and browse suggested accounts. **Bond: Like Memory Response** (ios, 2 screens, AI Tool, Social Networking) [Refero](https://refero.design/flows/14400) > The user is viewing a mobile conversation about a shared memory. The screen shows an image, a question from Fred Fox, and Danny Shnap’s response. The user taps the outline heart on the response card, and the icon changes to a filled pink heart to indicate the response has been liked. The flow starts on the conversation view and ends with the reaction visibly applied. The screenshots suggest a mobile device with an active user profile; authentication and permissions are not otherwise visible. Steps: Conversation View → Liked Response **Aboard: Filter documents by type and status** (web, 5 screens, Collaboration Tools, Paperwork) [Refero](https://refero.design/flows/797) > Logged-in user is viewing the Documents area of a product (left nav shows workspace and people sections). The main table lists document name, employee, created date and status. The top of the table has filter chips: a Documents-type selector and a Status selector. The flow shows the user opening the Documents-type dropdown, selecting the 'PitchDeck' type, opening the Status dropdown and selecting 'Approved'. Entry: user lands on Documents list with no specific filters (All). Exit: Documents table updated to reflect applied filters (result is an empty list in this example). Steps: Documents list (default) → Open document-type dropdown → Select 'PitchDeck' document type → Open status dropdown → Select 'Approved' status and view results **Cushion: Cancel Invitation** (web, 4 screens, Communication Tools, Collaboration Tools) [Refero](https://refero.design/flows/12534) > This is a workspace member-management modal opened from settings. The user is logged in as a workspace owner/admin and can view current owners, members, and invited users. The happy path starts on the Members screen, where the user opens the overflow menu for a pending invite, chooses Remove invitation, confirms the action in a destructive confirmation dialog, and sees the invited person removed from the list. The flow ends with a success toast confirming the invitation was cancelled. Steps: Members List → Invitation Actions Menu → Confirm Removal → Updated Members List **Cushion: Resolve Channel Post** (web, 6 screens, Communication Tools, Collaboration Tools) [Refero](https://refero.design/flows/12523) > Happy path starts in the General channel list, where the user opens a post marked for feedback, reviews the discussion, and chooses to resolve it. A modal appears asking for a resolution note; after the user enters the explanation and confirms, the post is marked as resolved and returns to the channel list with a resolved status badge. Assumptions: the user is already signed in, has access to the workspace and channel, and has permission to resolve the post. Steps: Channel Discussion List → Post Detail View → Resolve Confirmation → Submit Resolution → Resolved Post Detail → Channel List With Status **Сontra: Request recommendation from client** (web, 5 screens, Job Search) [Refero](https://refero.design/flows/4873) > Happy-path: A logged-in user on their Contra profile navigates to the Reviews section, opens the "Request a recommendation" flow, fills contact details for a previous client, proceeds to send the request, and returns to the Reviews tab where the request appears as pending. After the recommendation is approved/verified, the profile shows the new review and an updated rating. **WhatsApp: Announcement Replies** (web, 4 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/13211) > A logged-in user is viewing a community’s Announcements section in a desktop messaging interface. The entry point is the announcement list and an announcement post with a visible reply count. The user opens the replies panel for a specific announcement, reviews existing replies, types a short response, and sends it. The panel remains open and the new reply appears in the thread, while the list on the left reflects that the announcement has been replied to. ##### Task: approval workflow **District: Closed Support Rating** (ios, 2 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12841) > A mobile support chat ends after the user’s issue is resolved and the conversation is marked closed. The entry point is an active support conversation with recent agent messages. The exit point is a closed chat state where the user can rate the assistance or start a new chat if they still need help. **Introspect: Item Verification Share** (ios, 4 screens, Finance) [Refero](https://refero.design/flows/11951) > Happy path begins on an Item Overview screen for a saved item, with the user logged in on a mobile device. From here, the user opens the item’s overflow menu, chooses to verify/share the item, and is taken into the system share sheet. The user then selects Messages, which opens a new iMessage draft addressed to a contact with a prefilled prompt and link preview. The flow exits at message composition, ready to send. Steps: Item Overview → Item Actions Menu → Share Sheet → Message Draft **Lovable: QR Asset Label Implementation** (web, 12 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12336) > The user is working in an authenticated project-building workspace with an existing AssetWise asset-tracking project open in preview. They ask the assistant what to do next, select a suggested QR-code labeling feature, and submit the request. The assistant generates a detailed implementation plan covering QR codes, public scan pages, printable labels, files, and out-of-scope items. The user reviews the plan, opens its editing controls, and approves it. The system then processes the approved plan and returns a completed feature summary indicating that QR labels and a public scan page were added. Steps: Choose the Next Project Step → Select a Feature Request → Generate the Plan → View Plan Generation → Review the Proposed Plan → Inspect Technical Details → Open Plan Editing → Format Plan Content → Finalize Plan Review → Save and Approve → Process Approved Changes → Confirm Feature Added **Bond: Restaurant Recommendation Approval** (ios, 2 screens, AI Tool, Social Networking) [Refero](https://refero.design/flows/14397) > The user reviews an AI-generated restaurant recommendation for “Best food to get right now.” On the initial review screen, the user can inspect the restaurant details, copy the recommendation prompt, approve the suggestion, or regenerate it. After approval, the interface updates the primary action to “Finalize” and displays a small positive-feedback reaction from the associated profile. The user can then finalize the approved recommendation. The flow begins with a generated recommendation awaiting review and ends with the recommendation prepared for finalization. The user is assumed to be viewing the recommendation on a mobile device with access to the relevant profile or memory-based context; login state and permissions are not visible. Steps: Recommendation Review → Approved Recommendation **XChat: Message request acceptance** (ios, 2 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/11664) > Two-step mobile messaging sequence for handling an incoming message request from another user. Steps: Message Request (pending) → Accepted Conversation **Structured: Structured AI suggestions review** (ios, 3 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3289) > Happy path: The user views Structured AI's suggested plans for their day, can accept all suggestions at once, or give optional feedback about the suggestions. Steps: Review suggestions and Accept All → Open feedback modal → Enter and submit feedback **WhatsApp: Announcement Replies Thread** (ios, 3 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12679) > This is a mobile reply thread for an announcement message. The user starts with the reply panel open, sees the original announcement from Samuel, types a response in the bottom input, and sends it. After submission, the new reply appears in the thread with the sender’s name/avatar and timestamp. The user remains in the replies view and can close it with the X button. **Acctual: Manage approval rules (delete)** (web, 4 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8854) > Happy path: A signed-in user navigates to Settings → Approvals to remove an existing approval rule. From the approvals list the user opens an overflow menu on a rule, chooses Delete, confirms the deletion, and the rule is removed from the list with a success toast. Steps: Approvals list → Open rule actions → Confirm deletion → Rule removed and feedback **Aboard: Document Review and Approval** (web, 3 screens, Collaboration Tools, Paperwork) [Refero](https://refero.design/flows/834) > Two-column document review interface. Entry: user lands on Documents area (left column lists documents grouped under categories; right column is a preview area). Happy path: user selects a document from the left list -> the document opens in an embedded PDF viewer with thumbnails and toolbar -> an inline confirmation message and Approve button appear when the document is ready to approve -> user can select a different document from the list to open it. Assumption: user is signed in and has access to the Documents section; screenshots are on a desktop/laptop viewport. Steps: Document list — select document → Embedded PDF viewer with approval CTA → Switch to another document **Poppy: Voice Event Creation** (ios, 9 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12186) > A voice assistant-style interaction starts on a home screen with shortcut buttons and a microphone control. The user initiates voice input, grants required speech recognition and microphone permissions, дикtates an event request, reviews the interpreted event details, and then confirms creation. The flow ends with the created event summarized as a tappable item. Assumption: the user is on a mobile device and not previously denied the needed permissions. Steps: Home Dashboard → Voice Prompt → Speech Recognition Permission → Microphone Permission → Listening State → Speech Transcription → Request Refinement → Event Suggestion → Created Event Summary **Sora: Onboarding to First Feed** (ios, 45 screens, AI Tool, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/7819) > Happy path onboarding and first‑time setup for Sora. Entry: user opens the app and sees the branded landing screen. Primary choices offered: continue with Apple, continue with Google, sign up, or log in. The user selects Apple Sign In, completes system authentication (Face ID), then is routed through required setup screens: enter birthday (date picker), handle invite‑only gating (enter code or request notification), allow notifications, set profile visibility and pronouns, write cameo preferences, accept content advisory, and finally land on the main feed where a video is visible. Assumptions: user is on an iOS device with Face ID and system-level Apple Sign In available; user is not yet registered and is completing first‑time setup. Steps: Welcome / Auth choices → Sign in with Apple sheet → Face ID authorization → Sign-in completion → Enter your birthday (picker opened) → Birthday confirmed → Invite-only code entry → Notifications permission prompt → Profile privacy / Cameo settings (entry) → Pronouns selection popover → Custom pronoun input → Open Cameo preferences → Enter cameo instructions → Content advisory → Feed initial load (skeleton) → Feed with first video → Video interactions → Primary navigation available → Creation entry point → End state - active feed ##### Task: audit log activity **Linear: Issue Status Update** (ios, 3 screens, Collaboration Tools, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/6609) > Happy path: A signed-in user views an issue's detail screen, opens the Status selector by tapping the status chip, selects a different status from a bottom-sheet list, and returns to the issue detail screen where the status chip and activity log reflect the change. Steps: Issue detail — current status → Status selection sheet → Issue detail — updated status **Mindllama: Delete activity from day** (ios, 5 screens, Health & Wellness) [Refero](https://refero.design/flows/5506) > User is viewing a date-specific activity list (Apr 28, 2025) on a mobile device. From this screen the user can enter an edit mode, mark an individual activity for deletion, confirm the deletion via a system modal, and then see the list update (activity removed and activity count updated). **PayPal: Transaction Activity** (ios, 4 screens, Finance) [Refero](https://refero.design/flows/7270) > Happy path: A signed-in user opens the Activity tab to review recent transactions, uses the filter control to limit the list to a subset (example: Money in + Past 90 days), then returns to the Activity list where the results update and the active filters are shown as removable chips. Steps: Activity list (default) → Filter panel (initial) → Filter panel (selection changed) → Activity list (with active filters) **Steps: Filter Monthly Activities by Type** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/7518) > Happy-path: a signed-in mobile user opens the Activity tab to review workouts for a month (July 2025). The screen shows a scrollable list of activity cards (icon, type label, distance in km, time and date) and a floating filter button in the top-right. Tapping the filter button opens a small dropdown with options (All, Running, Walking). Choosing an option updates the list immediately: if there are items that match the filter they appear; if none do, the screen shows an empty-state message. The bottom navigation (Home, History, Activity, Settings) remains available throughout. Steps: Activity list (All) → Open filter menu → Apply 'Walking' filter → Change or clear filter **PayPal: Transaction filter and date-range selection** (ios, 9 screens, Finance) [Refero](https://refero.design/flows/7258) > Happy path: A signed-in user on the Activity tab opens the transaction filter drawer, adjusts type/payment/date options and/or a custom date range, applies the filters, and returns to the Activity list. The system loads results and shows either matching transactions or an empty state if nothing matches. Entry: Activity tab (logged-in user). Exit: Activity list updated (results or no-results state). Assumptions: user is authenticated, has permission to view transaction history, and the device is mobile (iPhone-like UI). Steps: Activity landing (empty state) → Filter screen (defaults) → Change filter radios → Open custom date range inputs → Select 'From' date → Select 'To' date → Apply filters (loading) → Activity list (filters applied) → No-results empty state **Napper: Sleep session event review** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/8776) > User opens the sleep overview (circular ring) to review a recorded sleep session and its detected interruptions. From the overview the user can reveal a bottom sheet list of detected events, open an individual event's detail card, confirm or act on that event, then return to the overview where the total "Asleep for" duration updates. Entry: user lands on the sleep overview screen. Exit: overview is visible with updated/confirmed event and recalculated total sleep duration. Assumptions: user is signed in on a mobile device, touch interactions available, the app has permission to show sleep/session data and detected events. Steps: Sleep overview (ring) → Bottom sheet — events list → Event detail card → Overview updated **Mindllama: Activity record deletion** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/5507) > Happy path: A signed-in user reviews activities for a specific date and deletes one of the listed activity records. **Foodllama: Weight Logs Management** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/11795) > A mobile weight-tracking screen showing a list of logged weights by date, with a back button in the top left, delete icons on each row, and a prominent 'Log weight' button at the bottom. The flow starts on the weight logs list and ends either with the user returning to the list after canceling deletion or with the selected entry removed after confirming deletion. Steps: Weight Logs List → Delete Confirmation → Updated Weight List **Phantom: Export Diagnostic Logs** (ios, 3 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3733) > Happy path: From the Security & Privacy settings screen the user taps Download App Logs, reads the explanation/assurance, then taps Download. The app generates a state log file and opens the iOS share sheet so the user can send or save the exported text document. Steps: Security & Privacy → Download App Logs (detail) → Share Sheet with Exported Log **TIDE: Sleep Recordings Manager** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/6767) > Happy path: The user opens the Sleep detail for a selected day, reviews recorded audio clips (sleep sounds/utterances), plays one to listen, enters an edit/confirm action for a clip, and confirms removal of that clip. Steps: Day view — recordings list → Playback active → Edit intention → Confirm removal and refresh list ##### Task: book GP appointment **District: Movie Booking Support Request** (ios, 7 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12840) > A user opens a movie-support conversation that already contains a booking summary. They confirm the booking, are prompted to describe the issue, open the attachment menu, and grant photo-library access. A booking confirmation containing a QR code and ticket details is shown as supporting context. The user types a request to replace cheese popcorn with salt, then sends the message with the booking image attached. The flow starts in an existing support chat and ends with the issue submitted to the support assistant. **Fresha: Add Appointment to Calendar** (ios, 3 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4974) > Happy-path: a signed-in mobile user views a confirmed appointment and chooses to add it to a calendar. Steps: Appointment details / Confirmed → Calendar selection modal → Device event editor (prefilled) **Fresha: Search & Book — Venue Listings** (ios, 2 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4966) > User performs a location-aware search for a service (Massage) with a selected date/time. The top of the screen shows a map and a persistent search bar (query + date/time). Below are filter chips (Sort, Price, Type) and a scrollable list of venue cards. Each card shows an image, name, rating, distance, available services with prices, and tappable time-slot buttons for booking. Happy path: user reviews a venue’s service and taps a time-slot to proceed into the booking flow. Steps: Venue list with time slots → No-results / Clear search state **Omio: Booking retrieval by reference** (ios, 4 screens, Travel & Booking) [Refero](https://refero.design/flows/5242) > Happy path: A user on the 'Your bookings' screen opens a "Retrieve booking" modal, enters a booking reference and the email used for the booking, and submits to fetch the booking so it appears in their upcoming bookings. Steps: Bookings list (Upcoming) → Retrieve booking modal (empty) → Enter details and submit → Lookup error (no booking found) **Longevity Deck: Add activity to daily timeline** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/9895) > Happy path: From the day timeline (Yesterday) the user taps an empty timeslot (11) to add an activity. The app opens an "Add Activity at 11" picker with search, category filters, and a scrollable list of activities. The user picks an activity (optionally adjusting duration or amount in the inline numeric control) and taps the circular + control to add it. The picker closes (or the list dismisses) and the selected activity appears as a colored card in the 11 slot on the timeline. Entry: day timeline with empty slot. Exit: timeline updated with new activity card. Assumptions: user is signed in and on the daily timeline view, editing is allowed, device shown is a phone in dark mode. Steps: Day timeline with empty slot → Activity picker — top of list → Activity picker — select activity → Updated timeline with activity card **Omio: Booking retrieval via Help Center** (ios, 4 screens, Travel & Booking) [Refero](https://refero.design/flows/5247) > Mobile Help Center flow for retrieving a booking or searching support content. Steps: Customer Help Center landing → Search field focused → Inline empty search feedback → Full search results – empty state **Airbnb: Photographer booking** (ios, 4 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6436) > Happy-path: A mobile user discovers a photographer listing, taps Reserve, picks a package, date, and time, then confirms booking and payment. Entry: provider detail card (listing page) with service cards and a sticky Reserve CTA. Exit: user reaches the Confirm & Pay screen with booking summary and payment method (ready to review/complete payment). Assumptions: Mobile (iOS-style) interface; user has access to a saved payment method (a card is visible on the confirm screen); availability is shown per-date and per-package. The UI uses a modal sheet for scheduling and horizontal time-chip selectors for available time slots. Steps: Provider listing & Reserve CTA → Schedule modal — pick date and time → Schedule modal — time confirmed → Confirm and pay **Saint Laurent: Store Appointment Booking** (web, 6 screens, Marketplace) [Refero](https://refero.design/flows/501) > Happy path: a shopper uses the store locator to find a nearby Saint Laurent boutique, opens the store details, and requests a private in-store appointment via a right-side booking panel. Steps: Store locator results → Store detail (Grant Avenue) → Open appointment panel → Select date — no hours available → Select alternate date and time → Submit appointment — error response **Fresha: Search availability by treatment, location, date, time** (web, 8 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/2323) > User starts on their Appointments screen (desktop, logged-in — profile avatar and an upcoming confirmed appointment are visible). From here they use the persistent search bar at the top to find new availability: they open the Treatment selector and pick "Tattoo & piercing", open Location and choose a saved address (Home), open the Date picker and choose a specific day (Dec 23), then set a custom time window (16:00–21:00). After applying filters and using Search, the interface displays search results on a split layout: a scrollable list of venues and available time slots on the left and an interactive map on the right. Steps: Appointments overview → Choose treatment category → Open location selector → Open date picker → Select a calendar date → Open time filter presets → Set custom time range → View filtered search results **Doctronic AI: Appointment Document Upload** (web, 4 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13702) > A logged-in user opens the Appointments area and views an upcoming video appointment with booking in progress. They open the Files section to review the file-upload prompt, add documents, and wait for the upload to complete. The uploaded PDF files then appear in the appointment's file list, where each file has a remove control. ##### Task: booking reservation **District: Restaurant Table Booking and Guest Invite** (ios, 19 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12862) > Happy path for booking a restaurant table and optionally inviting guests afterward. The user starts from a restaurant detail screen, opens the booking sheet, selects guest count, date, meal time, and an available offer, then reviews booking details and confirms the seat. After confirmation, the user can invite guests by sharing a link or choosing contacts, which may trigger contacts permission and contact selection. The flow ends on a confirmed booking screen with an option to pay the bill. **Airbnb: Reservation checkout and payment** (ios, 16 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6399) > Happy-path booking checkout for a short-term rental listing. **Omio: Flight extras & seat selection** (ios, 23 screens, Travel & Booking) [Refero](https://refero.design/flows/5237) > Mobile booking flow for a one-way Helsinki (HEL) → Stockholm (ARN) flight (1 adult). Steps: Search results / recommended flight → Outbound summary & fare carousel → Bags section (inline) → Open Bags modal → Bags modal - view included items → Add checked bag → Confirm bag selection → Booking summary updated → Seating options (choose reserve or auto-assign) → Seatmap (initial) → Select seat and preview cost → Seat details popup → Seating confirmed in summary → Extras & protections list → Passenger details CTA → Passenger details screen → Seatmap legend and seat types → Seat selection change (quantity/replace) → Confirm seat choice screen → Extras finalized / ready for payment **Omio: Bus ticket selection and passenger entry** (ios, 11 screens, Travel & Booking) [Refero](https://refero.design/flows/5183) > Happy path: A signed-in user searches for a bus from Barcelona to Madrid and selects a specific outbound journey from the results list. From the journey details screen they choose a class (Standard), pick a fare (Minimum Fare by default) and optionally enable the 'Cancel for any reason' insurance (Omio Flex). The interface shows fare terms, seat reservation availability and a running price summary. The user taps the primary CTA to go to passenger details; the app shows a brief loading state then presents a prefilled passenger-details form. Steps: Search results — journey list → Loading / details skeleton → Journey summary and classes → Expand trip leg details → Select your fare (fare cards visible) → Optional add-on — 'Cancel for any reason' → Seat reservation availability → Price summary & CTA → Loading skeleton (transition) → Passenger details — prefilled form → Ticket delivery confirmation and passenger selection **District: Find a Restaurant** (ios, 7 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12865) > Happy path for finding a restaurant from the Dining home screen. User starts on a location-based dining discovery home, opens the preference drawer, adjusts guests, date, and time, then submits to view matching restaurant results with available time slots. **Omio: Fare selection and add-on** (ios, 3 screens, Travel & Booking) [Refero](https://refero.design/flows/5181) > User is on the fare-selection screen for a previously chosen trip (Barcelona → Madrid, Wed Apr 9, 1 adult). The screen shows three horizontally scrollable fare cards (Minimum, Reduced, Standard). Tapping a card highlights it and updates the displayed fare terms and the header price summary. A “Cancel for any reason” add-on (toggle) and an Omio Flex information panel are shown below the fare cards. The user can change selection and optionally enable the cancel add-on before continuing to seat reservation or the next booking step. Assumptions: mobile portrait, user has already selected trip and passenger count, no account-specific controls are visible. Steps: Minimum Fare selected → Reduced Fare selected → Standard Fare selected **Omio: Seat selection** (ios, 9 screens, Travel & Booking) [Refero](https://refero.design/flows/5239) > Happy-path: a logged-in passenger moves from a booking's Seating options summary into an interactive seat map, selects a specific seat, reviews seat details, confirms the selection and returns to the Seating options summary with the chosen seat added. Steps: Seating options landing → Select seat — choose method → Seat map overview → Map — highlight special seats → Seat types legend → Selected seat indicated → Seat details modal → Confirm seat selection → Seating options summary (updated) **Volkshotel: Hourly work cabin reservation** (web, 20 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/7628) > Happy-path booking of an hourly "Work cabin" from a venue listing to confirmation. Entry: user browsing the Work Cabins section (desktop). Exit: confirmation screen with reservation details and confirmation number. Steps: Work cabins listing → Booking modal — select date/time → Calendar — month view → Calendar — date selected → Time picker — start time → Time/duration picker — end time options → Modal review — date/time/guests set → Loading availability → Select category (available cabins) → Cabin image gallery (lightbox) → Select rate and occupancy → Summary — booking breakdown → Preparing secure checkout → Contact details form — start → Country / phone selector open → Nationality selector list → Contact details form — populated → Terms and total review → Creating reservation → Confirmation **Trip: Hotel room selection** (web, 3 screens, Travel & Booking) [Refero](https://refero.design/flows/12080) > A hotel property detail and booking selection page on Trip.com. The user arrives from a hotel search context with destination, dates, and guests already filled in the search bar. The page lets the user review property information, photos, location, amenities, ratings, and available room offers, then choose a room and proceed to reserve. Steps: Property Overview → Room Rate Comparison → Offer Recheck **Understory: Booking Check-In** (web, 5 screens, Business, Scheduling Tools) [Refero](https://refero.design/flows/12721) > This admin booking detail screen lets a staff user review a customer’s reservation, open the ticket list, and mark the attendee as checked in. Entry is from a booking detail record for a specific experience and customer. The user is assumed to be logged in with permission to manage bookings. The happy path starts on the booking details view, continues through the Tickets modal, then a confirmation dialog, and ends with the booking and ticket shown as checked in, with a success toast confirming the action. Steps: Booking Details → Tickets Modal → Confirm Check-In → Checked-In Ticket → Updated Booking Details **Trip: Group Tour Booking** (web, 7 screens, Travel & Booking) [Refero](https://refero.design/flows/12096) > Happy-path group tour booking journey on Trip.com. The user starts on the Group Tours landing screen, searches for United States, reviews available routes, applies filters to narrow the results, opens a selected seven-day New York and East Coast tour, chooses an itinerary date and checks availability, reviews the included services and optional fees, enters traveler and contact information, agrees to the booking terms, submits the booking, and reaches a confirmation screen. The screenshots imply a desktop browser session; login is not visibly required before booking. The selected itinerary and pricing are carried through the booking steps, with the final payment step not shown. Steps: Group Tours Landing → Tour Search Results → Filtered Tour Results → Tour Details → Booking Selection → Traveler Information → Booking Submitted **Trip: Hotel Property Review and Room Selection** (web, 3 screens, Travel & Booking) [Refero](https://refero.design/flows/12081) > Happy-path hotel property browsing on Trip.com. The user enters a hotel detail listing for Ramada Hong Kong Harbour View, reviews photos, rating, amenities, and room options, optionally uses the share panel or copies the link, then chooses a room rate and taps Reserve. Steps: Property Overview → Share Property → Room Selection **District: Basketball Booking Payment** (web, 14 screens, Entertainment, Food & Drink, Travel & Booking) [Refero](https://refero.design/flows/13016) > The user reviews a basketball court booking, applies an available promotional code, selects a saved Visa card, enters the CVV, chooses a payment currency, and completes external card authorization. The flow begins on the booking review screen with the user details already populated and ends on a booking confirmation screen showing the confirmed reservation and confirmation code. The screenshots suggest the user is signed in and has a saved payment card available. ##### Task: budget spending insights **District: Restaurant Savings Calculator** (ios, 9 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12859) > User starts on a restaurant listing’s Offers tab and reviews available dining discounts for a specific date and meal period. They can open the sample bill panel, adjust guests, date, and meal type, and optionally enter a custom bill amount to recalculate savings. The experience is modal-driven on mobile, with entry from the offers list and exit via close controls back to the restaurant offers screen. **Pool: Place Details to Directions** (ios, 2 screens, Utility) [Refero](https://refero.design/flows/11980) > Happy path starts from a saved or viewed place detail screen for Citizens Bank Park, showing venue information, actions like call/website/tickets, and a directions button. The user can open navigation from the place detail view, which transitions into the maps interface focused on Citizens Bank Park and nearby points of interest. Steps: Place Details → Map Results **Foodllama: Nutrition Insight Completion** (ios, 7 screens, Health & Wellness) [Refero](https://refero.design/flows/11779) > This is a guided content experience in the Foodllama mobile interface. The user starts on a home/dashboard screen showing a featured insight card and a progress summary for today’s insights. Tapping or swiping into the card opens a full-screen lesson sequence with multiple steps, each presenting one nutrition tip with illustration, title, short explanation, and a progress indicator plus a Leave action. The sequence ends on a completion screen with celebratory confetti, a trophy graphic, and options to tell friends or finish. After completion, the user returns to the dashboard, where the insight count updates. Assumptions: user is already inside the app, on a mobile device, and viewing personalized insight content. Steps: Insight Card Dashboard → Intro Tip Screen → Carb Tip Screen → Strength Tip Screen → Completion Celebration → Return Dashboard → Updated Insight State **Binge: Viewing Statistics Breakdown** (ios, 3 screens, Entertainment) [Refero](https://refero.design/flows/12972) > A signed-in user views a personal viewing breakdown screen, then opens a display filter to switch between All, Movies, and Shows. The screen updates the chart and stats based on the selected category. Entry is from a stats or profile area; exit is via back navigation. Steps: Overview Stats → Display Filter Menu → Filtered Stats View **District: Restaurant Q&A Assistant** (ios, 7 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12846) > Happy path for an in-context restaurant question assistant. The user starts from a restaurant detail context and asks about Hotel Rishi. The interface shows the current topic, displays the user’s question as a highlighted prompt, and returns an AI-generated response with related action cards (Food Menu, Bar Menu, Beverage Menu). The user can rate the answer, ask a follow-up by text or voice, or submit a new query such as finding nearby breakfast options. The flow ends with a refreshed response and the follow-up input ready for another question. **Pinkllama: Consume and complete a recommended insight** (ios, 5 screens, Health & Wellness) [Refero](https://refero.design/flows/5659) > Happy-path: a logged-in mobile user browsing a "For You" feed of personalized cycle insights. The user taps a recommended card to open a short multi-step content piece (carousel style), swipes through the slides, and completes the content. On completion the app shows a celebratory confirmation with a share CTA; when the user dismisses or finishes, the feed updates the card as completed and increments the daily progress badge. Steps: Feed — For You recommendations → Insight detail — multi-step content → Completion — celebratory confirmation → Feed — card marked completed → Feed — next recommended insight **Linktree: Insights Dashboard and Content Assistant** (web, 2 screens, Business, Social Networking) [Refero](https://refero.design/flows/12411) > Logged-in user is viewing the Linktree Insights area from the left sidebar. The main dashboard shows performance cards such as most clicked links, visitor sources, lifetime totals, and social posting/connect prompts. On the right, an AI assistant drawer is open for generating personalized Instagram content. The user can review analytics, follow prompts to connect accounts or explore insights, and interact with the AI assistant. The flow starts on the Insights dashboard and ends with the user reviewing or continuing within the assistant panel or dashboard cards. Steps: Insights Overview → AI Prompt Refinement **Claude: Interactive Budget Dashboard Creation** (web, 3 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/1821) > Happy-path: A logged-in user (Robert) uses the chat assistant to convert a screenshot of a personal monthly budget into an interactive React dashboard component. Steps: Assistant Home / New Chat → Request Dashboard from Image → Review Generated Component and Preview **Spyglass: Audience Scaling Recommendations** (web, 10 screens, Business, Inspiration Board) [Refero](https://refero.design/flows/13462) > A logged-in user reviews an existing AI-generated advertising analysis, asks a follow-up question about target audience segments, and receives a detailed audience targeting and scaling recommendation. The flow begins in an existing conversation and ends with the assistant presenting campaign structure recommendations and asking whether to create performance-focused creative variants. Steps: Review Existing Creative Analysis → Inspect Voice Input → Activate Voice Recording → Enter Audience Question → Submit Follow-Up Request → Monitor Brand Research → Review Analysis Progress → Inspect Persona Results → Review Audience Strategy → Review Campaign Scaling Plan ##### Task: business account onboarding company verification **UGLYCASH: New user onboarding and verification** (ios, 19 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/11202) > End-to-end new-user onboarding for a trading app: marketing carousel -> sign-in options -> phone verification (WhatsApp/SMS) -> Apple sign-in fallback -> username + referral -> legal consent -> biometric enablement -> first-time home. Steps: Welcome carousel - follow traders → Welcome carousel - trade everything → Welcome carousel - trading clubs → Welcome carousel - custody ownership → Sign-in selection → Phone entry (channel chooser) → Phone entry (send via SMS) → Code entry (empty) → Code entry (auto-filled) → Linked-account conflict modal → Apple Sign In prompt → Apple signing in → Choose username → Username claimed (input filled) → Referral code entry → Accept terms and conditions → Enable Face ID prompt → Face ID OS permission → First-time Home (authenticated) **Blackbird: New user onboarding and wallet setup** (ios, 33 screens, Food & Drink) [Refero](https://refero.design/flows/9017) > Happy-path new user onboarding for Blackbird's payment wallet. Entry starts at the marketing splash where user taps 'Go', then they provide a phone number (country selector possible), receive an SMS verification code, optionally enter a referral, provide an email, add a card, and then walk through 3 short onboarding screens that explain how to tap the Blackbird puck and check out. The flow asks for location and notification permissions before landing on the authenticated wallet home. Steps: Welcome splash → Phone number entry (empty) → Phone number entry (filled) → Verification code - entry screen → Verification code - partial entry → Verification success -> Referral prompt → Email collection → Add card - input → Card added confirmation → Onboarding - Tap in → Onboarding - Puck image → Onboarding - Tell server → Onboarding - Checkout & leave → Wallet home (initial) + request location → System location permission → System notifications permission → Wallet home - populated → Check-in / Tap interaction → Payment/receipt preview → Final wallet state **Ada: Onboarding and Account Creation** (ios, 19 screens, Health & Wellness) [Refero](https://refero.design/flows/4094) > New-user onboarding that introduces the product (5 illustrated slides) and guides the user through creating an account. Steps: Onboarding - Intro (1/5) → Onboarding - Causes (2/5) → Onboarding - Next steps (3/5) → Onboarding - Get started (4/5) → Onboarding - Final slide (5/5) → Consent - Confirm you're ready (initial) → Consent - Checkboxes selected → Create account - Empty form → Create account - Inline validation errors → Create account - Corrected email → Create account - Password entry and show/hide → Manage your privacy - Default toggles → Manage your privacy - Agree or save choices → Notifications & tracking opt-in → Check your inbox - Verification sent → Tips to find your email → Verification email - External message → Email verification success (web) → Account verified → Welcome Home **Airbnb: Phone sign-in, verification, and onboarding** (ios, 32 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6385) > Happy-path sign-in and first-time onboarding. Entry: user taps Log in / Sign up from the app's bottom nav. The flow supports phone-based authentication (primary), with alternatives (email, Apple, Google, Facebook). The user selects country, enters a phone number, receives an SMS, chooses verification delivery if needed, enters the code, completes a short security check, and then is guided through permission prompts (notifications) and a short onboarding/feature tour before arriving at the main Explore/home screen. Assumptions: user is on a mobile device, not previously authenticated, has a reachable phone number, and allows keyboard/numeric input. Steps: Sign in / Choose method → Privacy / Cookie prompt → Country/Region selector → Phone number entry → Phone keyboard active → Code entry (prompt) → More verification options → Entering verification code (partial) → Entering verification code (filled) → Alternate sign-in (email) → Security check intro → Landing skeleton (post-auth) → Turn on notifications prompt → OS notification permission → Onboarding screen — intro → Onboarding — Experiences preview → Onboarding — Services preview → Pricing change snackbar → Explore / Home content → Final content with badge **Acctual: Company document verification** (web, 4 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8828) > Summary: This is a modal-based, multi-step document submission sequence for verifying a company's identity. Steps: Select document type and upload (Legal presence) → Upload ownership structure document → Upload company details / proof of address → Review uploaded documents and continue **Square: Merchant onboarding and account setup** (web, 38 screens, Business, Finance) [Refero](https://refero.design/flows/1952) > Happy-path merchant onboarding from marketing into the Square dashboard. Entry: visitor clicks through marketing/CTA to sign up. The flow collects account credentials, business entity details, business category, identity verification (SSN), payout preferences, and bank account details. It then shows a summary of chosen products, offers app download, and lands the user in the Square dashboard. Steps: Marketing / Signup entry → Create account (empty) → Create account (filled) → Submission loading → Business entity question → Entity type options → Business details form → Business details (filled) → Business category picker → Category search results → Full SSN requested → Verification failure / Retry → Choose payout frequency → Payout option selected → Link bank account (empty) → Link bank account (filled) → Selections overview → Welcome / Next steps → Download Point of Sale → Dashboard (first sign-in) **Nooka: Account creation, onboarding, and subscription** (ios, 50 screens, AI Tool, Education, Books) [Refero](https://refero.design/flows/8112) > End-to-end new user acquisition: from welcome -> sign-in / sign-up -> email verification -> short onboarding questionnaire -> subscription trial purchase -> return to main content. Steps: Welcome / Entry → Login (email) / Social alternatives → Sign Up — empty form → Sign Up — email input → Sign Up — password entry → Email Verification screen → Verification polling → Onboarding — Intro slide 1 → Onboarding — Question slide (choose friction) → Onboarding — Confirmed selection → Free Trial overview → Subscription modal — summary pricing → Subscription modal — Pro (default) yearly selected → Subscription modal — Plus tier selected → Subscription modal — Ultra tier selected → Native App Store confirmation → Purchase success modal → Subscription confirmation overlay → Tap to Join overlay (open mic prompt) → Home — authenticated content **Grok: Account creation and verification** (ios, 19 screens, AI Tool) [Refero](https://refero.design/flows/6076) > Happy-path new user onboarding that starts on the app's welcome screen and ends at the main Chat/Voice landing. Entry: app welcome / sign-in screen. Exit: authenticated user landing screen. Assumptions: user is on an iOS device, not signed in, and has network connectivity. Flow uses social sign-in (X/Google/Apple) or email sign-up; the screenshots follow the email sign-up path with verification, password entry, Cloudflare CAPTCHA, and an optional SuperGrok subscription upsell before landing in the app. Steps: Welcome / Sign-in prompt → Sign-in options including Email → accounts.x.ai — Log into your account → accounts.x.ai — Create account entry → Email input screen → Email typed (in progress) → Verify your email (code input) → Code validation error state → Code input with keyboard → Code accepted / proceed → Complete sign up — initial form → Filling name and password → Password validated (strength OK) → Cloudflare CAPTCHA challenge → Form ready — submit → SuperGrok upsell — monthly selected → SuperGrok upsell — yearly selected → Post-signup landing (empty state) → Main Chat/Voice interface **X: Account creation and onboarding** (ios, 41 screens, Social Networking, Entertainment, News, Video Streaming) [Refero](https://refero.design/flows/2903) > Happy-path sign-up and onboarding sequence for a new user. Entry from app launch (splash) → welcome screen offering Google/Apple/create options → create account form (name + phone or email + DOB) → verification (email/phone code) → suggested follows → interest selection (categories + subtopics) → follow additional recommended accounts → finished on the home timeline. Steps: Launch / Splash → Welcome / Sign-up options → Create account (empty form) → Name entry / Phone field activated → Phone entry with dial pad → Country code selection → Phone + DOB available (calendar trigger) → Date picker / Select DOB → Email entry (alternate path) → Email validation error → Verification code entry (sent) → Enter code and proceed → Suggested follows — single recommendation → Suggested follows acknowledged → Interest categories grid → Multiple interests selected → Interest subcategories (chips) → More subtopic selections → Follow recommended accounts list → Finalize onboarding — timeline **ElevenReader: New account onboarding** (ios, 21 screens, AI Tool, Books) [Refero](https://refero.design/flows/8941) > Happy-path onboarding and first-run experience for ElevenReader. Steps: Welcome / Auth choices → Create account form (initial) → Password requirements → Consent checkboxes → Submit sign up → Verification code entry → Autofill suggestion → Verification failure → Successful verification → Date of birth picker → Voice selection intro → Iconic vs in-house voices → Confirm voice and continue → Loading / transition → First-run Home → Notification permission prompt → Subscription upsell overlay → Plan selection → Purchase processing → Onboarding complete — Active Home **Turf: Phone Sign-In & Onboarding** (ios, 38 screens, Sports) [Refero](https://refero.design/flows/9419) > New-user sign-in and onboarding using phone verification. Steps: Welcome / Splash → Phone Number Input (empty) → Phone Number Input (filled) → Submitting Number (loading) → Verification Code Entry (empty) → Verification Code Entry (partial) → Verification Code Entry (processing) → Name Input (empty) → Name Entry (filled) → Contacts Permission Prompt → Contacts Shared Confirmation → Initial App Skeleton / Loading Home → Onboarding Card: Welcome → Onboarding Card: Get Ticket → Onboarding Card: Pick a Side → Onboarding Card: Win Real Cash → Onboarding Card: Coming Soon (Ask Coach) → Onboarding Card: You're Ready → Home: Ticket List Visible → Home: Interact With Ticket **PayPal: New account signup (email + phone verification)** (ios, 53 screens, Finance) [Refero](https://refero.design/flows/7224) > End-to-end new-account onboarding for a mobile payments product (PayPal). Steps: Welcome / Splash → Enter email (empty) → Email typed → Submitting email → Phone number entry (initial) → Phone input focused → Phone number entry (partial) → Phone number complete → Enter SMS code (modal) → SMS code entry in progress → Create password → Email verification entry → Email code typed → Email code validation (processing) → Notifications opt-in screen → OS notification permission → Contacts sync intro → OS contacts permission → Choose contacts to share → Completion and Home **ElevenLabs: Email sign-up & first-run onboarding** (ios, 12 screens, AI Tool) [Refero](https://refero.design/flows/10728) > Happy-path onboarding and first-run. Entry: user arrives at the app Home while logged out. The user chooses to create a new item which triggers the sign-in / sign-up sheet. They pick email sign-up, complete the form, receive a 6-digit verification code by email (possible iOS autofill), verify it, confirm age (18+), optionally answer a "Where did you hear about us?" question, then land back on the Home screen as a signed-in user. During first-run the OS permission prompts for notifications and app tracking appear. Exit: Home (signed-in). Assumptions: iOS device (keyboard autofill shown), user initially logged out, user has access to the provided email address, Terms/Privacy check is required to enable sign-up. Steps: Home — Logged out → Sign in / Sign up options → Sign up with email — empty form → Sign up with email — form filled → Enter verification code — empty → Verification email (external) → Enter verification code — filled → Age confirmation → Referral question → Home — First-run: Notifications prompt → App Tracking Transparency prompt → Home — Signed in / Ready **Intercom: New account signup and verification** (web, 17 screens, Business, Communication Tools) [Refero](https://refero.design/flows/674) > Happy-path new user onboarding for Intercom's 14-day free trial. Entry from a marketing/landing hero into a multi-step signup wizard: enter work email (or continue with Google), provide full name, create a password (minimum 10 chars), add company name, and select company size. After submission the system attempts automated account setup and sends a verification email with quick links to open common mail clients. Exit points: successful email-sent state (primary exit) or an error state requiring manual verification via chat. Assumptions: user is not signed in, on a desktop browser, and has access to the email account they entered. Steps: Marketing hero / CTA → Enter work email → Email entry confirmation → Collect full name → Name validation error → Valid full name entered → Create password (instructions) → Password validation error → Valid password entered → Collect company name → Company name confirmed → Select company size → Company size selected → Account setup in progress → Verification email sent → Waiting / transition state → Manual verification fallback **TravelPerk: New account signup** (web, 9 screens, Productivity, Travel & Booking) [Refero](https://refero.design/flows/5196) > Happy-path: an unauthenticated desktop user lands on the marketing homepage, opens the primary navigation, selects the Get started action and enters a three-step signup. Step 1 collects work email and personal name + acceptance of terms, step 2 collects company details (job title, phone, company name, company size, currency), and step 3 shows an email-verification prompt with quick links to open common mail clients. The flow ends at a sign-in screen (or waiting for the user to verify their email). Steps: Homepage — Hero → Primary navigation / action modal → Sign up — Step 1 (Basic info) → Sign up — Step 1 (Filled) → Sign up — Submitting (loading) → Sign up — Step 2 (Company details) → Sign up — Step 2 (Filled) → Email verification prompt → Sign in (post-signup / default auth screen) **Exactly.ai: Account signup to prompt composer** (web, 13 screens, Business, AI Tool) [Refero](https://refero.design/flows/5679) > Entry: Public marketing/home hero with a primary CTA to start. Happy path: user clicks the CTA, opens a centered Create account dialog, chooses to continue with email (or Google), enters name and email, receives a 6-digit verification code by email, submits the code, then completes a short 3-step profile onboarding (profession, how they heard about the product, primary use case). After completing onboarding the user lands in the image-creation workspace where template cards and a prompt input are available. Assumptions: user is on a desktop viewport, not signed in at start, has access to the provided email address and can receive verification codes. Steps: Marketing / Hero entry → Create account (initial) → Create account (form filled) → Enter verification code (blank) → OTP error feedback → Enter verification code (retry) → Profile setup — profession (1 of 3) → Profile setup — profession options continued → Profile setup — 'Other' text input → Profile setup — acquisition source (2 of 3) → Profile setup — primary use (3 of 3) → Onboarding complete — workspace revealed → Prompt entry and image creation **Fourmula.ai: Account Sign Up and Dashboard Entry** (web, 7 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/11833) > This happy path starts on a marketing landing page for an AI visual generation product, then moves into account creation and lands on the authenticated workspace dashboard. Steps: Marketing Landing → Create Account → Account Filled → Submitting Sign Up → Dashboard Empty State → Email Verified Notice → Ready to Create **Mercury: Formation document upload and validation** (web, 4 screens, Business, Finance) [Refero](https://refero.design/flows/11020) > Happy path: a logged-in user uploads a company's formation document from a desktop browser. The screen lists acceptable formation document types and provides a drag-and-drop upload area. After the user drops a file the system uploads it, runs automated checks (entity name, document type, filing date) and shows progress and status. If the document passes checks the user can continue with onboarding; if issues are found the system displays a warning with itemized problems and guidance. Steps: Document instructions and upload area → File added / upload started → Automated checks in progress → Validation results and issues ##### Task: buy crypto **Phantom: Buy Bitcoin with third‑party provider** (ios, 11 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3710) > Happy-path: User opens a Bitcoin asset detail, taps Buy and enters an amount (or selects a preset). The app validates the amount (enforces a $5 minimum), shows the USD ↔ BTC conversion and enables Buy once valid. The user then chooses a payment method (Credit & Debit Card), views partner providers and live quotes, picks a provider (Stripe), and is routed to an external Stripe/Link checkout where they must sign in/confirm payment. Entry: asset detail screen (logged-in user). Exit: redirected to external provider checkout (Stripe/Link) to complete payment. Assumption: user is logged in to the wallet, has at least one payment method available via partners, and network quotes update on a short timer. Steps: Asset overview → Buy screen — empty amount → Validation — below minimum → Preset selected — valid amount → Toggle input unit → Payment method — entry → Payment method options → Provider quotes list → Provider checkout (Stripe Link) — summary → Stripe Link — sign up or sign in → Return to app or finalize quote **Kann: In‑game Coin & Life Purchase** (ios, 6 screens, Education) [Refero](https://refero.design/flows/9159) > User opens the in‑game store to buy coins or lives. The screen lists a "Buy Lives" option (1 Life for 2 coins) and several coin packs with prices. The user taps a coin pack "Buy Coins" button -> the in‑app purchase begins, the UI shows a processing state, and the device displays the system App Store confirmation sheet (Double Click / side button). After the user confirms with the side button, the App Store reports success and the app shows a success alert. Steps: Store landing (coin + lives list) → Store — alternate copy state → Local processing state → Platform purchase confirmation (App Store) → Purchase success acknowledgment → Return to store (post‑purchase) **Reown: Wallet Receive & Activity** (web, 4 screens, Development Tools, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/5711) > Happy path: a user opens an embedded wallet modal (AppKit demo) to view an account, request a receive address (QR + copy), switch to a different account/address, and inspect recent transaction activity. Steps: Wallet overview (account selected) → Receive view (QR + copy) → Switch account / Selected account view → Activity tab (transaction history) **Twitch: Buy Twitch Bits** (web, 6 screens, Social Networking, Entertainment, Video Streaming) [Refero](https://refero.design/flows/13587) > A logged-in Twitch viewer purchases a 300 Bits package from a live channel. The user opens the Bits purchase selector, chooses the 300 Bits option, reviews the purchase in the checkout dialog, and attempts to pay by card. The first card submission fails validation, after which the user provides acceptable card details and proceeds through Revolut authentication. The payment completes successfully, and the Bits selector remains visible with the purchased 300 Bits balance reflected. The flow assumes the user is signed in, has access to a supported payment method, and can authorize a 3-D Secure payment. Steps: Select Bits Package → Review and Enter Payment → Correct Card Validation Error → Authorize Payment → Confirm Successful Purchase → View Updated Bits Balance ##### Task: calendar availability **Airbnb: Calendar View Switcher** (ios, 3 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6453) > Happy path: A signed-in mobile user opens the Calendar tab and sees a month grid showing per-day prices. They tap the calendar-view control in the header to open a bottom-sheet of view options (Year / Month / List), choose a different presentation (List), and confirm. The calendar updates from the month-grid to a list layout showing each date, availability status and price. Entry: Calendar tab selected. Exit: Calendar presented as a list of daily rows. Assumptions: user is authenticated and has permission to view the calendar; this is a mobile phone (portrait) view. Steps: Month grid overview → Presentation options modal → List view of dates **Mela: Share recipe to Messages** (ios, 5 screens, Food & Drink) [Refero](https://refero.design/flows/10367) > Happy-path: A user is viewing a recipe detail and uses the inline actions menu to share the recipe via Messages. Steps: Recipe detail view → Contextual actions menu → Share submenu opened → System share sheet → Message composer with recipe preview **Days: Create event title and optional calendar import** (ios, 4 screens, Scheduling Tools, Lifestyle) [Refero](https://refero.design/flows/7791) > Happy path for creating a new event starting on the Add event screen (iOS). Steps: Title entry (keyboard focused) → Calendar connect promotional card → System calendar permission modal → Imported suggestions and continue **Airbnb: Switch Calendar View** (ios, 4 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6452) > User is on the Calendar tab viewing a monthly price calendar (July). From this screen the user opens a bottom-sheet control to change how the calendar is displayed (Month / Year / List). They choose Year and confirm, which dismisses the sheet and navigates to a year overview (2025) composed of small month micro-calendars. Entry: Calendar tab (month view). Exit: Year overview. Assumptions: user is signed in and on a mobile device (iOS style chrome visible); user has permission to view calendar data. Visual cues: selected day is highlighted (red), each date shows a price, the modal dims the background while open. Steps: Monthly calendar (default) → Open 'Calendar view' modal (Month selected) → Change selection to Year in modal → Year overview displayed **Daylish: Day Timeline Filtering** (ios, 4 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/8021) > A mobile day-timeline where a user inspects a single day's events and uses a modal "Filter Timeline" sheet to show/hide calendars and view options. Entry: user opens the day timeline (logged in, has multiple calendars). Exit: timeline updates to reflect filter changes and the modal is dismissed. Steps: View Day Timeline → Open Filter Timeline (peek) → Adjust Filter Settings → Filtered Timeline Updated **Voicenotes: Calendar Streak Review** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14269) > The user opens the navigation drawer, selects the calendar option, and views a monthly activity calendar for October 2026. The system first loads the streak information, then displays the user’s selected date and global ranking. The user expands the highlights section to view a short summary of product capabilities. The flow begins in the main notes area with the navigation drawer open and ends with the expanded highlights content visible in a calendar modal. The screenshots suggest a mobile device and an authenticated user with access to personal activity data. **Joi: Calendar selection** (ios, 3 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/4390) > Happy path: A signed-in user opens the app Settings and navigates into Calendars to choose which calendar sources should be active. Steps: Settings - Integrations overview → Calendars - All enabled (list view) → Calendars - Toggle updates **CapWords: Daily review scheduler** (ios, 4 screens, AI Tool, Education) [Refero](https://refero.design/flows/6118) > Happy path: A user views their review summary and calendar, opens a modal to choose how many items they want to review per day, picks a preset amount, and confirms. Entry: review summary screen with statistics and calendar. Exit: modal closes and the selected daily review amount is applied (UI updates visible on main screen). Assumptions: user is on the review summary screen, signed-in state not shown, device is a tall mobile (iPhone-like). Steps: Review summary & calendar → Open 'Daily review amount' modal (5 selected) → Change preset to 10 → Confirm selection and return to calendar **Airbnb: Date-range pricing and rules editor** (ios, 11 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6451) > Host-facing calendar where a logged-in host selects a date range and updates per-night pricing and custom rules for those dates. Steps: Calendar overview (month) → Select date range → Open Custom settings → Open Minimum nights editor → Adjust Minimum nights (value = 1) → Adjust Minimum nights (value = 2) → Open price editor → View price breakdown → Compare similar listings (map) → Edit price and re-evaluate → Save changes and update calendar **Orbit: Edit Subscription Details** (ios, 4 screens, Finance) [Refero](https://refero.design/flows/9995) > User is on an Edit Subscription screen for an existing subscription (Apple TV+). They can change the subscription's payment date, billing cycle, free trial toggle, list/category, payment method, notification timing, and view billing history. The user opens a calendar date picker by tapping the Payment date field, chooses a new date, and the date field updates. The Save button (top-right) persists changes. Steps: Edit Subscription Overview → Open Calendar Picker → Select New Date → Confirm Updated Details **Apple Invites: Add Invite to Calendar** (ios, 4 screens, Scheduling Tools) [Refero](https://refero.design/flows/4042) > Happy path: A user sees an invite card (includes location map and shared album) and taps the calendar action to add the invite to their calendar. The system requests full calendar access to sync iCloud events. After granting access the app opens the New Event editor pre-filled with title, address, start/end times and the selected calendar. The user saves the event and is taken to the Event Details screen showing the scheduled event with alerts, URL, notes and calendar assignment. **Daylish: Edit scheduled item calendar** (ios, 6 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/8017) > Happy path: From the Today screen the user opens an in-progress item (Breakfast), invokes a contextual menu, chooses Edit Details, updates the Calendar assignment, saves the change and returns to the Today list where the item shows updated styling and a running countdown. Entry: Today / In Progress list (user already on the Today screen, device unlocked). Exit: Back to Today list with the event updated and visually refreshed. Assumptions: user is signed in, calendar integration is enabled, the event exists on the selected date, and edit permissions are available. Steps: Today list — in-progress card → Contextual action menu → Edit Details — timing & calendar → Calendar selection dropdown → Save edits → Today list — updated card **UNHINGED: Monthly calendar filtering** (ios, 6 screens, Health & Wellness) [Refero](https://refero.design/flows/6346) > Happy path: From the Stats tab the user views a compact month calendar (June 2025). They open a bottom filter sheet, choose one or more filter criteria (Mood, People, Tags), and the calendar updates to show only days that match the selected criteria. Selected filters appear as removable chips in a persistent filter bar; removing a chip updates the calendar. **Exoplan: Edit calendar event** (ios, 13 screens, Health & Wellness, Scheduling Tools) [Refero](https://refero.design/flows/7345) > Happy path for editing an existing calendar event on a mobile device (iPhone). Entry: user is on the Day view showing events for a date. Exit: user saves the edited event and returns to the Day view where the event appears with updated time and settings. Steps: Day view (schedule) → Event quick view → Edit modal — overview → Time picker → Repeat control → Select recurrence → Edit first notification → Set first notification → Add second notification → Change event calendar — open list → Select calendar → Save changes → Day view — confirm update **Steps: History - List / Calendar Toggle** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/7517) > User is on the History section of an activity tracker. Steps: History - List View → Open Presentation Toggle → Calendar - Month Grid (initial) → Calendar - Activity Indicators **Tripsy: Trip date selection** (ios, 5 screens, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/2331) > Happy path: A mobile user creates or edits a trip entry by entering a name, opening the date-picker, choosing a start date and an end date, confirming the selection, and returning to the edit screen where the selected date range and computed duration are displayed. **Daylish: Inline Event Edit — Peeking Ahead** (ios, 12 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/8016) > Happy-path: user creates/edits a single placeholder event from a day timeline using an inline card editor that appears in "Peeking Ahead" mode. Steps: Day timeline (Tomorrow) → Open inline editor (title focus) → Choose calendar (open selector) → Switch calendar to Home → Type new title → Title entry completed → Open start-time picker → Select PM and confirm start time → Open duration picker → Set duration to 2 hours → Save changes inline → Final timeline state **ROVR: Archive Date Filtering** (ios, 4 screens, Music Streaming) [Refero](https://refero.design/flows/13972) > The user starts in the Archive section with a searchable collection of audio episodes. They open the calendar filter, review September 2026, select September 9, and confirm the date. The archive then returns with the selected date shown in the search field and displays matching episodes. The flow assumes a mobile user is already viewing the Archive section; no account or permission state is visible. The flow begins at the unfiltered Archive listing and ends at filtered archive results for the selected date. Steps: Archive Listing → Calendar Open → Date Selected → Filtered Results **OKURI: Calendar Appearance Change** (ios, 5 screens, Utility) [Refero](https://refero.design/flows/11870) > The user starts on the OKURI calendar dashboard in the light appearance, opens the customization screen, selects the Dark sheet edition, confirms the system alert, and returns to the dashboard. The dashboard retains the displayed date, astronomical panels, and events while updating its colors and icon styling. The user is assumed to be viewing the calendar with access to the customization settings. Steps: Calendar Dashboard → Edition Selection → Change Confirmation → Dark Edition Applied → Dark Calendar Dashboard **Stardust: Cycle day logging** (ios, 21 screens, Health & Wellness) [Refero](https://refero.design/flows/5870) > Happy path: The user opens the app to the MONTH calendar view, taps a date in June to open a day summary card, then opens the full day editor to log or edit cycle, symptoms, temperature and journal entries. From the day editor the user can: edit period days (calendar overlay), add a numeric temperature reading, and add a journal entry using voice-to-text (requires two iOS permissions). The flow ends when the user saves/confirm entries and returns to the monthly calendar where icons and badges reflect logged data. Entry point: MONTH calendar (logged-in user, basic view). Exit point: back to MONTH calendar with updated data. Assumptions: user is authenticated; device is iOS (native permission dialogs shown); user has a cycle already recorded so cycle/day labels are visible; keyboard and system permission dialogs are available. Steps: Monthly calendar (Month view) → Compact day summary card → Full day editor (expanded) → Edit Period (calendar overlay) → Period selection hint → Period days selected → Day editor with symptoms visible → Symptoms edited / confirmed → Add Temperature → entry screen → Temperature numeric input → Confirm temperature → Open Journal editor → Speech recognition permission dialog → Microphone permission dialog → Microphone recording active → Transcription editing → Journal saved and UI updated → Day detail shows updated values → Return to monthly calendar → Final verification on calendar **Napper: Journal Calendar Entry** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/8809) > Happy path for creating a date-specific journal entry from a month calendar and returning to the month view with the new entry displayed. Steps: Month Calendar — Date Selected → Journal Editor — New Entry → Journal Editor — Save Entry → Month Calendar — Entry Saved **Exoplan: Natural Language Event Creation** (ios, 6 screens, Health & Wellness, Scheduling Tools) [Refero](https://refero.design/flows/7347) > Happy path: The user opens their daily calendar, taps the floating + button to invoke an in-context 'Magic Planner' composer, types a free‑form instruction (e.g., "Plan upcoming trip"), and submits it. The system shows a two‑stage progress state ("Loading context.." then "Adding to schedule...") while it processes the request and finally inserts a new event block into the day view. Steps: Day view (calendar) → Composer open — prompt shown → User types natural language → Submit — processing begins → Scheduling in progress → Calendar updated with new event **OKURI: OKURI First-Time Setup** (ios, 10 screens, Utility) [Refero](https://refero.design/flows/11863) > A first-time user opens OKURI and starts the setup experience. The system requests Calendar and Location permissions, then presents a short swipe tutorial explaining how to move between daily sheets. After the user completes the tutorial by swiping up, the main calendar view opens on May 13 with date, astronomical, lunar, and to-do information visible. The screenshots imply an iPhone or iOS-style device, and the permissions are expected to be granted for the full experience. Steps: Welcome → Calendar Permission → Location Permission → Swipe Tutorial Start → Sheet Flip Demonstration → Swipe Tutorial Navigation → Forward Flip Demonstration → Repeat Swipe Instruction → Setup Complete → Daily Calendar **Sky Guide: Date & Time Selector** (ios, 6 screens, Education, Reference) [Refero](https://refero.design/flows/8477) > Happy path for changing the observed date/time in a mobile star-chart viewer. Entry: user is viewing the live star map. The user opens the time/date control at the bottom of the screen, launches the calendar, selects a specific day, optionally opens the month/year wheel to jump to a different month or year, and confirms the selection. Exit: the sky render updates to the selected date/time and the timeline/time controls reflect the new selection. Assumptions: user is on a mobile device viewing the interactive star map with location access granted (so the map can render local sky), no authentication is required to change the date/time. Steps: Open time controls → Open month calendar → Select a day → Open month/year picker → Choose month and year → Apply selection and view sky ##### Task: calendar scheduling meeting **Cofounder: Daily Meeting Summary Test** (web, 5 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/7900) > Happy path: a user viewing a configured scheduled automation called "Daily Meeting Summary Email" starts a manual test of the scheduled trigger, watches the test run initialize, and inspects the run timeline as integrations (Calendar) and internal reasoning steps are executed. Steps: Plan view — Automation details → Test confirmation modal → History view — Test run initializing → History view — Run processing progress → Run timeline — Integration and reasoning details **OKURI: Schedule Team Standup** (ios, 5 screens, Utility) [Refero](https://refero.design/flows/11867) > The user starts on a daily calendar dashboard for Wednesday, May 13. They choose Add Event, enter a meeting title and location, review the default date and time settings, and save the event. The dashboard then displays the new event in the event list. The user selects the event to open its details, where they can review the location, schedule, calendar, alerts, and available edit or delete actions. The flow assumes the user has access to the calendar and the Work calendar is selected by default. Steps: Daily Calendar → New Event Form → Completed Event Form → Updated Daily Calendar → Event Details **Imprint: Commit & Schedule Reminder** (ios, 7 screens, Education) [Refero](https://refero.design/flows/5640) > Happy-path onboarding for committing to a learning streak and scheduling recurring reminders. Entry: user arrives at a 'Commit to learning' screen where they pick a streak length. After choosing a streak and tapping the primary CTA, the user is taken to a 'Schedule your next sessions' screen to pick a reminder time and optionally add the event to their calendar. If the user enables calendar integration and taps Schedule, the system prompts for device calendar permissions. Exit: user either grants calendar access (permission flow shown) or dismisses/declines and continues (not shown). Assumptions: user is signed into the app, running on an iOS device, and calendar permission has not yet been granted. The primary happy-path ends with the app requesting calendar access to add events. Steps: Choose streak length → Confirm streak selection → Open scheduling screen → Open native time picker (AM selection) → Change time to preferred value (PM) → Finalize scheduling → System calendar permission prompt **Calendly: Contact Meeting Booking** (web, 13 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/14036) > A logged-in user opens an existing contact record with no prior meetings, chooses to schedule a meeting, selects an event type, configures the meeting location, chooses an available date and time, and books the meeting. The process begins on the contact’s Meetings tab and ends when the contact record displays the newly scheduled meeting and a confirmation notification. The user appears to have permission to schedule meetings and access configured event types and availability. The flow is shown on a desktop device. **SavvyCal: New account setup and guided onboarding** (web, 27 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3346) > Summary: This is the new-user acquisition -> onboarding flow. Steps: Marketing landing / entry → Sign up (method selection) → Sign up (form completion) → Welcome — prefilled account settings → Edit scheduling URL (modal) → Edit URL — example placeholder → Edit URL — save new slug → Time zone selection (expanded list) → Time zone search results → Theme selection and time format → Usage type: individual vs team → Team size selection → Role selection (which best describes you) → Role selected (example) → Primary goal / prioritization → Primary goal selected → Connect your calendars → Wire up video conferencing → Set your working hours → Onboarding complete — Guided Setup dashboard **Homerun: Candidate Interview Scheduler** (web, 10 screens, Productivity, Communication Tools, Job Search) [Refero](https://refero.design/flows/7186) > Happy-path: A logged-in recruiter opens a candidate profile, uses the Schedule control to create a new interview event, picks a template/date/time/duration and location, adds or confirms attendees, inserts a video-call link and message, previews the email invitation, and sends it. Steps: Candidate profile — Schedule CTA → Schedule actions menu → Event creation modal — initial → Insert template → Set date → Select start time → Set duration → Choose location type → Add video-call link and message → Invitation preview and send **Wrike: In-chat meeting scheduling and GIF send** (web, 9 screens, Business, Productivity, Project Management, Collaboration Tools, Paperwork, Scheduling Tools) [Refero](https://refero.design/flows/1498) > Happy path: A logged-in user on the Inbox (Archive tab) opens the floating Wrike Help/Support chat, requests a sales call, walks through the embedded scheduler (select date → select time → confirm), sees an in-chat confirmation that an email was sent, then uses the GIF picker to insert and send a GIF. Steps: Inbox (Archive) — no archived notifications → Open chat thread → Request sales call → Select date → Select time / timezone info → Confirm meeting slot → Meeting confirmation in chat → Open GIF picker (GIPHY) → Insert and send GIF ##### Task: camera scan verify **Foodllama: Meal Logging and Scan Review** (ios, 5 screens, Health & Wellness) [Refero](https://refero.design/flows/11787) > Happy path for logging a meal in Foodllama and returning to the daily dashboard. The user starts on the main calorie summary screen, opens the camera-based add-food scanner, switches between Food/Barcode/Label capture modes as needed, captures or recognizes a meal/product, reviews the detected food entry and nutrition totals, then confirms the log. The flow assumes the user is already logged in, has permission to use the camera, and is on a mobile device. Steps: Daily Nutrition Dashboard → Food Camera Capture → Barcode Capture → Label Capture → Meal Review and Confirmation **Phantom: Send via QR** (ios, 10 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3706) > Happy path: A logged-in user on the wallet home screen taps Send, grants camera access, scans a recipient QR code, selects Bitcoin as the token, confirms the parsed address, then enters an amount to send. Steps: Wallet Home - Send Action → Camera Permission Prompt → Scanner - Initial View → Scanner - Framed Scan State → QR Preview / Decoded Content → Select Token → Send BTC - Recipient Prefilled → Enter Amount - Amount Input → Sat Protection Modal → Amount Validation & Error **District: Cinema Booking Payment** (ios, 25 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12837) > The user reviews a movie booking for Main Vaapas Aaunga at MovieMax Sion, Mumbai, including two seats and a popcorn add-on. They browse offers and food items, review the donation and payment summary, select card payment, add a Visa card using camera scanning, enter the CVV, and complete third-party card verification. The system then processes payment and confirms the theatre seats. The flow starts on the booking review screen with an active completion timer and ends with a confirmed booking containing a QR code. Steps: Booking Review → Offers and Food Selection → Donation and Payment Summary → Expanded Cost Details → Contact and Terms Review → Payment Method Selection → Add Card Form → Camera Permission → Card Scanner → Scanned Card Details → CVV Entry → CVV Processing → Payment Method Applied → Payment Gateway Loading → Issuer Authorization → Payment Confirmation Wait → Seat Confirmation → Theatre Processing Delay → Confirmation Countdown → Booking Confirmed **Chance AI: Image-based assistant answer and follow-ups** (ios, 4 screens, AI Tool) [Refero](https://refero.design/flows/7358) > Happy-path: User opens an image-based assistant result and is presented with specialized agent cards (e.g., OOTD Fitchecker, Product Value Recognizer), quick follow-up question suggestions, and actions (Share, Create Webpage). The user taps a suggested follow-up question, triggering the assistant to generate a long-form answer. The answer replaces the suggestion area and is scrollable with a collapsible handle; footer controls (feedback, Share, Create Webpage) and new follow-up suggestions remain available. Entry: image submitted / image-focused assistant view. Exit: user reads/scans answer and either shares it, asks another suggested question, or closes the view. Steps: Agent suggestions & quick actions → Select suggested follow-up / loading → Answer displayed and readable → Feedback, share, and next suggestions **Fuse: Recipient selection** (ios, 3 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8738) > Happy path: user opens the recipient selector to enter a destination address (explicit address or .sol domain). From the entry screen the user can paste an address (or trigger a scan) — tapping Paste triggers the iOS system clipboard permission dialog. If the user allows paste, the clipboard contents are inserted into the address field; the app performs a lookup and shows matching wallet(s) below the input. The user then selects the matching wallet as the recipient. Steps: Recipient entry → iOS clipboard permission → Match results / Choose wallet **BeReal.: In-Conversation Emoji Reactions** (ios, 3 screens, Social Networking) [Refero](https://refero.design/flows/4929) > Happy-path: A logged-in user opens a one-to-one conversation with "Sammy" (mobile, dark theme), reads the messages, reveals the emoji reaction options, selects an emoji reaction, and sees the reaction appear attached to the message. **Mela: Create and Save Recipe** (ios, 28 screens, Food & Drink) [Refero](https://refero.design/flows/10363) > The user starts in an empty Recipes section and opens the recipe actions menu. They choose to create a new recipe, add two photos from the device photo library, and enter recipe information including the title, source, category, servings, and preparation and cooking times. The recipe editor displays imported ingredients, instructions, notes, and nutrition content. After reviewing the completed recipe, the user exits to the Recipes list, where the saved recipe appears in the collection. The flow assumes the user is on a mobile device and has access to a photo library. The screenshots show an initially empty recipe collection and do not show an explicit Save tap, although the completed recipe appears saved in the final list. Steps: Empty Recipes → Recipe Actions → Blank Recipe Editor → Image Source Menu → Photo Selection → Selected Recipe Images → Recipe Title Entry → Completed Recipe Title → Source Entry → Recipe Metadata → Ingredients Review → Instructions Review → Notes Review → Nutrition Review → Recipe Detail View → Recipe Image Carousel → Recipe Overview → Ingredients and Instructions → Saved Recipe Content → Recipes List Confirmation **WhatsApp: WhatsApp QR Scanner** (ios, 4 screens, Social Networking, Collaboration Tools) [Refero](https://refero.design/flows/12607) > A camera-based WhatsApp QR scanner opens with a live scan frame and controls to close or toggle flash. The user can scan a QR code from another device or choose a code from the gallery/My code area. After a successful scan, the app identifies the WhatsApp account and presents a bottom sheet to add it to contacts. Steps: Scan Prompt → Camera Scan View → Code Detected → Account Found **LEGO® Builder: QR scanner onboarding and scan** (ios, 4 screens, Entertainment) [Refero](https://refero.design/flows/10128) > Happy path: From the app home screen the user opens the QR scanner via the QR icon in the search bar, grants camera permission, uses the in-app camera to align the printed QR code in the scanning frame, and receives a visual confirmation when the code is detected. Steps: Discover / Search header → Camera permission prompt → Scan QR code (camera view) → Scan success confirmation **Google Translate: Camera Text Translation** (ios, 6 screens, Reference, Utility) [Refero](https://refero.design/flows/12139) > A user starts on the Google Translate home screen with Thai and English selected, then opens the Camera feature. After granting camera permission and completing the camera-translation introduction, the camera recognizes English text in a résumé and displays a Spanish translation over the document. The user can then view translated text actions in a bottom sheet. The flow assumes an iPhone user who is not currently using the camera feature and has camera permission unset. Steps: Translate Home → Camera Permission → Camera Translation Introduction → Live Translation View → Detected Text Selection → Translated Text Actions **LEGO® Builder: Scan QR to open set** (ios, 6 screens, Entertainment) [Refero](https://refero.design/flows/10127) > Happy path: From the home Discover screen the user taps the QR icon in the search bar to scan a QR code printed on the paper instructions included with a LEGO set. The app requests camera permission (if not already granted). With permission, the built-in scanner opens, shows a scanning overlay and an instructional card, detects the QR, and displays a success confirmation modal with the identified set and two CTAs: 'Start Building' (primary) and 'Earn points' (secondary). Steps: Discover / Search header → Camera permission prompt → Scanner (instructions visible) → Scanning / detection state → Scan success (generic) → Scan success (set detail) **Endless Tools: 3D Text Material & Background Customization** (web, 7 screens, Design & Media Editing) [Refero](https://refero.design/flows/5156) > Happy path: A user customizes a 3D extruded wordmark by selecting an environment/preset, choosing a thumbnail style, opening the material/properties panel, and changing background color and lighting parameters. Steps: Initial Canvas → Open Preset Menu → Expand Category Thumbnails → Choose Preset Thumbnail → Open Properties Panel → Adjust Background Color and Intensity → Review Final Render **Matter: Sign-in to Queue** (web, 6 screens, Magazines & Newspapers, News) [Refero](https://refero.design/flows/9408) > This is the happy-path sign-in and authentication journey from the public marketing hero to an authenticated reading area (Queue). Steps: Marketing Hero → Sign-in Modal → Enter Email & Submit → Check Your Email Confirmation → Magic Link Email Received → Authenticated Queue **time2book: Edit Account Profile** (web, 3 screens, Business, Fitness, Productivity) [Refero](https://refero.design/flows/6312) > Happy path: A signed-in user opens the Settings area and selects the Account tab to update their public profile. From the Account screen they upload a profile photo and profile cover, enter or edit text fields (About, Full name, Phone, Username) and observe immediate previews of uploaded images. A confirmation toast appears indicating the profile was updated. **LinkedIn: New member onboarding and job preferences** (web, 20 screens, Business, Social Networking, Job Search) [Refero](https://refero.design/flows/810) > Happy-path onboarding for a new LinkedIn member who starts from the marketing landing area, creates an account, completes basic profile and job-preference setup, connects with suggested people, follows recommended sources, and lands on a personalized jobs feed. Steps: Entry - marketing landing → Create account - credentials (empty) → Create account - credentials (filled) → Name collection - validation error → Name collection - completed → Security verification - phone entry → Security verification - phone error → SMS code entry → Location - country & postal code → Profile - recent job title → Profile - recent company → Job seeking intent → Job preferences - titles & locations → Job alerts - toggle → Open to Work - recruiter sharing → Optional - mobile app QR → Connect - people suggestions → Follow - companies / courses / people → Onboarding complete - transitional state → Jobs hub - personalized feed ##### Task: care plan **Ada: Record a symptom entry** (ios, 12 screens, Health & Wellness) [Refero](https://refero.design/flows/4107) > Happy path: From a feedback prompt the user chooses to track symptoms, picks a symptom (Dizziness), answers a short guided questionnaire (trigger: standing up effect; description type), optionally sets current severity using a segmented color-coded slider and +/- controls, optionally adds freeform notes, saves the entry, then returns to a timeline where the new entry can be opened to view details. Steps: Feedback prompt → Choose symptom → Initial symptom questions → Select trigger effect → Select symptom description → Open severity control → Set severity (low) → Set severity (mid) → Set severity (high) → Open notes field → Enter free-text note → Save and view entry **Ada: Profile setup and symptom logging** (ios, 66 screens, Health & Wellness) [Refero](https://refero.design/flows/4095) > Happy path for adding a new user profile and logging a symptom (dizziness) into the symptom tracker. Steps: Home / Start assessment → Add profile — Intro & safety note → Ask for user name → Name input active → Ask sex assigned at birth → Why only two sexes (modal) → Ask date of birth → Ask smoking history → Ask high blood pressure history → Open symptom modal (Dizziness) → Dizziness — trigger & type → Dizziness — optional severity collapsed → Set severity — low value → Set severity — mid value → Set severity — high value → Optional notes — edit → Save symptom (network notice) → Symptom tracking timeline → Symptom detail summary → Finish / Done **Ada: Medical quality article walkthrough** (ios, 10 screens, Health & Wellness) [Refero](https://refero.design/flows/4104) > Happy path: A user finishes a symptom assessment and sees a high-priority recommendation ("Seek emergency care") with an inline "Tell me more" button. Tapping that link opens Ada’s external "Medical quality" article (about.ada.com) in a browser/webview. The user scrolls through the article: headline, supporting copy, visual illustrations, a "What to look for" section with accordion items (clinical governance, user experience, condition coverage, etc.), expands "Condition coverage" to view bullet points, continues to the "How Ada’s AI works" section, sees a CTA to contact the company, reviews the reference list (numbered DOIs), and reaches the footer with legal links and social icons. Steps: Symptom result with "Tell me more" → Article landing — Medical quality → Evidence section header → "What to look for" introduction → Accordion list of quality pillars → Expand 'Condition coverage' → How Ada's AI works → Contact CTA and Keep reading → Reference list (citations) → Footer and site navigation **Doctronic AI: Start Health Consultation** (web, 2 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13693) > The flow begins on a logged-in health assistant home screen with a collapsed navigation rail. The user expands the navigation to view available sections while remaining on the New consult screen. From here, the user can enter a health question and submit it. The screenshots show the entry state and expanded-navigation state; no question has been submitted yet. **MonoDesk: Conversational task planner** (web, 8 screens, Productivity, Project Management) [Refero](https://refero.design/flows/10852) > Happy-path: A logged-in user opens an in-product conversational assistant to request a personal productivity/growth plan. The assistant asks clarifying questions, offers quick-reply choices for focus areas, generates a week-based task plan, and then provides CTAs to add professional tasks or schedule the generated tasks into the product Planner (calendar). Entry: assistant chat screen. Exit: Planner view filtered to the user's selected project with the new tasks visible. Assumptions: user is authenticated, has at least one project (named Humanity) available, and the product can create tasks and update the Planner/calendar. Steps: Chat start / prompt → User submits goal → Assistant thinking → Assistant asks clarifying questions → Quick-reply focus selection → Assistant generates task plan → Task detail links and CTAs → Planner view confirmation and options **Longevity Deck: First Launch Introduction** (ios, 5 screens, Health & Wellness) [Refero](https://refero.design/flows/9879) > This is a first‑time user onboarding carousel that introduces the app’s core features, asks the user to acknowledge a medical disclaimer, then lands the user on the app’s main entry screen where they can begin setup (upload a photo, browse protocols, or connect health data). Steps: Intro: Personal Story → Feature: Connect Health Data → Feature: Track Daily Timeline → Legal: Medical Disclaimer → Main Entry: Setup & CTAs **Structured: Handwritten Planner Extraction** (ios, 5 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3293) > Happy-path: A user scans a handwritten weekly planner and reviews AI-extracted calendar items. From a modal titled "Structured AI" the user can preview extracted tasks, open a larger scan preview, add custom instructions, regenerate the extraction, and then review the updated list. Steps: Initial extraction preview → Detailed scan preview → Provide custom instructions → Regeneration / processing → Updated extraction results **Napper: Baby schedule & invite** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/8812) > Happy-path: A parent edits a baby's schedule and invites a caregiver. Steps: Schedule settings → Share baby profile (invite) → Caregiver added / list view **Sunsama: AI-Assisted Task Capture** (web, 7 screens, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/11601) > A logged-in user is viewing the Backlog and opens the Sunny assistant from the floating action button. The user selects the Braindump context, sends a natural-language request about planning a trip to Tokyo, and receives confirmation that Sunny added a corresponding task to the backlog for the next month. The flow begins in the Backlog workspace and ends with the new task visible in the appropriate time bucket. Steps: Open Sunny Assistant → Choose Braindump Context → Start Conversation → View Assistant Prompt → Enter Travel Idea → Send Travel Request → Confirm Backlog Task **Lovable: Plan Review and Approval** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12337) > A logged-in user is reviewing a proposed plan in a split-screen editor. The left side shows the task discussion and a plan card with a summary, while the right side shows the detailed plan content with a formatting toolbar. The user can inspect the proposal, optionally change the content type/format using the text style dropdown, and then either cancel or save and approve the plan. Steps: Plan Review → Text Style Selection → Approved Plan State **Care: Caregiver Review Submission** (web, 5 screens, Job Search, Lifestyle) [Refero](https://refero.design/flows/12231) > Happy path for leaving a review for a caregiver on Care.com. The user starts on a caregiver profile, opens the review form, selects an overall star rating, adds written feedback, answers whether they hired the caregiver through Care.com, submits the review, and reaches a confirmation screen with a call to action to continue browsing care options. Steps: Caregiver Profile → Start Review → Write Review → Review Ready to Submit → Submission Confirmed **Care: Pet Care Caregiver Search** (web, 12 screens, Job Search, Lifestyle) [Refero](https://refero.design/flows/12230) > Happy path for finding and narrowing pet care caregivers on Care.com. The user starts on the Pet care search results view with location, date, and basic search controls already set. They then refine the list using filter chips and dropdown panels (smart search, pay rate, pet type, rating, and professional skills), and the results list updates after each applied filter. The flow begins on the caregiver search results screen and ends on a filtered set of pet care profiles with a selected caregiver highlighted in the details panel, ready for contact. Steps: Results Overview → Smart Search Entry → Smart Search Applied → Pay Rate Filter → Pay Range Updated → Rate Filter Applied → Pet Type Selection → Pet Type Applied → Rating Filter → Rating Applied → Professional Skills Filter → Filtered Profile Review ##### Task: case management **Cohere: Account registration and onboarding** (web, 15 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3790) > Happy-path desktop web flow for a new user who discovers the product marketing site, creates an account, completes an onboarding questionnaire, and arrives at the product dashboard. Steps: Marketing landing → Create account — entry form → Password strength — weak → Password strength — improving → Password strength — good → Password strength — strong → Verify your email → Onboarding — name collection (step 1/3) → Onboarding — name confirmed → Onboarding — role selection (step 2/3) → Role selected → Onboarding — use case selection (step 3/3) → Use cases selected with custom note → Dashboard — welcome → Dashboard — modules and docs **Polar: Organization Review and Submission** (web, 4 screens, Business, Finance) [Refero](https://refero.design/flows/11406) > Desktop flow for submitting an organization's details for an automated compliance/use-case review. Steps: Account review form (initial) → Edit fields and add social link → Submission confirmation → Automated review in progress **Google Gemini: Access Gemini Help Center** (ios, 7 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/3226) > Happy-path: A signed-in user opens the account menu, chooses Help, then navigates into the Gemini Apps Help support site and reads help articles. Entry: user is already signed into a Google account on the device (profile avatar visible). Exit: user ends on a specific support article (’Use the Gemini web app’) and can browse related resources or give feedback. Assumptions: user is signed in, has network access, and taps controls on a mobile device using the system browser or an in‑app browser sheet. Steps: Open account menu → Help selection → Open Gemini Help hub → Browse popular topics → Open specific article → Read detailed sections → Access related resources & feedback **Fresha: Email & Social Sign-up** (ios, 17 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4946) > Happy-path and edge cases for a new-user sign-up started from the app's home/browse screen. Steps: Home / Browse (Logged out) → Sign-in Modal - Entry → Email Input Focus → Invalid Email Feedback → Valid Email Entered → Create Account - Name Entry (step 1) → Create Account - Name Autofill / Input → Create Account - Password & Mobile (step 2) → Country/Region Selector - List → Country Search / Select → Mobile Number Entry → Consent Checkboxes → Attempt Account Creation → Server Error (failure case) → iOS Notification Permission Prompt → Account Created - Profile Landing → Post-signup State & CTA **Omio: Checkout & ticket retrieval** (ios, 23 screens, Travel & Booking) [Refero](https://refero.design/flows/5185) > Happy-path checkout for a single passenger bus booking from Barcelona to Madrid. Steps: Booking summary & passenger → Seating options preview → Protection options list (collapsed details) → Protection details – first expansion → Select Cancellation Protection (state required) → Protection details expanded with coverage list → Protection selected confirmation → Switch protection choice → Select payment method → Choose Card and show card form → Card number input (numeric keyboard) → Card details filled & save option → Enter voucher code (collapsed) → Voucher typing → Voucher rejected error → Review cost summary & terms → Initiate payment (processing animation) → Payment success → Retrieve ticket → Booking confirmed modal **Days: Support chat conversation** (ios, 5 screens, Scheduling Tools, Lifestyle) [Refero](https://refero.design/flows/7818) > Happy path: A user opens the app's Support hub, chooses to contact support, and is routed into an in-app chat handled by an automated assistant (Days AI). The assistant shows an intro message and an input field. The user types a question, sends it, sees a sending/loading state, and receives an automated reply with instructions. Steps: Support hub (help topics) → Chat assistant intro → Compose message → Send message & pending state → Bot responds with instructions **Mindllama: Focus Session Setup** (ios, 27 screens, Health & Wellness) [Refero](https://refero.design/flows/5490) > Happy-path: user opens the main screen (large 'Focus 15 min' CTA) and configures a single focus session by selecting sounds, adjusting volumes, choosing a background scene, setting the session duration, picking a character, then saving the configuration as a preset. Entry: main/home screen with tabs (Sound, Background, 15 min, Character, Presets). Exit: a saved preset (optional) and the main screen updated with the selected session values. Assumptions: mobile device; user has the app open and can interact with touch controls; audio playback may be blocked by the device's silent mode (the app surfaces a modal in that case); sounds can be downloaded for offline use. Steps: Main screen (initial) → Scroll sound cards → Select / preview sound → Silent-mode modal appears → Dismiss silent-mode modal → Change selected sound → Open floating sound controls → Switch sound category → Ambient sound list (collapsed) → Select multiple ambient sounds → Adjust per-sound volume → Floating pill sync → Open Background tab → Pick a scene from grid → Open 15 min (time) tab → Switch to Hours mode → Adjust hour value → Open Character tab → Select a different character → Open Presets tab and create preset **Maze: New account onboarding** (web, 11 screens, Business, Communication Tools) [Refero](https://refero.design/flows/3602) > Happy-path new-user onboarding for Maze (desktop). Entry: marketing landing hero where user clicks CTA to create a free account. The flow collects credentials, basic profile (name, company, role, company size), use-case preferences, verifies email, offers an invite-by-link or email step, and lands the user on a time-limited trial with the first-run dashboard. Steps: Marketing hero → Get started → Signup form (empty) → Signup form (email entered) → Profile details form (empty) → Profile details filled → Use-case selection → Confirm your email → Email verified → Invite team → Invite team (email entered) → Trial welcome / Entitlement → First-time dashboard / Home **Reclaim AI: Reclaim account setup** (web, 24 screens, Collaboration Tools, Scheduling Tools) [Refero](https://refero.design/flows/4082) > Happy-path new user onboarding for a calendar-first AI scheduler (marketing → sign-up → guided setup → product launch). Entry: marketing/marketing-signup or invite link. The user signs up (Google), then completes a step-by-step setup: connect a primary work calendar, select use case, role and department, enable recurring habits, create scheduling link, opt into buffer times, enable calendar sync, optionally invite teammates, then waits for account provisioning and lands in the product with a setup guide. Steps: Marketing landing → Sign up prompt → Welcome / overview → Connect work calendar → Calendar connected confirmation → Select primary use → Use case selected (example) → Select role → Role selected (example) → Select department → Habit templates intro → Habit template selected → Create scheduling links → Auto-schedule buffer intro → Buffer options enabled → Personal calendar sync intro → Create a team prompt → Team invite review → Account provisioning → Onboarding complete — Setup guide **Matter: Manage Gmail Sources** (web, 3 screens, Magazines & Newspapers, News) [Refero](https://refero.design/flows/9397) > Happy path: A signed-in user who has already connected a Gmail account lands on the "Manage Gmail" area to review, search, and follow sources that were extracted from that Gmail account. The screen lists discovered senders (name, email) with a Follow/Following button per row. The user can type into the search input to filter the list; the UI updates to show matching senders or a "no results" message. Steps: Manage Gmail — Full list → Search — no matches → Search — filtered result ##### Task: code repository pull request **Wabi: Invite Collaborator** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12045) > Happy path for inviting another user to join a shared space from a collaboration screen. The user starts on a member management view showing current participants and invitation options. They can search for a person by username, tap an Invite action next to a matching result, and return to the main view where the member count updates. Steps: Member Overview → Search Active → Search Results → Invite Confirmed → Updated Member Count **Dropbox: Share File Request** (web, 6 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/13057) > A logged-in Dropbox user opens the File requests area, starts sharing an existing file request, selects a recipient by email, and sends the request. The flow begins on the File requests list and ends with the modal closing and a confirmation toast indicating that the Personal files request was shared. Steps: File Requests List → Share Request Dialog → Search Recipient → Select Recipient → Confirm Request Details → Share Confirmation **Blackbird: Referral Invite** (ios, 2 screens, Food & Drink) [Refero](https://refero.design/flows/9024) > A mobile referral/promotional modal that shows a scannable QR code, an explanatory headline and copy about the reward, the user's referral code, and two primary actions: Copy Code and Share Link. Steps: Referral Promo Display → Copy Confirmation **Linear: Project selector** (ios, 3 screens, Collaboration Tools, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/6614) > This is a mobile project-picker modal used to reassign an item to an existing project. Entry: user opens the "Project" picker from an item (implicit). The modal lists available projects and exposes a search field at the top. The happy path: open picker → type or scroll to find a project → tap a project to select it → close the picker. Assumptions: user is on a mobile device, keyboard appears when the search field is focused, user has permission to reassign the item, and there are multiple existing projects to choose from. Exit points: selecting a project (selection shown with a checkmark) or tapping the top-right close button to dismiss the modal. Steps: Project picker (initial list) → Search — empty results → Search — single match **Rork: Invite Collaborators** (web, 2 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12169) > A signed-in user opens an invite/collaboration panel for a project, adds collaborators by email or creates a shareable invite link, then gets confirmation when the link is created or copied. The flow starts from a project workspace and ends with invitation access being generated and ready to share. Steps: Invite Collaborators Panel → Link Copied Confirmation **Linear: Assign Project** (ios, 4 screens, Collaboration Tools, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/6613) > Happy path: A user viewing an issue in mobile detail opens the Project picker, searches for a target project (Pulse), selects it, and returns to the issue which now shows the updated project tag. Entry: issue detail screen where project is displayed as a pill. Exit: issue detail updated to show the new project. Assumptions: user is signed in, has permission to change the issue's project, and is on a mobile device (keyboard can appear). Steps: Issue detail (view) → Project selector (list) → Project search result → Issue detail (updated) **HYPE: Comment Discussion Panel** (ios, 2 screens, News, Lifestyle) [Refero](https://refero.design/flows/12248) > A mobile comments panel is opened over an underlying content detail screen. The user can view one existing comment, react to it with like/dislike, tap Reply, or type a new comment in the input field at the bottom. Entry appears to be from a content item with a comments trigger; exit is via Close or the back navigation on the underlying screen. Steps: Comments Open → Comment Liked **Podimo: Podcast Comments Thread** (ios, 2 screens, Entertainment) [Refero](https://refero.design/flows/12270) > A mobile comments drawer is open for a podcast episode/post titled “1×01 La partida con SoyUnaPringada y Dar…”. The user is viewing a threaded discussion with top-level comments, nested replies, like counts, and a text field to add a new comment. Assumption: the user is already inside the content and has permission to read and interact with comments. Steps: Comments View → Liked Comment State **Wabi: Recipe Comments** (ios, 4 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12066) > This is a mobile comment view for a shared collection called “Family recipe hub.” The user enters the screen from the comments tab of the collection, views existing comments and engagement counts, then taps the comment field to open the composer and type a response. A send action posts the comment and updates the comment count/list. Steps: Comments View → Comment Composer Open → Comment Drafting → Comment Submitted ##### Task: community forum **X: Community Scoped Search** (ios, 2 screens, Social Networking, Entertainment, News, Video Streaming) [Refero](https://refero.design/flows/3010) > Happy path: A mobile user browsing a public community (The Design Sphere) taps the search control, enters a query scoped to that community, and is taken to a search results screen. Entry: community profile/timeline. Exit: scoped search results screen showing either matches or an empty state with guidance. Assumptions: mobile device (iPhone-like status bar and bottom nav present), user can view the community but appears not to be a member (Join button visible), and search is scoped to the community tab. **YouTube: Community Post Deletion** (ios, 2 screens, Entertainment, Video Streaming, Social Networking) [Refero](https://refero.design/flows/2640) > Happy path: a signed-in channel owner views their channel's Community tab on mobile, locates a published post, opens the post options and chooses to delete it, then confirms deletion in a modal. Steps: Community - Post view → Delete confirmation dialog **HYPE: Article Comments** (ios, 5 screens, News, Lifestyle) [Refero](https://refero.design/flows/12246) > A user opens an article detail screen and interacts with the comments area in a bottom-sheet overlay. The entry point is the article page; the exit point is either returning to the article or leaving after posting a comment. Steps: Article Detail → Empty Comments State → Compose Comment → Submit Comment → Comment Thread **Longevity Deck: Air Purifier — Community Commenting** (ios, 6 screens, Health & Wellness) [Refero](https://refero.design/flows/9885) > Happy path: A signed-in mobile user browses the Discover carousel, taps the "Air Purifier" card to open its detailed view, opens the Community / Comments area, types and submits a comment, then returns to the Discover card where the Community badge and preview are updated. Entry: Discover carousel (signed-in user, avatar visible). Exit: Discover carousel card shows updated community count and preview. Assumptions: user is authenticated (avatar present), on a mobile device, and has permission to post comments. Steps: Discover — Card Browse → Card Detail — Open Community → Compose — Activate Input → Compose — Type Comment → Submit — Post Comment → Return — Card Summary Updated **Longevity Deck: Product Comments & Interactions** (ios, 2 screens, Health & Wellness) [Refero](https://refero.design/flows/9886) > This is a product-level comments screen for a device called “Air Purifier.” Entry point is the product detail / comments area; exit is returning to the previous screen (back arrow). The screen is optimized for a single-column mobile view and includes: product image + title at the top, a persistent disclaimer banner, a chronological comment list, inline actions for each comment (like/heart and an options menu), and a bottom text input with a send control. Assumptions: user is on a mobile device; the comment area is interactive (there is an input field and like control visible). Happy path: user reads existing comments, likes a comment (visual heart toggles and a count appears), optionally types a new comment in the bottom input and sends it, or leaves the comment area by tapping the back control. Steps: Comments - initial read state → Comment liked / updated state **Podimo: Delete Comment from Thread** (ios, 3 screens, Entertainment) [Refero](https://refero.design/flows/12273) > A logged-in user is viewing a podcast discussion thread on mobile and managing one of their own comments. The screen shows a scrollable list of comments and replies, with a comment composer fixed at the bottom. The user swipes a comment to reveal a delete action, taps delete, and receives a confirmation toast indicating the comment was removed. Entry is from a podcast episode discussion; exit is the same thread after deletion feedback appears. Steps: Comment Thread → Delete Action Revealed → Deletion Confirmed **Train Fitness: Comment list and composer** (ios, 2 screens, Health & Wellness, Social Networking) [Refero](https://refero.design/flows/7140) > A bottom-sheet comments panel showing a short list of comments and a persistent composer at the bottom. Entry: user opens the Comments panel (header shows total count). The panel lists comments with avatar, display name (tappable), timestamp, comment text, a Reply link, and a heart icon for likes. A composer with the user's avatar, placeholder text, and a send icon sits fixed at the bottom. Exit: user closes the panel or navigates away after reading, liking, replying, or posting a new comment. Steps: Comments list — initial → Comment liked — updated state **Longevity Deck: Product comments management** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/9888) > Mobile product detail -> comments area. Entry: user opens an individual product (Air Purifier) and scrolls to its comments section. The screen shows a product hero (image + title), a prominent disclaimer banner about medical advice, a comment list (avatar, name, timestamp, text, like icon, and overflow menu), and a persistent comment input at the bottom. Exit: after deleting the shown comment the list becomes an empty-state prompting the user to add the first comment. Assumptions: mobile portrait, user can view and post comments (input visible), at least one comment is present in the initial state. Steps: Comments list visible → Reveal delete action → Empty comments state **Wabi: Post Comments** (ios, 2 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12063) > This screen shows the comments section for a post titled “Psychoanalysis insights.” The user is viewing engagement on the post and can switch between Likes, Comments, and Remixes via the top tabs. The entry point is the comments tab on the post detail screen; the exit point is either staying in the comment thread, liking comments, or adding a new comment from the bottom composer. Assumption: the user is already logged in and viewing an existing post on a mobile device. Steps: Comments Overview → Comment Reaction State **Substack: Create and send first chat thread** (ios, 5 screens, News, Social Networking) [Refero](https://refero.design/flows/4501) > Happy-path flow for creating and sending the first chat thread in a publication's subscriber chat (mobile). Entry: user opens a publication's chat (logged-in author). Exit: a new thread with images and text is posted and visible in the chat; reply UI becomes available. Steps: Open publication chat → Compose new thread with attachments → First-message confirmation → Send in progress → Thread posted and visible **Sora: Comment reply to specific commenter** (ios, 4 screens, AI Tool, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/7829) > Happy path: A user opens a comment thread (sheet) that lists replies to a post, selects a specific comment to reply to, composes a reply using an inline composer that appears above the keyboard, and sends the reply. Steps: Replies list opened → Targeted reply composer opened → Compose reply → Reply posted and displayed **Substack: Comment thread interactions** (ios, 3 screens, News, Social Networking) [Refero](https://refero.design/flows/4466) > Happy-path: a signed-in user is viewing an article's comments list, opens the per-comment options menu, chooses an action from a bottom-sheet, and returns to the comments list with the UI updated or the chosen modal/OS sheet presented. Entry: Comments screen (header shows back and sort controls). Exit: options sheet dismissed and the comments list reflects the chosen action or remains unchanged if dismissed. Steps: Comments list → Comment options sheet → Post-action comments state ##### Task: compliance dashboard audit **YouTube Music: Report a comment** (ios, 5 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11336) > User is viewing the comments sheet for a video (comments are sorted by Top or Newest). From this sheet the user opens the per-comment actions menu, chooses to report a specific comment, picks a reason from a modal list, and receives a submission confirmation. Entry: comments sheet is open for a video. Exit: confirmation modal is dismissed and the user returns to the comment list. Steps: Open Comments and locate comment → Open per-comment actions → Choose report reason → Submit report → Acknowledge confirmation **ElevenLabs: Developer Usage Analytics** (web, 2 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/8713) > Happy path: A signed-in developer navigates to Developers → Usage to inspect API consumption and credit usage for a selected time window. The interface presents an Analytics panel with controls for metric selection (e.g., Credits), date range, granularity, log scale, cumulative toggle, breakdown/filter options, and an Export CSV action. The user can switch visualization modes (line vs cumulative area) and review related sections below (API Requests, High priority concurrent requests) which display aggregated values. Entry: user opens the Developers area (desktop layout) and selects the Usage tab. Exit: user either exports CSV, adjusts filters, or navigates back to another Developers tab (Overview, API Keys, Webhooks). Assumptions: user is authenticated with developer-level access; using a desktop or large-screen browser; has permission to view usage analytics and export data. Steps: Usage - Default (Per-interval Credits) → Usage - Cumulative Credits View **Runey: Invoice PDF Download** (web, 2 screens, Business, Productivity, Project Management) [Refero](https://refero.design/flows/11704) > A logged-in user is viewing a completed invoice inside a project workspace and downloads it as a PDF. The entry point is an invoice detail/preview screen showing invoice metadata, line items, totals, and attachments. The exit point is a successful download confirmation toast after the PDF export action. Steps: Invoice Preview → Download Success **Lovable: Generated Landing Page Review** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12309) > This flow shows a user reviewing an AI-generated SaaS landing page inside a split workspace. The user starts from a prompt describing the desired landing page, then inspects the generated result in a live preview on the right while the prompt/result card remains on the left. The user can bookmark the output to history and optionally give feedback via reaction controls. The flow appears to assume the user is already signed in or has access to the workspace and preview environment. It ends with the generated landing page visible and a confirmation message that the result was saved/bookmarked or feedback was submitted. Steps: Preview Generated Design → Bookmark Result → Submit Feedback **Foodllama: Nutrition and body stats review** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/11789) > Happy path starts on the main Foodllama dashboard, where the user can review daily calorie intake, macro distribution, streak status, and weekly trend data. From there, the user taps a profile/avatar area to open a detailed stats view with weight, calorie goal, and BMI information, then can move between avatar/profile variations and return to the dashboard. Steps: Dashboard Overview → Detailed Stats View → Avatar Selection State → Dashboard with Selected Avatar **Structured: Checklist modal interaction** (ios, 4 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3266) > Happy path: user opens a day in the calendar timeline, taps an onboarding checklist task to open a modal containing checklist items, selects some checklist items, and closes the modal. The timeline updates the task’s progress indicator to reflect the selections. Steps: Day timeline - onboarding task visible → Checklist modal opens → User checks checklist items → Modal dismissed and timeline updated **Structured: Daily timeline checklist** (ios, 2 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3258) > Single-day timeline view for a selected date (December 23, 2024). Steps: Day timeline — initial view → Item completed state **Tinder: Safety Basics help** (ios, 4 screens, Dating & Relationships) [Refero](https://refero.design/flows/3455) > Mobile (portrait) 'Safety Center' guide flow. Entry: user opens Safety Center and lands on the Guide tab (top tabs: Guide, Tools, Resources). The screen shows a personalized greeting, a list of safety sections presented as tappable cards (The Basics, quizzes, harassment help). Happy path: user taps the 'The Basics' card → navigates to a detail article with an illustration, brief intro, and scrollable content broken into Online and IRL sections. Exit: user scrolls to read content and then uses the back/close control to return to Safety Center. Assumptions: user is signed in (personal greeting and avatar visible) and on a mobile device. Steps: Safety Center — Guide list → Topic header — The Basics → Article content — Online safety → Article content — IRL safety **Foodllama: Daily Nutrition Review** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/11788) > This is a nutrition tracking dashboard for a logged-in mobile user. The user starts on the main daily summary screen, opens a bookmarks overlay to choose a saved meal, then returns to the dashboard where the selected meal is shown in the day’s log. Steps: Dashboard Overview → Bookmarks List → Logged Meal View **ElevenMusic: Music Creation Dashboard** (web, 2 screens, Music Streaming, Entertainment, AI Tool) [Refero](https://refero.design/flows/12569) > A logged-in user lands on a music creation home/dashboard with a left navigation rail, a central prompt-driven track card, and a persistent mini-player at the bottom. The user can review a generated track, adjust or remix it from the main card, or continue an existing project from the "Continue where you left off" section. Entry appears to be the Home area; exit points include publishing, opening Studio, or remixing/continuing a track. The interface assumes the user is authenticated and has at least one existing track/project. Steps: Home Dashboard → Continue Project **Rork: Vocabulary Dashboard** (web, 2 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12149) > Happy path starts on the Vocabulary home dashboard for a Spanish learning app, shown in a mobile device frame. The user is assumed to be logged in and viewing their personal learning summary. The screen provides quick access to core actions such as reviewing due cards, checking daily progress, and navigating to study, decks, or progress sections. The flow shown here is a static dashboard state, with the main exit point being entry into a review session or another bottom navigation destination. Steps: Home Dashboard → Dashboard Focus **Runey: Project Status Update** (web, 3 screens, Business, Productivity, Project Management) [Refero](https://refero.design/flows/11696) > This is a project overview dashboard for a project called “Website Redesign.” The user enters from a project dashboard view, with summary metrics, a task list, activity feed, and calendar. The main interaction shown is changing the project status from an “Active” dropdown to another status such as “Pending.” Assumes the user is already logged in and has permission to edit project settings/status. The flow ends with the project status updated and a confirmation message shown on the dashboard. Steps: Project Dashboard → Status Selection → Status Confirmation ##### Task: connect bank account open banking **PayPal: International bank transfer setup** (ios, 24 screens, Finance) [Refero](https://refero.design/flows/7244) > Happy path for sending money to a new recipient via bank deposit. Entry starts from the "Send" screen (user tapped "Send to someone new"). The user chooses a destination country, enters the sending amount and currency, selects a receiving option (Bank Deposit), reviews fees, provides recipient bank details and contact address, provides sender address, adds a payment method (debit/credit card) with billing address, and attempts to continue. Exit point is either advancing to the transfer confirmation step (not shown) or stopping the flow when a validation error requires correction or the user quits. Steps: Send entry / New recipient → Country selection → Country search → Select United States → Enter amount and currency → Open receiving options → Show fees for chosen option → Confirm receiving option and continue → Recipient bank details → Contact's details (recipient) → Contact's details — phone & email → Sender's details (address) → Payment method selection → Add new card — input fields → Billing address entry → Card validation fails → User attempts to exit flow → Card type selector (picker) → Card details populated (example) → Billing address completed and re-validate **Phantom: Create New Wallet Account** (ios, 7 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3695) > Happy path: From the Your Accounts list the user taps Add / Connect Wallet, selects Create New Account, types a custom account name, and confirms creation. The app then opens the new account's home (balances and actions). Returning to the accounts list shows the new account added and highlighted. Assumptions: user is authenticated and on a mobile device; permissions to create accounts are available; this is the standard "create new" path (not importing or connecting hardware). Steps: Your Accounts (start) → Add / Connect Wallet Options → Create Account (name input shown) → Enter Account Name → Create Account (confirm) → New Account Home → Your Accounts (new account added) **Fingerprint: Find and open account-related documentation** (web, 4 screens, Business, Development Tools) [Refero](https://refero.design/flows/11143) > Happy path: a user lands on the Fingerprint Docs home screen, opens the site-wide search/Ask AI modal, searches for "account" (or a related term), and selects an "Account Sharing" / "Account Takeover" article. The flow ends with the full documentation article loaded, including code examples and step-by-step instructions. Steps: Docs Home / Discover → Search Modal (empty) → Search Results → Documentation Article **Spoil me: Create wishlist and add first items** (ios, 37 screens, Fundraising) [Refero](https://refero.design/flows/5343) > Happy-path new-user onboarding that creates a Spoil.me wishlist (vanity link), verifies identity via an email magic link, imports product photos from the device, adds item details (price, crowdfunding toggle), and lands the user in their wishlist dashboard. Entry: first-time visitor (not signed in) at the hero screen. Exit: activated wishlist/profile screen showing created wishes. Steps: Hero - Choose vanity link → Typing vanity handle → Vanity handle entered → Email input - prompt → Email validation (invalid) → Email input - valid → Confirmation - check inbox → Resend active state → Open verification email (inbox) → Waiting for verification → Photo picker - select images → Photo selection confirm → Item details - edit → Alternate suggestions - browse images → Confirm item images → Add another wish - title input → In-app notifications prompt → OS-level notification permission → Optional payout connection (Stripe) → Wishlist dashboard / onboarding state **Mozi: Scan-to-Connect Contact** (ios, 4 screens, Social Networking, Travel & Booking) [Refero](https://refero.design/flows/4150) > Happy-path: A user scans another person's Mozi QR code with the iPhone camera, the system recognizes the code and offers to open the Mozi connection. The app loads the connection, shows a success confirmation modal, and navigates to the newly connected person's profile in 'My People'. Steps: Camera QR Detection → Open My People (Loading) → Connection Success Confirmation → Person Profile Opened **Binge: Apple Sign In in Settings** (ios, 5 screens, Entertainment) [Refero](https://refero.design/flows/12971) > Happy path in Settings on iPhone. The user starts on a settings screen with account sign-in and general preferences visible. They tap Sign in with Apple, review the Apple sign-in sheet, choose email-sharing options, and continue. After confirmation, the sheet closes and the settings screen updates to show the signed-in profile and account stats. Steps: Settings Overview → Apple Sign In Sheet → Signing In → Sign In Complete → Signed-In Settings **Revolut: Revolut Login and Welcome Onboarding** (web, 9 screens, Finance, Business) [Refero](https://refero.design/flows/76) > This flow starts from the Revolut marketing homepage for potential or returning users, moves to the login page where the user can enter their phone number or use alternative login methods, select their country code if needed, enter a passcode for security, and finally arrives at the Revolut web app onboarding welcome modal and main account home screen. Steps: Landing and Marketing → Login Screen with Phone Number Entry → Country Code Selection → Filtered Country Code Search → Phone Number Entered Ready to Continue → Passcode Entry Prompt → Passcode Entry Completion → Welcome Onboarding Modal → Main Home Dashboard ##### Task: contract management **Acctual: Create bill from uploaded invoice** (web, 2 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8845) > Happy path: A user creates a new bill by uploading a supplier invoice PDF. Steps: Start new bill — upload area → Preview and review extracted details **Google Gemini: Report legal content from Gemini chat** (ios, 18 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/3187) > Happy path: A signed-in user starts in a Gemini conversation, opens the message actions menu on an assistant reply, and chooses to report the content. The app links to Google Legal Help. The user selects a legal reason, taps Create request, completes the online ‘Report a legal removal issue’ form (country, personal details, contact email), attaches screenshots/files from the device photo picker, provides a written explanation, confirms the perjury declaration, types a signature (digital), and submits. The flow ends with a submission confirmation screen telling the user they will receive a confirmation email. Assumptions: user is signed in (avatar visible), using a mobile browser/webview on iOS, has the required evidence files in their photo library, and can receive email confirmations. Steps: Gemini conversation - message actions visible → Message options menu → Legal Help landing → Select reason to report → Create request CTA → Form – country selector (tap) → Form – country list (choose) → Form – fill identity fields → Allegedly infringing content section → Photo picker – browse library → Photo picker – confirm selection → Files attached in form → Explain why content is unlawful → Perjury declaration checkbox → Signature and date → Submit request → Submission success → End state / wait for confirmation **ElevenMusic: Generated Tracks Management** (ios, 4 screens, AI Tool, Entertainment, Music Streaming) [Refero](https://refero.design/flows/12467) > A mobile music-creation screen showing the user’s generated tracks under the Create section, with Drafts and Published tabs. The user can browse draft tracks, open each track’s overflow menu for actions, and trigger deletion. Entry is the Create tab; exit can be back to the draft list after canceling, or removal of the selected track after confirming delete. Steps: Drafts List → Track Actions Menu → Delete Confirmation → Updated Drafts List **Airbnb: Conversation - message compose, actions, edit** (ios, 19 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6431) > User navigates from the Messages list into a single conversation tied to a trip/listing, reads prior messages (including an expired booking invite), composes and sends new messages, interacts with individual message actions (react, copy, reply, report), and edits a previously-sent message. Steps: Messages list → Conversation thread (scrolled) → Conversation top + booking call-to-action → Open Details (trip & conversation actions) → Conversation actions list → Composer with keyboard open → Compose message draft → Send message → Host replies in thread → Long-press host message (actions) → Apply reaction to message → Reply to specific message (start) → Reply composer (typed) → Send reply (quoted) → Long-press user's own message (menu) → Select Edit (warning modal) → Composer pre-filled for editing → Send edited message → Inspect final thread state **Lovable: Workspace People Management** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12299) > This is a workspace member management section for a signed-in user who can view and manage people in a workspace. The entry point is the People area in the left navigation. The exit point is returning to the people list after generating or managing an invite link. Steps: People List → Invite Link Setup → Invite Link Management **Lovable: Workspace invite link management** (web, 2 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12300) > Happy path begins on the People section of a workspace settings area, with a modal open for generating and managing a workspace invite link. The user can choose a role, copy the link, and optionally regenerate or delete it. The flow assumes the user is already in the workspace, has permission to manage members, and is using a desktop interface. It ends after the invite link is successfully copied, with confirmation shown in the modal. Steps: Invite Link Modal → Copy Confirmation ##### Task: course enrolment university **WhatsApp: Add Community Member** (ios, 5 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12678) > Happy path in a community management area. The user starts on the Community info screen as the community owner, opens Add members, selects one contact, confirms the add, and returns to Community info where the member count updates and a success message appears. **Wabi: Invite Members** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12046) > This is a mobile collaboration/invite screen reached from a shared workspace or group context. The user can either share the space, invite others directly, search for users by username, or send an invite link. The happy path starts with the user already inside the invite interface and ends when matching users are found and the user can tap Invite next to a person. Assumption: the user is logged in and has permission to invite others. Steps: Invite Overview → Search Active → Searching → No Match Result → Matched Users **ALO: Prenatal Program Detail** (web, 2 screens, Fitness, Lifestyle) [Refero](https://refero.design/flows/12377) > A logged-in user opens a wellness program detail screen for a prenatal yoga series. The entry point is the program landing page with a hero image, program title, instructor, and tabs for Overview and Classes. The user can review the description, preview the series video, inspect instructor information, browse the full list of classes, view similar programs, and add a comment. In the second state, the user is shown as in-progress with completion tracking and a Stop Program control, indicating they have already started the series. The flow ends when the user leaves the program page, starts or continues a class from the series, schedules a class, or posts a comment. Assumption: the user is signed in and has access to the program content on desktop web. Steps: Program Overview → Program In Progress **District: Movie Seat Selection** (ios, 5 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12829) > A movie seat-selection screen for the film Main Vaapas Aaunga at Maison PVR Jio World Drive, BKC, Mumbai. The user enters with a specific showtime selected and reviews the seat map by row and pricing category. They select seats, optionally unlock and apply an offer from the booking banner, and then proceed to the next checkout step. Steps: Seat Map Review → Seat Selection Update → Offer Unlock State → Offer Applied Confirmation → Ready to Proceed **Binge: Entertainment Onboarding** (ios, 5 screens, Entertainment) [Refero](https://refero.design/flows/12954) > Happy-path onboarding for a media tracking service. Steps: Introductory Tracking → Rating Setup → Library Overview → Resume Watching → History Import **Imprint: New user onboarding and course starter** (ios, 54 screens, Education) [Refero](https://refero.design/flows/5634) > Happy-path onboarding and first-lesson entry for a mobile learning product. Entry: new user opens the app (not signed in). The user is shown a branded welcome screen, guided through 2–3 visual onboarding slides, then a short personalization questionnaire (referral source → demographics → topics). The flow includes an optional email opt-in, a push-permission prompt, and a course selection step. Exit: user selects a course and starts their first interactive lesson. Assumptions: user is on a modern smartphone, not signed in, able to interact with tappable cards and multi-select controls; progress is linear and reflected with a top progress bar. Steps: Welcome / Sign-up entry → Onboarding slide 1 — Concept intro → Onboarding slide 2 — Benefits → Personalization intro → Question — Referral source (part 1) → Question — Referral source (scroll) → Referral selection confirmed → Question — Gender / identity → Question — Age bracket → Question — Topics of interest (multi-select) → Email opt-in modal → Promotional slide — course count → Permission prompt — Push notifications → Course chooser intro → Course selection confirmation → Lesson tutorial overlay → Lesson viewer — card interface → Lesson content sequence → Interactive card / content detail → Lesson start / First content **Handshake: Event registration and confirmation** (web, 2 screens, Social Networking, Job Search) [Refero](https://refero.design/flows/8858) > Happy path: a logged-in student on desktop views an event detail for a virtual Microsoft career talk, uses the right-hand registration card to register, and receives an immediate confirmation state in that same right rail. Steps: Event Details (pre-registration) → Registration Confirmation ##### Task: credit score report **Lovable: Referral Invite Modal** (web, 2 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12350) > A centered referral modal opens over a dimmed dashboard/sidebar view. The user is shown a shareable invite link and QR code with instructions for earning credits through referrals. The intended entry point is a referral/share action from within the product, and the exit point is either closing the modal or copying the invite link. Steps: Referral Details → Link Copied State **Willow: Referral Card Sharing** (ios, 2 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/13241) > A logged-in user views a referral card screen for sharing an invite link or QR code. The entry point is likely from a rewards/referral section, and the exit points are back navigation, copying the referral link, downloading the card, or opening the system share sheet. The screen assumes the user already has an active account and referral code available. Steps: Referral Card View → Copy Confirmation **Train Fitness: User Profile Actions** (ios, 3 screens, Health & Wellness, Social Networking) [Refero](https://refero.design/flows/7143) > Happy path: A signed-in user views another person's profile feed (list of workout/activity cards). From the profile header the user opens the overflow (ellipsis) menu, chooses the Report User action, and receives immediate in-app feedback confirming the report. Steps: Profile Feed (Activity List) → Overflow Action Sheet → Confirmation Toast **GoFundMe: Report Content Submission** (web, 4 screens, Fundraising) [Refero](https://refero.design/flows/11736) > Happy path for reporting content from a public profile. The user starts on a profile/feed-style page, opens the report modal, selects a reason from a dropdown, and submits the report. The flow assumes the user is already logged in and has access to the report action. It ends with a confirmation modal indicating the report was received and will be reviewed by support. Steps: Profile Overview → Select Report Reason → Submit Report → Report Confirmation **Grammarly: Writing Performance Report** (web, 2 screens, Paperwork) [Refero](https://refero.design/flows/1878) > Happy path: a user working in the editor opens a centered Performance modal to inspect writing quality metrics (text score, word count, readability) and then exports a PDF report of those metrics. Entry: user is editing a document and the assistant/score panel is visible in the right-hand rail. Exit: the user either closes the modal after inspection or initiates (and waits for) a PDF report export. Assumptions: user is on a desktop-class viewport (editor with right sidebar visible) and has access to the performance/details control in the assistant; the document contains text and the modal can be opened from the sidebar. Steps: Open Performance modal → Export PDF report **Shop: Product Reporting** (web, 6 screens, Marketplace) [Refero](https://refero.design/flows/13269) > A user views a product detail screen and reports the listing. The user opens the product actions menu, selects the report option, chooses a report reason, optionally adds supporting details, and submits the report. The flow begins on a product detail view and ends with a confirmation toast indicating that the report was sent. Steps: Product Detail → Choose Report Reason → Select Other Reason → Add Report Details → Review Report Details → Report Confirmation **Tidal: Artist Credits and Track Liking** (web, 2 screens, Music Streaming) [Refero](https://refero.design/flows/3234) > Desktop, logged-in user (profile visible in left nav) is viewing a credits list for a single artist. The screen shows a header "Credits for [Artist]", a row-based table of tracks with columns (#, Title, Role, Artist, Album, Year, Time), and filter chips across the top (All, Songwriter, Producer, Engineer, Production team). A persistent mini player sits at the bottom showing the currently playing track. Steps: Browse artist credits → Like/favorite a track from list ##### Task: crypto wallet send receive **Fuse: Send SOL to external wallet** (ios, 14 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8734) > Happy-path: A logged-in user on the Wallet home screen uses the floating + button to start a Send to wallet flow. The user selects Send → To wallet, chooses Solana as the coin, selects or scans a recipient address, enters an amount using the numeric keypad, reviews the transfer (including on-chain fee sponsorship), authenticates, and sends. The flow exits back to the Wallet home with a transient Sending indicator and then a Sent confirmation; the displayed balances update. Assumptions: user is authenticated, has network connectivity, has a positive balance in the selected coin, and the device supports camera access and biometric authentication. Steps: Wallet Home → Quick Actions → Send Destination Choice → Select Coin → Choose Recipient (manual/QR) → Camera Permission Prompt → QR Scanner → Recipient Auto-filled → Enter Amount (keypad) → Amount Confirmation → Review Transfer → Authenticate & Send → Broadcasting / Sending State → Sent Confirmation & Updated Balance **Family: Token send initiation** (ios, 4 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/2672) > Happy path: A logged-in user on a mobile (iOS) Ethereum wallet lands on the Tokens view, opens the quick actions menu, selects Send, and is taken to an address entry screen where they can paste or scan a recipient address. The flow entry point is the Tokens list for a single wallet; the exit point is either successful navigation to the next send step (amount/confirm) or an inline validation error if the entered address is invalid (in the screenshots the app blocks sending to the user's own address). Assumptions: user is authenticated, Ethereum network selected, device is a mobile phone, and the wallet has no token balances (empty state shown). Steps: Tokens list / empty state → Quick actions bottom sheet → Send — recipient entry → Send — validation error for self-address **Copperx: Receive USDC via QR/address** (web, 5 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8898) > Happy-path flow for receiving USDC into a Copperx wallet. Steps: Dashboard — Get Paid entry → Get Paid modal — method selection → Receive payment — network selection → Deposit USDC — QR and address display → Dashboard — updated balance and transactions ##### Task: dark mode theme **GO Club: Theme selection and preview** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/6357) > Happy path: User opens the Preferences / Settings screen, inspects the Focus Mode and theme-related rows, opens the Theme picker, selects a theme (System / Dark / Light), and sees the app immediately apply the chosen theme. **Plane Finder: Appearance & Map Color Settings** (ios, 3 screens, Travel & Booking) [Refero](https://refero.design/flows/10415) > Mobile Settings → Customization entry point. User opens Settings and navigates to the 'Dark Mode' panel to choose app appearance behavior and map color. Entry: Settings > Customization (user visible list of customization options). Exit: selection applied — the app UI updates to reflect the chosen appearance (light/dark) and the chosen map color variant; user can also reset to dark-mode defaults. Steps: Settings — Customization → Dark Mode (light UI) — Choose Options → Dark Mode (dark UI) — Confirm Selection **Ground News: Night Mode selection** (ios, 3 screens, News) [Refero](https://refero.design/flows/7582) > Happy-path: From the app Settings > Preferences screen the user opens the Night Mode control, picks a display preference from a bottom-sheet (Light, Dark, System), and the app immediately applies the selected theme. Steps: Preferences list (Light/System shown) → Night Mode selection modal → Preferences list (Dark applied) **Steps: Appearance Mode Selection** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/7519) > Happy path: User opens the Display settings and changes the app's appearance/theme from the current value (System) to Dark. Steps: Display settings (Light) → Appearance options overlay → Display settings (Dark) **Mela: Change Appearance Theme** (ios, 3 screens, Food & Drink) [Refero](https://refero.design/flows/10386) > Happy path: From the General settings screen the user opens the Appearance option, chooses a theme (Automatic / Dark / Light), and returns to General where the chosen theme is shown and applied. Steps: General → Appearance Selection → General (Dark Mode Applied) **Comet: Theme & Appearance Settings** (ios, 4 screens, AI Tool, Utility) [Refero](https://refero.design/flows/10951) > Happy path: A logged-in user opens their account/settings panel, selects the Appearance option, chooses between System / Light / Dark modes, optionally toggles related switches (Force dark mode on websites, Autocorrect in search bar), then returns to the main settings which reflects the chosen appearance. Steps: Account & Settings list → Appearance options (light theme) → Appearance options (dark preview) → Account & Settings list (dark theme applied) **Turf: Appearance / Theme Selection** (ios, 4 screens, Sports) [Refero](https://refero.design/flows/9444) > Happy path: A logged-in user opens Settings, taps the Appearance row, chooses a theme from a bottom-sheet picker (System / Light / Dark), and the app updates immediately to reflect the chosen theme. Entry: Settings screen (user already signed in; name and phone shown). Exit: Settings reflects the selected appearance (label changed) and the visual styling of the Settings screen updates to match the chosen theme. Assumptions: mobile (iOS-style) device, user is authenticated, Settings modal is reachable from the main app and dismissal is via the top-right X or swipe down on the sheet. Steps: Settings Overview → Open Appearance Picker → Select Dark Mode → Confirm and Return **Blackbird: Change Appearance Mode** (ios, 5 screens, Food & Drink) [Refero](https://refero.design/flows/9039) > Happy path: A signed-in user opens their profile/overview, scrolls the settings list, opens the Appearance control (a bottom sheet), selects a different appearance option, and the app immediately applies the theme. Entry: user is on the Profile / Account overview screen (logged in) on a mobile device. Exit: the Appearance control is dismissed and the app’s theme is updated. Assumptions: user is authenticated, has permission to change local appearance settings, and the device supports light/dark theme rendering. Steps: Profile / Overview → Scroll Settings List → Appearance Modal (Default) → Select Dark Mode → Appearance Applied (Dark) **Adobe Photoshop: General settings - theme and interface** (ios, 3 screens, Design & Media Editing) [Refero](https://refero.design/flows/4833) > Happy path: From the Settings list the user opens the General section, changes the interface color theme from the system setting to Dark, and observes the UI switch to dark mode. **Reeder.: Display & Layout Settings** (ios, 2 screens, News) [Refero](https://refero.design/flows/3574) > Settings screen that lets a user adjust visual and layout preferences for the app on a mobile device. Steps: Display & Layout — Light theme → Display & Layout — Dark theme **Kin: Appearance Settings** (ios, 3 screens, AI Tool) [Refero](https://refero.design/flows/5447) > Happy path: From the main Settings screen the user opens the Appearance section to choose how the Kin interface looks. They can select Light, Dark, or Use system settings (default) and toggle a chat-specific option to highlight entities in messages. Selecting a theme immediately updates the app’s colors (demonstrated by the light → dark screenshots) and the choice is persisted. **RAD Weather: Weather Theme Selection** (ios, 5 screens, Weather) [Refero](https://refero.design/flows/13818) > The user starts on the New York weather dashboard in the Terminal dark theme. They open Settings, enter Appearance, and view the App Theme selector. The user selects the Equinox theme, which immediately changes the interface to a light teal appearance and returns to the weather dashboard. The flow assumes the user is already viewing a configured weather location and has access to appearance settings. It begins at the Now dashboard and ends with the same dashboard displayed using the newly selected theme. Steps: Weather Dashboard → Settings → App Theme Selection → Equinox Theme Applied → Updated Weather Dashboard **Vinyls: Appearance Settings** (ios, 4 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/8937) > Happy path: From the Settings screen the user opens the Appearance control, chooses one of three options (Light, Dark, System) from a bottom sheet, sees the app update immediately to preview the choice, then dismisses the sheet and returns to Settings. Steps: Settings list (Appearance row) → Appearance chooser (light preview) → Select Dark (apply preview) → Dismiss chooser and return **Substack: Appearance settings** (ios, 4 screens, News, Social Networking) [Refero](https://refero.design/flows/4507) > A single-screen Appearance settings list where the user reviews and changes visual preferences. Steps: Appearance — initial state → Change inbox layout / open icon picker → Select app icon (confirmation appears) → Confirm icon change and final state **DailyArt: Settings** (ios, 2 screens, Education, Magazines & Newspapers) [Refero](https://refero.design/flows/3338) > This is the app's Settings area accessed from the bottom navigation (Settings tab is selected). The screen lists navigable items (Account Details, Subscription, Language, Notifications, Widget, Legal) and several inline preference controls implemented as segmented buttons (App Icon, Units, Text Size, Theme). The user can tap list rows to open sub-screens or tap segmented buttons to change a preference immediately. The two screenshots show the same settings screen in light (first) and dark (second) themes; the Theme control indicates which theme is active. **Matter: Appearance Settings** (ios, 4 screens, Magazines & Newspapers, News) [Refero](https://refero.design/flows/9372) > Happy path: A signed-in user opens their Profile, navigates to Appearance, chooses a theme (System / Light / Dark), optionally picks an interface color swatch and an app icon, then returns to Profile where the selected appearance is applied. Steps: Profile (Light) → Appearance — Light Theme → Appearance — Dark Theme Selected → Profile (Dark) **Structured: Customization Settings** (ios, 3 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3302) > Happy-path summary: From the top-level Settings list the user opens the Customization section to change visual and feedback preferences. The user can toggle In‑App Sound and In‑App Haptics, pick a background color (System / Light / Dark), and see an immediate preview (a card mockup) at the top. Choosing a background option updates both the preview and the page appearance to the selected theme (example: Light preview vs Dark preview). **WhatsApp: List Theme Customization** (ios, 12 screens, Social Networking, Collaboration Tools) [Refero](https://refero.design/flows/12621) > The user starts on a New list screen for a list named “Friends,” with Dan Samson, Poppy, and Samuel already included. They open the list theme settings, preview available themes, adjust the chat bubble color and wallpaper, and return to the list configuration screen. The selected appearance uses a purple chat bubble and a blue scenic wallpaper. The user is assumed to have access to WhatsApp Plus theme benefits and is configuring the list on a mobile device. The flow ends with the customized theme displayed on the New list screen; the final checkmark suggests the user can save the list. Steps: New List → Theme Gallery → Theme Preview → Dark Theme Preview → Preview Adjustment → Selected Theme → Chat Bubble Colors → Wallpaper Gallery → Wallpaper Browsing → Wallpaper Preview → Updated Theme Settings → Configured New List **Agora 2: Theme & Display Settings** (ios, 3 screens, Entertainment, News) [Refero](https://refero.design/flows/11639) > This is the Appearance / Display settings screen on a mobile device where the user selects app theme (light/dark/colored accent), picks an accent color, and toggles several layout and audio preferences. Steps: Appearance panel (light) → Appearance panel (dark) → Accent applied (colored background) **Raycast: Settings — Theme Selector** (ios, 3 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/10199) > A Settings sheet presented as a rounded modal with grouped list items and a prominent logo. Steps: Settings — default (System) → Theme selection popover → Settings — Dark applied **Neuecast: Appearance Selection** (ios, 2 screens, News) [Refero](https://refero.design/flows/3361) > Happy path: User opens the app's Settings (Settings tab is active), navigates to Appearance, and chooses one of three mutually exclusive appearance options (System, Light, Dark). Selecting an option updates the selection state (checkmark) and immediately updates the visual preview area (Playback Controls) and the mini player at the bottom to reflect the chosen appearance. **Structured: Theme & Layout Customization** (ios, 2 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3300) > Happy path: a signed-in user opens Settings > Customization to change the app's primary color and layout density. The top of the screen shows a live preview card that reflects the selected theme color and layout. Below the preview the user can choose a primary color from a horizontal palette, tap Edit Palette to customize colors (opens another UI, not shown), and select one of three layout modes (Full, Simplified, Minimal). Changes update the preview immediately. **pillowtalk: Enable Night Mode** (ios, 3 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13502) > The user is in the Yap voice-capture workspace and opens the More menu to review optional modes. They select the Night option, which closes the menu and returns them to the Yap workspace with a darker orange visual treatment. The flow starts on the active Yap screen and ends with Night mode enabled; no recording, sending, or finishing action is shown. Steps: Yap Workspace → More Options → Night Mode Workspace ##### Task: dashboard overview **Stripe: Test Data Review** (web, 4 screens, Finance, Business) [Refero](https://refero.design/flows/72) > The sequence shows a user navigating through a payment platform’s test-mode administrative console. The flow starts on the Customers area with no customer records, moves to the Developers integration overview, opens the Events section with an expanded filter panel, and ends on the Payments list showing one incomplete payment. The user appears to be logged in with access to customer, developer, event, and payment data. No form submission or record creation is shown. Steps: Customer Records → Integration Overview → Event Log Filters → Payment Transactions **Runey: Business Dashboard Overview** (web, 4 screens, Business, Productivity, Project Management) [Refero](https://refero.design/flows/11711) > A logged-in user lands on a business performance dashboard that summarizes revenue, balances, activities, tasks, customers, and projects. From the top-right controls, the user can open the profile menu, switch appearance between light and dark mode, or access account actions. The flow begins on the main dashboard and exits through profile/account options or simply by reviewing the dashboard data. Steps: Dashboard Overview → Profile Menu Open → Dark Mode Enabled → Theme Menu in Dark Mode **Rork: Vocabu Home Dashboard** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12152) > A logged-in user lands on the main dashboard of a Spanish-learning mobile product called Vocabu. The flow starts on the home screen and ends with the user viewing an overview of progress, goals, and review cards, with access to study and progress sections via bottom navigation. The screenshots also show the same home screen in portrait and landscape orientation, indicating responsive behavior rather than a separate task state. Steps: Home Dashboard → Responsive Home View → Landscape Home View **Mapbox: Sign-in to Account Overview** (web, 4 screens, Development Tools, Map & Navigation) [Refero](https://refero.design/flows/9712) > Happy-path: an unauthenticated desktop user arrives at the Mapbox marketing/home hero, clicks a CTA to get started and reaches the Sign in form. The user enters credentials and signs in successfully, which redirects to the Account overview/dashboard. Steps: Marketing homepage (hero) → Sign in — empty form → Sign in — credentials entered → Account overview / dashboard **Shopify: Analytics dashboard tour** (web, 3 screens, Business) [Refero](https://refero.design/flows/5424) > A three-step in-product walkthrough (coach/popover) that appears on the Analytics > Overview screen for a logged-in store user on desktop. **Rork: Spanish Learning Dashboard** (web, 5 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12153) > Entry point is the main home screen of a Spanish learning mobile app, shown inside a device preview. The user is assumed to be logged in and viewing their personalized dashboard. The screen summarizes current status: greeting, streak, daily goal progress, review count, and level progress, with a primary prompt to start a review session. The flow also shows an edit iteration where a content block is selected and removed from the layout, updating the home screen structure. Exit point is either entering a review session or returning to the home dashboard after a layout change is applied. Steps: Dashboard Overview → Review Prompt Focus → Review Card Selection → Block Removal → Updated Dashboard **Runey: Account Login** (web, 5 screens, Business, Productivity, Project Management) [Refero](https://refero.design/flows/11680) > A visitor starts on Runey’s marketing landing screen, reviews the product positioning, and chooses to log in. They enter an email address and password, submit the form, and are taken to the authenticated business dashboard. The flow assumes the user already has an account and valid credentials. The final screen is a logged-in dashboard showing financial performance, activities, tasks, customers, and projects. Steps: Product Landing → Login Form → Completed Credentials → Authentication Processing → Business Dashboard **Gumroad: Seller Sign-in and Dashboard** (web, 4 screens, Business, Marketplace) [Refero](https://refero.design/flows/9235) > Happy path: an unauthenticated visitor arrives on the Gumroad marketing landing and chooses to sign in. The user is presented with an auth screen offering social sign-in buttons and an email/password form. After entering credentials and submitting, the user is authenticated and redirected to their Dashboard where getting-started tiles, sales summary, quick activity cards, and sidebar navigation are shown. Steps: Marketing Landing → Login - Method Selection → Login - Credentials Entry → Dashboard - Account Home **Fourthwall: Marketing to Creator Dashboard** (web, 4 screens, Business) [Refero](https://refero.design/flows/7278) > Happy path: An unauthenticated visitor discovers Fourthwall from the marketing homepage, uses the prominent CTA to sign in, authenticates (via social or email/password), and lands on the creator dashboard (Home). Steps: Marketing homepage → Login — method selection → Login — submit credentials → Creator dashboard (Home) **Fingerprint: Usage Insights Overview** (web, 3 screens, Business, Development Tools) [Refero](https://refero.design/flows/11147) > Happy path: A logged-in user (project owner or team member) lands on the project Overview where aggregated usage and health information is displayed. Steps: Overview landing → Metric selector opened → Requests per second selected **Appwrite: Project Overview — Integrations** (web, 2 screens, Development Tools) [Refero](https://refero.design/flows/6028) > This is the project-level dashboard for a project named "Oakline." Entry point: authenticated user opens the project dashboard (desktop/laptop). The screen presents top-level usage metric cards (Bandwidth, Requests, Database/Documents, Storage, Auth/Users, Functions/Executions) followed by an Integrations section with three tabs (Platforms, API keys, Dev keys) and a table of existing platform integrations. The primary call-to-action in the visible area is a prominent "Add platform" button. Exit points from this screen (not shown) include selecting a metric to drill into, switching tabs in Integrations, clicking the Add platform button to start adding a platform, or using the left navigation to go to other product areas (Auth, Databases, Functions, Messaging, Storage, Sites, Settings). Assumptions: user is logged in with access to the Oakline project and is using a desktop-class browser; project-level permissions allow viewing integrations and metrics. Steps: Overview — Expanded navigation → Overview — Collapsed navigation / narrow viewport **Fourmula.ai: Project Collection Management** (web, 5 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/11843) > A logged-in user views a dashboard of saved project collections in a grid layout. Each card shows a project preview, title, recency label, and a kebab menu for actions. The user can create a new project from the top-right button, open card actions, and confirm deletion through a modal. The flow starts on the project gallery dashboard and ends after one collection is removed and the dashboard updates to reflect the deletion. The screens assume a desktop viewport and an account with existing collections; a credit warning banner is visible but not acted on in this flow. Steps: Dashboard Overview → Card Selected → Action Menu → Delete Confirmation → Collection Removed **Hello Ivy: Workspace Overview Dashboard** (web, 3 screens, Productivity, Project Management, Collaboration Tools) [Refero](https://refero.design/flows/4397) > Overview screen for a workspace called "Office". The user is on the Overview (dashboard) view while signed in. The central area greets the user, shows a quick "+ Add task" input and a single project (Office #1) expanded to reveal one task. The UI includes a left navigation (workspace and project list) that changes from full to compact/collapsed across the three screenshots (responsive or manually collapsed). Top-right controls expose Filters and the current workspace. Entry: user is signed in and lands on the Office > Overview. Exit: user either creates a new task, opens a task/project, toggles filters, or navigates to another section (Calendar/Projects) — all preserved in the same Overview layout. Steps: Overview with full left navigation → Overview with compacted left navigation → Overview with collapsed/hidden left navigation and top controls **Glorify: Dashboard onboarding tour** (web, 6 screens, Design & Media Editing) [Refero](https://refero.design/flows/3777) > Happy-path: A first-run product tour shown on the logged-in user’s dashboard. The tour sequentially highlights key areas: My Files (left nav & file grid), My Uploads, Templates & Tutorials, Project organization tips, Workspace settings, and the Create design button. The user moves through the tour by clicking Next (or Back) on each tooltip; the interface highlights the corresponding UI area without navigating away from the dashboard. The flow ends when the user clicks the final Create design CTA (which opens the design editor) or dismisses the tour. Assumptions: user is signed in, on a desktop/larger viewport, and is viewing the default workspace/dashboard with at least one file thumbnail visible. Steps: My Files introduction → My Uploads overview → Templates & Tutorials → Organize with Projects → Workspace Settings → Create design CTA **Reown: Project authentication to workspace** (web, 5 screens, Development Tools, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/5728) > This sequence shows a user moving between a project dashboard and the authentication screens, ending at a workspace overview with top navigation tabs. Entry: the user begins on a project-level dashboard (left-hand navigation visible). The flow then presents the sign-in screen, the user fills email/password, receives an authentication error, and finally lands on a workspace dashboard that exposes project-level tools and usage. Exit: the user reaches the workspace (Dashboard/Analytics/Users/Domain/App ID/Paymaster) where they can continue working on the project. Assumptions: the flow is on a desktop browser; authentication is required to access project workspaces; the user may have been signed out or redirected to login before completing a protected action. Steps: Project Dashboard (left navigation) → Sign-in screen (empty) → Sign-in screen (credentials entered) → Authentication error shown → Workspace Dashboard (top tabs) **Care: Profile Photo Update** (web, 6 screens, Job Search, Lifestyle) [Refero](https://refero.design/flows/12238) > Logged-in Care.com member starts on the dashboard/home area and is prompted to improve the profile by adding a photo. The user opens the profile update prompt, uploads an image, adjusts its crop/size, confirms the upload, and returns to the household profile overview where the new photo is visible. Steps: Dashboard Prompt → Update Photo Prompt → Photo Upload → Crop and Position Photo → Saving Photo → Profile Overview **ShareWillow: Sign in and access plans** (web, 5 screens, Business, Scheduling Tools, Project Management) [Refero](https://refero.design/flows/2091) > Happy-path: A desktop user arrives at the marketing homepage, navigates to the sign-in screen, enters credentials, and is authenticated. After a brief loading state the user lands on the account dashboard which lists unpublished and live incentive plans and exposes actions to create a new plan. Steps: Marketing homepage → Sign-in form (empty) → Credentials entered → Authentication loading → Authenticated dashboard (Plans overview) **Riverside: Marketing-to-Studio sign-in** (web, 4 screens, Productivity, Entertainment, Collaboration Tools) [Refero](https://refero.design/flows/4648) > A desktop-first conversion path that takes an unauthenticated visitor from the product marketing homepage into their Riverside studio. Steps: Landing / Product Overview → Login Modal (empty) → Credentials Entered / Submit → Studios Dashboard **Navan: Sign-in to Admin Dashboard** (web, 8 screens, Travel & Booking) [Refero](https://refero.design/flows/9347) > Happy-path sign-in from marketing to admin. Steps: Marketing / Home Hero → Welcome — Email Entry → Email Confirmed (inline) → Password Input → Submit Credentials → Check Your Inbox — Verification Sent → Logged-in Confirmation → Navan Admin Dashboard **Enode: API credential generation** (web, 4 screens, Development Tools, Productivity) [Refero](https://refero.design/flows/4260) > Happy-path: A logged-in developer navigates from the Solstice tenant dashboard to the Settings > API credentials screen, generates a client secret for the displayed client ID, copies/saves the secret, and returns to the tenant dashboard which reflects the completed 'Access your API credentials' get‑started step. Entry: authenticated developer on tenant (Solstice) dashboard. Exit: tenant dashboard with the credential step marked completed and an active client ID / generated client secret stored by the user. Assumption: user has permission to view Settings and to generate client secrets; user is on a desktop browser. **Stellate: Sign-in to Organization Services** (web, 5 screens, Development Tools) [Refero](https://refero.design/flows/4445) > Happy path: An unauthenticated desktop user arrives from the marketing site, chooses a sign-in method (GitHub, GitLab, or email magic link), submits an email, receives confirmation that a magic link was sent, and lands in their organization’s Services dashboard where they can view and create services. Steps: Marketing landing / hero → Sign-in options → Email entry (pre-submit) → Magic link sent confirmation → Organization Services dashboard **Replo: Sign in to Projects** (web, 4 screens, Design & Media Editing, Business) [Refero](https://refero.design/flows/2052) > Happy-path summary: An unauthenticated visitor reads the Replo marketing/feature landing content, then chooses to sign in. The Sign In screen presents a credentials form (and Google SSO). After the user enters valid credentials and submits, they are taken to the Replo Projects dashboard where existing projects are shown as tiles and a 'New Replo Project' CTA is available. Steps: Marketing / Product Overview → Sign In — Empty Form → Sign In — Credentials Filled → Projects Dashboard (Authenticated) **Dopper: Business Account Sign-In and Dashboard** (web, 4 screens, Marketplace) [Refero](https://refero.design/flows/7538) > Happy path: A desktop user who is not signed in visits the homepage, navigates to the 'For Businesses' / Business Orders area, signs in with their business credentials, and lands on their account overview dashboard. Steps: Homepage / Entry → Business Orders Login (empty) → Business Login (credentials entered) → Account Overview Dashboard ##### Task: delete account **Fresha: Account deletion confirmation** (ios, 15 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4982) > Happy-path, mobile, logged-in user journey to permanently delete their account. Entry: user opens their Profile and navigates to Settings > Delete account. The flow is a 3–4 step modal-like wizard: (1) ask for a deletion reason (radio list + optional free-text), (2) show consequences and require explicit checkboxes acknowledging them, (3) request identity confirmation (email — password field implied) and final confirmation. Exit: user sees a success screen. Assumptions: user is authenticated and on a mobile device with touch input; the user has permission to delete their own account. Steps: Profile overview → Settings list → Delete reason prompt → Reason selection - continue → Free-text reason entry → Consequences acknowledgement → Validation on missing confirmation → Acknowledgements completed → Identity confirmation prompt → Email entry → Submit deletion → Success acknowledgement (minimal) → Success acknowledgement (with message) → Return to discovery → Progress and navigation affordances **Wispr Flow: Account Deletion** (ios, 5 screens, AI Tool, Utility, Productivity) [Refero](https://refero.design/flows/13690) > An authenticated user opens the Account settings screen and chooses to permanently delete their account. The system presents a confirmation dialog explaining that the action is permanent. After confirmation, a blocking loading state appears while the account is deleted, then the user is returned to the Flow sign-in screen. The sequence assumes the user is signed in and has access to Account settings on an iOS device. **komoot: Account deletion** (ios, 6 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8425) > Happy path: a logged-in user opens Settings → Profile Information, scrolls to Additional Information and taps Delete Account. The app shows a Delete account screen with a warning, the user's name/email, and a password field. The user enters their password, the Delete account button becomes active, and tapping it completes account deletion. The user is then returned to the initial onboarding / sign-in screen with social and email sign-in options. Steps: Settings - Account list → Profile Information (top) → Delete account (warning shown) → Delete account (password entered) → Post-deletion / Sign-in → Cancel or return **ElevenLabs: Account Deletion & Sign-Out** (ios, 6 screens, AI Tool) [Refero](https://refero.design/flows/10756) > Happy path: A signed-in user navigates to their account/profile settings, chooses the destructive 'Delete account' option, types their email to confirm, accepts a final confirmation modal, and is signed out / returned to the app's sign-in screen. Steps: Account overview → Deletion confirmation (entry) → Enter email to confirm → Final confirmation modal → Confirm deletion → Signed-out / Sign-in landing **Granola: Delete Account** (ios, 6 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14172) > A signed-in user opens their profile settings and initiates permanent account deletion. The system first presents the number of notes that will be lost, then shows a second irreversible-action confirmation. After the user confirms, a deletion progress state appears and the user is returned to the Granola sign-in screen. The flow is shown on a mobile device and assumes the user has permission to delete their own account. Steps: Profile Settings → Delete Account Entry → Notes Loss Confirmation → Final Confirmation → Deletion In Progress → Signed-Out Entry **Queue: Account Deletion Confirmation** (ios, 3 screens, Entertainment, News) [Refero](https://refero.design/flows/6242) > Mobile settings screen (logged-in) with legal links, account controls and app footer. Entry: user opens Settings > Account area. Happy path: user taps 'Delete Account...', a full-width bottom sheet confirmation appears with a prominent red 'Delete Account' button. After confirming, the UI returns to an unauthenticated/settings state that shows a 'Sign in with Apple' CTA. Assumptions: user is on an iOS mobile device, initially signed in and has access to account-level controls (Log Out, Delete Account). Steps: Settings — Account Options → Delete Confirmation (Bottom Sheet) → Post-Deletion / Signed-out Settings **Tripsy: Account Deletion** (ios, 3 screens, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/2324) > Happy path: An authenticated user opens Edit Profile to remove their account. From Edit Profile they tap the red Delete Account row, see a full-screen confirmation modal, and confirm deletion. After confirmation the modal closes and the app returns to the Settings screen showing a login/create-account prompt (user appears signed out). **Meta AI: Permanent Meta Account Deletion** (ios, 11 screens, AI Tool) [Refero](https://refero.design/flows/13404) > The flow begins in Meta Accounts Center for a logged-in user with multiple connected profiles. The user opens Manage accounts, selects their Meta account, and chooses Deactivation or deletion. They select permanent deletion, review prerequisites and informational warnings, continue through password re-entry, and confirm permanent deletion. The final screenshot shows a Meta loading/landing screen after the deletion request is submitted. The screenshots do not show a success message or deletion status confirmation. **Imprint: Account Deletion** (ios, 5 screens, Education) [Refero](https://refero.design/flows/5651) > User (already signed-in) opens Account Details on an iOS mobile device and deletes their account. **WishWish: Account Deletion** (ios, 4 screens, Entertainment, Kids) [Refero](https://refero.design/flows/14306) > The user starts on the Settings screen while signed in. They select Delete under Account deletion, review a confirmation dialog explaining that deletion is permanent, and confirm the action. The system closes the dialog, displays a success toast stating that the account has been deleted, and returns the user to the Settings screen. The final screenshot shows the product’s signed-out welcome and authentication screen, indicating the user is directed to sign in or create an account after deletion. The flow assumes a mobile device and an authenticated user with permission to delete their own account. Steps: Settings → Deletion Confirmation → Deletion Success → Authentication Entry **Canopi: Permanent Account Deletion** (ios, 3 screens, Scheduling Tools) [Refero](https://refero.design/flows/6075) > Happy path: A signed-in user opens their profile/settings, selects the "Delete Account" option, confirms the destructive action in a focused confirmation sheet, and is returned to the app's welcome/sign-in screen. Steps: Profile / Settings — Delete Entry → Confirm Permanent Deletion → Welcome / Signed-out State **Moises: Account deletion flow** (ios, 12 screens, AI Tool, Music Streaming) [Refero](https://refero.design/flows/6863) > Happy path: a logged-in user (email-authenticated, shown as Premium) navigates from Settings → Account info → Delete my account. The user selects a reason (or "Something else"), optionally types a free-form explanation, confirms they've cancelled any paid subscription via a checkbox, then taps Delete. The system prompts for the account password, shows a progress indicator while removing the account, and ends at a success screen with an option to create a new account. Steps: Open Settings → Account details → Choose deletion reason (list) → Select 'Something else' → Free-form feedback input (keyboard) → Deletion warnings & confirmation modal → Acknowledge subscription status → Trigger deletion → Re-authentication prompt → Enter password & confirm → Deletion in progress → Success confirmation → Post-deletion CTA **Artsy: Account Deletion & Confirmation** (ios, 5 screens, Marketplace) [Refero](https://refero.design/flows/7626) > Happy path: A logged-in user navigates from their Profile → Login & Security, taps "Delete My Account", reads consequences, provides an optional reason, confirms by submitting, and receives success or an error message. Steps: Login & Security (settings) → Delete — Warning & Reason Prompt → Reason Entry (keyboard visible) → Confirm & Submit → Validation Error / Blocking Message **Napper: Account management — delete account** (ios, 6 screens, Health & Wellness) [Refero](https://refero.design/flows/8816) > Happy path: a logged-in mobile user opens the account area, goes into Account details, and deletes their account. Steps: Account menu → Account details → Delete modal - instructions → Enter verification code → Final confirmation dialog → Deletion success **Twitch: Account Deletion** (web, 5 screens, Social Networking, Entertainment, Video Streaming) [Refero](https://refero.design/flows/13638) > An authenticated Twitch user starts in Profile settings, selects the account deletion option, reviews the deletion consequences, optionally enters a reason, and submits the request. The system briefly shows a loading state, then confirms that the account has been deleted and presents a link back to the front page. The flow assumes the user is signed in at the start and ends in a logged-out state. Steps: Profile Settings → Deletion Review → Deletion Form Completed → Deletion Processing → Deletion Confirmation **Acuity Scheduling: Account Deletion** (web, 8 screens, Business, Scheduling Tools) [Refero](https://refero.design/flows/9876) > Happy path for a signed-in user removing their account from the product. Entry: user opens Account Settings and navigates to the Security section. The user selects the Delete account control, is required to re-authenticate with their password, is shown a Delete account details screen, confirms deletion in a confirmation dialog, sees a deleting progress state, and then receives a final deleted confirmation screen with a Done action. Assumptions: user is already authenticated and has access to Account Settings; the interface requires re-authentication before showing deletion controls. Steps: Security settings overview → Re-authenticate (password entry) → Submit password → Delete account details → Delete action visible → Final confirmation modal → Deletion in progress → Deletion complete confirmation **Expedia: Account Deletion** (web, 6 screens, Travel & Booking) [Refero](https://refero.design/flows/3031) > Happy-path: A signed-in user navigates from their account settings into a multi-step account deletion flow, provides a reason, reviews consequences, confirms permanent deletion, and is shown an acknowledgement before being returned to the Expedia home/landing screen. **Rive: Delete account confirmation** (web, 3 screens, Design & Media Editing) [Refero](https://refero.design/flows/176) > Happy path: A logged-in user navigates to Account > Delete Account, reads the warning that files will no longer be accessible and that the account is recoverable for 90 days, enters their password, and submits the delete request. Steps: Delete Account — warning & password field → Enter password → Deletion error feedback **Bezi: Account Deletion** (web, 5 screens, Design & Media Editing) [Refero](https://refero.design/flows/208) > Happy path: A signed-in user opens their profile menu, navigates into Account settings, initiates account deletion, confirms by typing the required phrase, and completes the deletion which leads to an empty workspace/‘file not found’ state. **Preply: Account deletion confirmation** (web, 6 screens, Education) [Refero](https://refero.design/flows/9232) > Happy path: Logged-in user opens Settings → Delete account, types their email to confirm intent, clicks the primary Delete account control which opens a required reason-selection modal, chooses a reason and submits. The system shows an on-screen success confirmation and then redirects the user to the public homepage. Steps: Delete account landing → Email confirmation → Reason selection modal (open) → Reason selected and submit → Deletion success confirmation → Post-deletion redirect to homepage **Whereby: Account Deletion from Profile** (web, 3 screens, Productivity, Communication Tools) [Refero](https://refero.design/flows/1983) > Happy path: A logged-in user navigates to their Profile under Settings, initiates account deletion from the Delete account card, confirms the action in a modal, and is shown a final success screen stating the user has been deleted with a link back to the front page. Steps: Profile settings → Delete confirmation modal → Deletion success message **Craft: Account Deletion Confirmation** (web, 4 screens, Productivity, Project Management, Paperwork) [Refero](https://refero.design/flows/7970) > Logged-in user opens Account Settings (My Account) on a desktop-sized interface and uses the Danger Zone option to permanently delete their account. The user triggers a Delete Account modal that requires typing their email address as an explicit confirmation. When the email is provided and the user confirms, the app completes the deletion flow and returns the user to the product's sign-in/welcome screen. Steps: My Account (Settings) Overview → Delete Account Modal (Empty) → Delete Account Modal (Email Entered) → Welcome / Sign-in Screen **Spline: Account Deletion** (web, 3 screens, Design & Media Editing) [Refero](https://refero.design/flows/266) > Happy path: A logged-in user navigates to the Account section inside Settings to remove their personal account. From the Account Settings screen they click the Delete Account control, which opens a danger-confirmation modal that requires the user to type an explicit confirmation phrase. After the phrase is entered the Delete Account button becomes active and the user can submit the deletion. **Medium: Account deletion** (web, 4 screens, Social Networking) [Refero](https://refero.design/flows/11142) > Happy path: A logged-in user navigates to their Settings > Account and initiates permanent account deletion. The site opens a confirmation modal that explains consequences and requires the user to type the word "delete" to enable the final action. After confirmation the user is redirected to Medium's public landing (signed-out) view. Steps: Account settings → Deletion confirmation modal (empty) → Deletion confirmation modal (typed) → Post-deletion landing **Appwrite: Account Deletion** (web, 3 screens, Development Tools) [Refero](https://refero.design/flows/6061) > Happy path: An authenticated user visits their account Overview settings to permanently remove their account. From the Overview they locate the Delete account card and tap Delete. A modal overlay asks for confirmation. When the user confirms Delete, the system completes the action, signs the user out and redirects them to the Sign in screen with a success notification. Steps: Account settings — Overview → Delete confirmation modal → Signed-out — Sign in with success toast **Descript: Account deletion** (web, 5 screens, Collaboration Tools, Business) [Refero](https://refero.design/flows/2150) > Happy path: an authenticated user opens Settings → Account → Profile, initiates account deletion, confirms the irreversible action in a confirmation dialog, and the system begins account removal. The UI then shows a brief progress/notification that access to a shared drive was revoked and finishes by returning the user to the sign-in screen. **n8n: Account deletion (permanent) flow** (web, 4 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9564) > Happy path: A logged-in user navigates to Manage > Workspace, scrolls to the Danger zone and clicks Delete account. A modal appears asking for confirmation. A second confirmation modal prompts about exporting workflows before deletion. After the user confirms, the system deletes the account, shows a success toast and returns to the sign‑in screen. Steps: Manage > Workspace (Danger zone) → Confirm account deletion (first modal) → Export prompt / final confirmation → Sign‑in screen with success toast **ElevenLabs: Account Deletion Process** (web, 4 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/8718) > Happy path for a signed-in user to permanently delete their account from Profile settings. Entry: user opens Settings → Profile while authenticated. The user scrolls to and clicks the red "Delete Account" control. A blocking confirmation modal appears that warns about permanence and requires explicit confirmation by typing the account email. After entering the email and confirming, the system deletes the account and ends the session, returning the user to the Sign In screen. Assumptions: user is logged in with the shown email, has access to the Profile settings, and no additional multi-step verification is required beyond the email confirmation visible in the UI. Steps: Profile Settings → Deletion Confirmation Modal (Empty) → Deletion Confirmation Modal (Email Entered) → Signed Out — Authentication Entry **Tidal: Account Deletion** (web, 7 screens, Music Streaming) [Refero](https://refero.design/flows/3294) > Happy path: A logged-in user navigates from their Profile settings to permanently delete their account. The user clicks the Delete account control, proceeds through a 4-step confirmation sequence: an initial informational/consent screen, an explicit “are you sure” screen listing lost content, a verification code entry step (code sent to the user's email, with a resend countdown), and the final confirmation that triggers account deletion. Successful completion ends the session and redirects the user to the Sign up / Log in screen. Assumptions: user is authenticated, has access to the registered email address to receive the verification code, and is using a desktop/laptop layout (wide viewport). **[untitled]: Account deletion via typed confirmation** (web, 6 screens, Music Streaming) [Refero](https://refero.design/flows/5966) > Happy path: A logged-in user opens their profile/settings panel, opens the top-right "more" menu, and chooses to delete their account. The UI shows an initial soft-confirmation modal, then a stronger confirmation requiring the user to type their username. After entering the username and confirming, the interface exits the settings area and returns to the public marketing/home screen. Assumptions: user is logged in on a desktop/responsive web layout, has a visible username, and can access the more-menu in the profile/settings area. Steps: Profile / Account Settings → Open More Menu → Initial Delete Confirmation → Typed Username Prompt → Confirm Deletion (username entered) → Post-deletion / Public Home **Runey: Account Deletion** (web, 5 screens, Business, Productivity, Project Management) [Refero](https://refero.design/flows/11713) > A logged-in user opens the Profile settings, selects Delete Account, reviews the permanent deletion warning, types DELETE to confirm, and submits the request. The interface shows a deletion-in-progress state, then redirects the user to the sign-in screen with an account deletion success notification. Steps: Profile Settings → Deletion Warning → Confirm Deletion → Deletion Processing → Deletion Complete **MWM: Account Deletion** (web, 3 screens, AI Tool, Development Tools) [Refero](https://refero.design/flows/13405) > The user starts in the account Overview settings while signed in. They select the Delete account action, review an irreversible-action warning, and either cancel or confirm deletion. In the shown happy path, the user confirms by selecting Delete my account and is returned to the MWM AI welcome screen with Google, GitHub, and Apple sign-in options. The final screen indicates that the authenticated settings session has ended and the product is ready for sign-in or account creation. **IKEA: Account deletion (self-serve)** (web, 4 screens, Marketplace, Home & Decor) [Refero](https://refero.design/flows/8276) > Happy path: A logged-in IKEA Family user navigates to Account > Login & Security, selects Delete account, confirms by entering their password, and completes account deletion. Steps: Account settings — Login & Security → Delete confirmation — password required → Password entered — enable deletion → Deletion success — profile removed **Hashnode: Account Deletion Process** (web, 5 screens, Communication Tools, Development Tools, Business) [Refero](https://refero.design/flows/3950) > Happy path: a logged-in user opens Account management settings, chooses Delete account (Danger area), reads the warning modal, provides optional feedback, supplies their email and types the required verification phrase, then confirms deletion. After confirmation the user is redirected out of the authenticated area to the public landing (marketing) site. Steps: Account management → Delete confirmation (initial) → Delete confirmation (typing verification) → Delete confirmation (all inputs provided) → Post-deletion landing **Stellate: Account Deletion Confirmation** (web, 4 screens, Development Tools) [Refero](https://refero.design/flows/4446) > Happy path: a logged-in user navigates to their account settings, chooses the Danger zone's Delete account action, confirms deletion by typing their email into a modal, and is then taken to a public marketing/landing screen. Entry: authenticated user on User settings > Account tab. Exit: user ends up on the public marketing/landing content (screenshot shows site homepage). Assumptions: user is signed in with permission to remove their own account; device is desktop (wide layout). The backend outcome (actual account removal or logout) is not visible in the screenshots; only the UI transition to the landing content is shown. Steps: Account settings — Danger zone visible → Deletion confirmation modal (empty input) → Deletion confirmation modal (email entered) → Landing / marketing content **Visitors: Account deletion confirmation** (web, 5 screens, Business) [Refero](https://refero.design/flows/9878) > Happy path: a signed-in user navigates to their Account > General settings, opens the Delete account control, confirms deletion through a multi-step confirmation modal, and is returned to the login screen. Steps: Account settings — General → Delete confirmation modal (initial) → Delete confirmation — first tap → Delete confirmation — final tap → Post-deletion sign-in screen **Vimeo: Account deletion - privacy settings** (web, 4 screens, Entertainment, Video Streaming, Social Networking) [Refero](https://refero.design/flows/700) > Happy path: A logged-in user navigates to Privacy in Settings, chooses to delete their account, confirms by entering their password, and receives a final page confirming the account has been deleted. Steps: Privacy settings → Deletion confirmation (enter password) → Deletion confirmation (password entered) → Account deleted confirmation **Claude: Account deletion confirmation** (web, 3 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/1807) > Desktop settings -> Account screen where a user can export data or delete their account. Steps: Account settings → Delete confirmation modal → Main surface with usage notices **Steel: Account Deletion** (web, 5 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/14305) > The user starts in the Security settings of an authenticated account, selects the destructive Delete account action, reviews a confirmation dialog, and appears to confirm deletion by holding the delete button. The sequence ends at the Steel sign-in screen, indicating the user has been redirected out of the authenticated account. The screenshots do not show a success message or provide enough evidence to confirm whether account data was fully deleted. The user is assumed to be on a desktop browser and initially signed in with permission to manage the account. Steps: Security Settings → Deletion Confirmation → Hold to Confirm → Deletion Submission → Sign-In Screen **Dropbox: Delete Dropbox Account** (web, 4 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/13071) > A logged-in Dropbox user opens Personal account settings from the account area, chooses to delete the account, completes the required confirmation form (password, reason, optional details), and submits the permanent deletion request. The flow ends with a confirmation banner on the public Dropbox homepage indicating the account has been deleted. Assumption: the user is signed in and has permission to delete the account. Steps: Account Settings → Deletion Confirmation → Deletion Confirmation Filled → Deletion Completed ##### Task: direct debit **Givingli: Schedule & Pay for Gift** (ios, 8 screens, Lifestyle, Scheduling Tools) [Refero](https://refero.design/flows/8194) > User schedules a pre-configured gift to a selected recipient and completes payment. Steps: Gift review / recipient → Checkout with payment placeholder → Add payment modal opened → Start card entry → Numeric entry for card fields → Card & billing populated → Confirm card and return to checkout → Attempt schedule & payment failure **DailyArt: Checkout Payment Details** (ios, 2 screens, Education, Magazines & Newspapers) [Refero](https://refero.design/flows/3333) > Final payment step of a mobile checkout where the user provides payment details, chooses between credit/debit card or PayPal, accepts terms, and submits the order. Entry: user arrives at the payment section during checkout (mobile view). Happy path: user selects credit/debit card, fills card number, expiration date and CVC, checks the "I agree" checkbox, then taps Place Order to submit. Exit: payment submission is attempted (leading to order confirmation or error handling). Assumptions: mobile device (iOS status bar visible), user has reached the checkout summary and is ready to provide payment; network connectivity is available. The UI shows supported card logos, a PayPal option, privacy notice, and the required terms-and-conditions checkbox. Steps: Payment method & empty card fields → Filled card fields and consent checked **District: Booking payment selection** (web, 4 screens, Entertainment, Food & Drink, Travel & Booking) [Refero](https://refero.design/flows/13009) > A booking checkout payment modal opens over a dimmed booking review page. The user can choose among payment methods in a left-side menu, including UPI QR, card entry, and netbanking. Depending on the selection, the right panel shows either a QR code to scan, a form to add a new card, or a saved cards list. The flow assumes the user is already logged in and is completing payment on desktop web. It starts from booking review and ends when the user either scans the QR code, saves/uses a card, or selects a stored payment method. **Givingli: Add card and confirm payment** (ios, 6 screens, Lifestyle, Scheduling Tools) [Refero](https://refero.design/flows/8196) > Happy-path: user is on the Delivery / Checkout summary for a scheduled gift and needs to add a new credit card to complete the purchase. Steps: Delivery / Checkout summary → Payment modal — choose Link or card → Card entry — numeric keyboard → Card details filled → Confirm card entry → Delivery summary with saved card **Blackbird: Pass purchase & gift checkout** (ios, 15 screens, Food & Drink) [Refero](https://refero.design/flows/9028) > User views a single-pass product (Daily Scoop Club LA) then proceeds to an in-context checkout (BLACKBIRD PAY) to either buy for themselves or gift the pass. From the product card the user can open the checkout modal, toggle use of in‑product credit ($FLY), enter a promo code, or choose 'Gift this pass' which opens a recipient form and date picker. The user attempts to complete payment via a swipe-to-buy control; the attempt results in a card-declined error. Steps: Product detail / pass card → Checkout modal (initial) → Toggle $FLY on → Open promo code input → Enter promo code → Promo code applied state → Open Gift flow → Fill recipient & select duration → Open date picker → Select send date → Gift Checkout ready → Swipe-to-buy control → Processing state → Card declined error → Error dismissal / recovery **Blackbird: Blackbird tap-to-pay tutorial** (ios, 3 screens, Food & Drink) [Refero](https://refero.design/flows/9021) > Three‑screen onboarding that teaches a diner how to pay using a Blackbird puck at a restaurant. Entry: user opens the "How it works" onboarding within the app (device with NFC). Happy path: user reads each screen then advances with the circular arrowors CTA until tapping the final Continue button to exit onboarding. Assumptions: the user has the Blackbird app installed and is viewing an in‑app instructional carousel on a phone with NFC capability. Exit: onboarding marked complete and control returns to the main app (ready to use the payment feature). Steps: Tap in on arrival → Inform the server → Checkout confirmation & finish **District: Movie Ticket Booking** (web, 31 screens, Food & Drink, Travel & Booking, Entertainment) [Refero](https://refero.design/flows/13002) > A user in Mumbai browses the details of Spider-Man: Brand New Day, checks available cinemas and showtimes, applies filters, selects a screening, chooses two seats, and completes payment by card. The journey begins on the movie detail screen and ends with a confirmed booking containing a QR code and ticket information. The user is assumed to be able to browse without signing in; payment is completed using a saved or newly entered card. Steps: Movie Details → Browse Showtimes → View Seat Pricing → Filter by Format → Filter by Show Time → Filter by Price → Review Filtered Results → Open Seat Map → Inspect Seating Layout → Choose Seats → Select Payment Method → Add Card → Review Card Details → Choose Saved Card → Enter CVV → Submit Card Payment → Confirm Gateway Payment → Wait for Booking → Booking Confirmation → Access Digital Tickets ##### Task: dispute transaction chargeback **Fresha: Membership purchase and checkout** (ios, 7 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4956) > Happy-path purchase of a salon membership on a mobile device. Entry: user is on the salon's profile under the Buy tab. The user taps to buy a membership, selects a specific membership plan, reviews membership details and the list of services it can redeem, then proceeds to checkout and attempts to pay with a saved card. Exit: either successful purchase (not shown) or payment failure (shown). Assumptions: user is signed in, using a mobile device, and has at least one saved payment method (Visa •••• 4292). Steps: Buy tab — Membership entry → Select a membership → Review details (membership) → Redeemable services list → Review and checkout — summary → Payment attempt → Payment failure / error **PayPal: Activity transaction search** (web, 3 screens, Finance) [Refero](https://refero.design/flows/7382) > Happy-path user journey for locating transactions from the Activity section. Steps: Activity landing / default list → Search with numeric ID — no results → Search by keyword — results shown **Gumroad: Membership charge refund** (web, 3 screens, Business, Marketplace) [Refero](https://refero.design/flows/9249) > Happy path: Seller opens the Sales section, selects a specific sale (a membership) from the list, views the sale details in a right-hand detail panel, initiates a charge refund from the Charges section, confirms the refund in a modal, and observes the Charges entry update to show a Refunded state and related receipt options. Steps: Sales list and membership detail → Refund confirmation modal → Refund completed and updated charges **IKEA: Contact Support Chatbot Interaction** (web, 10 screens, Marketplace, Home & Decor) [Refero](https://refero.design/flows/8273) > A desktop customer support scenario starting on the Contact us landing area and continuing through the site’s chatbot. Steps: Contact us landing → Open chat — welcome message → Chat loading → Chatbot initial prompts → Select cancel order → Provide order number → Email confirmation → Confirm same email → Clarification and next steps → Chat end and survey ##### Task: doctor video consultation **District: Restaurant Question Assistant** (ios, 5 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12847) > Happy-path mobile assistant experience for asking a restaurant-related question and continuing with a follow-up. Entry is a restaurant detail or assistant overlay with an active query context (“Asking about Hotel Rishi”). The user asks a question by voice, the system transcribes and processes it, shows an answer with suggested next actions, and keeps a persistent follow-up input available. The flow ends when the user has a ready-to-send follow-up request, such as asking for nearby breakfast options. **Doctronic AI: Clinical Summary Feedback** (web, 4 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13701) > A user reviews an AI-generated clinical summary and provides product feedback. The flow begins on the summary view, where the user can rate how useful the summary was. The user then enters written feedback in an improvement form and submits it. After submission, the system confirms that the feedback was received. A chat input remains available throughout for asking additional questions. **WhatsApp: Switch to Video Call** (ios, 4 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12665) > Happy path for an ongoing 1:1 call with Samuel on a mobile device. The user starts on an active voice call, chooses to switch to video, confirms the change in a modal, and then reaches the video call state. **WhatsApp: Active Call Controls** (ios, 2 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12663) > An active voice/video call screen with a contact named Samuel. The user is already in an ongoing call; the screen shows the call timer, a large contact avatar placeholder, quick actions on the right, and a bottom control bar. **WhatsApp: Group Voice Conversation** (ios, 6 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12680) > Happy path for joining and using a group voice conversation inside a community chat titled "General." The entry point is an existing chat thread where the user can read messages, add members, or start talking. The user begins by holding the talk control, then transitions into a live voice room connection where participants are shown and call controls appear. The exit point is the expanded in-call voice interface with connected participants and controls for speaker, mute, reactions, and ending the session. **WhatsApp: WhatsApp Call Transition** (ios, 9 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12661) > Happy path for a WhatsApp conversation that transitions from chat to audio call and then to video call. **Weekrise: Embedded support chat conversation** (web, 7 screens, Collaboration Tools, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/1296) > Happy-path interaction with an embedded live-chat widget on a desktop landing screen. Steps: Open chat / initial greeting → Compose and send question → Email capture prompt → Offline / slow-reply fallback → Record voice message → Upload in progress → Audio playback available **Telegram: In-Chat Voice Call** (web, 4 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/571) > User is in a one-to-one conversation with a contact (Henry) on a desktop messaging client. From the open chat the user initiates a call using the header call control. A centered call card appears showing 'Ringing...' while the call is being placed and exposes inline controls (camera, screen, mute, decline). When the call connects a green active-call bar and a running timer appear; the call card updates to show the elapsed time and an End control. After ending the call the conversation is returned to and the chat thread (and left-hand conversation list) contains outgoing-call messages that show the call durations. Assumptions: user is signed in and in the specific one-to-one chat; device is desktop with enough screen width to show two-column layout (conversation list + chat). Entry: open chat with contact. Exit: chat updated with call records and user back to messaging view. Steps: Open Conversation → Ringing Call Card → Connected Call (Active) → Return to Chat with Call Record ##### Task: document editor collaboration **Linear: Attach document to draft** (ios, 5 screens, Collaboration Tools, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/6624) > Happy-path: A user is composing a note/task and attaches an existing document from the device Files picker, then returns to the composer with the file displayed as an attachment. Steps: Open composer / prepare to attach → Choose Attach file → Browse recent files → Confirm selection → Finalize or continue editing **Grok: Attach and summarize document** (ios, 9 screens, AI Tool) [Refero](https://refero.design/flows/6088) > Happy-path conversation where a user edits an image then attaches a PDF from iCloud and asks the assistant to summarize it. Steps: Conversation with image edits → Open attachment picker → Pick Files (iCloud) → File selected and attached → Compose instruction with keyboard → Submit request (processing) → Assistant returns summary (first view) → Scroll through full summary → Final state and next actions **Trip Way: Manage Documents** (ios, 3 screens, Travel & Booking) [Refero](https://refero.design/flows/8457) > Happy path: a user opens a modal-style Documents view that displays a scrollable 2-column grid of document thumbnails and filenames. The user can inspect the list, open a contextual action sheet for any tile (Edit or Delete), and use the Add Doc CTA to add more documents. Steps: Documents gallery → Document contextual actions → Gallery updated / modal state **Bezi: Signup & First-Time Onboarding** (web, 24 screens, Design & Media Editing) [Refero](https://refero.design/flows/178) > Happy path for a new desktop user who discovers the product, signs up with an email address, verifies their account with a code, provides their display name, and completes the in-app onboarding that includes creating a new file, taking an editor tour, and creating a team with invited collaborators. Steps: Marketing hero / Get started → Sign up modal (entry) → Credentials populated → Verification prompt (code error shown) → Verification prompt (correct code) → Verification in progress → Welcome — enter name → Workspace — Drafts with onboarding callout → Editor new file modal prompt → Editor first tour slide — share/export → Editor tour — play mode slide → Back to Drafts — create a team tooltip → Create a team modal (empty) → Create a team modal (ready) → Add collaborators modal → Add collaborators (filled) → Join community prompt → Team workspace (empty) → Onboarding completion toast → Recently opened (final state) ##### Task: document upload scan **Dropbox: Document scan and save** (ios, 20 screens, Business, Collaboration Tools, Productivity) [Refero](https://refero.design/flows/2425) > User creates a multi-page PDF from paper using an in-app document scanner, edits/crops pages, applies filters/brightness, arranges pages, sets file options (name, quality, save location) and saves/uploads the PDF into a folder. **ChatGPT: Attach Photo and Summarize Document** (ios, 9 screens, AI Tool) [Refero](https://refero.design/flows/7027) > Happy-path: A signed-in mobile user opens the chat composer, taps the plus button to attach content, chooses a file from the Photos/Files picker (Recents), and attaches a photographed/scanned page. The thumbnail appears inline in the message composer. The user types a request (“Give me summary”) and sends it. The assistant responds with a summarized transcription/summary shown as a message with standard message actions (like, share, thumbs up/down, etc.). Steps: Chat Home / Composer idle → Attachment sheet — Photos view → Attachment options expanded → Files picker — Recents → Composer with attached thumbnail → Typing instruction with keyboard → Send and message bubble created → Assistant reply — summary displayed → Post-response actions **ElevenReader: Create and listen to narrated text** (ios, 5 screens, AI Tool, Books) [Refero](https://refero.design/flows/8968) > Happy path: A logged-in user on the home feed selects the "Write text" upload option, types or pastes content, triggers conversion/import, and then listens to the generated narration while following highlighted text. Steps: Home / Upload entry → Blank text editor → Text entry completed → Import / processing → Reader with synchronized playback **ElevenReader: Scan & import document** (ios, 6 screens, AI Tool, Books) [Refero](https://refero.design/flows/8970) > Happy path: A logged-in user on the app Home screen chooses the 'Scan text' option to capture a physical document and convert it into a readable/listenable item. The app requests Camera permission, opens a camera preview for capture, automatically detects/crops the document, then uploads/loads the extracted text. The flow ends back on Home with an "Importing content" banner and the new item appearing in the Continue listening list. Steps: Home (Scan entry) → Camera permission prompt → Camera preview (document visible) → Auto-detect / crop confirmation → Processing / Loading → Import complete (Home) with banner **Dropbox: Document Scan Crop & Filter** (ios, 4 screens, Business, Collaboration Tools, Productivity) [Refero](https://refero.design/flows/2426) > User has already taken a photo of a paper document and is presented with an in-app scanning editor. The happy path: user adjusts the four corner handles to crop and correct perspective, chooses a scan filter (B&W, Whiteboard, Original), optionally adjusts a filter intensity/contrast slider, then taps Done to save/export or Cancel to discard. **ElevenReader: Create GenFM episode** (ios, 12 screens, AI Tool, Books) [Refero](https://refero.design/flows/8976) > Happy-path: A logged-in mobile user starts on the app Home screen and uses the central + action to import content and create an AI-generated episode (GenFM). They choose an import method (paste link / upload / scan / write), pick an episode format (Prime time or Bulletin), import a source (file or link), and submit. The system shows an import/creation modal and toasts indicating progress. The new episode is added to the user's Library with a “preparing” state, then updates to ready. The user opens the episode detail and plays it; the screen shows a synchronized transcript with line highlighting and standard playback controls. Steps: Home (entry) → Import content modal → Create an episode (GenFM) — entry → Create an episode — choose format → File picker — Recents → Home — Ready soon toast → Library — preparing state → Library — progress indicator → Library — ready notification → Library — new episode visible → Episode detail — playback UI → Playback — live transcript highlighting **ABY Journal: Scan handwritten journal entry** (ios, 11 screens, Health & Wellness, AI Tool, Lifestyle) [Refero](https://refero.design/flows/7994) > Happy-path: A user adds a new handwritten journal entry by choosing the Scan option from the entry composer, granting camera permission, photographing a paper page, confirming crop, and letting the app scan/parse the image into text and attach a media thumbnail. Entry is then inserted into the Timeline as a new card with parsed content and a media thumbnail. Steps: Timeline (Today) → Entry Composer (keyboard + options) → Scan Onboarding CTA → Camera Permission Prompt → Camera Viewfinder (initial) → Capture Preview → Crop / Confirm Framing → Scanning Progress → Parsed Entry Preview → Media Section in Composer → Timeline Updated / New Card **ABY Journal: Journal Scan Capture** (ios, 4 screens, Health & Wellness, AI Tool, Lifestyle) [Refero](https://refero.design/flows/7995) > Happy path: from the journal composer the user chooses the Scan option, sees a promotional modal that explains scanning, taps Get Started, grants camera permission, uses the in-app camera to photograph a handwritten journal page, then reviews/adjusts the crop and taps Scan to import the image into the journal entry. Steps: Scan Journal Modal → Camera Permission Request → Live Camera Capture → Review, Crop & Scan **Mela: Scan & Import Printed Recipe** (ios, 20 screens, Food & Drink) [Refero](https://refero.design/flows/10369) > Happy-path: A user in the recipe list opens the Recipes menu, chooses "Scan Recipe...", uses the phone camera to capture multiple pages of a printed recipe, reviews the captured images, then enters an OCR-assisted editor where they select text regions and map them to semantic fields (Title, Description/Notes, Ingredients, Instructions, Image). After the app processes the selections and populates a recipe editor, the user saves the recipe and returns to the recipe list with the new recipe visible. Steps: Recipes list → Recipes menu → Camera / scan: live view → Camera capture (in-progress) → Captured images preview → Image review tools → Save captures and proceed → Select text intro → Select title region → Assign text type menu → Select and tag description/notes → Select ingredients block → Select instructions block → Repeat for remaining pages/blocks → Start parsing OCR → Review parsed text → Final edit form → Save recipe → Saved recipe detail (optional) → Updated recipes list **Structured: Scan & Parse Schedule** (ios, 8 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3292) > User opens the Structured AI modal to convert a physical weekly planner into digital tasks. From the modal the user chooses to provide input via camera or photos, grants camera permission when requested, captures a photo of their paper schedule, reviews/retakes the photo, optionally adds instructions, and taps Generate. The system processes the image (shows a progress state) and returns a list of parsed events/tasks for the user to review and accept. **Mela: Document Photo Capture and Edit** (ios, 3 screens, Food & Drink) [Refero](https://refero.design/flows/10370) > Happy path: User opens the device camera in a document-capture orientation, frames a printed page using on-screen alignment guides, and taps the shutter to capture one or more photos. After each capture the interface shows a small preview and a prominent Save control; saved captures are collected into a review carousel. From the review screen the user can retake, edit (Adjust, Filters, Rotate), delete individual images, or finish (Done) to accept the images. Steps: Frame & Capture → Capture Preview → Review & Edit **Trip Way: Add and Import Documents** (ios, 4 screens, Travel & Booking) [Refero](https://refero.design/flows/8456) > Happy-path: A mobile Documents screen lets the user add new documents. Entry: user opens Documents (empty or existing list). The user taps the prominent Add Doc button, chooses a source (Files, Gallery, or Camera), picks one or more items in the system file picker, and confirms. Exit: the Documents screen updates to show the newly added items as thumbnails with filenames. Assumptions: user is on a mobile device (portrait), has permission to access files/gallery/camera, and the file picker is the system/native picker. Steps: Documents - Landing → Source Selection Menu → Files Picker - List View → Documents - Thumbnails (Imported) **Delphi: PDF Question Submission** (web, 6 screens, AI Tool) [Refero](https://refero.design/flows/13983) > A user in an existing conversation with Jason Jones attaches a PDF, waits for the upload to complete, adds a question about the document, and sends it. The assistant then begins processing the request. The flow starts in an open chat with an authenticated user and ends with the document and question submitted for an assistant response. The user appears to be on a desktop device with permission to send messages and upload files. **WhatsApp: Document Search in WhatsApp** (web, 4 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/13220) > A user opens the shared media/documents viewer in WhatsApp, reviews recent documents from chats, then uses the search field to find a specific document by sender, filename, or caption. The flow starts from a chat or media section and ends with either a filtered document result or a no-results state. **Cohere: Document Attachment in Conversation** (web, 4 screens, AI Tool, Business, Development Tools, Productivity) [Refero](https://refero.design/flows/3834) > Happy path: A signed-in user opens an existing conversation titled "Business Plan Outline for a Marketing Agency in French", reads assistant responses and the conversation's referenced documents, posts a request to summarize the documents, then attaches a new file (invoice PDF) from the message composer. Steps: Open conversation / view references → Compose summarization request → Attach file (upload starts) → Attachment ready in composer **WhatsApp: Send Documents in Chat** (web, 6 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/13184) > Happy path in a WhatsApp desktop conversation with Samuel. The user starts in an open chat, opens the attachment picker, selects one or more PDF documents, optionally adds a text caption, and sends the files. The conversation remains open throughout; the main exit point is the message appearing in the chat thread and the attachments showing as sent items in the message list. ##### Task: email inbox **Givingli: Send a Card to a Contact** (ios, 12 screens, Lifestyle, Scheduling Tools) [Refero](https://refero.design/flows/8216) > Happy-path: A logged-in user starts at an empty Inbox and chooses to send a digital card. They select one or more recipients from their contacts, pick a card from categories or popular cards, personalize the card (choose template and message), optionally add a gift, review delivery options (send now or schedule), then send. The flow ends at a send-confirmation screen with sharing options and a return to the Inbox where the sent item appears. Steps: Empty Inbox → Select Recipients (list) → Recipient confirmed → Card Gallery → Card Detail → Choose a Template → Card Editor (Message tab) → Add a Gift → Review & Delivery → Send Confirmation → Inbox Updated → Conversation Detail (recipient) **Fernand: Support tool sign in** (web, 4 screens, Business, Communication Tools) [Refero](https://refero.design/flows/11909) > Happy path for a support agent or founder starting from the public marketing site, moving through sign-in, and landing in the product’s main support workspace. Steps: Marketing homepage → Login screen → Login submission → Support dashboard **HYPE: Password Recovery and Reset** (ios, 7 screens, News, Lifestyle) [Refero](https://refero.design/flows/12243) > Happy path starts on the Hypebeast account login screen after a failed sign-in attempt. The user selects Forgot Password, enters an email or username, receives a reset email, opens the email link, sets a new password, and returns to the login screen with a success confirmation. Steps: Login Error State → Forgot Password Entry → Email Input While Typing → Recovery Sent Confirmation → Reset Email Inbox → New Password Form → Reset Success Return **Meta AI: Gmail Email Audit** (ios, 12 screens, AI Tool) [Refero](https://refero.design/flows/13354) > A user starts in Meta AI and asks it to identify active paid subscriptions and assess which are useful. Meta AI explains that it needs access to an email inbox, presents Gmail and Outlook connection options, and guides the user through Gmail authorization. After Gmail is connected, the user initiates a separate request to summarize unread Gmail messages. Meta AI reads the messages and returns a structured summary with priority items, recommended next actions, and follow-up options. The flow assumes the user is signed in to Meta AI and has access to a Google account, but Gmail is initially unconnected. **Linear: Passwordless email sign-in to Home** (ios, 10 screens, Collaboration Tools, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/6603) > Happy path for a passwordless email sign-in that starts at a branded welcome splash and ends at the authenticated Home screen. The user chooses email login from a webview, enters their email, receives a temporary verification code in their inbox, copies/pastes or types the code back into the webview, submits it, then the app loads and prompts for push notifications. Steps: Welcome / Get started → Choose sign-in method → Enter email address → Check inbox / open email → Return to code entry → Submit verification code → Spinner / loading → Notification permission prompt → Home (initial state with overlay) → Home (ready) **XChat: Sign‑up and Chat Onboarding** (ios, 25 screens, Communication Tools, Social Networking) [Refero](https://refero.design/flows/11642) > Happy-path sign-up and chat onboarding for a mobile X Chat experience. Steps: Branded splash / Sign in CTA → Sign-in options → Birthday picker (initial) → Birthday picked & confirm → Enter email (empty) → Enter email (filled) → Enter verification code (empty) → Email with code → Code entry with suggestion → Code submitted (full) → Choose a password (empty) → Choose a password (filled) → Chat onboarding summary → Create passcode (enter) → Create passcode (partial) → Confirm passcode (empty) → Confirm passcode (partial) → Confirm passcode (invalid) → Notification permission prompt → Empty Chat inbox **Kin: Onboarding to Chat & Reminder Setup** (ios, 34 screens, AI Tool) [Refero](https://refero.design/flows/5431) > Happy‑path new user onboarding: user starts at an introductory hero screen and taps Get started. They answer a referral source question, then provide a display name and (optional) email. The app validates the email inline. After submitting, the user enters the conversational interface with an AI assistant (Kin). The user provides personal details via typed text or voice, the assistant proposes a weekly reminder, the user approves it, and the assistant confirms the reminder and continues the conversation. Steps: Intro hero → Referral source question → Referral selection confirmation → Name and optional email form → Email inline validation (error) → Correcting email input → Form submitted (valid email) → Assistant welcome screen → Typing first user message → Assistant proposes a plan → Offer voice input (microphone) → Microphone permission modal → Recording started → Voice recording in progress → User confirms via text → Reminder card with actions → Reminder approved confirmation → Assistant continues conversation → Conversations list → Home / welcome dashboard **Daze: Account creation & friend onboarding** (ios, 36 screens, Social Networking) [Refero](https://refero.design/flows/9114) > Happy-path new account onboarding for a social messaging product (Daze). Entry: user opens the app and sees a welcome splash. Flow collects birthdate, then asks for phone number or email; the user chooses email, enters it, and confirms via a 6-digit code. After verification, the app shows benefit screens (notifications, contact sync) and requests system permissions for notifications and contacts. The app then offers options to share contacts (select or share all), lets the user add up to 3 close friends (by contact or username), and finishes at the main Messages list. Steps: Welcome splash → Birthday picker (initial) → Birthday picker (select date) → Birthday confirm → Phone number entry → Email entry (keyboard) → Email typed → Confirm email modal → Email verification (enter code) → Verification email (external) → Feature intro: notifications → System notification permission → Contacts sync intro → System contacts permission → Contacts share options → Add 3 best friends (list) → Search username to add friend → Friend(s) added preview → Messages list (initial) → Onboarding complete / home state **komoot: Password recovery via reset email** (ios, 4 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8422) > Happy-path recovery after a failed login attempt on a mobile device. Entry: user lands on the login screen with an email prefilled and an "Incorrect password" inline error. From there the user taps "Forgot password?" to open a Reset password screen (email address prefilled). They submit the email, receive an in-app confirmation modal "Email Sent", then check their inbox and receive an email containing a temporary password and instructions to change it. Exit: user has a temporary password in email to use to sign in and update their password in account settings. Assumptions: user is on a mobile device, not currently authenticated, and has access to the listed email inbox. Steps: Sign-in with inline error → Reset password (enter/confirm email) → In-app confirmation modal → Received reset email with temporary password **Netflix: New account signup and onboarding** (ios, 41 screens, Entertainment, Video Streaming) [Refero](https://refero.design/flows/10554) > Happy path for a new user creating a Netflix account on mobile (web flow that redirects to app). Entry: marketing hero / promotional screen. Exit: user finishes onboarding and is prompted to open/download the Netflix app to start watching. Steps: Marketing Hero / Call to Action → Account Entry Choice → Email Collection (empty) → Email Input (autofill visible) → Send Sign-up Link (passwordless option) → Check Inbox / Next Options → Create Password Instead → Profile Details (name & DOB) → Language Preferences → Content Preference Grid Intro → Pick Favorites (selecting) → Finish Selections → Processing / Personalization Loading → App Download Prompt → Final Marketing / Confirmation → Email input validation state → Send Link confirmation with checkbox → Password creation validation → Language multi-select confirmation → Completed onboarding redirect **Cushion: Email Login to Inbox** (web, 6 screens, Communication Tools, Collaboration Tools) [Refero](https://refero.design/flows/12506) > Happy path starts on the Cushion marketing/landing page for a signed-out visitor. The user can either choose the login entry point from the header or hero CTA, then move to a minimalist authentication screen where they enter an email address and request a one-time login link. After submission, the system confirms the link was sent and, once authenticated, takes the user into the workspace inbox. The exit point is the logged-in inbox view with navigation, message lists, and an empty-state/read-state mix visible. Assumptions: user is on desktop, not already logged in, and email-link authentication is available. Steps: Marketing Landing → Log In Start → Email Entered → Sending Login Link → Link Sent Confirmation → Inbox Home **Fernand: Send New Support Email** (web, 7 screens, Business, Communication Tools) [Refero](https://refero.design/flows/11894) > Happy path for sending a new support email from an inbox/inbox-style messaging workspace. The user starts on the Inbox list, opens the compose modal, enters or selects recipients, adds a subject and message, optionally formats selected text, then sends the email. After sending, a confirmation toast appears and the new conversation is added to the inbox list. Assumptions: user is already logged in, has access to the inbox, and can send from the shown support address on desktop. Steps: Inbox Overview → Open Composer → Recipient Entry → Draft Message → Edit Formatting → Send Confirmation → Inbox Update **Jace AI: Jace Trial Onboarding** (web, 35 screens, Communication Tools, AI Tool) [Refero](https://refero.design/flows/10210) > Happy path from the public Jace marketing experience into a new-user onboarding journey. The user starts logged out, selects a free trial, reviews social proof and privacy information, signs up with Google, grants requested Google permissions, answers a short discovery question, optionally invites a teammate, and enters the Jace inbox. The final state shows the connected inbox being prepared and then populated with categorized email. The flow assumes the user has a Google account and is willing to grant access to Gmail and related Google data. Steps: Product Introduction → Social Proof → Privacy Assurance → Account Signup → Google Account Selection → Initial Google Consent → Discovery Question → Google Permission Review → Permission Authorization → Team Invitation → Add Teammate → Send Invitation → Invitation Confirmation → Welcome Preparation → Inbox Synchronization → Initial Inbox → Prepared Workspace → Signup Testimonial Carousel → Authorization Loading → Permission Selection State **Jace AI: Snooze an inbox message** (web, 5 screens, Communication Tools, AI Tool) [Refero](https://refero.design/flows/10216) > Happy path: a signed-in user in the Main inbox suspends a single message until a later time. Steps: Main inbox overview → Reveal quick actions → Open time input dialog → Enter and select time → Confirmation and removal from inbox **Jace AI: Find messages by contact** (web, 3 screens, Communication Tools, AI Tool) [Refero](https://refero.design/flows/10223) > Happy path: A signed-in user on desktop is viewing their main inbox list, uses the top search bar to find messages by typing a short query, selects a contact suggestion from the autocomplete dropdown, and is shown a filtered list of emails matching that contact. Steps: Inbox Overview → Search Autocomplete → Search Results (Contact) **Skiff: Select and Move Messages to Trash** (web, 4 screens, Communication Tools, Scheduling Tools, Paperwork) [Refero](https://refero.design/flows/1887) > Happy path: a signed-in user is viewing an alias inbox with two messages, selects both messages using the list checkboxes, and deletes/moves them to Trash via the list toolbar. Steps: Inbox with unread messages → Select messages → Execute delete (move to Trash) → Empty inbox state after action **Lemni: Inbox Conversation Triage** (web, 3 screens, Business, Productivity, AI Tool) [Refero](https://refero.design/flows/4594) > Three-column inbox workspace for triaging and acting on conversations. Steps: Open conversation in inbox → Adjust list order and filters → Run quick actions and view details ##### Task: empty state first use **ten ten: Invite first friend** (ios, 4 screens, Communication Tools, Social Networking) [Refero](https://refero.design/flows/2266) > Happy path: A new user with no friends on the service is prompted to add at least one friend. From the empty state they tap the central plus CTA to open an action sheet with three invite options (add by pin, share my pin, invite via Instagram). Selecting “share my pin” opens the iOS share / Messages composer with a prefilled link and the user's PIN. After sending the invite (or returning), the main screen moves to a waiting state that indicates a friend request is pending. Entry: user is signed in, has zero friends, on an iPhone. Exit: app shows a pending/waiting-for-acceptance state. Assumptions: device-level share/messages are available; the app can generate and display a short alphanumeric PIN; the app returns to the waiting state after initiating an invite. **On: Country / Region selector** (ios, 4 screens, Marketplace) [Refero](https://refero.design/flows/9608) > This is a mobile Country / Region selector where the user can scan an alphabetized list or use the search box to filter and pick a country. Entry: user navigates into the Country/Region screen (from settings or onboarding). Happy path: user opens the screen, optionally types in the search box, filtered results appear, user taps a radio item to select a country, and the selection is reflected in the list (and the user can go back). Exit: user returns to the previous screen with the updated selection. Assumptions: user is on a mobile device (iOS-style keyboard shown), has permission to change their profile/region, and the screen is reachable from a settings or onboarding flow. The UI shows current selection via a filled radio control and supports incremental search, clear (x) and empty-state messaging. Steps: Country list (default) → Search input focused → No results state → Filtered results and selection **Fresha: Country/Region Selector** (ios, 2 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4948) > A modal Country/Region selector that lets a user search and choose a country from a list. Entry: user opens the country/region selector (e.g., from a profile, settings, or form field). Exit: user either selects a country (keeps the modal open or closes it depending on parent behavior) or dismisses the modal with the close (X) control. Assumptions: user is on a mobile device and the selector is presented as an overlaid modal. Keyboard appears when the search input is focused. The interface presents each country with a flag, name, and a single-choice radio indicator. Steps: Search Focused with Keyboard → Filtered Results Display **Kickstarter: Project search and empty state** (ios, 2 screens, Fundraising) [Refero](https://refero.design/flows/4634) > Happy path: The user opens the Search tab (bottom navigation) and sees a list of popular projects with a persistent search input at the top. From this listing the user can tap the search field, enter a query (in this case a numeric string) and submit. When the query returns no matches the interface shows an empty-state screen with a clear “No Results” message and a Cancel control. Steps: Popular Projects List → No Results **Willow: Willow onboarding introduction** (ios, 2 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/13228) > A two-step onboarding sequence introduces Willow as a custom keyboard and explains how to switch it on and off, then explains that it can be used for dictation in any phone app. The flow starts on an instructional screen with a Back button, a Skip option, progress indicators, and a primary Next button. It ends on a second instructional screen with the same navigation pattern and a primary call to action labeled Ready to get started. The screenshots assume the user is on a mobile device and is not required to be logged in or enter data during this flow. Steps: Custom Keyboard Introduction → Dictation Introduction **PayPal: Country/Region Selector** (ios, 3 screens, Finance) [Refero](https://refero.design/flows/7233) > A mobile "Country or Region" selection modal that presents an alphabetized list of countries with flags and a search field. Steps: Country list (default) → Search input active (typed query) → Filtered results and selection **Fresha: Country/Region selector** (ios, 2 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4947) > Mobile Country/Region selection overlay. Entry: user opens a Country/Region picker (modal) from a prior context (not shown). Happy path: user either scrolls and taps a listed country or types into the search input to filter the list, then closes the modal. The modal is shown on an iPhone-sized viewport with the system keyboard. Assumptions: user may or may not have an existing selection; the modal remains visible while the user interacts with the list or search input. Exit: user leaves the modal by selecting a country or explicitly closing it. **Verse: User search and no-results create prompt** (ios, 2 screens, Social Networking, Design & Media Editing) [Refero](https://refero.design/flows/4007) > Mobile in-app user search flow (happy path). Entry: user opens the search interface (back arrow visible) and sees a list of suggested/recent user profiles and a search field. The keyboard is shown for text entry. The user types a query (name, handle, or numeric id). Exit points: selecting a listed profile to view their profile, or — if the query returns no matches — tapping the primary CTA to create a new verse. Assumptions: user is signed in on an iOS device; keyboard is active; searching can accept alphanumeric and numeric input. **raum.: Home Location Setup** (ios, 5 screens, Travel & Booking) [Refero](https://refero.design/flows/13479) > The user launches Raum and begins onboarding. They tap the splash screen, respond to the operating system’s location permission prompt, and arrive at a location setup screen. The interface suggests the United States as the user’s home country, which the user accepts. The system then opens an interactive map showing the home region highlighted and prompts the user to tap another place they have visited. The final screenshot shows the map zoomed into the northeastern United States and Canada with city labels, a progress indicator, settings control, and search control. The flow assumes a first-time user on an iPhone and does not show account creation or authentication. Steps: Welcome Splash → Location Permission → Home Country Selection → Visited Places Map → Map Exploration Guidance **DailyArt: Billing country selection** (ios, 3 screens, Education, Magazines & Newspapers) [Refero](https://refero.design/flows/3330) > User is filling the Billing details form on a mobile browser (checkout). The user scrolls to the Country / Region field, opens the country selector (a dropdown with a searchable input), types into the search field and expects to select a matching country. Entry: billing details form (Country / Region collapsed). Exit (happy path): user selects a country from the list, the Country / Region field is populated and the dropdown closes. Assumptions: mobile device (iPhone), browser context (secure checkout), form marks Country / Region as required (asterisk). **Verse: Explore Search and Create Verse** (ios, 3 screens, Social Networking, Design & Media Editing) [Refero](https://refero.design/flows/3976) > Happy path: A logged-in user browsing the Explore feed opens search, types a query (or selects a suggestion), and either lands on search results or an empty-result state that prompts them to create new content. **Showcase: Discover & Search Content** (ios, 3 screens, Entertainment) [Refero](https://refero.design/flows/7508) > Happy-path: A mobile user opens the discovery/search area, taps the persistent search pill at the bottom, types a query, and sees either a populated results grid or an empty-state message. Steps: Discovery / Search landing → Empty results (numeric query) → Search results (keyword query) **Mocha: New Project Creation** (web, 2 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/11606) > Happy path: authenticated user visits the Dashboard > All projects listing and encounters an empty-state. The interface shows guidance text and two prominent CTAs to start a new project (central purple button and top-right button). Steps: Dashboard — All projects (empty state) → Empty-state focused (alternate view) **Rork: App Project Creation Dashboard** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12146) > Logged-in landing/dashboard screen for creating a new mobile app project through a chat-style input. The user starts on an empty projects view, optionally adds references or settings via the plus menu, enters an app idea, and can submit it to begin project creation. The visible exit state is the first project being created and appearing in the projects area. Assumptions: user is authenticated, has access to the dashboard, and is on desktop. Steps: Empty Dashboard → Input Options Menu → Prompt Ready to Submit **MyMind: Create New Space** (web, 6 screens, Productivity, Reference, Inspiration Board) [Refero](https://refero.design/flows/13318) > A user creates a new space from the All Spaces view. They open the creation dialog, enter a name, continue to the color-selection step, choose a color, and save. The system opens the newly created empty space and confirms creation. The user is assumed to be signed in and viewing the existing Spaces area on a desktop device. Steps: All Spaces → Enter Space Name → Continue Creation → Choose Space Color → Save Space → New Empty Space **Runway: Empty Projects → Image Library** (web, 2 screens, AI Tool) [Refero](https://refero.design/flows/1328) > Logged-in user lands on the Projects area and sees an empty-state dashboard showing they have no projects. The page provides two prominent ways to start: a large primary CTA in the main panel (“Create your first project”) and a “New Project” button in the header. The left navigation exposes related sections (Assets, Images, Videos) and shows the account plan and project limit. From this state the user navigates to the image/asset library where a masonry grid of images is displayed with a top toolbar for controls and a small promotional CTA to be featured. **Mocha: Apps Dashboard — Empty State and Account Menu** (web, 4 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/11605) > A calm onboarding sequence for a user landing on the 'All apps' dashboard with no projects yet. The user can create their first project from the prominent CTA, open the account menu to access documentation and settings, and switch the interface appearance from Light to Dark. Entry: signed-in user lands on the dashboard (no projects). Exit: user either opens the 'Create new app' flow or changes theme and closes the account menu. Steps: Empty dashboard (Light) → Account menu opened (Light) → Account menu — Switch to Dark → Empty dashboard (Dark) **Melius: Create Project Workspace** (web, 7 screens, AI Tool, Design & Media Editing) [Refero](https://refero.design/flows/14064) > The flow begins on the logged-in Melius Home screen. The user selects New project, enters a project name in a modal, and submits it. The system displays a brief loading state while connecting nodes, then opens the newly created project canvas. The flow assumes the user has permission to create projects and is using a desktop browser. It ends at an empty canvas ready for the user to start imagining or generating content. **Aboard: Document category creation** (web, 3 screens, Collaboration Tools, Paperwork) [Refero](https://refero.design/flows/796) > Happy path: a logged-in desktop user opens Documents > Overview, adds a new category inline, and sees the category created with an empty state. Steps: Documents overview (before add) → Enter new category name → New category created (empty state) **PamPam: Account menu to empty-state transition** (web, 4 screens, Map & Navigation, AI Tool) [Refero](https://refero.design/flows/4762) > Happy path: A logged-in user views their maps dashboard (left rail with map list and primary 'Make a map' CTA) and opens the profile/settings dropdown. They choose an account-level option from the dropdown which triggers a transient processing modal. After processing the app navigates to an empty/404 state that offers a single recovery action. **Bezi: Create new scene** (web, 4 screens, Design & Media Editing) [Refero](https://refero.design/flows/179) > Happy path: A logged-in team member opens the Team #1 workspace, clicks the New file button, chooses a scene type (Empty 3D / AR / VR or a gallery template) from the New Scene modal, and lands in the 3D editor with an initialized scene (grid, default camera/lighting and scene hierarchy). Steps: Team workspace (no projects) → New Scene modal — choose type → Selection confirmation (preset highlighted) → Editor with new scene **Hashnode: Search with empty-state transition** (web, 3 screens, Communication Tools, Development Tools, Business) [Refero](https://refero.design/flows/3948) > Happy-path summary: The user invokes the site-wide search (via the search control), which opens a centered search modal with a focused input. The user types a query; the modal shows category tabs (Top, Latest, Tags, People, Blogs) and an inline results area. If there are no matches, the modal displays an empty-state message and a keyboard hint to view all results. The user can press Enter (or otherwise request full results), which transitions to a full-page search results view that shows the same empty-state message plus an illustrative graphic. Steps: Open Search Modal → Type Query — Inline Results → Full Results Page — Empty State **Aboard: Form Creation to Draft Editor** (web, 5 screens, Collaboration Tools, Paperwork) [Refero](https://refero.design/flows/820) > Happy path for creating a new form from the Forms area when there are no existing forms. Steps: Forms empty state → New form modal - initial → Template dropdown open → Template selected and ready → Draft form editor **Spline: Create first team project** (web, 4 screens, Design & Media Editing) [Refero](https://refero.design/flows/260) > Happy path: A logged-in team member creates the team's first project from an empty team projects view. Entry: user is on the team workspace projects area (no projects exist). Exit: a new project is created and visible as a project card; its detail area shows the empty-file state and a prompt to add a new file. Assumptions: the user is signed in, is viewing a specific team (has permission to create projects), and is on a desktop or large-screen device (layout indicates wide-screen). **Mapbox: Get Started Preferences** (web, 2 screens, Development Tools, Map & Navigation) [Refero](https://refero.design/flows/9690) > A lightweight onboarding questionnaire presented as a modal on the Account overview screen. Steps: Onboarding modal appears → Complete form and enable submit **Reown: Cloud onboarding carousel** (web, 4 screens, Development Tools, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/5721) > A modal-based onboarding carousel that appears over the Projects area of the product. Entry: user is on the Projects screen (logged-in state) and the welcome/onboarding modal opens automatically. The modal provides a short multi-step introduction to 'Reown Cloud' features (Discover Cloud, SDK installation, WalletGuide listing, Analytics). Navigation is via Next/Previous controls, a Skip link, and a close (X). Exit: user either dismisses the modal (X or Skip), follows a CTA (Discover Cloud or Create Project), or navigates through all slides and uses the final CTA. Steps: Welcome modal → SDKs highlight → WalletGuide listing → Analytics CTA **Portrait: Portrait Editor Setup** (web, 3 screens, Social Networking) [Refero](https://refero.design/flows/11793) > This is a portrait-creation editor shown in a logged-out or anonymous draft state. The user enters a name and description, builds a portrait by adding content frames from the bottom toolbar, optionally previews the result, and can publish from the top-right button. The sequence shown starts with an empty portrait canvas and moves to the user filling in the profile/avatar area while the main canvas remains empty. Steps: Empty Portrait Draft → Draft in Progress → Profile Image Added **WhatsApp: Region Search Selector** (web, 3 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/13217) > A region selection modal is opened over the Find channels screen. The user can search for a country/region, review matching results, or see an empty state when no match exists. The flow starts from the channel discovery area and ends with the user narrowing the region list by typing a query and selecting a visible region option, or clearing/changing the query to find matches. **Apple Maps: Recent Places Management** (web, 2 screens, Map & Navigation) [Refero](https://refero.design/flows/12477) > This is a map navigation interface focused on the user’s recent places history. The entry state shows a left sidebar with Search, Guides, and Directions, plus a Recents panel listing previously visited or searched locations. The user can review recent destinations, remove an individual item, or clear the full list. The exit state shown is an empty Recents panel after the history has been cleared, while the map remains visible on the right. Assumption: the user is already in the map experience and has prior recent locations saved. Steps: Review Recents → Empty Recents **Leonardo.Ai: AI Creation Intro Guide** (web, 6 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/12905) > This is a guided introduction inside an AI image generation workspace. The entry point is the main creation screen with a prompt field, generation controls in a left sidebar, and a primary Generate action. The user is assumed to be logged in and viewing the creation interface. The tour moves through key controls in sequence: prompt box, generate button, model selector, references/elements, prompt generator, and the intro guide/info button. The exit point is finishing the tour, returning the user to the main creation screen with the guide available for later revisit. Steps: Prompt Box → Generate Images → Models → References and Elements → Prompt Generator → Intro Guide **Missive: Help Center search and empty-result state** (web, 3 screens, Business, Collaboration Tools, Communication Tools) [Refero](https://refero.design/flows/6243) > Desktop Help Center entry -> user searches for a term -> search overlay shows result state. Entry: user lands on the Help Center main screen which lists categorized guide tiles and a prominent search input. Exit: user either opens a guide, closes the search modal, refines the query, or follows the 'can't find your answer' support link. Assumptions: desktop layout, public help center (user not required to be signed in to view), and the search control triggers a focused overlay for results. Steps: Help Center landing → Search query entered → Search overlay — empty results **Microsoft Copilot: Start new Copilot conversation** (web, 4 screens, AI Tool) [Refero](https://refero.design/flows/5520) > Happy-path: A signed-in desktop user reads an article that includes an inline "Message Copilot" composer at the bottom. From that context they open the global composer/menu and select to start a new conversation. The flow transitions from contextual (article-scoped) assistance to a blank, standalone Copilot composer with suggested task chips. **Linear: Project statuses management** (web, 4 screens, Business, Productivity, Project Management, Collaboration Tools) [Refero](https://refero.design/flows/6699) > Happy path: A user with permissions navigates to the Project statuses settings (via the left navigation under Projects → Statuses), views the grouped workflow buckets (Backlog, Planned, In Progress, Completed, Canceled), and adds a new status to the In Progress group. The user clicks the group's plus button, an inline row expands into a Name + Description input with Cancel/Create actions, the user fills the fields and clicks Create, and the new status is inserted into the In Progress group and displayed with its icon and description. Steps: Statuses overview → Open inline create form → Fill name and description → Create status and update list **Tango: Submit Button Step Review** (web, 3 screens, Business, Productivity, Project Management, AI Tool) [Refero](https://refero.design/flows/5820) > Happy path: A user is reviewing a single guide step inside a guide editor (Draft) titled "Submit Spark Survey Responses." The left rail shows the step list with the current step "Click on Submit" selected. The main canvas displays a screenshot that highlights a Submit button and a contextual callout / transcript box below it. The three screens show three visual states of that same step (info state, markers visible, success state). Steps: Step shows info callout → Markers/icons visible → Success/completed state ##### Task: error state recovery retry **Tripsy: Password recovery from failed login** (ios, 4 screens, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/2325) > Happy path: A signed-out user attempts to log in with an email and password, receives a failed-login alert, chooses the "Forgot Password?" option, confirms or edits their account email in a prompt, and then receives a confirmation that reset instructions were sent. **X: Unsent DM with Retry/Delete** (ios, 2 screens, Social Networking, Entertainment, News, Video Streaming) [Refero](https://refero.design/flows/2920) > Happy path: A logged-in (but unverified) user opens a conversation with an account, types a message containing a link, and attempts to send it. The message fails to deliver and is shown with a 'Not sent' status and an error indicator. The user opens a context menu on the unsent message to either retry sending or delete it. Steps: Conversation - Send Attempt → Unsent Message Context Menu **Dropbox: Password recovery via email** (ios, 5 screens, Business, Collaboration Tools, Productivity) [Refero](https://refero.design/flows/2416) > Happy path: A signed-out mobile user arrives at the Sign in screen (offers Apple/Google and email/password methods). From there the user selects "Having trouble signing in?" which opens an action sheet. They choose "Forgot password?" to go to the Forgot password form, enter their email and tap Send. The app shows a sending overlay then displays a confirmation alert that a recovery email was sent (if an account exists). **Instagram: Direct message audio calling** (ios, 3 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/2883) > Happy path: A logged-in mobile user sees a missed audio call in their messages list, opens the conversation, and uses the in-thread control or header call control to place a return call. The system attempts the call, then reports 'No Answer' and presents retry and cancel options. Entry: Messages list (user logged in). Exit: Call attempt ends with 'No Answer' state (user can Call Again or Cancel). Assumptions: user is authenticated on a mobile device and has access to the messaging/conversation and in-app calling features. Steps: Messages list with missed call → Conversation thread with missed call bubble → Call attempt result (No Answer) **Instagram: Account recovery — Find account** (ios, 5 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/2900) > Happy path: A user attempts to sign in from the app's main login screen. After entering credentials, the attempt fails with an "Incorrect Password" system alert. The user taps the "Find your account" link to start account recovery, is taken to an account lookup screen, enters an email/username, and taps "Find account." The system responds with a "No account found" alert. Steps: Login attempt: incorrect password → Edit password (reveal and retype) → Open account recovery form → Enter identifier and submit → No account found alert **X: Password recovery via confirmation code** (ios, 7 screens, Social Networking, Entertainment, News, Video Streaming) [Refero](https://refero.design/flows/2902) > User attempts to sign in but enters an incorrect password, then uses the "Forgot password" path to recover their account. The flow starts at the password entry screen (assumes user is signed out and an identifier is pre-filled) and proceeds through: password error → find account (identifier confirmed) → choose verification method (email or phone) → send code → enter code → code rejected. Exit points: successful login (not shown) or cancel/back to previous screens. Steps: Password entry → Password error alert → Find account (identifier confirmation) → Choose verification method → Code sent — code entry (idle) → Code entry and submit → Code validation error **Dropbox: Recover deleted file** (ios, 5 screens, Business, Collaboration Tools, Productivity) [Refero](https://refero.design/flows/2437) > Happy path: A signed-in mobile user opens the Deleted files area, views a recently deleted PDF, chooses to restore it, and receives progress + success feedback. Entry: user navigates to Deleted files (already authenticated). Exit: file is removed from Deleted files and a success message is shown; Deleted files list becomes empty. Assumptions: user is logged in with permission to restore their files, working on a mobile device; only one file is shown in the list during this flow. **X: Login and Recovery** (ios, 14 screens, Social Networking, Entertainment, News, Video Streaming) [Refero](https://refero.design/flows/2901) > Happy path: A new or returning user lands on the app splash and welcome screens, chooses to log in, enters an identifier (phone, email, or username), and — if the account exists — is prompted for a password. If the password is forgotten or incorrect, the user can start the password recovery flow: confirm account, choose a delivery method (email or SMS) for a confirmation code, receive the code, enter it, and proceed to reset the password. Steps: Splash → Welcome / Sign up options → Identifier entry (empty) → Identifier entered (ready to submit) → Account not found error → Password entry prompt → Password input and login attempt → Wrong password error → Find account for recovery → Choose confirmation delivery method → Code sent – enter code → Code entry attempt → Incorrect code error → Back/Retry states and next actions **222: Resume invites in Account settings** (ios, 4 screens, Lifestyle, Dating & Relationships) [Refero](https://refero.design/flows/9329) > Happy path: an authenticated mobile user opens the Account screen and re-enables (resumes) receiving invites. Steps: Account settings (paused state) → Confirm resume modal → Success acknowledgement → Account settings (active state) **Picnic: Troubleshooting hub** (ios, 2 screens, Productivity, Utility) [Refero](https://refero.design/flows/11510) > A mobile Troubleshooting hub that exposes diagnostic and recovery actions. Steps: Troubleshooting overview → Copy confirmation **Roots: Resume global blocking** (ios, 3 screens, Productivity) [Refero](https://refero.design/flows/5756) > User navigates the Blocking hub (assumed logged-in state). The screen initially shows a global paused state with a prominent banner and a primary CTA to resume blocking, followed by lists of rules grouped as Currently Paused, Upcoming, and Recommended Blocking. Tapping the resume CTA opens a confirmation bottom sheet. Confirming resumes blocking and updates the main screen to reflect an active/currently blocking state. Steps: Blocking (Paused) Overview → Confirm Resume Bottom Sheet → Blocking (Active) Updated **FocusFlight: Pause and Give Up Confirmation** (ios, 6 screens, Productivity) [Refero](https://refero.design/flows/8607) > Happy-path (attempted): User is viewing a map-based navigation/Guidance screen showing route, time remaining and distance remaining. They pause guidance which opens a minimal overlay. From there they open an options menu that lists actions (Give up, Airplane Mode, Keep Screen Awake). The user selects Give up which requires a long press to confirm. A hold-to-confirm control appears and fills as the user holds. In this recording the confirmation fails and the final screen shows a failure message with a retry-oriented CTA. Steps: Map - Guidance Active → Paused Overlay → Options List → Hold-to-Give-Up Hint → Hold Progress / Confirmation → Failure Feedback **Volv: Email password recovery** (ios, 4 screens, AI Tool, News) [Refero](https://refero.design/flows/2153) > This is a small mobile auth sequence that starts at an email login screen and branches to a ‘Forgot password’ recovery path. **Breathwrk: Password Recovery via Email** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/7741) > Happy-path: A signed-out user on the Login screen chooses the Reset Password path, enters their email, and receives an on-screen confirmation that reset instructions were sent. Steps: Login → Reset Password — ready → Enter Email → Confirmation **TikTok: Password reset via email** (ios, 5 screens, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/2746) > Happy-path summary: A signed-out user attempts to log in, encounters an account error, and uses the "Forgot password?" flow to request a reset via email. The user is presented with a choice of reset methods, selects Email, types their email address, submits it, and is prompted to enter a 6-digit verification code sent to that address. Steps: Enter password (error) → Choose reset method → Reset - enter email → Reset - submit email → Verify email - enter code **SSENSE: Password recovery and reset** (ios, 10 screens, Marketplace) [Refero](https://refero.design/flows/7113) > Happy path for a user who cannot log in and needs to reset their password. Entry: user attempts sign in (logged-out state). The UI shows a login error, offers a "Forgot Password" action, sends a reset email, the user follows the email link which opens a web reset form, they enter a new password (with strength meter and validation), submit it, and are redirected into their signed-in account page (Account Details). Steps: Login attempt with error → Forgot password form → Sending state → Email sent confirmation → Open reset link (reset form) → Password entry with validation → Progressive strength feedback → Submit new password → Password reset success (signed-in) → Account details (post-reset) **Synthesia.io: Password Recovery** (web, 7 screens, AI Tool) [Refero](https://refero.design/flows/13712) > The user recovers access to their Synthesia workspace after an unsuccessful sign-in attempt. Starting from the sign-in modal, they open password recovery, submit their email address, enter and confirm a new password, correct a password mismatch validation error, and continue into the authenticated workspace dashboard. The flow assumes the user has an existing account and access to the submitted email address. **Grammarly: Password recovery (email)** (web, 4 screens, Paperwork) [Refero](https://refero.design/flows/1867) > Happy path: A signed-out user attempts to sign in, encounters an authentication error, navigates to the password-reset form, submits their email, and receives an on-screen confirmation that reset instructions will be sent if the email is associated with an account. Steps: Sign in (authentication error) → Password Reset (form shown) → Password Reset (email entered) → Password Reset (confirmation) **Acctual: Manage contact record** (web, 4 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8849) > A desktop contact management sequence where a signed-in user reviews the list of saved contacts and opens a right-side slide-over to edit a contact's details and attempt to archive it. Steps: Contacts list → Open contact editor (slide-over) → Archive confirmation → Archive failed (error toast) **Patreon: Email Code Sign-In** (web, 5 screens, Social Networking, Fundraising) [Refero](https://refero.design/flows/1268) > Happy path: a signed-out user attempts to sign in with an email and password, receives a password mismatch, then switches to a temporary (one-time) email code. The user requests the code, checks their email, enters the code on the verification screen, and (on success) would be signed in and redirected to the app. **Replo: Password Recovery to Projects** (web, 8 screens, Design & Media Editing, Business) [Refero](https://refero.design/flows/2053) > Happy-path password recovery and immediate entry into the user's Projects area. Steps: Sign In (error shown) → Reset Password (empty) → Reset Password (email entered) → Reset Confirmation → Enter New Password (initial) → Enter New Password (confirmation) → Post-auth Loading → Projects Dashboard (signed in) **Jasper AI: Restore Deleted Content** (web, 3 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/5311) > Happy path: A logged-in user opens the Trash view (accessible from the left navigation) to restore a previously deleted content item. The Trash screen shows a banner that items are permanently deleted after 30 days, a searchable list of trashed items with metadata (name, creator avatar/initials, deletion date), and per-item actions (ellipsis). The user uses the item actions to restore an item. After restoring, the item is removed from the Trash list and a success toast appears. Steps: Trash list overview → Trigger restore action → Restore confirmation **Dropbox: Resend Member Invite** (web, 3 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/13061) > This is a members administration screen in an admin console, focused on managing invited users. The flow starts on the Members list with an invited member already visible and filtered by invited status. The user can view member status, manage access, use row actions such as Resend invite, and receive confirmation when the invitation is resent. Assumptions: the user is already signed in with admin permissions and is viewing the team members section on desktop. Steps: Members List → Invitation Action Selected → Invite Sent Confirmation ##### Task: expense management receipts **Cashcaded: Receipt Sharing Preparation** (ios, 2 screens, Finance) [Refero](https://refero.design/flows/14136) > The user opens a Share sheet for a purchase record and can copy the displayed social-media tag before sharing the receipt image. The flow begins on a mobile device with a share modal already open over a purchase detail screen and ends with confirmation that the tag was copied. The user appears to be viewing an existing receipt or transaction; no login, permissions, or account state are visible. The final Share button remains available but is not shown as selected. Steps: Share Preview → Copy Confirmation **Loowner: Add new item with photo and receipt** (ios, 13 screens, Marketplace) [Refero](https://refero.design/flows/6597) > Happy-path for adding a new owned item from an empty-states entry point. User starts on the "All your stuff" empty screen and taps the primary CTA to add an item. They pick the purchase date, enter the cost using a large numeric keypad, add optional photo(s)/receipt (invoking camera/gallery and permission prompt if needed), enter a name, then save. After save the app shows a success confirmation and returns to home where the new item appears in Quick access and in the item detail view. Steps: Empty state / Entry → New Item — Choose purchase date → New Item — Enter cost (empty) → New Item — Enter cost (value entered) → New Item — Name & images screen (keyboard open) → New Item — Confirmed name preview → Add photo — camera permission → Add photo — choose source → Image attached preview → Save item → Home updated — Quick access → Open item detail → Item detail — edit / add more **Runey: Receipt Creation** (web, 7 screens, Business, Productivity, Project Management) [Refero](https://refero.design/flows/11707) > Happy path begins on the Invoices dashboard, where the user opens the New menu and chooses to create a receipt. The user enters the receipt editor, selects a client, picks or confirms dates, adds a line item, reviews payment details, and creates the receipt. The flow ends on a completed receipt view with a success confirmation. Steps: Invoices Dashboard → Choose Receipt Option → Receipt Editor → Client Selection → Date Selection → Line Item Entry → Created Receipt **GlossGenius: Send sale receipt to client** (web, 3 screens, Business) [Refero](https://refero.design/flows/9964) > Happy path: A logged-in user (desktop) opens the Clients section, selects a client with a recorded sale, views the sale details in a right-hand panel, and sends a receipt to the client via a contact method (text or email). Steps: Clients list & Sale details → Choose contact method → Receipt sent confirmation ##### Task: export download data **Comet: Safari data import** (ios, 3 screens, AI Tool, Utility) [Refero](https://refero.design/flows/10952) > Happy path: a user follows step-by-step instructions to export bookmarks, history and passwords from Safari on iOS, taps the in‑app button to open the iOS Files picker, selects the exported .zip, and sees a confirmation screen with counts for each imported category. Steps: Export instructions → Files picker — select .zip → Import complete confirmation **Steps: Download Steps QR** (ios, 2 screens, Health & Wellness) [Refero](https://refero.design/flows/7523) > User opens a settings area (background shows items like Privacy Policy, Leave a Review, Restart Onboarding, Re-sync Apple Health data). From there a bottom-sheet modal titled "Scan to download Steps" appears containing a QR code and two action buttons: "Copy QR Code" and "Copy Link". Steps: Show QR bottom sheet → Copy link confirmation **Wrike: Resources export configuration** (web, 2 screens, Business, Productivity, Project Management, Collaboration Tools, Paperwork, Scheduling Tools) [Refero](https://refero.design/flows/1519) > Happy-path: A signed-in user viewing the project/space "Refero" navigates to the Resources view (a timeline-style resource allocation grid showing assignees and hours). From this view the user opens an Export dialog to configure export options (date range, effort units, mode, time granularity, and whether to include only filtered tasks or all tasks) and triggers an export. Steps: Resources overview (timeline) → Export resources view modal **Riverside: Recording Asset Export** (web, 2 screens, Productivity, Entertainment, Collaboration Tools) [Refero](https://refero.design/flows/4672) > User views a project's recorded session assets (audio/video tracks and transcript) and chooses export or download options for individual tracks or all assets. Steps: Recording files overview → Item overflow — transcript & subtitle options **Parallel: Executive Profile Enrichment** (web, 4 screens, AI Tool) [Refero](https://refero.design/flows/9067) > User starts on an Enrichment session page that is processing a list of executive inputs (names + companies) and populating structured output columns (education, job_title, etc.). Steps: Session Table — Processing → Open Session Settings → Inspect Expanded Cell → Review Completed Results **Chargetrip: Workspace analytics overview** (web, 3 screens, Development Tools, Map & Navigation, Transport) [Refero](https://refero.design/flows/9267) > Summary: The flow is the Analytics area for a workspace. Steps: Metrics grid (default charts) → Interact with charts and change date range → List / table view and export **Zendesk: Talk configuration and reporting** (web, 5 screens, Business) [Refero](https://refero.design/flows/1401) > Happy path: An admin configures the Talk Professional product by adjusting call routing and agent settings, then manages greetings (create/edit/assign) and reviews historical usage and charges. Steps: Talk settings → Greetings list → Greeting editor (inline) → History overview → Usage charges and call history **Aboard: Create and view Time-off report** (web, 5 screens, Collaboration Tools, Paperwork) [Refero](https://refero.design/flows/814) > Happy path: a user in Analytics > Reports creates a new Time-off report, names it (example: "September"), opens the report detail which shows a tabular list of recorded time-off entries, and then returns to the Reports list where the new report appears under the Time-off category. Steps: Reports overview (Analytics) → Open New time-off report modal → Enter report name → Report detail — table view → Reports list updated **Family: Connect Wallet Theme Builder** (web, 35 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/6171) > Happy path: User arrives at the 'Try It Out' documentation/demo and opens the theme selector to choose or create a custom theme for the Connect Wallet modal. They select "Build Your Own" which opens the Theme Builder. In the Theme Builder they iterate on fonts, text color, and other general settings while seeing a live mobile preview of the Connect Wallet modal. The user toggles demo settings (preview URL, language, disclaimer), downloads the theme JSON, then opens the advanced editor to fine‑tune CSS variables while keeping the preview visible. Steps: Try It Out landing → Open theme presets → Theme Builder initial view → Components preview grid → Open font selector → Alternate font list → Apply decorative font → Switch to another font (Monrope) → Open text color picker → Apply new text color → Open Demo Settings → Enable disclaimer text → Mobile preview with disclaimer → Prepare to export theme → Download confirmation toast → Open advanced editor → Inspect theme variables → Return to simple editor affordance → Edit additional variables → Final preview and code sync **Parallel: Profile enrichment setup** (web, 16 screens, AI Tool) [Refero](https://refero.design/flows/9064) > Happy-path: a logged-in user opens the Enrichment area, picks a pre-built template from the Library (or starts from scratch / uploads CSV), loads sample input data into a spreadsheet-like editor, configures output columns (manually or via Ask AI), starts an enrichment run and monitors progress, then reviews and downloads the enriched results. Steps: Enrichment start screen → Enrichment library → Spreadsheet editor with sample inputs → Add output columns → Column header actions → Reorder/rename columns → Open Ask AI modal (blank) → Edit AI prompt → Columns updated/prepared → Enqueue/run session → Progress indicator → Results beginning to populate → Table options and session actions → Review partially populated rows → All rows populated (final results) → Export or use results via API **Aboard: Create Device Report** (web, 5 screens, Collaboration Tools, Paperwork) [Refero](https://refero.design/flows/817) > Happy path: a logged-in user with access to Analytics > Reports creates a new Devices report, names it, opens the report editor to view/add rows, then returns to the Reports list where the new report appears. Steps: Reports list (Devices empty) → New device report modal (empty) → New device report modal (name entered) → Report editor (table view) → Reports list (new report visible) **Typeform: Account deletion with email MFA** (web, 8 screens, Productivity, Communication Tools) [Refero](https://refero.design/flows/5601) > Happy path: A logged-in user navigates to their account settings, chooses to delete their account, is required to re-authenticate and set up/confirm an email-based multi-factor method, enters a verification code sent to their email, confirms permanent deletion, and sees a final success/feedback screen. **The Washington Post: Birthday front-page lookup** (web, 5 screens, Magazines & Newspapers, News) [Refero](https://refero.design/flows/11200) > Happy path: A signed-in user opens their My Post profile and uses the "What made the front page the day you were born?" feature to enter a birthdate, submit it, and view the scanned front page for that date. Steps: My Post dashboard (entry) → Birthday date entry (empty) → Birthday entry confirmed → Loading / date header → Front-page result with export options **Linear: Members management** (web, 3 screens, Business, Productivity, Project Management, Collaboration Tools) [Refero](https://refero.design/flows/6642) > Admin-style members management interface showing a list of workspace/team members with search and status filters, plus actions to export or invite. Entry: user lands on Members screen (already signed in and has member-management privileges because Invite and Export controls are visible). Happy path: user scans the All members view (Active + Invited), opens the status filter, selects "Members" to show only active users, and the list updates to hide invited entries. Exit: filtered members list displayed (or user can proceed to Invite or Export). Assumptions: user is authenticated and has permission to view/manage members (Invite and Export CSV buttons available). Device: desktop/larger screen. Steps: Members list — All → Status filter dropdown → Members list — Members filter applied **Teal: Signup to Resume Builder Onboarding** (web, 34 screens, Productivity, Job Search) [Refero](https://refero.design/flows/2682) > Happy-path conversion from marketing landing to account creation, quick resume import, and first-time guided tour inside the resume builder. Steps: Marketing Hero → Signup Options → Email Prefill → Name Entry → Password Entry → Submit Signup → Welcome Modal → Resume Import - File → Resume Import - LinkedIn → Resume Import - Paste → Import Success / CTA → Resume Builder - First Tour Overlay → Tour - Preview Mode → Tour - Guidance Mode → Tour - Analysis Mode → Tour - Job Matching → Tour - Design Mode → Tour - Cover Letter → Tour - Export & Menu → Tour End / Resume Editor Ready ##### Task: eye test vision **Nextdoor: Rate Ask Answer** (web, 2 screens, Communication Tools) [Refero](https://refero.design/flows/11439) > A logged-in neighbor reviews an AI-generated answer in the Ask section and rates it as helpful. The user starts on an answer about places to learn hobbies, selects the thumbs-up feedback control, and receives a confirmation toast. The answer remains visible, and the selected feedback state is highlighted. Steps: Review Answer → Feedback Confirmed **Preply: Pre-class device check** (web, 6 screens, Education) [Refero](https://refero.design/flows/9214) > Happy path: A learner arrives at a scheduled lesson landing screen and runs the built-in audio & video diagnostics before entering the classroom. From the landing area the user opens a modal that guides them to record a short (5-second) test video, watches a recording progress indicator, then confirms two simple checks: can they see themselves and can they hear the playback. The user can rerun the test and then view the consolidated "Test your devices" panel which shows device check items (Camera and mic, Network and device). Steps: Lesson landing with test prompt → Start test modal → Recording in progress → Playback – visual confirmation → Playback – audio confirmation → Device checklist overview **Duolingo: Practice Word Question** (web, 8 screens, Education) [Refero](https://refero.design/flows/11827) > Happy-path practice sequence in the Duolingo English Test practice area. The user starts on the Practice dashboard, launches a practice test, waits through a loading state, answers a timed word-recognition question, receives either incorrect or correct feedback, advances through the question, and returns to the Practice dashboard with progress updated. Steps: Practice Dashboard → Loading State → Word Decision Question → Incorrect Feedback → Next Question → Correct Feedback → Practice Complete → Practice Dashboard Return **TikTok: Visual product search camera** (ios, 3 screens, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/2973) > Mobile visual-search entry point that lets a user take a photo or upload an existing image to find visually similar products. Entry: user opens the visual search camera view. Primary actions: take a photo (shutter), upload from photo library (Upload), or read contextual help (Help). Secondary controls include Flash toggle and a Close (X). Help appears as a two-page bottom sheet with short guidance for taking or uploading images. Exit: after capturing or selecting an image the flow should proceed to a results/search step (not shown). Assumptions: device has camera/photo permissions, product is shown centered on a plain background, user is on a mobile device. Steps: Camera / Visual Search → Help — Take a photo → Help — Upload a photo **Chance AI: Visual Product Lookup** (ios, 7 screens, AI Tool) [Refero](https://refero.design/flows/7365) > Happy-path: A user discovers the "Search by Seeing" feature on the app home feed, opens the in-app camera experience (AI Snaps), selects a product recognizer snap, points the camera at a product, captures an image, and is taken to a generated result page containing a review, photo gallery, price cards, expert commentary and suggested follow-up questions. Entry: Home feed. Exit: Detailed result/article card with share and follow-up options. Assumptions: User is on a mobile device, has camera access, and has network connectivity. No special permissions or account state are visible/required in the screenshots. Steps: Home — Search by Seeing → Camera — AI Snap Selector → Camera — Capture → Results — Article Header → Results — Visual Gallery & Price Cards → Results — Expert Review Content → Results — Actions & Suggested Questions **Podimo: Search Episodes in Show** (ios, 4 screens, Entertainment) [Refero](https://refero.design/flows/12262) > Mobile show episode list with an in-show search experience. Entry is from a show’s episode list, where the user taps the search icon to open a dedicated search screen for that show. The user types a query in the episode-title field and the list filters in real time. Results count updates as matches change, and an empty state appears when no episodes match. Steps: Episode List → Search Entry → No Results → Filtered Results **Apple Maps: Place Details and Street View** (web, 3 screens, Map & Navigation) [Refero](https://refero.design/flows/12472) > A user opens a specific place listing in Maps, reviews its details in a side panel, and then switches to Street View to inspect the venue’s exterior and surrounding street context. The entry point is a selected place marker or search result on the map; the exit point is the place shown in Street View with the option to close or continue exploring. Steps: Place Details Opened → Street View Preview → Return to Map Context **Rork: Analytics Review and Date Range Change** (web, 4 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12160) > Happy path for viewing analytics in a dark-mode web dashboard. The user is already signed in and on the Analytics section of the workspace. They review activity metrics, change the date range using a dropdown, briefly see a loading state while the data refreshes, and then return to the updated analytics view. The screen also shows a revenue integration card with a connect action, but no connection is completed in this sequence. Steps: Analytics Overview → Date Range Menu → Loading State → Updated Analytics View ##### Task: feedback rating review **Apple Books: Book Reviews and Rating** (ios, 5 screens, Books) [Refero](https://refero.design/flows/11920) > This is a book reviews and rating interaction. The entry point is the Customer Reviews screen, where the user can view the overall rating summary, browse existing reviews, sort reviews, or start writing/editing their own review. Tapping the star row opens the review editor. In the editor, the user can set a star rating, enter a title and written review, then save via the checkmark. The user can also cancel with the close button or remove an existing rating/review using the delete action. Assumptions: the user is already on the book’s reviews section and has permission to rate/review the book. Steps: Reviews Overview → Rated State → Start Review Editor → Writing Review → Completed Review **Apple Books: Customer Reviews** (ios, 3 screens, Books) [Refero](https://refero.design/flows/11914) > A mobile reviews screen for a book, entered from a previous book/details context. The top area shows an overall rating summary with star distribution and total rating count. Below it is a reviews section with a Sort control that opens a sorting menu. The user can browse reviews and change ordering without leaving the screen. Steps: Reviews Overview → Sort Menu → Sorted Reviews **pushr: Positive Rating Response** (ios, 3 screens, Health & Wellness, Fitness) [Refero](https://refero.design/flows/13792) > A user starts in the Settings screen and selects “rate us” from the Feedback section. A bottom-sheet prompt asks whether they are enjoying Push. Selecting “yes” displays a confirmation message with a “continue” button. The screenshots end at this confirmation step; the destination of “continue” is not shown. The user is assumed to be on a mobile device and able to access Settings. Steps: Settings Feedback → Satisfaction Prompt → Feedback Confirmation **District: Restaurant Reviews and Ratings** (ios, 2 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12860) > This screen sequence shows a restaurant’s review section for “G Punjab Fine Dine & Bar.” The user enters from the venue’s ratings area and can inspect the overall score, rating distribution, category scores, sort/filter reviews, read individual comments, and start writing a new review. Assumptions: the user is on a mobile device and is viewing a public venue profile; no login state is shown, though leaving a review may require it. The flow ends at the review list with the persistent “Leave a review” action available. **District: Restaurant Review Submission** (ios, 17 screens, Entertainment, Food & Drink, Lifestyle, Travel _… 173477 more characters in the zip._ ### research/experience/REFERO-FLOWS-2.md `research/experience/REFERO-FLOWS-2.md` · sha256 `16141d6c58761db48570ce70a34a4ffa345a87c6093b7bc52dc74109be8dfec2` #### Customer experience (all applications) research: real product flows from Refero, part 2 of 2 **What this is.** Real screens and user flows from existing products, recorded by Refero (a library of real app and website screenshots, each journey described step by step). They show the floor (what users already expect) and good customer experience from any industry worth borrowing. **Statement class.** Every flow is an OBSERVATION by Refero of a real product, searched 2026-10-31. Descriptions are Refero's wording. Product names belong to their owners. **How it was gathered.** Semantic searches for this application's tasks on iOS and web. Results from other industries are kept on purpose: they are often the best customer experience to learn from. Refero does not hold many incumbents (and few government, enterprise or NHS systems), so this is a sample, not a census. **Products in this sample:** Airbnb (34), WhatsApp (33), District (29), Dropbox (26), PayPal (22), Instagram (20), Fresha (19), Lovable (18), Mercury (15), komoot (15), TikTok (15), Meta AI (15), ElevenLabs (14), Linear (14), ElevenMusic (14), X (13), ElevenReader (13), On (13), Nextdoor (13), Omio (13), Runey (11), Riverside (11), Willow (10), Faire (10), Cushion (10), Tinder (10), Artsy (10), Fernand (10), Pinterest (10), Substack (10), Structured (10), Intercom (9), GoFundMe (9), YouTube Music (9), IKEA (9), Napper (9), Tripsy (9), Rork (9), XChat (9), Missive (9), Ada (8), Wabi (8), DailyArt (8), HYPE (8), Sora (8), Anam (8), Foodllama (8), Zendesk (8), Voicenotes (8), Podimo (8), OpenAI (8), Trip (8), ChatGPT (8), Blackbird (7), Phantom (7), Revolut (7), Comet (7), Train Fitness (7), Google Gemini (7), Apple Books (7), Twitch (7), Family (7), Binge (7), Wealthsimple (7), Raycast (7), Doctronic AI (7), Longevity Deck (7), Leonardo.Ai (7), Melius (7), Manus (7), Asos (7), Brilliant (7), Maze (7), YouTube (7), Aboard (7), SSENSE (7), Threads (7), Grok (6), Givingli (6), Fingerprint (6), Adobe Photoshop (6), Shopify (6), Mapbox (6), Instacart (6), Duolingo (6), Chance AI (6), Vinyls (6), Handshake (6), Hello Ivy (6), Vimeo (6), Synthesia.io (6), Canva (6), DailyArt Magazine (6), Сontra (6), Medium (6), Tango (6), HBO Max (6), Steps (6), Acctual (5), Google Translate (5), TIDAL (5), Granola (5), Grammarly (5), Shop (5), Loowner (5), Teal (5), GlossGenius (5), Reown (5), Ground News (5), Cohere (5), Poppy (5), Mozi (5), Apple Maps (5), Exoplan (5), Care (5), Steel (5), Roots (5), Rox (5), Bond (5), Jace AI (5), Skiff (5), OKURI (5), Figma (5), Trip Way (5), Bezi (5), Dropset (5), Genie (5), LEGO® Builder (5), BeReal. (5), Reeder. (5), ABY Journal (5), Daylish (5), Jasper AI (5), Verse (5), RAD Weather (5), Mela (5), UGLYCASH (4), Fourmula.ai (4), Fuse (4), Introspect (4), Waterllama (4), The Outsiders (4), Cake Equity (4), Pi (4), Pinkllama (4), Linktree (4), Apple TV (4), Product Hunt (4), Chargetrip (4), fal (4), ALO (4), Nike (4), Factory (4), Imprint (4), Mindllama (4), Matter (4), Netflix (4), Wrike (4), Microsoft Teams (4), Cursor (4), Whereby (4), TIDE (4), Glorify (4), Kickstarter (4), Microsoft Copilot (4), 222 (4), Replo (4), Airtable (4), Spotify (4), LinkedIn (4), Delphi (4), Wispr Flow (4), Stellate (4), Parallel (4), DoorDash (4), Square (3), Nooka (3), Turf (3), TravelPerk (3), Spoil me (3), Literal.club (3), AVNIER (3), Copperx (3), Gumroad (3), UNHINGED (3), Carbon (3), Understory (3), Claude (3), Kin (3), Appwrite (3), n8n (3), Kit (3), Dribbble (3), Digg (3), Saint Laurent (3), Telegram (3), Remy (3), Preply (3), The Leap (3), MonoDesk (3), Miles (3), TidyCal (3), Programa (3), Polar (3), Exa (3), Acne Studios (3), West Elm (3), Microsoft (3), Reclaim AI (3), Apple Invites (3), Daze (3), Calendly (3), PocketShelf (3), Joi (3), Canopi (3), Runway (3), Aesop (3), Tripadvisor (3), ManyChat (3), Hashnode (3), Typeform (3), Peak Performance (3), Stardust (3), Visitors (3), Mocha (3), Spline (3), ZARA (3), The Washington Post (3), Julienne (3), Navan (3), adidas (3), Box Box Club (3), Luminar (3), pushr (3), Pop Site (3), MWM (3), pillowtalk (3), Abode (3), Wise (2), Cashcaded (2), Pool (2), Tidal (2), Nudea (2), Kann (2), Spyglass (2), Haptic (2), Weekrise (2), Shazam (2), Cofounder (2), Sunsama (2), Prose (2), time2book (2), Zara Home (2), ROVR (2), Dopper (2), Loom (2), Days (2), Lemni (2), Memotron (2), Epidemic Sound (2), Volv (2), SoundCloud (2), FocusPomo (2), Agora 2 (2), Tella (2), Modyfi (2), Descript (2), Cal (2), Luminary (2), Pattern Brands (2), Acuity Scheduling (2), Fourthwall (2), Onefinestay (2), Breathwrk (2), raum. (2), Showcase (2), PamPam (2), Ozone (2), Polaroid (2), Picnic (2), FocusFlight (2), Patreon (2), Playground (2), Enode (2), Plane Finder (2), Dub (2), Ballpark (2), Drinkit (2), Peerlist (2), Navia (2), Rive (2), [untitled] (2), Stripe (2), Moises (2), Expedia (2), Telescope (2), voilà (2), Sky Guide (2), Endless Tools (2), Urban Outfitters (2), Brink (2), Clipchamp (2), Exactly.ai (1), Apple Stocks (1), Denim (1), Allbirds (1), Butter (1), Farfetch (1), Make (1), Column (1), Yelp (1), Stemz (1), Clime (1), SavvyCal (1), Homerun (1), Shuttle (1), Play (1), Gizmo (1), MD Vinyl (1), Savee (1), Vocabulary (1), Moonlitt (1), Endless (1), Soundtrap (1), TwelveLabs (1), WURRD (1), Lummi (1), ARKET (1), Douze (1), ten ten (1), MyMind (1), Portrait (1), Womp (1), Rarible (1), VSCO (1), Skyscanner (1), Macaw (1), Touch Time (1), (Not Boring) Weather (1), Unwind (1), Moonly (1), (Not Boring) Habits (1), Miro (1), Prime Video (1), Pitch (1), Doji (1), Athleats (1), Scape (1), Apple Podcasts (1), B150 (1), Stocktwits (1), Project Indigo (1), StrideGate (1), tone (1), Queue (1), WishWish (1), Craft (1), Open (1), Marathon (1), Oppie (1), Weather mini (1), pico cam (1), Pryzm (1), one year (1), (Not Boring) Camera (1), ShareWillow (1), Uber (1), Resend (1), LEGO (1), Volkshotel (1), CapWords (1), Orbit (1), Variant (1), Apple Games (1), Excalidraw (1), Dona (1), GO Club (1), Neuecast (1), PlayStation (1), Wittl (1), Hello Klean (1), Xbox (1), Checker (1), Tilt (1), Slite (1), Subo (1), CocoonWeaver (1), Uber Eats (1), New Balance (1) Tasks in this part: log in passkey biometric to workflow automation builder. Part 1 is REFERO-FLOWS-1.md. ##### Task: log in passkey biometric **XChat: Username sign-in to chat** (ios, 7 screens, Communication Tools, Social Networking) [Refero](https://refero.design/flows/11644) > Happy-path sign-in flow that takes an unauthenticated mobile user from a branded splash screen through selection of a sign-in method, optional username/password entry, an optional passkey (Face ID) enrollment prompt, and finally into the chat inbox. Steps: Splash with primary CTA → Sign-in method selection → Enter username → Enter password (input) → Enter password (confirmation) → Passkey / Face ID enrollment prompt → Chat inbox (authenticated) **Breathwrk: Onboard & Sign In** (ios, 5 screens, Health & Wellness) [Refero](https://refero.design/flows/7744) > Happy-path flow for a new user who opens Breathwrk, reviews the marketing welcome screen, chooses a sign-in method (Sign in with Apple shown), authenticates using the platform modal and biometric, and lands on the app home screen with an option to start a short onboarding tutorial. Steps: Welcome / Landing → Authentication Choice → Sign in with Apple Sheet → Biometric / Signing In → Authenticated Home / Quickstart **XChat: Password recovery and reset** (ios, 8 screens, Communication Tools, Social Networking) [Refero](https://refero.design/flows/11645) > Happy path: A user tries to sign in, encounters an incorrect password error, uses the "Forgot password" path to request a recovery code by email, retrieves a six-digit code from their inbox, enters it in the app, then sets a new password. After the password is accepted the app offers an optional passkey (Face ID) setup. Steps: Enter password → Choose recovery method → Code entry (waiting state) → Check email for code → Code entry (filled) → Reset your password (enter) → Reset confirmation (continue) → Optional passkey offer **Comet: Sign in from Settings** (ios, 4 screens, AI Tool, Utility) [Refero](https://refero.design/flows/10953) > Happy path: A signed-out user opens the app settings and taps the top "Sign in" row. A bottom-sheet sign-in modal from an identity provider (Perplexity) appears offering multiple authentication options (Apple, Google, Email, existing account). The user picks an option (in the screenshots they continue with Google), the modal shows a loading state, and after successful authentication the settings screen updates to a signed-in view that displays the user avatar, name and email plus an "Account settings" button and a new "Sync" row. Entry: settings screen (signed-out). Exit: settings screen (signed-in, account details visible). Assumptions: mobile device (iPhone-like status bar visible), user initially signed out, network available, sign-in handled via third-party providers in a bottom-sheet modal. Steps: Settings — Signed out → Sign-in modal — Provider selection → Authentication — Loading state → Settings — Signed in (profile shown) **Fresha: Sign in to personal profile** (ios, 4 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4980) > Happy path: a signed-out mobile user opens the account modal (likely via the bottom navigation profile icon), chooses a sign-in method or types an email, confirms identity with a password, and lands on their personal profile with wallet balance and account options. Steps: Sign in or sign up (entry modal) → Email entry (keyboard visible) → Password entry (Welcome back) → Authenticated profile landing **ChatGPT: Sign-in to Chat** (ios, 8 screens, AI Tool) [Refero](https://refero.design/flows/7040) > Happy-path sign-in using Apple SSO that falls back to email/password. Entry: user launches the ChatGPT mobile app and is presented with a bottom sheet of sign-in options. Exit: user is authenticated and lands on the ChatGPT home screen with the chat composer available. Assumptions: user is on an iOS device, not currently signed in, and the app uses system Apple sign-in which opens an OS-level permission prompt then an in-app browser/webview (auth.openai.com) to complete authentication. Alternative sign-in buttons (Google, Microsoft, phone, manual sign up) are present as visible alternatives. Steps: Welcome / Sign-in Options → System Authorization Prompt → Auth Landing — Choose / Enter Credentials → Email Input → Password Entry (empty) → Password Entry (filled) → Authentication / Loading → Authenticated Home — Chat Composer **On: Sign in and enter home** (ios, 4 screens, Marketplace) [Refero](https://refero.design/flows/9610) > Happy path: an unauthenticated mobile user lands on a branded welcome screen and chooses to sign in. Tapping "Log in" opens a rounded top modal (bottom sheet) that contains email/password fields, password visibility toggle, a primary "Continue" CTA, "Forgot password?" link and social sign-in buttons (Google, Apple). The user fills credentials and taps Continue; on successful authentication the app navigates to a personalized home feed that greets the user by name and shows content cards and a bottom navigation bar. Steps: Welcome / Entry → Login modal - empty state → Login modal - credentials entered → Authenticated home feed **Vinyls: Onboarding and Sign-In** (ios, 7 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/8913) > Happy-path onboarding and authentication flow for a vinyl-collection app on an iPhone. Entry: first-run onboarding / welcome screen (user is logged out). The user can swipe through feature slides or choose a sign-in method using the prominent CTAs: Sign In With Apple or Login with Code. Selecting Login with Code opens a bottom sheet that offers QR-based sign-in (Read QR Code). Selecting Sign In With Apple launches the system Apple authentication sheet. After successful auth the app shows a welcome message and the main home screen with import and add-vinyl actions. Assumptions: iOS device, camera permission required for QR scanning, user has no active session. Steps: Welcome / Sign-in entry → Onboarding — Apple Watch feature → Onboarding — Preview Tracks → Onboarding — Calibrate Turntable → Login with Code — bottom sheet → Sign in with Apple — system auth sheet → Post-auth Home / Dashboard **HYPE: Hype App Sign In Entry** (ios, 6 screens, News, Lifestyle) [Refero](https://refero.design/flows/12242) > Happy path starts on an unauthenticated welcome screen for the Hype App. The user can choose to log in, sign up, or continue as a guest. If they tap Login, the app shows an iOS account permission prompt, then opens the HYPE account login page in a browser-style web view. After entering credentials and submitting, the app returns to the main HYPE content feed. The flow assumes a mobile iPhone device and a user who is either signing in with an existing account or opting out via guest access; no logged-in state is assumed at entry. Steps: Welcome Screen → Sign-In Permission → Login Form Start → Login Form Filled → Authentication Loading → Home Feed **Stardust: Apple Sign-in to Cycle Dashboard** (ios, 10 screens, Health & Wellness) [Refero](https://refero.design/flows/5905) > Happy-path sign-in and launch flow for a mobile cycle-tracking product. Entry: user opens the app (cold start). Primary entry points shown: (A) Sign up with Apple button on the splash, (B) Sign in with an existing account link that opens a login bottom sheet. The flow demonstrates choosing "Continue with Apple", interacting with the iOS Sign in with Apple modal, showing in-app and system progress states, then a branded animation before landing on the main dashboard (cycle view). Assumptions: user is on an iPhone, not previously signed in to this app session, standard iOS Sign in with Apple integration available, network connectivity present, and the app accepts Apple as a sign-in provider. Steps: Splash — Sign up / Sign in entry → Login bottom sheet → iOS Sign in with Apple modal (account chooser) → Apple modal — Signing in state → In-app 'Completing your sign-in' overlay → In-app 'Signing you in' confirmation → Brand reveal / transition animation → Initial authenticated home (stars background) → Cycle overview — Circular widget → Full dashboard / main navigation **Brilliant: Sign-in and entry to home** (ios, 4 screens, Education) [Refero](https://refero.design/flows/4341) > Happy-path: user opens the app on a mobile device (logged out) and is presented with a promotional onboarding/welcome screen with a primary Continue CTA and a secondary Log in link. Tapping Log in navigates to a sign-in screen that offers three social sign-in buttons (Apple, Google, Facebook) and an email/password form. The user focuses the email field, enters credentials and taps the Log in button. Successful authentication transitions the user to the authenticated home feed (personalized “For you” content with cards and bottom navigation). Steps: Onboarding / Welcome → Sign-in options (keyboard visible) → Email entry and credential submission → Authenticated Home / Feed **Nooka: Sign-in to Join Live Session** (ios, 6 screens, AI Tool, Education, Books) [Refero](https://refero.design/flows/8124) > Happy path: an unauthenticated user launches the app, is presented with a branded welcome screen offering social and email sign-in options. Selecting email/sign-in navigates to a Login screen where the user can enter credentials or continue with Apple/Google. After submitting credentials the UI shows a loading state and then returns to the app home content. On the content screen the user is prompted to "Tap to Join" a live session (open mic / transcript). Tapping joins the session and reveals live audio UI and playback controls. Steps: Welcome / Sign-in Options → Email Login (initial) → Email Login (email entered) → Authentication In-Progress → Content — Tap to Join Overlay → Live Session / Playback UI **Raycast: Sign-in to Main** (ios, 6 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/10201) > Happy path: A logged-out mobile user opens the product and is presented with a branded welcome screen with two primary calls-to-action: Log in or Create Account. Tapping Log in navigates to an authentication screen that offers email-based login (email + optional password / magic link) and three social/OAuth providers (Apple, GitHub, Google). After entering an email and tapping Log in the UI shows a loading/authenticating state. On success the user is shown an onboarding card that highlights “Raycast AI”; tapping Next dismisses the card and takes the user to the main/home view containing a search field, quick tool icons (AI, Notes, Quicklinks, Snippets) and a favorites list with a persistent quick input bar. Assumptions: user is on a mobile device, starting logged out, and the device supports the shown OAuth providers. Steps: Welcome → Authentication Options → Email Entry & Log in → Authenticating / Loading → Onboarding Card (Raycast AI) → Main / Home (Search & Favorites) **Luminary: Onboarding to Authenticated Home** (ios, 4 screens, Entertainment) [Refero](https://refero.design/flows/4733) > Happy path: A logged-out mobile user opens the app, sees a marketing / onboarding hero and taps Get Started to sign in. They are presented with social sign-in buttons (Apple, Facebook, Google) and an alternative email/password form. The user either authenticates via a social provider or enters email + password and submits. On successful authentication the app navigates to the personalized Home/Discovery area (recommended show, play trailer CTA, latest episodes) and activates the bottom navigation. Assumptions: user on a modern smartphone (iOS-like chrome visible), network access available, user is initially logged out and has or can create credentials with a supported social provider or email/password. **Vinyls: QR Sign-In and Onboarding** (ios, 5 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/8939) > Happy path: a new or returning user starts on an onboarding screen explaining a feature (step 7/9). They choose to authenticate using a QR code. A bottom-sheet explains the QR login option and offers a Read QR Code CTA. Tapping that opens the camera scanner; the app reads the QR and immediately authenticates, showing a welcome toast and advancing the onboarding. After authentication the user lands on the main, personalized home screen (recent vinyls, global charts). Steps: Onboarding: Calibrate your Turntable → Bottom-sheet: Login with Code → QR Scanner (Camera View) → Welcome Confirmation + Next Onboarding Card → Authenticated Home: Your recent Vinyls **PayPal: Password recovery and verification** (ios, 9 screens, Finance) [Refero](https://refero.design/flows/7269) > Happy path for a user who has forgotten or entered an incorrect password and needs to reset it. Steps: Login attempt (error shown) → Quick security check - choose method → Quick security check - phone option selected → Enter security code (empty) → Enter security code (typing) → Create a new password (empty) → Passwords validated (match) → Password creation confirmation → Back to login (ready to sign in) **Instagram: Sign in to Instagram** (ios, 6 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/2834) > Happy path: user opens the app -> sees splash screen -> is presented with a login form (username/email/phone + password) -> enters credentials (or selects saved credentials) -> taps Log in -> a loading state displays -> user is taken to the Instagram feed. Steps: Launch / Splash → Sign-in form (empty) → Username populated → Password entry with keyboard → Submitting / Loading → Home feed (success) **Dropset: Apple sign-in and profile setup** (ios, 4 screens, Health & Wellness, Fitness) [Refero](https://refero.design/flows/6551) > Happy-path sign-in and initial profile setup for a chat feature on an iOS device. Entry: user opens the Sign in screen while logged out. The user taps the 'Sign in with Apple' CTA which invokes the system Apple authentication sheet. After selecting an Apple account and confirming sign-in, the app returns to a Profile screen where the user sees and can edit their chat handle, picture, social links, bio, and notification preferences. Assumptions: user is on an iPhone, not signed in to the product, and has at least one Apple ID available on the device. The device's standard Apple authentication sheet is used for account selection/consent. Steps: Sign in (landing) → Apple account selection → Authentication confirmation → Profile (post-sign-in) **Voicenotes: Email Sign In** (ios, 7 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14274) > Happy path for an existing user signing in to Voicenotes with email and password. The user starts on the public welcome screen, accepts the Privacy Policy and Terms of Service, selects Log in, chooses email authentication, enters an email address, provides a password, and continues into the authenticated Voicenotes workspace. The flow is shown on a mobile device. The user is initially logged out and has permission to access the account associated with the submitted credentials. The final screen is the notes workspace with existing content visible. **Family: Sign-in to Wallets Dashboard** (web, 7 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/6208) > Happy-path sign-in and entry into an authenticated wallet dashboard starting from the public marketing landing. Entry: an unauthenticated visitor lands on a marketing/home overview for Family. Exit: an authenticated user lands on the 'Wallets' dashboard showing their wallets and a primary CTA to create a new wallet. Assumptions: user is on a desktop browser, not signed in at start, and can receive/enter an email address and password (alternate passkey option is shown but not used in the happy path). Steps: Marketing Landing → Welcome — Identifier Entry (empty) → Identifier Submitted → Choose Password Authentication → Enter Password (empty) → Password Submitted → Wallets Dashboard **Visitors: Passwordless sign-in to analytics** (web, 7 screens, Business) [Refero](https://refero.design/flows/9877) > Happy-path sign-in using a passwordless (email verification code) flow that takes a visitor from the public product marketing entry through authentication and into the analytics dashboard. Steps: Marketing / Entry → Sign-in (empty) → Email submitted → Verification prompt (empty) → Email with code → Verification input (filled) → Analytics dashboard (post-login) **Meta AI: Meta AI Sign In** (web, 8 screens, AI Tool) [Refero](https://refero.design/flows/12772) > Happy path starts on the Meta AI landing screen where the user can begin a prompt or choose a suggested starter. The user then signs in using a saved profile, completes email code verification, and optionally saves login info. The flow exits on the authenticated Meta AI home screen with the sidebar, recent chats, and prompt composer available. Steps: Landing Screen → Login Entry → Account Selection → Email Code Prompt → Code Submitted → Verification Loading → Save Login Prompt → Authenticated Home **Leonardo.Ai: Password Reset Recovery** (web, 4 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/12781) > This is a password recovery and reset sequence for Leonardo.ai by Canva. The user starts on a sign-in screen with an incorrect password state, moves to a reset form where a verification code and new password are entered, and ends on a success confirmation screen with a continue action. Steps: Sign In Error → Reset Verification → Password Submitted → Reset Complete **Fingerprint: Sign-in and Getting Started Onboarding** (web, 5 screens, Business, Development Tools) [Refero](https://refero.design/flows/11177) > Happy path: an anonymous visitor on the public marketing site opens the product and chooses to sign in. A centered authentication modal offers OAuth providers and an email/password option. The user enters an email and password and logs in successfully, landing on the product "Getting started" dashboard that confirms required setup tasks and shows navigation to other areas. Steps: Marketing entry → Authentication modal (choose method) → Email entry → Password entry and sign-in → Getting started dashboard **Сontra: Passwordless Email Sign-in to Dashboard** (web, 8 screens, Job Search) [Refero](https://refero.design/flows/4815) > Happy-path passwordless sign-in using an emailed verification code, starting from the public project gallery and ending at the authenticated user dashboard. **Lovable: Lovable Sign-In and Builder Entry** (web, 6 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12208) > Happy path from the marketing homepage into authenticated product experience. The user starts on Lovable’s public landing page, enters a prompt from the hero input, is asked to log in, authenticates with email/password or a provider, and then lands in the workspace dashboard where they can continue prompting and start or manage projects. Steps: Landing Page → Login Modal → Email Entry → Login Form → Authenticated Login View → Builder Dashboard **Linktree: Password Recovery and Reset** (web, 10 screens, Business, Social Networking) [Refero](https://refero.design/flows/12385) > Happy path starts at the Linktree sign-in screen and continues through the password recovery request, email confirmation, password reset form, and final success state. The user is assumed to be logged out and already has access to the email address or username tied to the account. The main exit point is the confirmation screen that indicates the password has been updated, with a link back to log in. Steps: Login Error → Forgot Password → Send Reset Email → Sending State → Email Confirmation → New Password Entry → Password Mismatch → Reset Password Submit → Reset in Progress → Reset Success **Medium: Passwordless email sign-in to personalized feed** (web, 5 screens, Social Networking) [Refero](https://refero.design/flows/11082) > Happy path: An unauthenticated desktop user lands on Medium's promotional homepage and chooses to start reading. A sign-in modal appears offering a saved account and a 'Sign in with email' button (passwordless). The system sends a one-time code (or magic link) to the user's email and prompts the user to check their inbox. The user enters the 6-digit code in the modal and submits. On successful verification the modal closes and the site loads the personalized, signed-in home feed (left navigation, center article list, right rail recommendations). Steps: Landing / Start reading CTA → Welcome back — account selection → Check your inbox — code entry (empty) → Enter code and submit → Authenticated home feed **Pi: Account sign-in and status** (web, 2 screens, AI Tool) [Refero](https://refero.design/flows/4703) > Happy-path: A signed-out user opens the Profile > Account area (desktop layout) and is presented with multiple sign-in options (Google, Facebook, Apple, or phone). After completing an authentication method the Account panel updates to a signed-in state that confirms the user identity, displays a conversation count, and exposes actions to manage history or sign out. **Pi: Pi onboarding and chat entry** (web, 3 screens, AI Tool) [Refero](https://refero.design/flows/4710) > Happy-path: an unauthenticated user opens the Pi welcome screen, taps Next (or Login) to reach an account sign-in prompt, chooses a sign-in method (Google, Facebook, Apple, or phone) and completes authentication, then arrives at the personalized Pi dashboard/chat where they can read previous messages, interact with the assistant via the input field, and access conversation history or discovery cards. **Lovable: Password Recovery and Reset** (web, 8 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12209) > Happy path plus error recovery for account access. The flow starts on a sign-in screen where the user can log in with social providers or email and password. If credentials are invalid, an error message appears and the user can choose forgot password. A reset-password modal sends a reset link to the user’s email, followed by an inbox email containing the reset URL. The reset link opens a password creation screen where the user enters and confirms a new password, then submits it. A success toast confirms the password was updated, and the flow ends at the main workspace/home dashboard. Steps: Login Attempt → Request Reset Link → Reset Request Confirmed → Reset Email → New Password Entry → Submit New Password → Password Updated → Workspace Home **Polar: Passwordless Sign-in to Admin Overview** (web, 6 screens, Business, Finance) [Refero](https://refero.design/flows/11380) > Happy-path: an anonymous visitor on a desktop lands on the product marketing homepage, opens the sign-in modal, authenticates via email (passwordless) using a 6-character verification code, and arrives at the authenticated admin dashboard (test-mode). Steps: Marketing landing → Sign-in modal (provider options) → Email submitted → Verification code prompt → Code entered, submit → Authenticated dashboard Overview **Linear: Sign-in to project workspace** (web, 6 screens, Business, Productivity, Project Management, Collaboration Tools) [Refero](https://refero.design/flows/6690) > A public marketing homepage leads into the product sign-in flow and ends with the authenticated project workspace (issues list). Steps: Marketing homepage (hero) → Sign-in method selection → Email address entry → Check email instruction → Manual code entry → Authenticated project workspace ##### Task: map location nearby **Apple Maps: Search and Directions in Maps** (web, 3 screens, Map & Navigation) [Refero](https://refero.design/flows/12479) > This is a desktop map navigation experience in Apple Maps. The flow starts on the main Maps screen with a search panel and nearby category shortcuts. The user enters a place name into search, reviews matching location suggestions, then switches to Directions to build a route from My Location to a selected destination. The flow ends with a route preview shown both in the sidebar and on the map. Assumptions: the user is already on the Maps interface, with location access available and no authentication required for these actions. Steps: Search Home → Search Suggestions → Route Preview **Apple Maps: Map Search and Place Details** (web, 3 screens, Map & Navigation) [Refero](https://refero.design/flows/12478) > This is a map-based search experience. The user starts on the main map with a search panel and nearby category shortcuts, enters a location or place name, reviews matching results in a list alongside the map, then opens a specific place detail card. Assumption: the user is already in Maps and can use search, browse categories, and select results on desktop. Steps: Search Start → Search Results → Place Details **Nike: Locate & view Nike store** (web, 4 screens, Marketplace, Sports) [Refero](https://refero.design/flows/2017) > This flow is the 'Find a Nike Store' discovery path on desktop. Steps: Store locator — initial results → Search by location — map recenters → Focus a specific store from list or map → Store detail page **Aesop: Store locator and store detail** (web, 5 screens, Marketplace) [Refero](https://refero.design/flows/5132) > Happy path: A desktop user opens the brand header and navigates to the Stores section, uses the locator to search by city or postcode (or uses "See nearby stores"), reviews matching stores presented as a list alongside a map, selects a store from the list or map to reveal a side detail panel, and then opens the full store profile page for full contact details, opening hours, images and a larger map. Steps: Locator entry → Search input filled → Results — list and map → Quick store detail panel → Full store profile **Apple Maps: Gas Station Search Refinement** (web, 10 screens, Map & Navigation) [Refero](https://refero.design/flows/12480) > The user searches for gas stations in the current New York map area, reviews the matching locations, and refines the results with hours, chain, and amenity filters. After applying filters, the user updates the map search area, reviews the refreshed results, opens the amenities filter, and views sorting options. The flow assumes the user is already in Maps with the search interface available and a map region selected. Steps: Gas Station Results → Open Search Filters → Select Station Criteria → Apply Search Filters → Review Chain Filter → Confirm Chain Selection → Search Updated Map Area → Review Refreshed Results → Open Amenities Filter → Choose Result Sorting **Fresha: Salon detail listing** (web, 2 screens, Travel & Booking, Health & Wellness) [Refero](https://refero.design/flows/2321) > This is a desktop salon detail page showing a single salon’s profile (hero gallery, rating, address, services with prices and durations, team, reviews, about blurb, map, opening times, other locations and nearby venues). **adidas: Store Locator with Map and Details** (web, 6 screens, Marketplace, Sports) [Refero](https://refero.design/flows/10230) > User opens the Store Finder (desktop) to locate a retail/pickup location. The interface displays a large Google map on the right and a left-hand results/search column. The user can type a city, ZIP, or address into the search box; an auto-suggest dropdown appears. Selecting a suggestion recenters the map and updates the left list. If stores exist near the selected location, the list shows matching stores (name, address, distance, open/closed status); otherwise a "no pickup locations" message appears. Selecting a store from the list or map reveals a detail panel with hours, phone, address and store features. Steps: Store Finder landing → Type search (auto-suggest) → Select city with no stores → Select alternative city with results → Select store from list → Open store detail panel **Urban Outfitters: Store Locator - Search & Map Interaction** (web, 3 screens, Marketplace) [Refero](https://refero.design/flows/539) > Desktop store-locator interaction for Urban Outfitters. Happy path: user enters an address, selects a search radius, optionally toggles one or more store-feature checkboxes, then clicks Search. The page returns a count and list of matching stores in a left column and plots corresponding markers on a Google map on the right. The user can click a map marker (or a list item) to open a store info bubble with hours/contact details, and can switch the map between Map and Satellite tiles. **DoorDash: Pickup discovery and merchant inspection** (web, 2 screens, Food & Drink, Food Delivery) [Refero](https://refero.design/flows/10145) > User arrives at a location-centered restaurant discovery screen (location set to 300 Post St). The interface presents a left-hand, scrollable list of nearby merchants with quick actions and a right-side interactive map showing restaurant markers. Top controls include a global search, address, and a Delivery/Pickup toggle; additional filters (Open now, Pickup, DashPass, rating) appear above the list. Happy path: user filters for pickup, scans the list or map, taps a merchant either in the list or on the map, inspects the merchant card (hours, sample menu items, prices), and chooses pickup. Steps: Browse filtered merchant list → Open merchant card on map **LEGO: Find LEGO Store Locator** (web, 2 screens, Marketplace, Entertainment, Kids) [Refero](https://refero.design/flows/1801) > Happy-path: a shopper lands on the LEGO Store locator landing screen, enters a location (city, zip code or country) in the prominent search bar, and is shown search results on a map plus a list of store cards. Steps: Store Locator Landing → Search Results & Map **IKEA: Select collection store** (web, 5 screens, Marketplace, Home & Decor) [Refero](https://refero.design/flows/8256) > Happy path: A shopper on the Checkout » Delivery or collect section opens a right-side modal to pick a preferred IKEA store for collection. The modal supports two views (list and map). The user can pick a store from the list or by tapping a pin on the map, mark it as selected, then confirm. After confirming, the modal closes and the selected store appears in the Delivery or collect area; the system immediately displays product availability information for that store and offers actions (e.g., Save for later). Steps: Open store selector — list view → Switch to map view → Open store callout on map → Mark store as selected → Apply selection and show availability **Instacart: Recipe store selection** (web, 3 screens, Food & Drink, Marketplace, Food Delivery) [Refero](https://refero.design/flows/10096) > User is on a desktop recipe detail view that lists ingredients mapped to grocery products. Entry: recipe detail opened for "Slow-Cooker Shrimp in Purgatory." The recipe view shows an inline store selector and a green "Add X ingredients" CTA. The user opens the store chooser, selects a different retailer, and the ingredient tiles update to reflect that retailer's products, prices, and availability. Exit: updated ingredient list with the new store selected and the CTA count updated (user can then review quantities or tap the CTA). Assumptions: desktop/laptop viewport, a delivery ZIP/location is set (94102 shown), a default store is selected at entry (Safeway in the first screen), the user has a visible cart indicator in the header. Steps: Recipe detail (default store) → Choose Store modal → Recipe detail (updated store) **DoorDash: Pickup results filtering** (web, 6 screens, Food & Drink, Food Delivery) [Refero](https://refero.design/flows/10144) > Happy path: A signed-in (or anonymous) user searching for pickup options at a specific address sees a map on the right and a scrolling list of stores on the left. The interface exposes quick filter pills across the top (Pickup, DashPass, rating, price). The user opens the rating filter, adjusts the minimum rating, applies it, then opens the price filter and selects price tiers. After applying both filters the left list and the map update to show only stores that match the selected Pickup/DashPass/rating/price criteria. Entry: search results for location with Pickup mode active. Exit: filtered list + map showing matching pickup stores (user may then pick a store to view its menu). Assumptions: user location/address is set (300 Post St shown), the user has pickup selected, and the interface is on a desktop browser/large viewport. Steps: Search results (map + list) with Pickup selected → Rating filter modal (default value visible) → Rating adjusted and applied → Price filter modal opened → Price tiers selected → Filtered pickup results displayed **ZARA: Store Locator - search and select store** (web, 3 screens, Marketplace) [Refero](https://refero.design/flows/1606) > Happy path: A user arrives at the Store Locator page, types a location (city/postcode/address) into the search field, optionally toggles the 'Clothes Collection Programme' filter, and clicks Search. The interface returns a list of matching stores on the left and a synchronized map on the right. The user selects a store from the list (radio control) to view its address and program details; the map shows a marker and an info popup for the selected store. Steps: Store Locator - initial → Location entered and filter selected → Search results list and map **Care: Activity Search and Filtering** (web, 7 screens, Job Search, Lifestyle) [Refero](https://refero.design/flows/12232) > User starts on Care.com’s activities/camps search results for San Francisco, CA 94108. They refine the listings using filter chips for camp type, activity category, date range, activity format, and child age. The results list and map update in place as filters are applied. Steps: Browse Recommendations → Date Filter → Apply Date Selection → Activity Type Filter → Category Filter → Age Filter → Filtered Results **Airbnb: Map-side Listing Discovery** (web, 3 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/5997) > Desktop, logged-in user homepage that combines a scrollable list of results (homes, experiences, services) on the left with an interactive map on the right. The happy path: user scans recommended homes and experiences, opens the map's place search, and selects a point-of-interest marker to view a place card. Steps: Home list with map → Map search overlay → Point-of-interest (POI) detail card **Airbnb: Private Chef booking** (web, 19 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/5990) > Happy path: a signed-in user discovers Services, sets location, dates and a service type (Chefs), browses filtered results, opens a listing, picks a menu/option, provides a service address, schedules date/time/guest count, and completes the booking by paying. Steps: Services landing → Location quick action (Nearby) → Location autocomplete suggestions → Location selected → Open date picker → Select date range (preset) → Open service type menu → Select Chefs category → Run search and view filtered results → Open listing detail → Open image preview modal → View full gallery → Initiate booking (Reserve) → Address entry prompt → Address autocomplete → Confirm address → Open scheduling modal → Adjust date selection → Confirm booking details and pay **Nextdoor: Publish Neighborhood Poll** (web, 12 screens, Communication Tools) [Refero](https://refero.design/flows/11425) > A logged-in Nextdoor member starts from the neighborhood home feed, opens the post composer, adds a location, attaches two photos, explores additional creation options, and converts the draft into a poll. The user sets the poll to have no end date, fills in the question and two answer choices, and publishes it. The new post appears at the top of the home feed with its location, images, and selectable poll responses. Steps: Neighborhood Feed → Open Post Composer → Choose Post Location → Search for Address → Attach Location → Add Photos → Open Creation Options → Configure Poll → Select Poll Duration → Complete Poll Content → Publish Post → View Published Poll **Tripadvisor: Restaurant discovery and reservation** (web, 9 screens, Travel & Booking) [Refero](https://refero.design/flows/7072) > Happy-path: A signed or anonymous user arrives at the Restaurants hub, uses the large hero search field to look for a city (example: Lisbon), selects the suggested city, and lands on a long list of restaurants for that city. From the listing the user can open the inline reservation controls (date, time, guests), apply filters or sorting, and then open a restaurant detail page to view photos, hours, contact info and reviews. Steps: Restaurants landing / Hero search → Search suggestions panel → Type-ahead results (Lisbon) → Restaurant listing (initial) → Open date picker → Select time → Choose guest count → Filtered & updated listing → Restaurant detail page **Tripadvisor: City hotel search and filtering** (web, 15 screens, Travel & Booking) [Refero](https://refero.design/flows/7062) > Happy-path: A user lands on the London "Hotels and Places to Stay" landing area, uses the hero search bar (location + dates), scans curated hotel collections, opens the Filters modal to narrow results (price range, traveler rating, amenities, property types, distance and neighborhoods), applies filters, views a filtered list with an interactive map, and finally opens an individual hotel detail page. Entry: city-level landing. Exit: hotel detail with rates and reviews. Assumptions: user is on desktop, not necessarily signed in, and has selected or is able to select check-in dates in the header search control. Steps: City landing with hero search → Search results overview (expanded calendar) → Open Filters modal → Adjust price range and rating → Quick amenities view → Full amenities selector → Amenities applied summary → Select property type, class and style → Set distance and neighborhoods → Filter summary (selected chips) → Filtered list with map (neighborhood focus) → Paginated/long-scrolling result list → Open Filters modal from results (alternate view) → Select a hotel to view details → Hotel detail: reviews and additional info ##### Task: marketplace listing **Fresha: Help center search and results** (ios, 2 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4979) > Mobile help center entry screen with a prominent search field, quick suggestion chips, and category cards (Getting started, Knowledge base, Fresha academy). Happy path: user types a query or taps a suggested topic → the app shows a Search results screen with a ranked list of matching help articles. Steps: Help center home → Search results list **Family: Explore — Suggested Sites** (ios, 2 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/2674) > Happy-path: User opens the browser's Explore (compass) tab on a mobile device and sees a grid of suggested Ethereum/web3 sites. From here the user can: tap a site icon to open that site, use the search bar to look for other sites, or tap the question (?) help control to read a short explanation of what "Suggested" means. Steps: Suggested sites grid → Suggested Apps info modal **Airbnb: Map-based filtered search** (ios, 21 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6389) > Mobile map-based property search where the user opens a full-screen filters panel, selects multiple filter categories (recommended amenities, place type, price range, rooms/beds, amenities, booking options, accessibility), and applies them to narrow results. **voilà: Map Place Picker** (ios, 3 screens, Travel & Booking) [Refero](https://refero.design/flows/9289) > Mobile map-based place discovery inside an itinerary step (Ubud). Entry: user views the 'Ubud' step (top selector) with a map and a draggable bottom sheet. The bottom sheet is opened to the 'Places' tab and shows category filter chips and a horizontal list/card for a highlighted place (Moksa Ubud). The user can toggle category chips (Restaurant, Accommodation, Activity, Coffee, Shopping) to change which markers appear on the map. Exit point: user either taps a place card (to view details / add to itinerary) or navigates back via the top-left back button. Assumptions: mobile device, map centered on Ubud, bottom sheet and map are both interactive; no assumptions made about authentication or background services. Steps: Open map with Places tab → Enable additional category filters → Switch or refine category selection **Airbnb: Share listing to Messages** (ios, 3 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6402) > Happy path: A mobile user views a place detail (Katalina's in Harrison West) and uses the on-screen share control to send the listing to someone via the device's native share sheet. From the share sheet the user selects Messages which opens the system Messages composer containing an auto-generated link preview for the listing. **Spoil me: Add Listing Image** (ios, 5 screens, Fundraising) [Refero](https://refero.design/flows/5345) > Happy path: A user creates or edits a product/listing entry by entering a title and price, optionally enabling a "Crowdfunding" toggle, then adding one or more images from the device. Steps: Listing form with image placeholder → Action sheet for image source → iOS Photo Picker (initial) → iOS Photo Picker (selection) → Form with image preview and confirm **Binge: Where to Watch Providers** (ios, 3 screens, Entertainment) [Refero](https://refero.design/flows/12959) > A mobile movie-discovery screen shows where a title can be watched across multiple providers. The user enters from a movie detail context and can return via the back button or open a sort/filter menu from the overflow icon. The menu lets the user change the provider order by popularity or alphabetical and filter the list by availability type (All, Rent, Buy, Subscribe). After applying a sort or filter, the provider grid refreshes to show the matching services in the chosen order. Steps: Provider List → Sort and Filter Menu → Filtered Provider List **komoot: Places Selector** (ios, 3 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8390) > Happy path: From the map search entry the user opens a list of place categories to show on the map, optionally focuses the search input to type a keyword, then selects a search result. Steps: Category List → Activate Search → Search Results **On: Refine product listing by filters** (ios, 15 screens, Marketplace) [Refero](https://refero.design/flows/9596) > Happy-path: User browsing a product listing (shoes) opens an on-screen filter modal, applies multiple filters (sort, sizes, price range, colors, product models, activities, technology), reviews applied filters and then applies them to update the product grid. Entry: product listing grid (mobile). Exit: filtered product listing with applied filter chips and updated result count. Assumptions: mobile device (portrait), user is browsing catalogue (not necessarily signed in), user can open a modal filter overlay from the listing and apply multiple simultaneous filters; filter state persists until cleared. Steps: Product listing (start) → Filter modal (entry) → Change sort option → Sizes grid (open) → Applied size chips appear → Confirm multi-selection of sizes → Price slider default → Adjust price range → Color swatches (open) → Select colors → Product model list (partial) → Extended model selection → Activity selection → Technology selection and review → Review applied filters & apply **Airbnb: Listing Save & Share** (ios, 3 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6394) > Happy-path: A signed-in user views a short-term rental listing, optionally saves it to a wishlist, then shares the listing using the device's native share sheet and chooses Messages to send a rich link preview. **TikTok: Find and open place profile** (ios, 3 screens, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/2784) > Happy path: A mobile user opens the global search, sees suggested queries, enters or selects a location term, switches to the Places category, selects one of the place results, and lands on that place's detail view which shows the address, a prominent Add to Favorites CTA and a grid of user posts. Steps: Search suggestions → Places results list → Place detail and posts **Airbnb: Save Listing to Wishlist** (ios, 3 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6393) > Happy path: Signed-in user views a property listing and taps the heart icon to save it. A bottom-sheet modal appears letting the user pick an existing wishlist or create a new one. After choosing an existing wishlist, the modal closes and a temporary confirmation (toast) is shown indicating the listing was saved; the heart icon updates to a filled state. **Airbnb: Save listing to new wishlist** (ios, 4 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6387) > Happy-path: A signed-in user browsing the Explore tab discovers a listing they want to save. They tap the listing’s heart icon, which opens a bottom-sheet modal to create a new wishlist. The modal contains a single text field with a 50-character limit and two actions: Cancel and Create (disabled until text is entered). The user types a name, taps Create, and receives in-context feedback: the heart becomes filled and a transient 'Saved to ' confirmation appears with an option to change the wishlist. Entry: Explore tab (listing visible). Exit: modal dismissed and listing saved. Assumptions: user is on a mobile device (iPhone/iOS), has permission to create/save wishlists, and the listing card exposes a tappable heart affordance. **Airbnb: Map area search and filter** (ios, 4 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6390) > Mobile browse experience for finding short‑term rentals in a map area. Entry: user is on the Explore list of homes (top search pill shows 'Homes in Map area'). The user opens the Filters sheet, reviews and edits selected filters, then applies filters to update the results and views the results on the map (bottom sheet shows result count and listing cards). Exit: user either taps a listing to view details or closes the map/filters. Assumptions: mobile (iPhone) context; user has search scope already set (map area, date set to "Any weekend", 1 guest); user can open/close a bottom sheet for Filters and Map and the product shows a dynamic result count. **Asos: Brand search to product listing** (ios, 4 screens, Marketplace) [Refero](https://refero.design/flows/2816) > Happy path: User starts on the Sportswear category landing (mobile). From the header search control they open the search input, optionally use recent searches or the visual search (camera/photos), type a query (example: "nike"), pick an autocomplete suggestion or submit the query, and lands on a product-grid of Nike results. Steps: Sportswear landing → Search input with recent & visual options → Autocomplete suggestions → Search results — Nike product grid **Instagram: Map-based place discovery** (ios, 4 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/2862) > User explores a map populated with circular photo markers (location stories). Tapping a marker opens a draggable bottom sheet with that place's name, distance, address, category/hours and a media gallery. From the sheet the user can pull up to view more posts, or choose a ‘Search this area’ / category action which triggers a loading state and then populates a ‘Popular nearby’ region view with category chips and a list of places. Steps: Map with place markers and expanded sheet → Collapsed sheet with 'Search this area' affordance → Loading area search → Region view: Popular nearby with categories **Airbnb: Review Search & Results** (ios, 4 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6401) > User is viewing the reviews section for a specific listing on a mobile device. From the listing's reviews header (overall score + subratings) the user can open a searchable list of reviews, type queries into the search field, and see matching review items or a "no results" message. **Microsoft: Marketplace Product Discovery** (web, 3 screens, Business, Marketplace) [Refero](https://refero.design/flows/9795) > User arrives on the Microsoft Marketplace home (desktop, not signed in). The page presents a prominent search bar, a 'See all products' call-to-action, a 'Featured solutions' carousel/grid, category/industry pills, customer stories, and a global footer. From the home screen the user can open the left navigation to browse product categories, or click a featured category or 'See all products' to go to a category results page. The flow ends at a category results/listing page showing a two-column layout: left column with filters/navigation and right column with a grid of product cards and pagination. Assumption: user is on a desktop/laptop viewport and has not signed in (Sign in CTA visible). Steps: Marketplace Home → Main Navigation Drawer → Category Results — AI Apps and Agents **Nextdoor: Marketplace Category Filtering** (web, 4 screens, Communication Tools) [Refero](https://refero.design/flows/11430) > A logged-in user browses the For Sale & Free marketplace, opens the category filter, selects Home decor, and opens the distance filter to review available radius options. The final screen shows Home decor results with the existing 15-mile distance setting and other filters still available. The flow begins on the All listings view and ends with filtered marketplace results; no item is opened or purchased. Steps: Browse All Listings → Open Category Filter → Open Distance Filter → Review Filtered Results **Nextdoor: Copy Listing for Sharing** (web, 3 screens, Communication Tools) [Refero](https://refero.design/flows/11437) > A logged-in user browses the For Sale & Free listings and shares a listing by opening its share menu and copying the listing information to the clipboard. The flow begins on the listings view and ends with an on-screen confirmation that the content was copied successfully. Steps: Browse Listings → Open Share Options → Copy Listing **Shopify: Install Klaviyo from marketplace** (web, 4 screens, Business) [Refero](https://refero.design/flows/5418) > Happy path: A signed-in merchant on desktop opens the Shopify admin global search, types a query to find a marketing integration, selects the suggested integration result, and lands on the integration's listing in the Shopify marketplace. **Intercom: Install Clearbit Reveal integration** (web, 7 screens, Business, Communication Tools) [Refero](https://refero.design/flows/683) > Happy-path flow for discovering and installing a third‑party integration (Clearbit Reveal) from an in‑product App Store. Steps: Marketplace browse → App detail modal → Scroll details & related apps → Fullscreen media lightbox → Initial permission consent → Expanded permission details → Installed apps list **Nextdoor: Delete Marketplace Listing** (web, 4 screens, Communication Tools) [Refero](https://refero.design/flows/11435) > A logged-in seller manages their For Sale & Free listings. From the Your listings section, the user opens the overflow menu for a listing, chooses Delete, confirms the deletion in a modal, and returns to the listings view. The deleted listing is removed from the Selling list and a confirmation notification appears. Steps: Your Listings → Listing Actions → Confirm Deletion → Deletion Complete **Airtable: Add Extension from Marketplace** (web, 5 screens, Project Management, Collaboration Tools, Scheduling Tools, Paperwork, Productivity) [Refero](https://refero.design/flows/1581) > Happy path: A logged-in user in the "Creative Asset Development" base navigates from the main Data/Kanban workspace to the Extensions area, opens the Extensions Marketplace, views the Base schema extension details, and is shown an upgrade prompt because their current plan is Free. Steps: Extensions drawer open → Marketplace landing → Marketplace browsing (repeat) → Extension detail view → Upgrade plan dialog **Nextdoor: Share Listing in Chat** (web, 6 screens, Communication Tools) [Refero](https://refero.design/flows/11431) > A logged-in Nextdoor user shares a For Sale & Free listing with a neighborhood chat. From the listing detail panel, the user opens the chat-sharing option, searches for a recipient or group, selects the "san francisco" chat, and creates the conversation. The system opens the chat drawer with a listing link prefilled, and the user sends it. The flow starts on the listing detail view and ends with the listing link displayed as a sent message in the selected chat. Steps: Listing Detail → Search Chat Recipient → Select Chat → Create Conversation → Review Chat Message → Message Sent **Faire: Image Search Product Browse** (web, 4 screens, Business, Marketplace) [Refero](https://refero.design/flows/13090) > Happy path starts on a marketplace search results page using image search. The user can refine the query via the top search bar, browse a grid of matching products, and use filters, saved searches, or pagination to explore more results. It assumes the user is already on the marketplace and can browse without logging in; saving items may require an account. The flow ends when the user finds a relevant product, saves it, or continues to another search or category. Steps: Image Search Results → Search Refinement → Autocompleted Search → Updated Product Results **Intercom: Installed Integrations Browse** (web, 3 screens, Business, Communication Tools) [Refero](https://refero.design/flows/684) > User starts at the marketplace's Installed Integrations screen (desktop view) and can browse installed items or search the marketplace using the top search field. From this entry point the user performs a search that first returns an empty state, then revises the query (or clears it) and receives a populated results grid of available integrations. Assumptions: user is signed in with permission to view marketplace listings; this is a desktop viewport; left-hand navigation and category filters are persistent across steps. Steps: Installed integrations overview → Search - empty results → Search - populated results **Nextdoor: Create and Publish Listing** (web, 9 screens, Communication Tools) [Refero](https://refero.design/flows/11434) > A logged-in Nextdoor user creates and publishes a For Sale & Free listing. The user starts from the Selling section with no existing listings, begins a new listing, adds a photo, reviews or edits the suggested title, description, and price, optionally adds more context, and publishes the item. The new listing then appears under Selling and can be opened in a public listing detail view. The flow assumes the user is authorized to post in the selected neighborhood and has at least one item photo available. Steps: Empty Selling Inventory → Add Listing Photos → Process Uploaded Image → Enter Item Details → Confirm Listing Details → Add Posting Context → Published Selling Listing → View Listing Details → Browse Related Listings **Faire: Trending Products Browsing** (web, 3 screens, Business, Marketplace) [Refero](https://refero.design/flows/13087) > A marketplace browsing screen for trending products, with a left-side filter panel and a multi-column product grid. The user enters from a general product discovery area and can refine results using filter chips, sidebar filters, search, and sort controls. A product detail view or a similar-products view can be reached by selecting an item or related suggestion; pagination supports browsing more results. Steps: Trending Products Listing → Refined Product Results → Similar Products View **Nextdoor: Contact Listing Seller** (web, 3 screens, Communication Tools) [Refero](https://refero.design/flows/11432) > A logged-in user browses the For Sale & Free listings and opens a listing for “The New Archies Comic Book Collection.” From the listing detail panel, the user sends a prefilled availability question to the seller. The interface confirms that the message was sent, then allows the user to open the conversation and view the ongoing chat in a side panel. Steps: Listing Detail → Message Sent Confirmation → Seller Conversation **Faire: Kitchen & Tabletop Product Browsing** (web, 4 screens, Business, Marketplace) [Refero](https://refero.design/flows/13075) > Happy-path browsing experience in a product marketplace category listing. The user enters from a broader marketplace navigation area into a category such as Kitchen & tabletop, reviews the product grid, optionally applies filters and sort options, and can continue paging through results. Steps: Category Listing View → Filter and Sort Controls → Category Refinement → Paginated Results Browsing **Faire: Seller Storefront Browsing** (web, 3 screens, Business, Marketplace) [Refero](https://refero.design/flows/13082) > A shopper is viewing a seller storefront for DAYLA on a marketplace, with the seller header, promotional banner, product listings, and a persistent cart sidebar visible. The entry point is the seller’s storefront page. The user can browse products, follow the seller, add items to a list, or proceed to cart/checkout from the right-side cart summary. Assumption: the user is already signed in and has items in the cart, with the storefront loaded on desktop. Steps: Storefront Overview → Seller Follow State → Cart Review and Checkout **Shop: Lamp Product Search** (web, 4 screens, Marketplace) [Refero](https://refero.design/flows/13274) > A shopper starts on the Deals view and opens the global search field. The interface first displays suggested searches, then shows matching suggestions as the shopper types “lamp.” After submitting the query, the shopper is taken to a results view containing related shops, product listings, filters, and sorting controls. The flow assumes the shopper can access the marketplace and is not required to complete checkout. Steps: Open Search → View Search Suggestions → Enter Lamp Query → Review Search Results **Nextdoor: Mark Listing Sold** (web, 3 screens, Communication Tools) [Refero](https://refero.design/flows/11436) > A logged-in seller manages an active For Sale & Free listing. From the Selling tab, the user selects Mark sold for the Modern Brown Chair, chooses where the item was sold in a confirmation dialog, and returns to the listings view. The listing is updated to show Sold, and its action changes to Mark available. Steps: Selling Listings → Sale Details → Sold Listing ##### Task: medication reminder **Poppy: Create Location Reminder** (ios, 14 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12188) > The user starts on Poppy’s assistant dashboard and opens the Create shortcut. They choose the Reminder type, select or edit a suggested reminder prompt, configure it to trigger when arriving at a saved place, and choose Home as the location. On the final step, they keep the suggested name “Spend Check,” configure the recurring reminder’s end date, and create it. The dashboard updates to show five active items. The user then opens Inbox to review saved reminders. The screenshots show an authenticated user on a mobile device; some intermediate screens represent selectable template and date-picker states. Steps: Assistant Dashboard → Choose Creation Type → Enter Reminder Prompt → Select Commute Template → Select Spend Check Template → Choose Reminder Trigger → Enter Arrival Location → Confirm Home Location → Name Reminder → Open End Date Picker → Select Reminder End Date → Adjust End Date → Reminder Created → Review Reminder Inbox **Twitch: Set Stream Reminder** (web, 2 screens, Social Networking, Entertainment, Video Streaming) [Refero](https://refero.design/flows/13608) > The user searches for “discoelysium” and reviews matching Twitch results. They locate the DiscoElysium513 channel’s scheduled stream and select its reminder control. The interface confirms the reminder is set without leaving the search results. The flow assumes the user is viewing Twitch in a desktop browser and can access channel reminder controls. Steps: Search Channel Results → Reminder Confirmation **Kit: Manage Recommendations** (web, 4 screens, Business, Communication Tools) [Refero](https://refero.design/flows/2264) > Logged-in desktop web flow for managing creators in the "My Recommendations" list. Steps: Recommendations list (My Recommendations) → Open per-creator actions → Confirm stop recommending → Return to recommendations list ##### Task: meeting notes transcription **Wabi: Voice Input Transcription** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12057) > This mobile chat-based experience starts in a conversation view where the user can review a proposed concept and use the voice input control to speak a follow-up request. If microphone access is not yet granted, the system shows an OS permission prompt. After the user allows access, the recording state appears at the bottom with an active audio indicator. When the user stops recording, the system switches to transcription mode and attempts to convert speech to text. The flow ends either with the transcribed input ready for use or, in the shown case, an error toast indicating the data could not be read in the correct format. Assumptions: the user is on a mobile device, inside a chat interface, and interacting as an existing user with access to the voice input feature. Steps: Chat Review → Microphone Permission → Recording → Transcribing → Transcription Error **Handshake: Document Upload & Preview** (web, 7 screens, Social Networking, Job Search) [Refero](https://refero.design/flows/8772) > Happy path: A logged-in user navigates to the Documents section, uploads a PDF (resume), selects a document type, optionally marks visibility and adds a description, then saves. The system shows a saved notification and displays a document detail view with an embedded PDF preview. The user is returned to a Documents list that now contains the uploaded file. Steps: Empty Documents - CTA → Upload Form Displayed → Document Type Selection → File Selected & Fields Filled → Document Detail (Preview - Loading) → Embedded PDF Viewer → Documents List Updated **Memotron: Create interview with audio and transcription** (web, 8 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/8325) > Happy path for creating a new note titled "Interview" using an audio capture input, linking that note to existing collections or nodes, recording audio, saving it, and viewing the generated transcription. Entry: user opens a new capture/note composer (assumed authenticated). Exit: saved note with attached audio file and a generated transcription shown in a playback/transcription viewer. Assumptions: user is signed in, has permission to create notes and link to collections/nodes, microphone access is available, and the device supports audio recording and playback (desktop or laptop). Steps: Composer - choose input type → Audio composer - ready to record → Search collections - results shown → Search audio node - preview → Collections attached → Recording in progress → Saving / processing → Playback & transcription viewer **Figma: Summarize team sticky notes** (web, 5 screens, Design & Media Editing, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/5020) > Happy path: A user working inside a collaborative infinite canvas selects a framed area containing colored sticky-note cards titled "Team Roles & Responsibilities," uses a built-in summarize/organize command from the floating toolbar/context menu, and inserts the generated summary as a text/document object on the board. Steps: View roles board → Select frame → Open organize menu → Place summary placeholder → Insert generated summary ##### Task: mental health questionnaire mood check **Ada: Symptom assessment feedback and rating prompt** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/4106) > Happy-path: After completing a symptom assessment the user is asked a single-question feedback prompt (Was this symptom assessment helpful?). The user selects an option (Yes / No / I'm not sure). If the user selects the positive path the flow asks for optional written suggestions in a multi-line text area; the keyboard appears for typing and a Send button becomes available when text is entered (Skip is available to bypass text entry). After sending (or skipping), the user sees a thank-you/acknowledgement screen with a clear call-to-action to rate the product in the store and a secondary dismissal option. Steps: Quick feedback prompt → Request optional written feedback → Text entry and send → Acknowledgement and rating CTA **Meta AI: Chat Response Feedback** (web, 3 screens, AI Tool) [Refero](https://refero.design/flows/12737) > Logged-in chat interface with a left navigation sidebar and a conversation area. The user asks the assistant for an unusual fact, receives a response, then uses the message feedback controls to mark it positively. After the feedback action, the interface shows a confirmation that the feedback was submitted and the reaction state updates to indicate the response was liked. Entry is from an existing chat thread; exit is the completed feedback state on the assistant message. Steps: Assistant Reply → Positive Rating → Feedback Confirmation **Train Fitness: Feed sentiment prompt** (ios, 2 screens, Health & Wellness, Social Networking) [Refero](https://refero.design/flows/7160) > Happy path: a signed-in mobile user opens the Feed tab and sees a large hero card asking for quick sentiment feedback (two prominent buttons: Good / Bad). The user taps one of the buttons and an ephemeral confirmation toast appears at the top of the screen reading “Feedback Submitted.” The feed content (activity post with metrics and muscle map) remains visible and interactive. Steps: Feed — Feedback Prompt → Feed — Confirmation + Post **Podimo: Live Poll Participation** (ios, 2 screens, Entertainment) [Refero](https://refero.design/flows/12276) > A logged-in mobile user is viewing a content detail screen for Casa Jedet with comments and a live poll. The entry point is the content card/detail view; the exit point is either closing the screen or staying after voting. Steps: Poll Preview → Poll Results **Tinder: Profile quiz setup (Ask Me About)** (ios, 6 screens, Dating & Relationships) [Refero](https://refero.design/flows/3449) > Happy path: A signed-in user opens the Edit Info screen and taps "Add quiz" on the "Going Out" row under the "ASK ME ABOUT" section. A modal bottom sheet presents a sequence of short multiple-choice questions (2 options per question). The user selects one option per question; the modal advances through all questions and then shows a compact results card summarizing the chosen answers. The user taps Done to close the modal — the parent list updates to show the result summary next to the "Going Out" row. Entry: Edit Info screen (logged-in user). Exit: Bottom sheet closed, Edit Info list updated with quiz summary. Assumptions: user is signed in and allowed to edit profile fields; device is a phone (vertical modal bottom sheet); selections are single-choice per question and recorded locally or persisted when Done is tapped. Steps: Edit Info — Ask Me About list → Question 1 — Select where you can be found → Question 2 — Dress style preference → Question 3 — Arrival timing → Question 4 — Exit strategy → Finalize — Close results and update list **Teal: Work-style assessment** (web, 21 screens, Productivity, Job Search) [Refero](https://refero.design/flows/2730) > Happy-path: A signed-in user on the platform launches a short, 17-question work-style survey from a company/dashboard view and completes it. Each question is presented as a bipolar adjective pair (e.g., Competitive — Cooperative) with a five-position slider. The user moves the slider (or taps a position), then navigates forward with Next; from question 2 onward a Previous button appears. The final step asks the user to choose one of three career-growth card options and press Complete. On completion the user is taken to a results dashboard that shows a video, a textual summary of the user's work style, a 360 invite CTA, a right-hand 'Dimensions' sidebar visualizing each bipolar dimension as miniature sliders, and expanded content sections (Career Insights, Start & Finish, Results & People, Sequence Pattern, Job Searching). Steps: Dashboard — Work Styles entry → Question 1 — Competitive vs Cooperative (initial) → Question 1 — Competitive vs Cooperative (selected) → Question 2 — Laid-back vs Assertive → Question 3 — Peaceful vs Defiant → Question 4 — Agreeable vs Pioneering → Question 5 — Animated vs Soft-spoken → Question 6 — Cautious vs Adventurous → Question 7 — Convincing vs Restrained → Question 8 — Self-controlled vs High-spirited → Question 10 — Spontaneous vs Planned → Question 11 — Methodical vs Enthusiastic → Question 12 — Skeptical vs Trusting → Question 13 — Light-hearted vs Serious → Question 14 — Impatient vs Patient → Question 15 — Lenient vs Strict → Question 16 — Questioning vs Accepting → Question 17 — Career growth options (card selection) → Results — Work Style dashboard → Results — Expanded details and CTAs **Duolingo: Friend update reaction** (web, 4 screens, Education) [Refero](https://refero.design/flows/11829) > Happy path starts on the user’s profile dashboard and moves to a friend updates view where the user reacts to a friend’s achievement post. Steps: Profile dashboard → Friend updates → Reaction picker → Reaction confirmed **IKEA: Chatbot post-conversation survey** (web, 5 screens, Marketplace, Home & Decor) [Refero](https://refero.design/flows/8274) > Happy path: A user browsing the IKEA 'Contact us' section interacts with the site chat widget after a conversation has ended to submit a feedback survey. Steps: Contact page with post-chat survey → User selects a star rating → User enters written feedback → Submit survey and show success modal → Acknowledge and return to chat state **OpenAI: Submit feedback on assistant message** (web, 6 screens, Business, Development Tools, AI Tool) [Refero](https://refero.design/flows/4197) > Happy-path for submitting feedback on a single assistant response in the Playground chat. Steps: View assistant response → Consent prompt appears → Allow consent and open feedback entry → Enter feedback text → Submit feedback → Final confirmation state ##### Task: messaging inbox **Kickstarter: Inbox message search** (ios, 3 screens, Fundraising) [Refero](https://refero.design/flows/4638) > User is on the Profile tab viewing their Inbox (mobile). The entry point is the Inbox list showing message cards with sender, subject, snippet, date and an unread indicator. The happy path: user taps the search control in the inbox header, the input field gains focus and the keyboard appears, the user types a query and the message list is filtered live to show matching results; a clear (x) control is provided to reset the query. Exit point: user either clears the query or navigates back to the previous screen. Assumptions: user is signed in, on a mobile device, and has at least one message in the inbox. Steps: Inbox list (default) → Search input focus → Live filter results **Instagram: Account Switch and Messages** (ios, 3 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/2889) > Happy path: A signed-in user on their personal profile opens the account picker from the bottom-right profile avatar, selects a secondary account, and lands in that account's Direct Messages (DM) inbox. Steps: Profile (current account) → Account picker → Direct Messages (Primary tab) **Kickstarter: Inbox message search** (ios, 3 screens, Fundraising) [Refero](https://refero.design/flows/4637) > Happy path: a logged-in user on a mobile device opens their Profile > Inbox, taps the search control, types a query and the inbox list filters in real time. Steps: Inbox list (initial view) → Search field focused → Enter query and view filtering **Airbnb: Inbox feedback submission** (ios, 6 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6428) > Happy-path flow for submitting product feedback from the Messages (inbox) area on a mobile device. Entry: user is on the Messages list (logged-in state) and opens messaging settings. Exit: user sees a confirmation screen and closes it (returns to Messages). Assumptions: user is authenticated, on a mobile phone, and can open the messaging settings sheet. The flow uses a bottom sheet to surface settings, then a modal sequence to collect feedback category, free-text details, and an optional bug flag, then shows a submission confirmation. **Artsy: Gallery Conversation Reply** (ios, 8 screens, Marketplace) [Refero](https://refero.design/flows/7612) > Happy-path for reading an unread inquiry and replying to a gallery in a mobile conversation view. Steps: Inbox list / Unread item → Conversation opened / Read message → Artwork card inline → View prior messages → Focus composer (no text) → Typing message (enabled send) → Send message → Message appended / Confirmation **Tinder: Unmatch Conversation** (ios, 4 screens, Dating & Relationships) [Refero](https://refero.design/flows/3471) > Happy path: A logged-in user opens the Messages tab, locates a conversation, reveals inline actions by swiping left, chooses Unmatch, confirms the irreversible action in a modal, and sees the conversation removed from the inbox. Steps: Messages List → Reveal Conversation Actions → Confirm Unmatch → Conversation Removed **Airbnb: Inbox message management** (ios, 3 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6455) > This is a single-screen messages inbox where the user views conversation previews and can take quick actions by swiping a conversation. Entry: user opens the Messages tab (bottom nav) already signed in. The default 'All' filter is active with other filters available (Hosting, Traveling, Support). Exit: user either opens a conversation (not shown) or performs a quick action (Star or Archive) on a conversation, which updates the conversation row to show the new state (starred). Steps: Messages inbox (default) → Reveal quick actions via swipe → Apply quick action and show state **Airbnb: Messages inbox (conversation list)** (ios, 3 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6454) > Compact mobile messages inbox (iPhone) showing a list of conversations with filter pills across the top. Entry: user opens the Messages tab from the bottom navigation (Messages is selected). Exit: user either opens a conversation (navigates to the chat view), performs a contextual action (Unread) via swipe, or navigates to search/settings. Steps: Inbox — default list → Contextual swipe action → Post-action / refreshed list **Instagram: Post ephemeral note** (ios, 6 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/2853) > Happy-path: A signed-in user opens their Messages inbox, taps the profile avatar + to create a short ephemeral "note", types text in a 60-character composer, optionally changes the audience (Followers you follow back or Close Friends), shares the note, sees a confirmation toast, and returns to the inbox where the note bubble appears above their avatar. Entry: Messages/inbox screen. Exit: Messages/inbox with the note visible on the avatar. Assumptions: user is logged in on a mobile device, has at least one follower and a Close Friends list option, and network connectivity for posting the note. Steps: Inbox - Start note → Open composer → Type note → Open audience selector → Select audience → Share and confirm **XChat: Start new conversation** (ios, 5 screens, Communication Tools, Social Networking) [Refero](https://refero.design/flows/11647) > Happy path: A logged-in user with an empty inbox starts a new one-to-one conversation. Entry: empty Chat screen. The user taps the compose/search control to open a New chat sheet, types a username/handle, selects the matching contact from the filtered results, and is taken into a new conversation view. Exit: the conversation is created and visible in the chat list with a preview and timestamp. Assumptions: user is authenticated and has permission to message other users; contacts/handles are discoverable via the app's search. Steps: Empty inbox → New chat — contact list → Search input (keyboard active) → Search results — choose recipient → Inbox updated with new conversation **XChat: Pin conversation** (ios, 3 screens, Communication Tools, Social Networking) [Refero](https://refero.design/flows/11662) > Happy path: From the chat list the user swipes a conversation to reveal quick actions, taps the pin action, and the conversation shows a pin indicator. Steps: Conversation list → Reveal quick actions → Toggle pin and confirm **Daze: Message interaction and sticker insertion** (ios, 112 screens, Social Networking) [Refero](https://refero.design/flows/9121) > Happy-path for reading a conversation, interacting with a specific message (react, open context menu, delete), then inserting a GIF/sticker from the GIF/Image picker and editing its placement, and finally viewing the contact's chat settings. Entry: user is on the app's Messages list, logged in. Exit: sticker/GIF inserted into the message composer (ready to send) and contact panel inspected. Steps: Conversations list (empty state) → Conversations list (preview) → Open conversation → Message selected (reaction bar visible) → Context menu opened → Alternate reaction selection → Message annotated with reaction → Context menu on different message → Delete confirmation → Open GIF/Image picker → Search GIFs (keyboard) → Enter search term → GIF/Image results → Sticker placement editor → Adjust sticker order / duplicate → Return to chat with sticker placed → Open contact/profile panel → Browse contact media → Close contact panel → Final ready state **HYPE: Share Article by Message** (ios, 3 screens, News, Lifestyle) [Refero](https://refero.design/flows/12245) > Happy path for sharing a Hypebeast article from the HYPE feed into a new iMessage conversation. The user starts on the feed, opens the system share sheet from an article card, chooses Messages, and lands in a prefilled iMessage composer with the link preview attached and the recipient selected. Steps: Feed Browsing → Share Sheet → Message Composition **Substack: Send photo in direct message** (ios, 4 screens, News, Social Networking) [Refero](https://refero.design/flows/4496) > Happy-path: A logged-in user in a one-to-one conversation attaches and sends a photo from the device photo library. Entry: user is in a direct message thread with another person. Exit: the selected photo is uploaded and displayed inline in the conversation thread. Assumptions: user is authenticated, has granted photo-library access (and optionally location access), has network connectivity, and the device uses the system photo picker. Steps: Conversation view → Photo library (system picker) → Composer with image preview → Sent image displayed in thread **CocoonWeaver: First-time Recording and Transcription** (ios, 12 screens, Productivity) [Refero](https://refero.design/flows/4605) > Happy-path for first-time user onboarding and creating a voice note in CocoonWeaver on iPhone. Steps: Welcome: Cocoon: Ideas → Privacy & Categories → Example: Gratitude Content → Brand + Primary CTA → In-app Microphone Rationale → System Microphone Permission → Ready-to-Record Home → Recording Active (start) → Recording Visual Feedback → Post-recording: Finish Prompt → Transcription Loading → Inbox: Transcribed Note **TikTok: Direct Message — Stickers & Reactions** (ios, 13 screens, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/2799) > Happy-path: a logged-in user opens Inbox, selects an existing conversation, types and sends a text message, reacts to their own message, views the recipient's replies, opens the Stickers panel, browses/creates/adds a sticker pack and sends stickers, then composes and sends a text reply that references a sticker. Entry: Inbox (logged-in). Exit: Conversation updated with new sticker and text messages; composer open. Assumptions: user is authenticated, has a conversation with the contact, has permission to add sticker packs (store access) and the device is iOS (visual keyboard + status bar). Steps: Inbox / Conversation list → Compose message → Message sent with reaction affordance → Recipient reply appears → Continue conversation → Open Stickers tray → Create sticker onboarding → Browse sticker candidates → Message actions (long-press) → Sticker pack preview → Send sticker into conversation → Compose text referencing sticker → Conversation updated with sticker + text **Genie: Conversation Photo Reimagining** (ios, 18 screens, AI Tool, Social Networking) [Refero](https://refero.design/flows/12433) > The flow begins in a friends inbox with an existing conversation from Danny Shnap. The user opens the conversation, enters the camera through a swipe gesture, captures a photo of two gloves, and edits it using annotations and a lighting preset. They then enter a detailed visual prompt and submit it for reimagining. After processing, the transformed image is returned to the conversation and sent as a shared image. The conversation continues with a question-and-answer exchange. The screenshots imply a logged-in user on a mobile device with camera access and messaging permissions. Steps: Friends Inbox → Open Conversation → Camera Gesture Hint → Camera Opening → Camera Ready → Capture Subject → Review Captured Image → Mark Image Subjects → Describe Edit → Choose Lighting → Compose Transformation Prompt → Submit Reimagining → Processing Feedback → View Reimagined Image → Share Image in Chat → Open Shared Image → Reply to Conversation → Continue Conversation **Cake Equity: Inbox message — react and confirm** (web, 2 screens, Business, Collaboration Tools, Productivity) [Refero](https://refero.design/flows/6156) > Happy path: a logged-in user in the company portal opens an Inbox item, reads the message content, and optionally selects an emoji reaction. The interface shows the message header (sender display name and date), the message body, an emoji reaction bar with counts, and a legal/disclaimer footer. After the user clicks an emoji, the reaction count updates and a small confirmation toast appears. Steps: Message detail (read) → React to message (confirm) **Dribbble: Messages: Read, Reply, Delete Conversation** (web, 4 screens, Communication Tools, Job Search) [Refero](https://refero.design/flows/807) > Happy path: a signed-in desktop user opens Messages, switches between Archive and Inbox, opens a single conversation, composes or replies, and triggers deletion which opens a confirmation modal. Steps: Archive (empty) → Inbox list → Conversation view → Delete confirmation modal **Handshake: Report and archive conversation** (web, 5 screens, Social Networking, Job Search) [Refero](https://refero.design/flows/8888) > Desktop inbox flow where a logged-in user reviews a conversation, opens the conversation actions menu, chooses to flag (report) the conversation, provides a reason and optional details in a modal, then submits the report. Submission results in the conversation being archived and removed from the active thread view; a confirmation toast briefly appears. Steps: Open conversation in Inbox → Open actions menu → Open report dialog → Provide reason and details → Conversation archived and confirmation **Missive: Shared Account Setup** (web, 20 screens, Business, Collaboration Tools, Communication Tools) [Refero](https://refero.design/flows/6251) > Happy-path: An admin in Settings > Organizations > Message-sharing opens the Add shared account flow, selects a channel type (email or other), configures whether the account is shared, chooses where incoming messages should be routed (team inbox or users' regular inboxes), sets sharing recipients and auto-archive behavior, creates a custom channel (name + type), configures message/recipient permissions, adds one or more aliases (ID + name) and optional webhook data, and connects the account. On success the new account appears in the Accounts list and is marked as shared. **Peerlist: Start and send direct message** (web, 4 screens, Job Search, Business, Social Networking) [Refero](https://refero.design/flows/1186) > Happy-path: A signed-in user opens the Inbox, creates a new conversation, selects a contact via a centered search modal, and sends a message. **Instagram: Delete a direct message** (web, 3 screens, Social Networking, Entertainment) [Refero](https://refero.design/flows/888) > Happy-path: A logged-in user on a desktop browser opens an individual or group direct message thread, uses the conversation details to choose "Delete chat", confirms the permanent deletion in a modal, and then sees the conversation removed from the messages list with the inbox empty state displayed. Steps: Open conversation and access details → Confirm permanent deletion → Post-deletion inbox and empty state **Nextdoor: Reply to Existing Chat** (web, 4 screens, Communication Tools) [Refero](https://refero.design/flows/11448) > A logged-in Nextdoor member opens the Chats area, selects an existing conversation with San Francisco Chronicle, reviews the previous message, types a response, and sends it. The flow begins in the Chats inbox with no conversation selected and ends with the new message displayed in the conversation and the thread preview updated. Steps: Chats Inbox → Conversation Details → Compose Reply → Sent Reply **ManyChat: Inbox message reply** (web, 4 screens, Business, Productivity, Communication Tools) [Refero](https://refero.design/flows/10351) > Happy-path: A logged-in user opens the Inbox, selects a specific Instagram conversation from the middle conversation list, composes a reply in the right-hand composer, and sends it. Steps: Open conversation → Compose reply → Confirm send shortcut → Send message and observe updates **Fernand: Message Options and Deletion** (web, 5 screens, Business, Communication Tools) [Refero](https://refero.design/flows/11882) > A support agent is viewing a customer conversation in an inbox-style workspace. The entry point is an open conversation thread with a message selected; the agent can inspect the message, open its overflow menu, and choose actions like edit or delete. If delete is chosen, the interface asks for confirmation before removing the message. The flow exits back to the conversation view, where the thread is updated, and the reply/note composer remains available. Assumptions: the user is logged in as a support agent on desktop and has permission to manage messages in the thread. Steps: Open Message Actions → Choose Delete → Confirm Deletion → Conversation After Action → Onboarding Conversation View **Fernand: Conversation Search** (web, 5 screens, Business, Communication Tools) [Refero](https://refero.design/flows/11895) > This is a support inbox search and filtering interaction. The user starts on a conversation list, opens the global search field, reviews the operator hints for filtering by from/is/tag/assigned/after, enters a numeric query that returns no matches, then adjusts the query to a tag-based filter that returns matching conversations. Assumptions: the user is already logged in and viewing the inbox on desktop, with access to search and conversation filters. Steps: Inbox Overview → Search Operator Help → No Match Result → Tag Filter Entry → Filtered Results ##### Task: multi-step form wizard **Chance AI: Image-to-Insight (camera → article)** (ios, 14 screens, AI Tool) [Refero](https://refero.design/flows/7353) > Happy-path: A user discovers the app's 'Search by Seeing' feature on the home/discovery screen, taps the camera CTA to start a visual search, grants camera permission, captures a photo, and waits while a multi-step visual analysis completes. When the analysis finishes the app presents a long-form, scrollable result (title, descriptive text, images, charts, hashtags, external links and suggested specialist agents). The user can scroll through content, open suggested agents or share/create a webpage. Entry: discovery/home screen. Exit: user views or interacts with the generated article and supporting actions (share, open agent, open sources). Assumptions: user is on a mobile device, not yet granted camera permission, and has network connectivity to perform image analysis. Steps: Discovery / Camera CTA → Camera Permission Prompt → Camera + Agent Picker → Captured Image Preview → Analysis Start / Thinking State → Progress: Initial Scan → Progress: Contextual Analysis → Progress: Synthesis & Finalizing → Result: Title + Article Header → Result: Media Gallery & Body Copy → Result: Interactive Indicators & Charts → Suggested Specialist Agents → Actions, Sources & Sharing → Exit / Close Controls **Family: Report Bug via Support** (ios, 13 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/2645) > Happy path: A signed-in user on the Wallets screen opens the support/help modal, chooses to report a bug, selects relevant product areas, fills a bug subject and detailed description, attaches one or more media files, provides contact details (name + email) and the wallet context, then submits the report. Steps: Open Support Modal → Choose Areas (list) → Select Areas → Open Bug Report Form → Type Subject & Description → Open Attach Media → Add First Attachment → Add Additional Attachments → Open Reporter Details → Fill Contact Details → Submit Report → Success Confirmation → Return to Wallets **Dropset: Workout exercise editor** (ios, 7 screens, Health & Wellness, Fitness) [Refero](https://refero.design/flows/6534) > Happy path: A user working through an in-progress workout opens the Add exercises control, chooses to create a custom exercise, names it, selects how it should be tracked (Sets / Duration / Intervals / etc.), confirms, and returns to the active workout where the new exercise appears in the list and a bottom sheet shows a quick action for the next exercise. Entry: in-progress workout list. Exit: workout list updated with new custom exercise and contextual bottom-sheet action. Assumptions: user is in an active workout (logged in state not shown), device is a modern phone in portrait, keyboard interactions are possible. Steps: In-progress workout list → Add exercise menu → Create exercise — name input → Create exercise — name entered → Select tracking type → Confirm tracking type → Updated workout with quick action **Airbnb: Account deletion** (ios, 15 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6460) > Happy path: A logged-in user navigates from Account settings → Privacy → Data privacy → Delete my account, completes a short form (country/state + optional reason), confirms the destructive action, verifies ownership via phone (SMS or call), and submits a deletion request. Entry: Account settings screen. Exit: Deletion request confirmation screen and user is deactivated/logged out. Assumptions: user is authenticated, has phone or email on file, and the app can send SMS/phone calls for verification. Steps: Account settings (list) → Privacy overview → Privacy toggles and Data privacy → Delete account — form (country dropdown) → Country selection list → State selection → Reason selection (optional) → Reason choice → Delete CTA and confirmation prompt → Confirm delete — choose verification → Enter verification code (SMS) → Submit deletion request → Deletion request received → Signed-out / Login or sign-up modal → Sign-in options **Artsy: Add Artwork to Collection** (ios, 7 screens, Marketplace) [Refero](https://refero.design/flows/7619) > Happy path: a signed-in user on a mobile device opens their Profile (Artworks tab) and taps the Add action to add a new piece to their collection. They search for or pick an existing artist, choose an artwork (or skip selecting an existing record), fill or confirm artwork metadata (title, medium, year, materials, rarity, location, notes), add a photo, then save. The system shows a saving progress state and returns the user to their Profile with the new artwork visible in the Artworks list. Steps: Profile — Artworks Tab → Select an Artist → Select an Artwork → Add Details — Metadata Entry (partial) → Add Details — Location, Notes & Photos → Saving Artwork (Progress) → Profile — New Artwork Visible **Julienne: Account signup, Pro subscription and onboarding** (ios, 17 screens, Food & Drink) [Refero](https://refero.design/flows/6475) > Happy-path: a first-time or signed-out user opens the app, is prompted to sign up or continue as guest, chooses Sign in with Apple, confirms sharing/hiding their email, and completes account creation. The user then opens the Pro subscription sheet, selects a plan and starts a free trial. The App Store purchase sheet appears for system-level confirmation (side button). After confirmation the app shows a success message and a thank-you / welcome screen. The user is returned to the app and a short onboarding sequence collects allergy and dietary preference data. Steps: Home + Sign up prompt → Sign in with Apple sheet → Apple sign-in confirmation → Pro promotion open → Plan selection state → Prompt to confirm subscription → Confirm with side button → Purchase success alert → Thank-you / Welcome → Discover screen (Pro badge) → Onboarding — Allergies intro → Allergy selection (cards) → More allergies + free-text → Free-text entry → Dietary preferences list → Onboarding complete → Discover content (personalized) **Trip Way: Location Item Creation** (ios, 2 screens, Travel & Booking) [Refero](https://refero.design/flows/8463) > Mobile creation form for a location/place entry. Steps: Initial create form (empty link / image) → Form with pasted link and selected image **District: Edit Profile Details** (ios, 3 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12812) > Happy path for editing a user profile on a mobile screen. The user opens Edit Profile, reviews existing personal information, selects a gender from a dropdown under Additional details, optionally fills other fields such as anniversary, then saves changes with Update profile. Steps: Open Gender Menu → Select Gender → Review Updated Field **Canopi: Create Multi‑Day Event Template** (ios, 10 screens, Scheduling Tools) [Refero](https://refero.design/flows/6065) > Happy path: User creates a multi-day event template called "Chasing Horizons." Entry point is a template canvas with placeholder cards. The user opens a date editor, switches to Multi‑Day, selects a start date and an end date from a calendar, optionally adds start/end times using a time picker, enters an event title, and saves the template. Exit point is the canvas populated with a new card preview of the event (title, date range, times). Assumptions: user is on a mobile device, is allowed to create templates (no extra permissions required), and the template canvas supports adding multiple cards. All interactions happen in stacked modal sheets over the canvas. Steps: Template Canvas (start) → Date & Time Editor (modal) → Calendar — pick a date → Start Date populated → Select End Date → Date Range Confirmed → Time Picker → Times Added & Title field active → Enter Event Title → Save Template & Canvas Preview **Clipchamp: Submit feedback slide-over** (web, 5 screens, Design & Media Editing) [Refero](https://refero.design/flows/10907) > Happy path: From the Clipchamp home view the user opens a right-side feedback slide-over, selects a feedback classification (compliment, problem, suggestion), completes the feedback form (text, optional screenshot/recording/attachment, contact preference), submits it, views a brief progress indicator, then sees a thank-you confirmation. Steps: Feedback entry → Feedback form (empty) → Feedback form (filled) → Submitting state → Submission confirmation **Mercury: Recipient banking and upload** (web, 4 screens, Business, Finance) [Refero](https://refero.design/flows/11040) > Happy path: a user is on the first step (Recipient) of a multi-step payment/onboarding flow. They review or select an existing recipient, choose ACH as the payment method, enter routing and account numbers (the routing number triggers a detected bank pill), and optionally upload a supporting document (invoice). Steps: Enter recipient bank details → Reveal upload affordance → Upload in progress → Document preview and attachment **Glorify: Subscription cancellation and downgrade** (web, 12 screens, Design & Media Editing) [Refero](https://refero.design/flows/3782) > Happy path: A signed-in workspace admin opens Billing in Workspace settings and initiates cancellation of a paid trial/plan. The product displays an initial confirmation modal, then launches a multi-step cancellation questionnaire (reasons, optional comment, confirmation questions and an offer to stay). The user proceeds through the form, confirms they still want to cancel, and submits. The flow ends with a thank-you/processing message and the workspace billing view updated to the Free plan with no payment method on file. Steps: Billing settings (Workspace) → Cancel Trial confirmation dialog → Cancellation form — intro → Select cancellation reason → Choose 'Other' reason → Enter free-text feedback → Confirm intent to cancel (binary) → User confirms decision (YES) → Retention offer (discount) prompt → User declines retention (chooses to continue cancel) → Cancellation submitted — thank-you message → Billing settings — Free plan state **PlayStation: PlayStation Account Signup and Verification** (web, 19 screens, Marketplace, Entertainment) [Refero](https://refero.design/flows/7719) > Happy-path signup and verification for a PlayStation / Sony account initiated from the PlayStation storefront. Steps: Homepage (Entry) → Modal: Welcome / Start → Select Country & Language → Provide Date of Birth (empty) → Provide Date of Birth (filled) → Enter Email & Password (empty) → Enter Email & Password (filled) → Residential Address (empty) → Residential Address (filled) → Profile Setup (Online ID & Name) → Profile Setup (filled) → Marketing Preferences & Agreements → Account Created Confirmation → Update Account Prompt → Personalized Content Consent → Processing / Loading → Verify Email Prompt → Email Verified Confirmation → Return to Homepage (Exit) **Faire: Retailer Account Deactivation** (web, 7 screens, Business, Marketplace) [Refero](https://refero.design/flows/13097) > Happy-path account deactivation in a retailer settings area. The user starts on Security & privacy, opens the deactivate account form, selects a reason, adds an optional comment, confirms the email for the account, and submits the deactivation request. The flow ends with a success modal confirming the account has been deactivated. Steps: Security & Privacy → Deactivation Reason → Reason Selected → Email Confirmation Step → Confirm Email → Deactivate Account → Deactivation Confirmed **LinkedIn: Easy Apply to a Job** (web, 9 screens, Business, Social Networking, Job Search) [Refero](https://refero.design/flows/811) > Happy path: A logged-in candidate browsing job listings opens a job detail and chooses "Easy Apply." A modal guides them through a multi-step application: confirm contact info, upload/select a resume, answer employer questions, review the completed application, and submit. The flow shows progress in a horizontal bar and allows saving/discarding a partial application. Steps: Job detail / Easy Apply entry → Contact info → Resume upload → Additional questions (initial) → Validation errors shown → Save application confirmation → Review application → Final review and submit → Submission confirmation **Wrike: Trial signup and workspace onboarding** (web, 28 screens, Business, Productivity, Project Management, Collaboration Tools, Paperwork, Scheduling Tools) [Refero](https://refero.design/flows/1495) > Happy-path from marketing hero CTA to an in-product workspace with an interactive onboarding tour. Entry: user lands on marketing site (hero CTA) and enters an email to start a two-week trial. They progress through an email validation state, a multi-step signup modal (name, phone, company, company size, team/role, invite teammates), then a workspace provisioning loading screen, followed by first-time in-product modals and a short guided tour that highlights Table view features and allows adding fields and structuring tasks. Exit: user lands in the populated project table with assignees and can continue using the product. Steps: Marketing landing (expanded) → Hero CTA — enter email → Email validation error → Email accepted → Signup modal — initial → Signup modal — filled → Company size question — options → Company size — selection → Team & role inputs — blank → Team & role inputs — filled → Invite teammates → Preparing workspace (loading) → Welcome trial modal inside workspace → Workspace with assistant sidebar → Trial reminder modal → Contextual tooltip — Table view → Tour step — add tasks → Tour step — create custom field → Tour step — restructure hierarchy → Post-tour workspace state **Wittl: Job Description Image Upload** (web, 3 screens, Job Search) [Refero](https://refero.design/flows/5365) > Happy path: an employer creates a new job and adds a header/hero image in the Job description step of a multi-step job creation wizard. Steps: Job Description — Add Image → Uploading Image → Image Uploaded Preview **Сontra: Create a Service Listing** (web, 15 screens, Job Search) [Refero](https://refero.design/flows/4874) > Happy path: a logged-in freelancer opens their profile > Services tab, starts "Create a service", picks a category, fills the multi-step listing form (Overview → Deliverables → Description → Price → Image), uploads an image and publishes the service. Entry: user profile Services tab. Exit: published service visible on the user's Services listing and profile. Assumption: user is logged in, has a completed profile and permission to create services. **Wealthsimple: Account creation and phone verification** (web, 15 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8494) > Happy-path desktop registration that starts from a public marketing landing and ends at the authenticated dashboard. Entry: user clicks sign-up from the marketing hero. The sign-up is a multi-step form on the right column (preferred name, email, password) followed by phone collection and SMS OTP verification. Assumptions: user is on a desktop-browser, not signed in, and has access to a valid mobile number for SMS. Exit point: authenticated dashboard showing suggested accounts. Steps: Marketing hero / entry → Sign up — initial form → Password validation — too short → Password suggestion — weak → Password strength — strong → Show password control → Password visible state (final) → Enter phone number — initial → Phone validation — unsupported number → Phone validation — accepted → OTP entry — prompt → OTP validation — invalid code → OTP validation — accepted → Transition / loading → Authenticated dashboard **Maze: Feedback submission** (web, 5 screens, Business, Communication Tools) [Refero](https://refero.design/flows/3692) > Happy-path: an invited participant opens the survey, reads a short intro, and clicks Get started. They are presented with a multi-step feedback form with a progress indicator. Step 1 asks the reason for feedback (single-choice list); after choosing an option they continue to an optional/extra-comments text field. After submitting, they see a thank-you / completion screen. Steps: Welcome / Intro → Select reason for feedback → Confirm selection / Enable continue → Optional additional feedback → Thank you / Completion **Mercury: Invite and onboard team member** (web, 10 screens, Business, Finance) [Refero](https://refero.design/flows/11071) > Happy path: An admin (already signed in) opens the Users list from Settings, clicks Invite → Invite a user, completes a three-step modal (User Details → Assign Role → Issue Card), reviews the invitation, and sends it. The flow ends with an on-screen confirmation and the invited user appearing in the Users list with a right-hand detail panel showing their pending invite, assigned role and card options. Entry: Settings → Users (signed-in admin). Exit: Confirmation screen and return to Users list with the new invite added. Assumptions: the actor is logged in with permission to invite users; using a desktop browser (layout indicates wide-screen). Steps: Users list → Invite menu → Invite — User Details (start) → Invite — User Details (filled) → Assign role — preview (Employee) → Assign role — Admin selected → Issue card options → Review and send invite → Invite sent confirmation → Users list with invite detail panel **Navan: Add company credit card** (web, 8 screens, Travel & Booking) [Refero](https://refero.design/flows/9399) > Happy-path flow for an admin to add a company credit card from an account setup area and have it appear in the organization's payment methods list. Steps: Intro: Add a company card → Credit card details — blank → Credit card details — filled → Billing address — blank → Billing address — filled → Confirm payment method — review → Confirm payment method — validation error → Payments list — card saved **ElevenReader: Publish manuscript to ElevenReader** (web, 20 screens, Books, AI Tool) [Refero](https://refero.design/flows/9195) > Happy-path author publishing flow for preparing a written manuscript as an ElevenReader title and submitting it for review. Entry: user is on a draft/read editor with a completed manuscript and clicks through a multi-step publisher wizard. Exit: confirmation modal + Bookshelf view showing the submitted title in pending review. Assumptions: user is authenticated (account/profile indicated in header), using a desktop browser, has at least one uploaded manuscript or sample chapter available, and can upload images and fill text fields. Steps: Manuscript editor (start) → Author profile (blank) → Author image uploaded → Author language selection → Complete author profile fields → Book details — title and cover placeholder → Cover image upload preview → Enter subtitle and edition → Book description and copyright → Series and author assignment → Genre and category selection → Distribution details → Payouts and agreements (start) → Payouts — opt-out and accept terms → Finalize and preview audio (initial) → Open voice selector → Select recommended voice → Confirm submission → Submission confirmation modal → Bookshelf — book pending review ##### Task: music player playlist **MD Vinyl: Music Player Onboarding** (ios, 3 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/12538) > A 3-step in-product onboarding tour for a music player interface. It starts on the Now Playing screen, then introduces the Library, then the Queue. The user is assumed to be in the music player with no special permissions required. Each step uses a tooltip anchored to the bottom control bar and a Next button, ending with a primary Start to Explore action that exits the tour and lets the user use the player normally. Steps: Now Playing Intro → Library Intro → Queue Intro **Epidemic Sound: Select and play stem mix** (web, 4 screens, Music Streaming) [Refero](https://refero.design/flows/5109) > Happy-path: A user visits the 'Laid-back Lofi' playlist page, opens the mix selector in the persistent bottom player, chooses the Melody-only stem, and plays the track. Steps: Playlist overview with player → Open mix selector → Select Melody stem → Play selected Melody mix ##### Task: notifications inbox **Gizmo: Manage conversation from inbox** (ios, 3 screens, Entertainment, Social Networking, AI Tool) [Refero](https://refero.design/flows/10036) > User lands on a dark-themed chat inbox showing recent conversations (entry point: chat list). From the list the user opens a contextual actions menu for a specific conversation and chooses Delete. The UI immediately removes that conversation from the list (exit point: updated chat list). Assumptions: user is authenticated and on a mobile device; chat list is the primary inbox view and actions are available per conversation. Steps: Chat inbox → Conversation contextual actions → Updated chat inbox **Substack: Appearance & Visual Preferences** (ios, 3 screens, News, Social Networking) [Refero](https://refero.design/flows/4506) > Happy path: a logged-in user opens the Settings list, chooses 'Appearance', adjusts visual and layout preferences (auto-play, theme, inbox density, default tab), and observes immediate UI feedback. Steps: Settings list → Appearance (light mode) → Appearance (dark mode applied) **Savee: Email & Social Login Onboarding** (ios, 13 screens, Social Networking, Inspiration Board, Reference) [Refero](https://refero.design/flows/5267) > Happy path sign-in and first-run onboarding for a new user on an iOS device. Entry: user launches the app (not authenticated). The user can choose social sign-in (Apple/Google) or email sign-in. If they choose email, they enter their email/username, receive a magic-login link by email, open that link which completes authentication, then the app shows a post-login welcome/onboarding carousel and an optional push-permission prompt. Exit: user finishes the carousel and lands on the main content feed. Assumptions: user is on a mobile device, network connectivity is available, user has access to the email account they provide, and the device can open links back into the app. Steps: Welcome / Sign-in choices → Email / Username entry → Validation error shown → Email corrected / Ready to continue → Email sent confirmation → Login email contents → Opening link / Loading → Authentication success + Notifications prompt → Onboarding carousel — welcome → Onboarding carousel — saving interaction → Onboarding carousel — following people → Onboarding carousel — premium prompt → Main content feed **Artsy: Account Signup and Onboarding** (ios, 33 screens, Marketplace) [Refero](https://refero.design/flows/7591) > Happy-path: An unauthenticated iOS user opens the app, taps to sign up with email, completes a three-step account creation (email → password → name & consents), then completes a short personalization step (choose a content category). The app shows short loading/splash screens, surfaces a curated collection (Curators' Picks), asks for notification permission, and lands the user in the logged-in Home feed. Entry: app launch on iOS with no account. Exit: user is logged in and viewing the For You / Home feed. Assumptions: mobile (iPhone) viewport, network available, user creating a new account (not social sign-in). Steps: Landing / Sign up prompt → Email input (keyboard visible) → Email submit / Loading → Password entry (rules shown) → Password typing and validation → Full name + consents (initial) → Full name validation error → Full name entry and consents checked → Welcome splash (account creation) → Onboarding preference selection → Preference submit / Loading → Curators' Picks preview → Explore More / Curators' Picks expanded → Post-onboarding splash → Home feed initial load → In-app notification prompt (soft) → System notification permission (iOS) → Home feed with notifications handled → Finalized personalized feed → Optional: Reopen permission or onboarding flows **Cushion: Inbox Management** (web, 5 screens, Communication Tools, Collaboration Tools) [Refero](https://refero.design/flows/12509) > Logged-in user views the Inbox in a workspace messaging product. The flow starts with unread messages visible in the Active inbox and a top-right overflow menu for bulk actions. The user can open the menu, choose grouping or bulk actions, and then see the inbox update. The flow ends with the inbox cleared of unread items and an empty-state message indicating everything has been sorted. Steps: Inbox View → More Menu Open → Bulk Action Selection → Inbox Cleared → Completed State **Skiff: Archive message from inbox** (web, 2 screens, Communication Tools, Scheduling Tools, Paperwork) [Refero](https://refero.design/flows/1893) > Happy path: A logged-in user on a desktop mailbox workspace views their Inbox and archives a thread. Steps: Inbox list (pre-archive) → Archive confirmation and list update **Fernand: Support Inbox Review** (web, 2 screens, Business, Communication Tools) [Refero](https://refero.design/flows/11893) > This is a support inbox screen in a web-based messaging tool. The user is viewing the conversation list with tabs for Inbox, Waiting for answer, and All, plus toolbar actions in the top-right. The flow appears to start with a populated inbox list showing a conversation card and status badges, then changes to a filtered state where no conversations match the selected filters. Assumptions: the user is already logged in, has access to the inbox, and is using a desktop device. The visible toast indicates a system update about snoozed conversations. Steps: Inbox List → Empty Filter State **Nextdoor: Remove Chat** (web, 4 screens, Communication Tools) [Refero](https://refero.design/flows/11447) > A logged-in user removes an existing conversation from the Chats area. The user opens the conversation’s overflow menu, selects the removal action, confirms the decision in a modal, and returns to the chat list. The removed conversation disappears from the list, the conversation pane switches to a general empty state, and a temporary confirmation banner offers an Undo action. Steps: Open Chat Actions → Choose Remove Chat → Confirm Removal → View Removal Confirmation **Missive: Schedule Snooze for Email** (web, 4 screens, Business, Communication Tools, Collaboration Tools) [Refero](https://refero.design/flows/6281) > Happy path: A user selects an email in the inbox, opens the message actions menu, chooses Snooze, picks a preset or custom snooze time, optionally toggles related options, confirms the snooze, and sees a success toast. Steps: Open message actions → Open snooze presets → Adjust snooze options → Snooze confirmation and feedback ##### Task: notifications settings **Fingerprint: Manage email notifications** (web, 3 screens, Business, Development Tools) [Refero](https://refero.design/flows/11176) > Happy path: A signed-in user navigates to Settings → My notifications to review and edit email notification preferences for their account/workspaces. The page displays three notification types (Summary, Status changes, Usage spikes) and frequency controls for the Summary. The user disables the "Status changes" notification, clicks Save, and receives a success confirmation. Steps: Notification settings overview → Edit notification preference → Save confirmation **Instacart: Flyer notifications settings** (web, 3 screens, Food & Drink, Marketplace, Food Delivery) [Refero](https://refero.design/flows/10095) > Happy path: A signed-in user browsing the Flyers section opens the notification manager, reviews available retailers, enables notifications for selected stores, and receives an immediate confirmation. Steps: Flyers listing → Manage flyer notifications (initial) → Manage flyer notifications (confirmation) **Reown: Theme and Notifications Settings** (web, 4 screens, Development Tools, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/5737) > Happy path: A signed-in user opens the Account > General section to change the visual theme and confirm notification preference. Steps: General settings overview → Open theme options → Theme applied (dark preview) → Update email notifications **Threads: Manage profile notifications** (web, 4 screens, Social Networking, Entertainment) [Refero](https://refero.design/flows/10661) > Happy path: A logged-in user viewing a public profile (The New York Times) opens that profile's notification options, selects notification types (Threads and Replies), and receives immediate confirmation. Steps: Profile feed (notification control available) → Notification options modal (initial state) → Select notification types → Confirmation and return to feed **Twitch: Stream Notification Preference** (web, 4 screens, Social Networking, Entertainment, Video Streaming) [Refero](https://refero.design/flows/13586) > A logged-in Twitch viewer is watching a live channel and opens the channel notification menu beneath the stream controls. The menu presents four notification preferences: Personalized, Always, Go Live Only, and Off. The viewer changes the selection from Personalized to Go Live Only. The selected preference is visibly marked, and the menu remains open in the final state. No additional confirmation step is shown. Steps: Open Notification Menu → Review Notification Options → Choose Live Alerts → Confirm Selected Preference **Pitch: Manage account settings** (web, 12 screens, Productivity, Collaboration Tools, Design & Media Editing) [Refero](https://refero.design/flows/698) > Happy-path: A signed-in desktop user opens their workspace, accesses Account settings from the workspace menu, and iteratively updates personal account details: avatar, display name, email preferences, and presentation notification subscriptions. Steps: Workspace landing (Recents) → Account settings modal (overview) → Change account avatar (initial) → Change account avatar (preview) → Save confirmation / secondary state → Avatar dialog (post-save) → Change account name (entry) → Change account name (edit) → Email preferences → Notifications (overview) → Presentation subscriptions (expanded) → Subscription update confirmation **Hello Ivy: Pause Notifications Temporarily** (web, 3 screens, Productivity, Project Management, Collaboration Tools) [Refero](https://refero.design/flows/4443) > Happy path: A signed-in user opens the Notifications overlay from a project/task board to temporarily pause incoming notifications. They open the Notifications panel, click the 'Pause notifications' control, select a duration from the dropdown (30 minutes, 1 hour, 2 hours, 4 hours, 8 hours, or Until tomorrow), and the UI updates to show notifications are paused. Entry: user is on the Tasks kanban-style board and clicks the notifications icon. Exit: notifications are paused and the Notifications panel shows a paused state/confirmation. Steps: Open Notifications → Open Pause Duration Menu → Confirm Paused State **Pinterest: Chat feedback on help article** (web, 2 screens, Inspiration Board, Reference, Social Networking) [Refero](https://refero.design/flows/1078) > Happy path: A user reads the Help Center article titled “Edit notification settings” and interacts with a floating chat-feedback widget anchored to the right side of the screen. The widget asks the user to rate their chat experience (1–5), pick descriptive tags about what went well, answer whether their issue was resolved (Yes/No), optionally add freeform feedback, and then submit. **Zendesk: Personal settings editor** (web, 2 screens, Business) [Refero](https://refero.design/flows/1395) > User is in the Settings > Personal area of a web dashboard. **Medium: Writer Follow Modal** (web, 3 screens, Social Networking) [Refero](https://refero.design/flows/11108) > A centered "Writers to follow" modal overlays the page (desktop). It lists recommended authors with avatar, short bio, and a right-aligned primary action button for each (Follow). The happy path: user opens this modal, taps Follow on one or more recommended writers, sees an immediate success confirmation and the Follow button change to a Following state that exposes notification and unfollow controls. Steps: Recommended Writers List → Follow Confirmation and Controls → Updated List State **Twitch: Follow Alert Customization** (web, 6 screens, Social Networking, Entertainment, Video Streaming) [Refero](https://refero.design/flows/13617) > A logged-in creator edits an existing “New Follow” alert variant in the Alerts dashboard. The creator changes the alert message, selects a font, adjusts the text color using the color picker, reviews the live preview, and can save or discard the changes. The flow begins with the alert variant editor open and ends with the customized preview ready to be saved; no save action is shown. Steps: Review Alert Variant → Update Alert Message → Choose Alert Font → Apply Font Selection → Adjust Text Color → Review Pending Changes **Whereby: Pre-call Settings Setup** (web, 6 screens, Productivity, Communication Tools) [Refero](https://refero.design/flows/1972) > Happy-path: A signed-in user configures audio, video and meeting preferences from the pre-join (pre-call) screen, then returns to join the meeting. Steps: Pre-call lobby → Settings — Video & Audio → Settings — Effects → Settings — Notifications → Settings — General → Settings — Advanced **TidyCal: Create Booking Type** (web, 19 screens, Business, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/12490) > An admin starts on the Booking types list, opens the New booking type menu, and creates a new service booking type by editing details across multiple sections. The happy path assumes the user is logged in with scheduling permissions and is configuring an existing account from desktop. The flow ends back on the Booking types list with the new booking type saved and visible in the list. Steps: Booking Types List → Choose Booking Type → Basics Setup → Basic Details Filled → Availability Settings → Availability Applied → Booking Restrictions → Multi-Booking Options → Package Setup → Advanced Calendar → Guest Settings → Attendee Questions → Generated Questions → Question Set Complete → Notifications Settings → Custom Calendar Invitation → Visibility Settings → Save Confirmation → Updated Booking Types List **Fernand: Account Details Settings** (web, 4 screens, Business, Communication Tools) [Refero](https://refero.design/flows/11896) > Happy path for editing personal account settings in a workspace settings area. The user enters the Account details screen from the left navigation under Personal settings, updates profile fields such as name, optionally changes the profile image via the avatar menu, adjusts notification and timezone preferences, toggles email subscription, and saves the changes. Steps: Edit Account Details → Avatar Options → Avatar Selection State → Saved Confirmation **GoFundMe: Fundraiser Settings and Delete Confirmation** (web, 5 screens, Fundraising) [Refero](https://refero.design/flows/11749) > Logged-in fundraiser owner opens Edit fundraiser and lands on the Settings tab. The screen lets the user adjust fundraiser page options, notification preferences, and sharing-related AI content settings. If the user chooses to delete the fundraiser, a confirmation modal appears requiring acknowledgment before the delete action proceeds. After confirmation, the user is taken back to the Fundraisers area where a success banner confirms the fundraiser was deleted and the deleted item appears in the list. Steps: Edit Fundraiser Settings → Delete Confirmation → Ready to Delete → Deletion in Progress → Deleted Fundraiser List **Runey: Update Profile Picture** (web, 2 screens, Business, Productivity, Project Management) [Refero](https://refero.design/flows/11682) > An authenticated user updates their profile picture from the Profile section of Settings. The flow starts with the existing personal information form and an empty/default avatar. The user selects a new image through the upload control, after which the avatar is replaced and a confirmation notification appears. Other account settings remain unchanged. Steps: Profile Settings → Profile Picture Updated **Fingerprint: Create environment with usage notification** (web, 6 screens, Business, Development Tools) [Refero](https://refero.design/flows/11170) > Happy path: a logged-in user with Settings access opens Settings > Environments, clicks New environment, fills name and optional description, enables monthly usage notification, sets a numeric API-call threshold, and creates the environment. Steps: Environments overview → Create environment modal (empty) → Fill form and enable notifications → Validation on missing threshold → Enter threshold value → Environment created and listed **Fingerprint: Rename workspace** (web, 4 screens, Business, Development Tools) [Refero](https://refero.design/flows/11169) > Happy path: A signed-in user with permissions opens Settings → General (Workspace settings), clicks the Change name control, edits the workspace name in a modal dialog, and confirms the change. The modal closes and the new name is shown in the workspace card along with a success toast. Steps: Workspace settings overview → Rename modal (pre-filled) → Rename modal (edited) → Confirmation and updated settings **fal: Delete Model Set** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/13527) > The user manages saved model sets in a modal. They choose to delete the custom “Gimli” model set, review a warning that the deletion is permanent, and confirm the action. The system removes the set from the list and displays a success notification. The user is assumed to be in an editing or management state with access to model sets. Steps: Model Set Management → Deletion Confirmation → Deletion Complete **Cal: Automated Reminder Setup** (web, 5 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3483) > Happy path: A logged-in user opens the Workflows section, creates a new reminder workflow (or chooses a template), configures the trigger (time relative to an event) and action (send email/SMS), saves the workflow, and returns to the Workflows list where the new workflow appears. Steps: Workflows Hub → Workflow Editor — Initial → Workflow Editor — Adjust Trigger → Save Workflow → Workflows List — Confirm **Cushion: Workspace Logo Upload** (web, 5 screens, Communication Tools, Collaboration Tools) [Refero](https://refero.design/flows/12533) > Happy path for updating a workspace logo from Workspace settings. The user starts on the General section of Workspace settings, opens the logo upload control, selects an image file in the upload modal, waits while the file uploads, and returns to the settings screen where the new logo is shown and a success toast confirms the change. Steps: Workspace Settings → Select File → Uploading → Upload Complete → Updated Logo Preview **X: Block participant in Direct Messages** (web, 2 screens, Social Networking) [Refero](https://refero.design/flows/992) > Happy path: A logged-in user opens Messages, selects a conversation, navigates to the conversation info panel, taps the "Block" action for the other participant, and confirms the block in a modal. Steps: Conversation info → Block confirmation modal **Cursor: Configure repository routing rule** (web, 4 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9952) > Happy path: A signed-in user opens the Cloud Agents settings area to add a repository routing rule. From the Cloud Agents overview they expand the Repository routing section, add a new rule by entering a description and choosing a repository, submit the rule, and see a success toast while the new rule appears in the routing list. Entry: user navigates to Settings > Cloud Agents while authenticated. Exit: the new routing rule is persisted and shown in the Repository routing list. Assumptions: user is logged in, has permission to edit Cloud Agents settings, and has at least one connected repository available in the repository selector. Device: desktop browser (wide layout). Steps: Cloud Agents settings overview → Enter routing rule details → Creation confirmation → Routing rule listed **Synthesia.io: Profile Picture Update** (web, 3 screens, AI Tool) [Refero](https://refero.design/flows/13761) > The user is already signed in to a workspace and opens the Account information section within User settings. They upload a new profile picture using the profile picture control. After the upload completes, the placeholder is replaced with the uploaded image and a success notification appears. The flow ends with the updated profile picture visible in the account settings. Steps: Account Information → Profile Picture Upload → Upload Confirmation **Navan: Create travel policy** (web, 9 screens, Travel & Booking) [Refero](https://refero.design/flows/9394) > Happy-path: A logged-in admin is guided through a step-by-step policy creation wizard to set booking notifications, flight spend rules, additional flight settings, hotel spend rules, and hotel settings, then reviews and saves the overall travel policy. Entry: Admin clicks Start policy setup from the landing hero. Exit: Admin saves the policy (or continues to car policy) and is returned to the main configuration area where the new policy is listed and can be edited. Assumptions: user is authenticated as an admin with permission to create company travel policies, using a desktop/laptop viewport (two-column layout). Steps: Welcome / Start policy setup → Booking notifications → Flight spend - traveler preview → Flight spend - admin modal → Other flight policy settings → Hotel spend policy → Other hotel policy settings → Travel policy overview → Configuration list - policy applied **ManyChat: Account deletion confirmation** (web, 5 screens, Business, Productivity, Communication Tools) [Refero](https://refero.design/flows/10361) > Happy path: A signed-in user navigates to Settings → General and initiates permanent account deletion. The interface opens a confirmation modal that requires the user to type the word “Delete” to enable the final Delete action. After confirming, the app closes the modal and returns the user to an accounts overview / channel list where success toasts appear, then the toasts disappear leaving an empty-account state with an Add New Account CTA. Entry: user is inside account Settings (logged in). Exit: user lands on the accounts overview (no channel) with the deleted account removed from the list. Steps: Settings - General → Confirmation modal (empty) → Confirmation modal (typed) → Accounts overview with toasts → Final empty-state ##### Task: offline mode sync **Willow: Offline Mode Setup** (ios, 9 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/13250) > Happy path for downloading and enabling an offline dictation model. The user enters an Offline mode settings screen, reviews what offline mode does, starts a model download, waits through download and extraction states, and ends with offline mode enabled and the downloaded model available. Steps: Offline Mode Settings → Offline Mode Explanation → Download Start → Download Progress → Download Near Completion → File Extraction → Model Loading → Download Complete → Offline Mode Enabled **Subo: iCloud Sync Status Check** (ios, 3 screens, Finance) [Refero](https://refero.design/flows/8591) > Happy path: The user opens the iCloud Sync Status screen, verifies the current sync state, optionally taps the 'Check for changes' control to force a manual check, watches a transient 'checking' / 'syncing' state, and then sees the final success state with updated information. Steps: Status — Synced → Checking for updates → Status — Synced (Updated) **Comet: Browser Sync Setup** (ios, 3 screens, AI Tool, Utility) [Refero](https://refero.design/flows/10949) > Happy path: A logged-in user (avatar visible in the header) discovers a prompt to enable cross-device browser sync, opens the Browser sync onboarding, and starts a new sync. The onboarding explains requirements, shows the user's account card, and offers a single prominent CTA to start a new sync. Tapping that CTA navigates to a settings/detail sheet that lists connected devices and sync options (Add device, Sync everything toggle, Customize), plus actions to leave or delete the sync. Steps: Home — sync prompt → Browser sync intro → Sync settings and devices **Comet: Browser sync settings** (ios, 3 screens, AI Tool, Utility) [Refero](https://refero.design/flows/10950) > Happy path: A signed-in user opens Browser sync settings, reviews devices and sync options, chooses to permanently delete the sync chain, confirms the deletion, and sees a post-deletion state instructing them to relaunch the product to re-enable sync. Steps: Browser sync overview → Delete confirmation → Post-deletion acknowledgement **Cashcaded: Offline Setup and Apple Pay Introduction** (ios, 5 screens, Finance) [Refero](https://refero.design/flows/14118) > A mobile onboarding sequence helps the user enable offline intelligence, then introduces Apple Pay tracking automation. The user starts on the first onboarding card, taps Download, waits while required resources are downloaded, and reaches an “All set” confirmation. The onboarding then presents the next card, where the user can choose “Show me how” or “Skip.” The screenshots show an iPhone-sized experience with the main product interface dimmed behind the onboarding sheet. No authentication or account permissions are visible; the user is assumed to be on a device that supports the illustrated offline and Apple Pay features. Steps: Offline Intelligence Introduction → Resource Downloading → Download Progress → Setup Complete → Apple Pay Tracking Introduction **TikTok: Manage Offline Videos** (ios, 3 screens, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/2942) > User views an in-context 'Offline videos' settings panel while a set of videos are downloading, then clears all downloaded videos and is shown an empty offline library. Steps: Offline videos settings (downloading) → Confirm clear all downloads → Offline videos empty state **komoot: Manage Offline Regions** (ios, 3 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8411) > Happy path: a signed-in user opens Offline Maps, views the Downloaded tab, removes a downloaded region (Bay Area), and returns to an empty Downloaded view with a call-to-action to download a new region. Entry: user navigates to Offline Maps (Profile tab > Offline Maps). Exit: Downloaded tab shows no regions and a CTA prompting first download. Assumptions: user is logged in, has at least one downloaded region, device has enough permissions to delete local data, and network state is not required for deletion. Steps: Downloaded Regions List → Region Actions — Delete Option → Empty Downloaded State / CTA **YouTube Music: Manage offline downloads** (ios, 3 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11378) > Happy path: user opens the "Downloads & storage" settings on a mobile (dark theme) device to review and remove offline downloads. From the settings list they tap "Clear downloads", confirm the action in a modal, and the app removes downloaded content and updates displayed storage usage and toggles. Steps: Downloads & storage overview → Clear downloads confirmation → Post‑clear settings state **TikTok: Offline videos manager** (ios, 5 screens, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/2808) > Happy path: A signed-in user navigates from the main Settings and privacy list to the Cache & Cellular section, opens the Offline videos setting, chooses how many videos to download, and starts the background download. The UI shows a bottom sheet/modal for choosing quantity, then a downloading state with progress and controls. Steps: Settings and privacy - landing → Cache & Cellular / Support section → Choose download quantity → Downloading state (initial) → Downloading state (in progress) **TIDE: Save & Download Meditation** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/6756) > Happy path: A signed-in user views a meditation detail screen, opens the share/options sheet, chooses to download the item, acknowledges the download info, and returns to the detail screen with a saved/downloaded state visible. Steps: Meditation Detail → Share / Action Sheet → Download Confirmation Modal → Saved State Return **Adobe Photoshop: Immediate cloud save** (ios, 3 screens, Design & Media Editing) [Refero](https://refero.design/flows/4818) > Happy path: The user is working in an image editor with a document open, opens the settings sheet, chooses the Save now action to push the file to cloud storage, and receives an on-screen confirmation. **Brink: Download and Save Event Recording** (ios, 3 screens, News) [Refero](https://refero.design/flows/10927) > User is viewing a vertically scrolling list of recorded events (each item shows artwork, title, date, short description, a play-duration pill on the left and a download control on the right). Entry: user is on the recordings list screen. Happy path: user taps the Download control for one event → download control immediately switches to an in-progress state (“Saving”) → a transient top toast appears confirming “Downloaded” → the download control becomes a persistent Saved state (green with check). Exit: the list remains visible with that item showing the Saved state. Assumptions: user is on a mobile phone in dark mode, has network/storage available, and is permitted to save content locally. The screenshots show only the visual states and microinteractions (no explicit progress bar or percent complete is shown). Steps: Recordings List — Download idle → Download in progress → Download completed — confirmation and saved state **YouTube: Smart downloads onboarding and management** (ios, 4 screens, Entertainment, Video Streaming, Social Networking) [Refero](https://refero.design/flows/2571) > Summary: The flow shows the Downloads area for a signed-in user (accessed from the Library section). Entry: user opens Downloads and sees an empty state with a Smart downloads promotional card. Happy path: user taps TURN ON to enable Smart downloads → the UI shows a loading state while recommended videos are fetched → Smart downloads begins; individual items show download progress (or 'waiting to download') → items complete and the list displays downloaded thumbnails and metadata with updated storage used. Exit: user remains in Downloads where they can open or manage downloaded videos or tap the settings gear to change Smart downloads limits. Assumptions: user is signed in, the device has Wi‑Fi available (note text in screen references Wi‑Fi), and there is available device storage. Steps: Empty state + Smart downloads CTA → Loading recommended downloads → Smart downloads in-progress list → Completed downloads list **komoot: Save and download route** (ios, 6 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8351) > Happy-path for saving a public route to the user's account and optionally downloading it for offline use. Steps: Route detail overview → Overflow actions menu → Save modal (name input) → Toggle offline save → Saving in progress → Saved-route confirmation **Waterllama: Apple Health Sync Setup** (ios, 7 screens, Health & Wellness) [Refero](https://refero.design/flows/11536) > Happy path: A signed-in user navigates from the app's settings list to enable two-way synchronization with Apple Health. The user taps the 'Apple Health Sync' row, views a short product-modal explaining sync, and follows a two-step onboarding screen which then triggers the iOS Health permission sheet. After granting permission, the app shows the sync settings screen with a toggle and an optional 'Sync all-time data now' action. The user starts a full data sync, sees a progress modal, and receives a success confirmation when finished. Steps: Settings - Health Row → In-app Intro Modal → Allow Health Access Instructions → iOS Health Permission Sheet → Sync Settings Screen → Sync Progress Modal → Sync Completion Confirmation **YouTube Music: Single track download and feedback** (ios, 3 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11339) > Happy-path for downloading a single track from the in-player song screen. Entry: user is on the song detail player (song playing/paused) for "cellophane" by FKA twigs. Exit: the track is saved for offline use and the UI reflects a Downloaded state. Assumptions: user is signed in and has permission to download (subscription or entitlement not shown); network is available; the player screen contains primary playback controls, metadata, and an inline Download control. Steps: Song detail — ready to download → Downloading — transient feedback → Downloaded — final confirmation **Apple TV: Movie download and completion** (ios, 4 screens, Entertainment, Video Streaming) [Refero](https://refero.design/flows/11095) > Happy path: a signed‑in iPhone user is viewing a movie detail screen (Store tab). From this screen they can start playback or initiate a download for offline viewing. Tapping the download control starts a download; the control switches to an in‑progress symbol, then updates to a completed state when finished. The rest of the detail content (title, synopsis, Play button, metadata, Trailers list, bottom navigation) remains accessible throughout. Steps: Movie detail (idle) → Download tapped — starting → Downloading (in‑progress) → Download complete **YouTube Music: Download single track from album** (ios, 4 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11342) > Happy-path: A signed-in user opens an album detail screen, views the tracklist, initiates an offline download for a specific track, and watches the UI transition from idle → downloading → downloaded. Entry: user is on the album detail screen (album header, play controls, tracklist, and mini-player visible). Exit: the selected track is marked as downloaded (check icon displayed in the track row) and the mini-player continues to show the playing track. Assumptions: the user is on a mobile device and has permission to download (signed-in / has required feature enabled). Steps: Album detail — tracklist view → Initiate per-track download → Download progress feedback → Download complete — persistent state **Netflix: Downloads Settings** (ios, 3 screens, Entertainment, Video Streaming) [Refero](https://refero.design/flows/10594) > Happy path: A signed-in mobile user opens the app's settings sheet, navigates to the Downloads section, taps the 'Delete All Downloads' option, confirms deletion in a modal, and returns to the settings screen with the storage indicator updated. Steps: Settings — Downloads → Confirm Delete Modal → Downloads Updated **Google Translate: Downloaded Languages Management** (ios, 3 screens, Reference, Utility) [Refero](https://refero.design/flows/12144) > This is a language management screen for offline translation. The user opens a list of downloaded translation languages, can remove an existing downloaded language, and can browse and download additional languages from the available list. Steps: Downloaded Languages List → Remove Confirmation → Updated Downloaded List **komoot: Manage Offline Data** (ios, 3 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8420) > Happy path: user opens the Data Usage screen (assumed logged into the app with permission to manage local data) to free up space by removing saved individual routes. From the Data Usage screen the user taps the active Clear button for "Routes saved offline", sees a confirmation modal, confirms deletion, and returns to the Data Usage screen where the route data shows as cleared and the Clear action is disabled. Steps: Data Usage — overview → Confirm clear routes → Data Usage — updated **Podimo: Settings and Downloads Management** (ios, 3 screens, Entertainment) [Refero](https://refero.design/flows/12282) > A mobile settings screen for a media app, accessed from the Library tab. The user can manage notifications, download/storage behavior, and streaming quality. The flow starts on the Settings screen and includes opening a delete confirmation dialog for downloaded content, then returning to the settings view after dismissal or completion. Steps: Settings Overview → Delete Download Confirmation → Updated Settings View **YouTube Music: Podcast show detail and saved-filtered episodes** (ios, 2 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11354) > User is on a podcast/show detail screen (hosted by Complex) on a mobile device. The screen displays large cover artwork, show title and description, primary actions (share, save/bookmark, more), and a horizontal filter row (Latest, In progress, Unplayed, Downloaded, Played). The episode list appears below and a persistent mini-player is anchored to the bottom. Entry: user navigated to the show detail from content or search. Exit: user either filters the episodes (Downloaded), plays an episode via the list or mini-player, or navigates back via the system back button. Steps: Show detail — Save action → Filter episodes — Downloaded **YouTube Music: Configure Smart downloads** (ios, 4 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11368) > User is in the Downloads area of a music player on a mobile device (portrait). A Smart downloads card appears at the top letting the user pick how many tracks the system should automatically download overnight when on Wi‑Fi. The happy path shown: open Downloads → read Smart downloads card → adjust the slider to change count/storage estimate → confirm by tapping Turn on → Smart downloads card dismisses and the downloads list is visible. Assumptions: user is signed in, on a phone, has a music library with downloaded content, and has Wi‑Fi/network indicators visible. Steps: Downloads with Smart downloads card (initial) → Adjust slider to change download quantity → Enable Smart downloads → Browse downloads list after enabling ##### Task: open bank account identity verification **Dropbox: Vault setup and biometric enablement** (ios, 9 screens, Business, Collaboration Tools, Productivity) [Refero](https://refero.design/flows/2467) > Happy path for setting up a protected “Vault” inside an existing account. Entry: user taps the Vault feature (onboarding card or account feature). Flow creates a vault, prompts the user to set a numeric PIN, then offers to enable Face ID for quick unlocking. Exit: user lands in the empty Vault with action tiles (Scan document, Upload photos, Create folder, Take photos). Steps: Vault introduction → Creating vault (progress) → Set PIN (empty) → Set PIN (partial) → Set PIN (multiple digits entered) → In-app Face ID offer → System Face ID permission dialog → Face ID authentication (animated) → Vault landing (empty state) **PayPal: Sign-in and biometric enablement** (ios, 7 screens, Finance) [Refero](https://refero.design/flows/7268) > Happy-path mobile sign-in for an existing PayPal account. Entry: user opens the app (logged out). Exit: user lands on the account home/dashboard (logged in). Assumptions: user has an existing account and network connectivity; device can show the on-screen keyboard and supports Face ID/biometric prompts. The flow: enter email (or switch to phone), tap Next → enter password → tap Login → brief setup/loading → optional Face ID prompt → arrive at account home. Steps: Enter email / identifier → Email typed with keyboard visible → Password entry screen (email shown) → Submit credentials / loading → Biometric prompt (Face ID) → Landing on account home → Password focused state (alternate layout) ##### Task: order tracking delivery status **Drinkit: Cart review and edit** (ios, 3 screens, Food & Drink) [Refero](https://refero.design/flows/11007) > Mobile cart/checkout sheet for a specific location (West Hollywood). Entry: user opens the in-app cart for a selected store. The sheet lists line items with images, sizes, modifiers, per-item price, and +/- quantity controls. A prominent red banner indicates the store is currently closed (opens at 7:00 AM), and the bottom CTA is disabled with the label "Not open yet." There is a payment-method selector and a large right-aligned total that updates as items or quantities change. Additional interactions include revealing a delete action (red trash) for a line item and viewing nutritional summary chips. Exit: user either edits the cart (change quantities, remove items, change payment) or leaves; checkout cannot be completed while the store is closed. Steps: Cart overview (store closed) → Edit item and reveal delete → Nutrition & final summary **Podimo: Audiobook Detail and Download Management** (ios, 4 screens, Entertainment) [Refero](https://refero.design/flows/12277) > Happy path on an audiobook detail screen. The user arrives at a title’s detail view (Ansia), reviews metadata such as format, author, narrator, rating, and description, then can start playback, add the title to their library, or download it for offline use. The screen updates inline to show download progress and completion, and a confirmation banner appears after downloading. Steps: Audiobook Details → Download Progress → Download Progress Update → Download Complete **222: Notification Delivery Selection** (ios, 3 screens, Lifestyle, Dating & Relationships) [Refero](https://refero.design/flows/9326) > Simple, single-screen preference editor where the user sets the delivery method for two notification categories. Entry: user is on the "notification preferences" screen (likely reached from a settings area). The screen lists categories ("invites" and "experience updates") each with a current delivery label and a disclosure control. The happy path: user taps a category row, a rounded popover appears with two delivery options, the user selects an option, the popover closes and the row label updates to reflect the new choice. Exit point: user can use the back button in the top-left to leave the screen. Assumptions: user is on a mobile device and is already at the notification preferences screen; categories shown are editable inline via the popover control. Steps: Notification Categories → Open Delivery Options → Selection Applied **On: Restock notification signup** (ios, 4 screens, Marketplace) [Refero](https://refero.design/flows/9594) > Happy path: On the Cloudsurfer Next product detail (mobile, portrait) the user opens the size selector, sees some sizes marked out-of-stock with a bell icon, requests a restock notification for an out-of-stock size, confirms via a modal, and receives a success confirmation. Steps: Open size selector → Tap notify bell for out-of-stock size → Confirm notification signup → Return to size list with updated state **ZARA: Check Order Status via Chat** (web, 7 screens, Marketplace) [Refero](https://refero.design/flows/1607) > Happy-path of a guest user using the Zara customer service chat widget to check an order status. Steps: Customer Service landing → Chat opened — greeting and top choices → Order submenu → Bot requests order number → User enters order number → Assistant returns status guidance → Close chat confirmation **Shopify: Purchase Order — Mark as Ordered** (web, 3 screens, Business) [Refero](https://refero.design/flows/5402) > Happy path: A merchant reviews a draft purchase order, confirms they want to place the order, and the system updates the purchase order status to Ordered so the merchant can later receive incoming inventory. **DailyArt Magazine: Account Order Management** (web, 2 screens, Education, Magazines & Newspapers, Marketplace) [Refero](https://refero.design/flows/3529) > Happy path: A logged-in user visits their My Account dashboard and navigates to the Orders list to act on a failed order. Steps: Orders List → My Account Dashboard **On: Order Cancellation** (web, 4 screens, Marketplace, Fitness, Lifestyle) [Refero](https://refero.design/flows/13834) > A logged-in customer views an order in their account, chooses to cancel an order that is still in the Order Placed state, confirms the request, and receives confirmation that the cancellation request was received. The order status then updates to Cancelled, with the timeline, messaging, and available action changing accordingly. Steps: Order Details → Cancellation Confirmation → Cancellation Request Received → Cancelled Order **Shop: Delete Order History Item** (web, 3 screens, Marketplace) [Refero](https://refero.design/flows/13285) > An authenticated customer is viewing the Orders section of their account. They open the order actions menu for an existing order, choose Delete, and the order is removed from the list. The Orders section then displays an empty state with an "Order deleted" confirmation message. The flow starts on the populated Orders view and ends after the deletion is completed. Steps: Orders List → Order Actions Menu → Deletion Confirmation **Lemni: Conversation status update** (web, 3 screens, AI Tool, Business, Productivity) [Refero](https://refero.design/flows/4588) > Happy path: a support agent opens a conversation from the inbox, inspects the message and contact details in the right-side details panel, and updates the conversation status from Open to In Progress. Entry: agent is signed in and viewing the Inbox. Exit: the Details panel reflects the updated status. Steps: Inbox → Open conversation → Open status control → Select new status **IKEA: Select delivery or collection method** (web, 5 screens, Marketplace, Home & Decor) [Refero](https://refero.design/flows/8257) > Happy-path: A shopper on the IKEA checkout Delivery or Collect step chooses how they'd like to receive their order (store collection or a delivery option), reviews the order summary, optionally opens the shopping bag overlay to edit quantities/remove items, and then proceeds by clicking Continue (or using Express Checkout). Steps: Delivery or collect selection → Open shopping bag overlay → Remove an item from the bag → Verify updated cart → Return to delivery selection and continue **Uber Eats: Delivery & address configuration** (web, 9 screens, Food & Drink, Food Delivery) [Refero](https://refero.design/flows/431) > User is on the restaurant checkout screen and configures how they want their order delivered or picked up. **New Balance: Checkout and Order Confirmation** (web, 8 screens, Marketplace, Sports) [Refero](https://refero.design/flows/11860) > Happy-path ecommerce checkout for a single item. The user starts in the bag/cart with one product and proceeds through shipping details, delivery method selection, payment entry, order review, external payment authorization, and final order confirmation. Steps: Bag Summary → Shipping Details → Shipping Method Selection → Payment Entry → Payment Validation → Order Review → Payment Authorization → Order Confirmation **On: Product Purchase Checkout** (web, 6 screens, Marketplace, Fitness, Lifestyle) [Refero](https://refero.design/flows/13828) > The flow begins with a single product already in the user's shopping bag. The user reviews the item and order summary, then proceeds to checkout. At checkout, the user reviews or confirms delivery information, selects a shipping method, enters credit card details, and submits payment. The payment is verified through an external Revolut authorization prompt. After successful authorization and processing, the user reaches an order confirmation screen with delivery, payment, and order details. The user is assumed to be an existing customer with delivery information prefilled and is using a desktop browser. Steps: Shopping Bag Review → Delivery and Payment Entry → Completed Payment Details → Payment Authorization → Payment Processing → Order Confirmation **Programa: Update Project Status** (web, 4 screens, Project Management, Business) [Refero](https://refero.design/flows/13411) > The user is viewing the Overview for Project #254 and changes its status from Planned to Active. The flow starts on the project overview with the project currently marked Planned. The user opens the status selector, reviews the available states, and selects Active. The project overview remains visible and the status badge updates to Active. The user appears to be logged in with access to manage the project. **IKEA: Back-in-stock notification signup** (web, 4 screens, Marketplace, Home & Decor) [Refero](https://refero.design/flows/8254) > Happy-path: A desktop user viewing their Shopping Bag sees an item that is unavailable for a selected fulfilment option and chooses to sign up to be notified when it is back in stock. Steps: Shopping Bag - cart overview → Notification modal - empty state → Notification modal - filled form → Shopping Bag - confirmation state ##### Task: pay invoice bill **District: Dining Save and Reserve** (ios, 3 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12857) > Happy path for a dining savings experience. The user starts from a dining discovery/home screen, then opens an in-context help or promo overlay that explains two options: reserve a table or pay a bill. The user can switch between the two tabs, view the instructional content, and proceed to either book a table or pay the restaurant bill. **District: Add Card for Booking Payment** (ios, 9 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12835) > The user is reviewing a movie booking and needs to select a payment method before completing payment. Starting on the booking review screen, the user opens the payment method selector, chooses to add a credit or debit card, and enters card details using the card-scanning option. The system requests camera access, scans the card, auto-fills the card number and expiry date, and enables the Add Card action. After the card is added, the user returns to the booking review screen, where the Visa card is selected and the Pay now action is available. The flow assumes the user has an active booking, is within the displayed payment timer, and has permission to add a payment card. The screenshots end before payment authorization or booking confirmation. Steps: Booking Review → Payment Method Selection → Card Entry → Camera Permission → Scan Card → Review Card Details → Add Card Processing → Updated Booking Review → Card Ready for Payment ##### Task: payment checkout **TikTok: Checkout — add address & payment** (ios, 25 screens, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/2783) > Happy path: user moves from the cart to checkout, adds a shipping address, selects/creates a payment method, optionally adds seller notes per item, reviews shipping methods and totals, then places the order. Steps: Cart overview → Checkout loading → Checkout main (no address) → Add address (empty form) → User enters address → State selection (picker) → State scrolled selection → Phone country picker → Address partial validation → Save address → Add new card modal → CVV helper modal → Checkout updated with address & card → Payment method modal → Shipping method list → Add order note (open) → Order note saved → Order summary review → Place order attempt → Payment failure & exit **Blackbird: Promo Code Checkout** (ios, 4 screens, Food & Drink) [Refero](https://refero.design/flows/9029) > Happy path: User reviews the Blackbird Pay checkout for a one-time purchase, opens the Promo Code entry, types a promo code, applies it, and returns to the checkout where the promo is shown and the order total is updated. Steps: Checkout Summary — rewards active → Promo Code Input (empty) → Enter Promo Code → Checkout Summary — promo added **SSENSE: Add Credit Card During Checkout** (ios, 6 screens, Marketplace) [Refero](https://refero.design/flows/7110) > Happy path for adding a new credit card during checkout on a mobile device and returning to the order summary to place the order. Steps: Checkout summary (no payment) → Select payment method → Add credit card — name entry → Add credit card — card data entry → Processing → Checkout summary (payment selected) **On: Checkout: payment, summary & promo** (ios, 5 screens, Marketplace) [Refero](https://refero.design/flows/9604) > Compact mobile checkout screen where a signed-in user reviews payment options, applies a discount code, reviews the order summary and totals, and completes payment. Steps: Payment selection and Pay CTA → Order summary expanded → Discount entry (keyboard active) → Applying discount (loading) → Discount invalid error **On: Checkout and Payment Completion** (ios, 11 screens, Marketplace) [Refero](https://refero.design/flows/9601) > Happy-path checkout starting from the shopping bag. Steps: Bag review / Checkout CTA → Checkout entry & delivery form → Enter recipient name → Address validation (empty/invalid) → Type address to trigger suggestions → Address suggestions list → Address auto-fill verified → Select shipping method → Payment method selection → Review order & Pay → Payment error / Verification failed **voilà: Buy Indonesia 1GB Data** (ios, 6 screens, Travel & Booking) [Refero](https://refero.design/flows/9295) > Happy path: A user browsing available eSIM/data plans for Indonesia selects a plan (1 GB) -> views plan details (validity, activation delay, operator) -> initiates payment -> enters card details (or chooses 'Link') -> submits payment -> sees processing and then a success confirmation. Steps: Plan list → Plan details → Payment sheet (method selection) → Card entry → Submitting payment → Payment confirmation **District: Event Ticket Booking** (ios, 10 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12871) > A logged-in user starts on the event details screen for “Sip and Act at 3 Art House” in Mumbai, selects two regular tickets, reviews the booking, pays using the available District Money balance, and waits for confirmation. The system then displays the confirmed booking and mobile QR ticket, along with venue and directions information. The flow is shown on a mobile device and ends on the confirmed ticket details screen. **Asos: Mobile Checkout: Bag to Payment** (ios, 33 screens, Marketplace) [Refero](https://refero.design/flows/2824) > Happy-path mobile checkout starting from the bag, proceeding through checkout summary (delivery destination, bag preview), entering promo/voucher codes, choosing shipping (postal or click & collect), adding a shipping address if needed, selecting a payment method (credit card / PayPal / pay-later options), entering payment details and validating them, and finally returning to checkout totals with the Place Order CTA. Steps: Bag - Review items → Checkout Summary (top) → Shipping address selection → Payment summary & Place Order → Codes & E-Gift Cards (promo tab) → Voucher / Gift Card tab → Promo code entry (keyboard) → Promo code invalid error → Promo code applying (loading) → Add postal address (contact) → Address finder / manual entry hint → Payment section - method list → Order totals and Place Order (with warning) → Sales tax information → Payment method selected (PayPal) → Add Payment - method options → Credit/Debit card form (empty) → Card number entry (numeric keypad) → Card validation error → Exit checkout confirmation **SSENSE: Checkout — add address & card** (ios, 16 screens, Marketplace) [Refero](https://refero.design/flows/7108) > Happy path: user opens the shopping bag (2 items) and proceeds to Checkout. On Checkout they add a new shipping address using an address-autocomplete, save it (including phone), the order recalculates for the new destination, then the user adds a credit card and sets the billing address (same as shipping). They return to Checkout where the chosen shipping, delivery option and payment method are displayed, then attempt to place the order. The flow ends with a validation error (invalid card) shown. Entry: Shopping Bag (with 'Go to Checkout'). Exit: attempted Place Order (error shown). Assumptions: mobile portrait, user has items in bag; the interface supports address autocomplete and multiple payment options. Steps: Shopping Bag → Checkout (empty shipping) → New Shipping Address (blank form) → Street address autocomplete → Address fields auto-filled → Phone entry & save options → Checkout — destination update notice → Checkout (shipping populated) → Select payment method list → Add Credit Card (name) → Add Credit Card (card details) → Processing card (loading) → Checkout — payment selected → Order review with totals → Payment error — invalid card → Post-error state / next steps **TikTok: Add and select card at checkout** (ios, 7 screens, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/2971) > Happy-path summary: From the checkout screen the user adds a new credit/debit card using the camera scanner (or manual entry), corrects any invalid/missing fields, optionally saves the card and uses the billing address toggle, then returns to checkout where the new card is selected and the user places the order. **Asos: Add Credit/Debit Card** (ios, 5 screens, Marketplace) [Refero](https://refero.design/flows/2936) > Happy-path summary: From an empty payment-wallet state the user chooses a payment option, selects Credit/Debit Card, fills the card entry form (card number, expiry month/year, name on card, CVV), optionally toggles "Save for future use", and taps "Use This Card". The system validates the inputs and either saves the card and returns to the previous context (successful add) or shows inline validation errors (invalid data). Steps: Select Payment Type → Card Details Form (initial) → Enter Card Number and CVV → Select Expiry Date → Submit and Validate **Omio: Payment selection and card entry** (ios, 4 screens, Travel & Booking) [Refero](https://refero.design/flows/5186) > Happy path: The user reaches the booking checkout payment screen (Barcelona › Madrid, Wed Apr 9) on an iPhone. They choose a payment method (Apple Pay, Card, or PayPal). If Card is selected the inline card form expands with a prefilled cardholder name, then the user enters card number, expiration date and CVV using the numeric keyboard. The user optionally checks "Save this payment method for future bookings". Optional accordions for billing address and voucher code remain collapsed. Steps: Choose payment method → Expand card form → Enter card number → Complete card details and save **DailyArt Magazine: Guest Billing & Payment Checkout** (web, 3 screens, Education, Magazines & Newspapers, Marketplace) [Refero](https://refero.design/flows/3526) > Desktop checkout flow for completing an order (happy path). Steps: Checkout — Billing & Order Summary → Country Selector Expanded → VAT Input with Inline Feedback **Brilliant: Start 7-day Premium trial** (web, 5 screens, Education) [Refero](https://refero.design/flows/4348) > Happy-path flow for starting a 7-day premium free trial and completing payment details. Steps: Trial details and CTA → Checkout — payment method selection → Checkout — billing details expanded → Inline validation — card error state → Trial activation confirmation **AVNIER: Cart to checkout** (web, 5 screens, Marketplace, Lifestyle) [Refero](https://refero.design/flows/12007) > Happy path starts on the cart page and continues into checkout. The user is assumed to have items already added to the cart and to be proceeding on desktop. From the cart, they can review items, change quantities, remove products, or continue to checkout. In checkout, they enter contact, shipping, and payment details, choose a shipping method and payment method, review the order summary, and submit payment. The flow ends with payment submission, with an error state shown if the card is declined. Steps: Cart review → Checkout entry → Order details filled → Field validation → Payment declined **Acne Studios: Checkout with gift wrap and Klarna payment** (web, 5 screens, Marketplace) [Refero](https://refero.design/flows/4384) > Happy-path checkout on desktop: user starts from a product detail (gift card) and adds it to the bag, reviews the cart in a right-side drawer, then proceeds to the checkout / delivery details screen where they enter or confirm shipping address, select a delivery method, optionally add gift wrapping, choose a payment method (card, Alipay, or Klarna shown), agree to terms and submit payment. Steps: Product detail and cart drawer → Delivery details & payment selection → Gift wrapping drawer → Confirm details and choose Klarna → Payment error / validation **Nike: Checkout Purchase** (web, 9 screens, Marketplace, Sports) [Refero](https://refero.design/flows/2041) > Desktop checkout happy path starting from the shopping bag (user is signed in). Entry: user begins on the Bag screen with items, a promo applied and a Checkout button. The user chooses delivery (Pick Up or Ship), provides or selects a shipping address / pickup point, selects shipping speed, selects a payment method, and completes an order review. Exit point: user is redirected to the external PayPal payment flow (or proceeds to an internal card payment) to finalize payment. Assumptions: user is logged in (account name visible), on a desktop browser, items and promo are already in the bag, taxes and shipping are calculated after address/speed selection. Steps: Cart Review → Delivery Mode Selection → Pickup Points List → Ship Form - Contact & Address → Address Saved → Select Shipping Speed → Delivery Summary & Payment Reveal → Payment Method Selection → Order Review & Submit **Pattern Brands: Guest Checkout with Shipping and Payment** (web, 6 screens, Marketplace) [Refero](https://refero.design/flows/1716) > Happy-path checkout on desktop using a two-column layout: left column walks the user through Information → Shipping → Payment; right column is a persistent order summary with discount entry and cost breakdown. Steps: Cart — Out of stock notice → Information — Contact & Shipping Form → Information — Discount validation & SMS opt-in → Shipping — Method selection → Payment — Enter payment details → Payment — Different billing address & save info **Peak Performance: Desktop checkout - 3 step purchase** (web, 10 screens, Marketplace) [Refero](https://refero.design/flows/9487) > Happy-path desktop checkout for a logged-in user who starts from the mini cart (slide-out), proceeds through a three-step checkout (1. Personal information, 2. Delivery, 3. Payment), and finishes on the payment submission screen. Steps: Cart slide-out → Checkout - Personal Information (start) → Order details expanded → Order summary updated after removal → Address entry completed → Delivery options selection → Delivery confirmed - proceed to Payment → Payment options presented → Credit card entry with validation → Place order (submit) **SSENSE: Cart Checkout — Address & Payment** (web, 5 screens, Marketplace) [Refero](https://refero.design/flows/7122) > Happy-path summary: A signed-in customer reviews items in their shopping bag, proceeds to the checkout screen, fills in shipping details, selects a shipping method and a payment method, then attempts to place the order. Steps: Shopping Bag (Cart Review) → Checkout — Empty Shipping Form → Checkout — Inline Address Validation → Checkout — Address Corrected → Payment Authorization Error **DailyArt Magazine: Checkout Billing & Payment** (web, 3 screens, Education, Magazines & Newspapers, Marketplace) [Refero](https://refero.design/flows/3527) > Happy-path checkout for a single-product purchase. Steps: Checkout — Billing & Order Summary → Country Selector — List Open → Country Selector — Search/Filter ##### Task: payroll **Mercury: Schedule Payment Amount & Recurrence** (web, 4 screens, Business, Finance) [Refero](https://refero.design/flows/11041) > This is the 'Amount' step of a multi-step payment scheduling flow where the user confirms the amount, chooses the sending account, sets the send/start date, and optionally configures a recurring payment. Entry: user arrives here after selecting a recipient. Exit: user clicks Next to continue to categorization/details/review steps. Assumptions: user is authenticated, has at least one funding account and a recipient already selected, and is on a desktop or large-screen device. The UI shows estimated arrival timing and allows toggling and configuring recurrence before moving forward. Steps: Confirm Amount & Select Source → Pick Send / Start Date → Enable Recurring Payment → Configure Recurrence End & Preview ##### Task: permissions access request **Google Gemini: Save full-size image to library** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/3212) > User opens a full-screen image preview and attempts to download the high-resolution file to the device. The app shows a transient download status, triggers the system photo-library permission sheet (iOS style) if access isn't granted, and then completes the save with a confirmation toast. Entry: user is viewing an image in the app's preview. Exit: image is saved to device photo library and a success message is shown. **Willow: Voice Dictation Onboarding** (ios, 6 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/13230) > A first-time onboarding exercise introduces Willow's voice dictation feature. The user starts from the dictation tutorial, grants microphone access through an in-product prompt and the iOS permission dialog, then records speech. After stopping the recording, Willow displays the transcribed text in the editor and offers a Continue action. The user is assumed to be on an iPhone and viewing this onboarding step; microphone permission may not yet be granted. Steps: Start Dictation Tutorial → Review Microphone Request → Grant System Permission → Record Voice Input → Finish Recording → Review Transcription **StrideGate: Onboarding, Permissions & Purchase** (ios, 22 screens, Health & Wellness) [Refero](https://refero.design/flows/7202) > Happy-path onboarding -> permissions -> app-selection -> subscription -> final setup. Entry: fresh install or first-run onboarding (user not yet granted permissions or set goal). Exit: user has granted Screen Time and Health access, selected apps to block, chosen a paid plan and landed in main dashboard. Assumptions: user is on iOS, has Screen Time and Health available, Face ID or Side Button purchase confirmation available, and network connectivity for App Store purchase. Steps: Welcome → Feature - Touch Grass First → Set Daily Goal - default → Set Daily Goal - alternative selection → Grant Permission Intro → System - Screen Time Prompt → System - Allow Access (Face ID) → Screen Time Approved → Select Apps - prompt → System - Select Apps to Lock → Apps Selected Confirmation → Choose Plan - pricing list → Confirm Plan & Continue → App Store Confirmation → Purchase Success → Request Health Access → Permission Granted - Select Apps CTA → Purchase Processing Feedback → Final Health Toggle Confirmation → Main Dashboard - Home **Bond: Find Friends from Contacts** (ios, 6 screens, AI Tool, Social Networking) [Refero](https://refero.design/flows/14375) > The user enters a Bond onboarding experience focused on finding friends. They advance from an introductory screen, respond to the iOS contacts permission prompt, choose how contacts should be shared, review contacts discovered from the device, and send connection requests. The flow begins on an iOS device with Bond open and no visible contacts permission decision made. It assumes the user can grant contacts access and that Bond can retrieve contact records. The flow ends on the discovered-contacts list with selected connection requests shown as “Requested.”. Steps: Friend Discovery Introduction → Contacts Permission Request → Contact Sharing Choice → Discovered Contacts → Contact Profile Loading → Connection Requests Sent **tone: Save Edited Image** (ios, 3 screens, Design & Media Editing) [Refero](https://refero.design/flows/13865) > The flow begins in an image editor with an edited floral image and export controls visible. The user selects Export, which triggers an operating-system permission prompt asking to allow the tool to add photos and videos to the photo library. After the user selects Allow, the prompt closes and the editor displays a confirmation message stating that the image was saved to the photo library. The flow assumes the user is on a mobile device, has an image open for editing, and has not yet granted the required photo-library permission. The flow ends after a successful save confirmation. Steps: Image Editor → Photo Library Permission → Save Confirmation **Ground News: Enable Local News** (ios, 4 screens, News) [Refero](https://refero.design/flows/7571) > Happy path: A user on the News tab elects to enable Local News. From the Local News hero screen they tap the primary CTA to get started. The app shows an in‑app prompt asking the user to provide their device location (with an alternative to skip). If the user consents, the system presents the OS-level location permission dialog. After granting permission the app resolves the device's location and displays a populated local-news feed for the detected city. Steps: Local News Hero → In-app Location Request → OS Permission Prompt → Populated Local News Feed **komoot: Take and Save Photo During Navigation** (ios, 6 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8380) > Happy-path summary: From the map screen the user opens an embedded camera to take a picture, reviews the captured photo, and chooses to save it. Because the app needs permission to write to the system Photos library, the UI first shows an in‑app rationale then triggers the platform (iOS) permission sheet. After granting permission, the user returns to the map navigation view. Entry: user is on the map (navigation/elevation) screen. Exit: navigation/map screen with active route header. Assumptions: user is signed in/have route loaded; device is iOS; camera and location hardware permissions are available (camera already allowed by the OS since camera view launches). Steps: Map - Current Elevation → Camera Capture → Photo Review → In‑App Photo Access Rationale → System Photo Library Permission → Map - Follow the Map (Return) **DailyArt: Save image to Photos** (ios, 3 screens, Education, Magazines & Newspapers) [Refero](https://refero.design/flows/3319) > Happy path: From a full-screen artwork viewer on an iPhone, the user taps the Download/Save control, iOS asks for Photos access (system permission dialog). If the user allows access, the app saves the image to the Photos library and shows a temporary success toast. **Remy: Calendar event enablement** (ios, 4 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/9765) > Happy path: user enables in-app event toggles to add circadian‑based items to their device calendar. Tapping a toggle triggers the OS calendar permission request. After granting access the app opens the device's New Event editor prefilled (title, start/end times, repeat, calendar). The user saves the event and returns to the app, where the list updates to reflect the item(s) added and exposes controls to delete or remove items from the calendar. Steps: Calendar setup — select events → OS calendar permission prompt → Native New Event editor (prefilled) → Updated setup list — manage added items **BeReal.: Save BeReal to Photos** (ios, 4 screens, Social Networking) [Refero](https://refero.design/flows/4903) > Happy-path: A logged-in iPhone user views a BeReal in their Friends feed and saves that post to the device photo library. Entry: user is on the 'My Friends' feed viewing a single BeReal with the caption input and action icons visible. Exit: the image is written to the Photos library and a transient 'Saved!' confirmation is shown. Assumptions: user is on iOS (dark mode), signed in, and has not previously granted Photos permission (so the system permission prompt appears). **YouTube Music: Voice Search to Artist Results** (ios, 6 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11362) > Happy path: A signed-in iOS user opens the YouTube Music search, types a query, then switches to voice search. The app requests microphone permission; after allowing, the app displays a full-screen listening UI, captures spoken input, shows intermediate recognition, and finally performs the search returning the artist results page. Entry: search screen with keyboard visible. Exit: artist results displayed with playback controls. Assumption: user is on a mobile device (iPhone), logged into YouTube Music and on the app’s search tab. Steps: Text Search & Suggestions → Microphone Permission Prompt → Speak Now (Start Recording) → Live Partial Transcript → Final Transcription Detected → Artist Results & Playback Controls **pushr: Pushr First Workout Setup** (ios, 31 screens, Health & Wellness, Fitness) [Refero](https://refero.design/flows/13783) > A first-time iPhone user is introduced to pushr, learns how push-up tracking works, grants camera, optional location, and Health permissions, and completes an initial tracked workout. After finishing, the user sees an encouragement message, a review request, and a pushr PRO subscription offer. The happy path enables a three-day free trial through Apple purchase confirmation, then returns the user to a PRO welcome message and the main workout dashboard. Steps: Welcome → Tracking Explanation → Camera Permission Explanation → Camera Access Request → Location Permission Explanation → Location Access Request → Health Sync Explanation → Health Access Selection → Health Access Confirmation → Setup Processing → Active Workout → Workout Result → Review Request → PRO Offer → Yearly Plan View → Free Trial Selection → Apple Subscription Confirmation → Purchase Success → PRO Welcome → Workout Dashboard **ChatGPT: Create Image from Photos** (ios, 14 screens, AI Tool) [Refero](https://refero.design/flows/7020) > Happy path: A signed-in mobile user (iOS) opens the chat composer, taps the Add (+) button, selects Photos, grants the app access to the photo library (or uses the limited picker), selects multiple images, writes a prompt and sends. The app processes the images (requests notification permission to notify when done) and finally returns a generated image based on the provided photos. Steps: Main Chat Composer → Attachment Options → Photo Access Additional Request → System Photo Picker → Composer with Photo Thumbnails → Submit Request → Processing Card & Notification Prompt → User Grants/Dismisses Notifications → Processing Placeholder → Generated Image Display → Alternative Permission Paths (optional) → Count & Selection Feedback → Composer Post-Send State → Final Result and Reactions **Patreon: Privacy Erasure Request** (web, 5 screens, Social Networking, Fundraising) [Refero](https://refero.design/flows/1295) > Happy-path: a logged-in user opens Patreon’s Privacy Center (privacy policy content visible) and initiates a data privacy action. From the policy/landing they open a privacy controls modal offering two choices (Download or Erase). The user selects Erase, grants required account permissions on an authorization/consent screen, then reviews an erase confirmation modal and confirms. The system returns a success state inside the modal and updates the header with an active request indicator. **Calendly: Gmail Connection and Email Composer** (web, 7 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/14034) > Happy path for connecting a Gmail account to a contact-management workspace and opening a new email composer. The user starts logged in to the workspace with access to Contacts and chooses to connect Gmail. They select a Google account, review and approve the requested email permissions, then return to the workspace and see a success confirmation. The final state is a New email composer using the connected Gmail account. The flow assumes a desktop browser, an authenticated workspace user, access to the selected Google account, and permission to authorize Gmail access. **Stripe: Mailchimp installation and permissions** (web, 3 screens, Finance, Business) [Refero](https://refero.design/flows/1316) > Happy path: A logged-in Stripe account admin visits the Stripe App Marketplace Mailchimp listing, clicks Install, reviews the permission request in a right-side confirmation modal, confirms installation, and sees a success confirmation with an option to view the installed integration. **Steel: Connect GitHub Account** (web, 3 screens, AI Tool, Development Tools) [Refero](https://refero.design/flows/14304) > The user starts in the profile settings area with GitHub listed as a disconnected connected account. They select the add control for GitHub, review an authorization request on GitHub, and approve access for Steel.dev. After authorization, they return to profile settings and see GitHub marked as connected. The flow assumes the user is already signed in to Steel.dev and GitHub, has permission to manage their account connections, and is using a browser capable of redirecting between services. **Whereby: New account signup and room creation** (web, 12 screens, Productivity, Communication Tools) [Refero](https://refero.design/flows/1985) > End-to-end new-user onboarding that starts from the marketing homepage and guides a visitor to create an account, verify their email (magic link or manual code), choose a plan, create a persistent meeting room name, and grant camera/microphone permissions to enter the room. Steps: Landing / Marketing → Sign-up form (start) → Validation error (name) → Sign-up form (ready) → Check your email (magic link prompt) → Enter verification code → Verification in progress / code entered → Choose a plan → Create meeting room (prompt) → Room name conflict → Room name accepted → Camera & microphone permission **Runey: Project Sharing Settings** (web, 4 screens, Business, Productivity, Project Management) [Refero](https://refero.design/flows/11695) > Happy path starts on a project dashboard for a project called 'Website Redesign.' The user opens the Share action, which launches a centered sharing modal over a dimmed dashboard. In the modal, the user can enable a public link, copy the generated URL, and control what the client can see or do through visibility and permission toggles. The flow ends with feedback confirming either that the public link is enabled or that the link was copied to the clipboard. Steps: Project Dashboard → Share Link Setup → Permission Configuration → Link Copied Confirmation **Parallel: Add Profile Video** (web, 13 screens, Job Search) [Refero](https://refero.design/flows/1037) > Happy path: A signed-in user opens their Profile > About Me editor, chooses to add a short profile video, grants camera & mic permission, records (with an optional practice preview), reviews the clip, confirms upload, waits for upload/sync to finish, and returns to the About Me editor where the video is shown as a thumbnail with edit/remove options. Entry: user is in the Profile modal (About Me tab). Exit: About Me tab shows the uploaded video and controls. Assumptions: user is logged in, using a desktop browser with camera/microphone available and browser prompts supported. Steps: About Me editor (Profile modal) → Recorder intro / preview → Permission request prompt → Recorder initialisation → Awaiting browser permission (third-party dialog) → Recorder ready state → Practice / pre-record preview → Recording in progress (timer) → Recording continued (countdown) → Review / Send confirmation → Upload progress → Post-upload syncing → About Me editor with uploaded video ##### Task: policy management approval **Yelp: Report a business post** (web, 3 screens, Food & Drink) [Refero](https://refero.design/flows/1783) > Happy path: User views a business photo inside a lightbox/preview and uses the inline report control to flag the post. The UI shows a confirmation dialog asking the user to confirm the report; after confirming, the UI displays a submission acknowledgement. Steps: View photo with post details → Confirm report → Acknowledge submission **YouTube: Shorts comment moderation** (ios, 4 screens, Entertainment, Video Streaming, Social Networking) [Refero](https://refero.design/flows/2606) > Happy-path for setting comment preferences while adding details to a YouTube Short. Steps: Add details (Short upload) → Comments settings list → Confirmation modal — Allow all comments → Explanatory dialog — Hold-for-review details **Apple Invites: Event Guest Management** (ios, 4 screens, Scheduling Tools) [Refero](https://refero.design/flows/4043) > Happy path: A host opens an upcoming event, opens the guest management / invite sheet, chooses or shares an invite link (public or one-time), reviews a pending access request, and accepts it so the requester becomes an attendee. Steps: Event Card — Access Requests → Invite & Guest Management Panel (Initial) → Accepting a Pending Request → Guest List Updated — Confirmed Attendee **Hello Ivy: Asset Approval** (web, 3 screens, Productivity, Project Management, Collaboration Tools) [Refero](https://refero.design/flows/4526) > Happy path: A reviewer opens an asset in a viewer, inspects the image and comments area, opens the "Set Status" control, and marks the asset as "Approved." Entry point: reviewer has an asset open in the viewer (top bar shows filename and version). Exit point: status label updates to show "Approved" and the dropdown closes; the viewer and comments area remain visible. Assumptions: reviewer is signed in and has permission to set review status, the asset and its version are already loaded, and the interface is viewed on a desktop or wide viewport showing both the asset and the comments pane. Steps: Open Asset Viewer → Open Status Dropdown → Select Approved Status **Mercury: Bulk Payment Approval** (web, 4 screens, Business, Finance) [Refero](https://refero.design/flows/11063) > Happy path: A user with approval privileges opens the Payments area on desktop, navigates to the 'Needs Approval' tab, selects multiple pending payment requests, clicks to review, confirms approval in a modal, and receives a success confirmation while the list updates to reflect the change. Steps: Needs Approval list → Select requests → Confirm approval modal → Approval result ##### Task: prescription refill repeat **The Leap: Referral sharing center** (web, 2 screens, Business, Entertainment) [Refero](https://refero.design/flows/4569) > Referral center for a logged-in merchant where the user obtains and copies their personal referral/store link to share. Steps: Referral overview and copy affordances → Confirm link copied **The Leap: Referral Link Copy and Share** (web, 2 screens, Business, Entertainment) [Refero](https://refero.design/flows/4568) > Logged-in merchant visits the Referrals area (left navigation). The screen shows referral KPIs (total referrals, paid referrals, total earned) and a prominent CTA to copy a personal referral link. Happy path: user clicks Copy referral link (or the inline `Copy to share` button), the app copies the referral URL to the clipboard and shows a success toast. Steps: Referral overview → Copy confirmation **ElevenReader: Referral link sharing** (web, 2 screens, Books, AI Tool) [Refero](https://refero.design/flows/9188) > Happy path: A signed-in desktop user navigates to the Explore section and sees a promotional "Refer a friend" card with a referral URL and a Copy link button. The user copies the link and receives immediate visual confirmation. Steps: Explore — Referral card visible → Copy link confirmation **GlossGenius: Sales refund handling** (web, 6 screens, Business) [Refero](https://refero.design/flows/9978) > Happy-path: A logged-in user on desktop opens the Payments -> Sales history list, selects a completed appointment to view details in a right-hand panel, opens the appointment's action menu, issues a full refund via an 'Issue refund' modal, confirms success, and then sees the transaction updated with a refund line. Steps: Sales list → Completed appointment panel → Overflow action menu → Issue refund modal → Refund completed confirmation → Updated transaction details **Butter: Share referral modal** (web, 2 screens, Design & Media Editing) [Refero](https://refero.design/flows/3035) > A desktop modal that lets a user share a referral/gift link with a friend. Entry: user triggers a "Gift/Share" action which opens this modal. The modal displays a referral message, a prefilled share link, a 'Copy Link' button, and a close (X) control. Exit: user closes the modal or leaves it open after copying the link. Steps: Referral modal (initial) → Referral link copied confirmation **ALO: Refer a Friend Sharing** (web, 3 screens, Fitness, Lifestyle) [Refero](https://refero.design/flows/12381) > Logged-in user opens a referral sharing dialog from the site, copies an invite link or chooses another sharing method, then closes the dialog or continues sharing. Steps: Home Screen → Invite Link Modal → Copied Confirmation **ElevenMusic: Invite Friends Referral** (web, 3 screens, Music Streaming, Entertainment, AI Tool) [Refero](https://refero.design/flows/12593) > Happy path starts in the ElevenMusic Studio dashboard with the account menu open. The user selects the invite-friends option, which opens a centered referral modal over a blurred background. The user copies the invite link and sees a confirmation state on the copy control. Steps: Studio Dashboard Menu → Referral Dialog → Copy Confirmed **On: Order Details Review** (web, 2 screens, Marketplace, Fitness, Lifestyle) [Refero](https://refero.design/flows/13833) > The flow begins on a logged-in customer’s account Orders section, displaying the details of a placed order. The user reviews the order number, delivery estimate, item, status timeline, shipping address, payment summary, and available cancellation action. In the shown interaction, the user selects the copy icon beside the order number. The system copies the order number and confirms the action with a toast notification. The page also includes recommended products and the site footer below the order information. Steps: Order Details → Order Number Copied **Instacart: Referral invite** (web, 2 screens, Food & Drink, Marketplace, Food Delivery) [Refero](https://refero.design/flows/10080) > A post-order referral invite modal presented as an overlay on the order details screen (desktop). Entry: user taps an "Invite"/referral CTA from the order details (not shown) or the system auto-displays the referral modal after order completion. The modal offers a $10-for-you/$10-for-friend referral incentive, shows a large illustration, and provides two primary actions: share via a selected channel (WhatsApp in the screenshots) or copy the referral code/link. Assumptions: user is signed in and on a desktop browser; the modal is an overlay blocking interaction with the page behind it. Exit: user closes the modal via the Close control after sharing or copying the code. Steps: Referral modal — initial state → Referral modal — success/confirmation **GlossGenius: Referral link copy and confirmation** (web, 2 screens, Business) [Refero](https://refero.design/flows/9982) > Happy path: A logged-in desktop user (business account) opens the Referrals area from the left navigation and lands on a referrals overview. The hero presents a personalized referral link and a prominent Copy control. The user copies the link and receives immediate visual confirmation (button state change and toast). The page includes supporting content under “Here’s how it works” explaining sharing options, benefit details, and community access. Steps: Referral overview (initial) → Copy confirmation **Cofounder: Referral Link & Copy** (web, 2 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/7905) > This is a two-step happy-path for the Referral Program screen on a desktop device. Steps: Referral Overview → Copy Confirmation **ElevenMusic: Track Published Confirmation** (web, 2 screens, Music Streaming, Entertainment, AI Tool) [Refero](https://refero.design/flows/12588) > A publishing confirmation modal appears after a music track is successfully published. The modal shows the track artwork, publishing label, track title, and a success message indicating the track was added to an album. The user can copy a shareable track link, dismiss the modal with Done, or continue to the album with View Album. Assumptions: the user is already signed in and has just completed a publish action on a desktop web experience. Steps: Publish Success → Copy Link Confirmed ##### Task: pricing plans compare **one year: Subscription Purchase** (ios, 7 screens, Health & Wellness) [Refero](https://refero.design/flows/10992) > A mobile subscription purchase flow presented as a horizontally scrollable carousel of pricing cards. Entry: user opens the subscription screen from the app (dotted journal background visible). The user browses plan cards (Yearly / Monthly) and taps the CTA to start a free trial. The system then presents the iOS App Store purchase sheet for authentication/confirmation. After user confirms (side button/double click), the app shows a loading state, then an App Store success alert. Exit: user returns to the main screen with the subscription active. Steps: Subscription landing → Yearly plan detail → Monthly plan detail (alternate) → App Store confirmation sheet → Purchase in progress → Purchase success alert → Return to main screen **Nooka: Nooka Trial & Subscription Onboarding** (ios, 8 screens, AI Tool, Education, Books) [Refero](https://refero.design/flows/8113) > A mobile onboarding path that introduces a user to a 7-day free trial and captures a paid subscription via the App Store. Steps: Welcome / Pact Sealed → Free Trial Overview → Price Summary Pitch → Plan Selector — Pro (default) → Plan Selector — Plus → Plan Selector — Ultra → iOS Purchase Confirmation → Purchase Success & Dismiss **HBO Max: HBO Max Subscription and Personalization** (ios, 46 screens, Entertainment, Video Streaming) [Refero](https://refero.design/flows/10268) > A mobile HBO Max signup and personalization journey. The user starts logged out on the streaming service welcome screen, selects a subscription plan, enters an email address, creates an account, chooses content preferences across three onboarding steps, and lands on the personalized Home screen. The flow assumes the user has access to a valid email address and completes the required account details. The visible happy path ends after personalization is applied and the user reaches Home. Steps: Welcome → Browse Monthly Plans → Select Standard Plan → Review Additional Plans → Confirm Plan → Enter Email → Email Entered → Create Account Form → Complete Account Details → Set Password → Choose Genres and Brands → Pick Series → Confirm Series Selections → Feedback Prompt → Continue Series Setup → Pick Movies → Review More Movies → Continue Movie Selection → Complete Preferences → Personalized Home **Picnic: Picnic Photo Organization Onboarding** (ios, 22 screens, Productivity, Utility) [Refero](https://refero.design/flows/11490) > This flow begins with a first-time Picnic launch and introduces the product's photo-organization value proposition. The user accepts the initial terms, reviews permission explanations, grants Photo Library and notification access through native iOS dialogs, and views a short visual onboarding sequence showing photo years. The product then reports the number of photos ready to sort and presents a premium trial offer. The user reviews available plans, selects a subscription option, confirms the purchase through the App Store, and reaches the photo-sorting interface. The final screens demonstrate swipe-based actions for archiving or keeping photos. The flow assumes an iPhone user who has not completed onboarding and has photos available in the device library. The screenshots show a successful weekly subscription purchase and entry into the sorting experience. Steps: Welcome → Product Introduction → Permission Explanation → Photo Library Permission → Notification Permission → Photo History Preview 2021 → Photo History Preview 2023 → Photo History Preview 2024 → Photo History Preview 2025 → Sorting Ready → Premium Trial Offer → Plan Selection → Yearly Plan Option → App Store Purchase Review → Purchase Processing → Purchase Success → Photo Sorting Viewer → Archive Gesture Tutorial → Keep Gesture Tutorial → Photo Comparison **Foodllama: Foodllama Free Trial Subscription** (ios, 15 screens, Health & Wellness) [Refero](https://refero.design/flows/11776) > A mobile iOS subscription acquisition journey for Foodllama. The user starts on a free-trial introduction screen, reviews reminder and pricing information, explores benefits, testimonials, company information, and FAQs, then encounters a limited-time discounted offer. Selecting the trial or special offer opens the Apple purchase sheet, where the user confirms the subscription with the device side button. The flow ends with a native purchase-success confirmation. The user is assumed to be on an iOS device with an Apple account configured for purchases. Steps: Free Trial Introduction → Trial Reminder Setup → Annual Trial Plan → Weekly Plan Comparison → Subscription Credentials → Product Benefits → Ratings and Testimonials → Company Information → Subscription FAQ → Billing and Policy Details → Limited-Time Offer → Offer Exit Confirmation → Apple Subscription Confirmation → Purchase Processing → Purchase Success **Carbon: Subscription Purchase** (ios, 5 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/7784) > User is presented with a two‑card pricing screen (Carbon Pro vs Carbon Ultra). The screen allows the user to toggle which plan is selected by tapping a card; the selected plan is visually highlighted and a subscription row below shows the plan name, trial length and monthly price. The primary call-to-action starts a 1-week free trial and triggers the device's native purchase sheet. The native purchase flow requires OS-level authentication (side button double-click / Face ID). After successful authentication the store returns a confirmation and the app displays a success message. Entry: user arrives at the pricing tab (bottom nav 'pro' selected). Exit: native purchase flow completes and the app shows a purchase success alert. Assumptions: user is on a modern iPhone with an Apple account signed in, has network connectivity, and the app uses the platform store flow for subscriptions. Steps: Pricing screen — Carbon Pro selected → Select Carbon Ultra → Native purchase sheet presented → Authenticate purchase → Purchase success confirmation **(Not Boring) Camera: Subscription purchase paywall** (ios, 5 screens, Utility) [Refero](https://refero.design/flows/7681) > Happy path: user encounters a fullscreen subscription paywall that lists benefits, opens the bottom sheet to choose a plan (single-app yearly/lifetime or a multi-app “Super” bundle), selects a plan/term, and confirms payment using the platform purchase sheet. Steps: Paywall overview → Plan type + term chooser → Bundle (Super) tab → Initiate purchase → Platform purchase confirmation **Vinyls: Subscription Purchase and Confirmation** (ios, 7 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/8916) > User enters a subscription offer screen that lists feature benefits and two pricing options (Annual — highlighted by default — and Monthly). The user can switch plan selection and tap Subscribe. Tapping Subscribe opens the iOS/App Store purchase sheet which requires device-side confirmation (double-click). After confirming, the UI shows a loading state, then a system success alert. The app displays an in-app confirmation toast and returns to content with premium features unlocked. Entry: user navigates to subscription/checkout from content. Exit: purchase completed and premium content available. Steps: Offer Overview and Default Selection → Change Plan Selection → Open System Purchase Sheet → Confirm with Device → System Purchase Success Alert → In-App Confirmation Toast → Return to Content with Premium Access **Zendesk: Subscription purchase and customization** (web, 5 screens, Business) [Refero](https://refero.design/flows/1397) > Happy-path for upgrading/purchasing a Zendesk subscription from the Admin Center. **Endless Tools: Subscription and Account Management** (web, 3 screens, Design & Media Editing) [Refero](https://refero.design/flows/5195) > Happy-path summary: The user lands on the subscription/pricing screen (desktop layout) where monthly and yearly billing options and three plan cards (Free, $12.99, $24.99) are presented. From here the user can either start a plan (Start button on a plan card) or access account controls visible on the same screen. If the user initiates a destructive account action, a confirm modal appears. If the user chooses a paid plan’s Start action while not signed in, the interface surfaces a centered registration modal to create an account and begin the 7-day free trial. Steps: Choose subscription plan → Account deletion confirmation → Register for free trial **Wealthsimple: Self-serve 2024 tax setup** (web, 6 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8507) > Happy-path: A logged-in user chooses to self-file their 2024 tax return, creates a taxpayer profile, selects a pricing plan, and enters personal and residency information in the tax form workspace. Steps: Filing method selection → Choose taxpayer (profile) or year → Create profile — name input (empty) → Create profile — name input (filled) → Select plan and compare features → Tax return workspace — fill personal & residency info **Microsoft: Microsoft 365 plan recommender** (web, 11 screens, Business, Marketplace) [Refero](https://refero.design/flows/9794) > Happy-path summary: A signed-in desktop user starts from the Microsoft 365 marketing entry, navigates to a short guided assessment (“Find the right plan for your business”), answers seven multiple-choice questions about company size, role, devices, features, offline needs, IT support, and security concerns, then receives a side-by-side comparison of recommended Microsoft 365 business plans with one plan visually emphasized. Steps: Marketing landing (entry) → Assessment introduction → Question 1 — Company size (options) → Question 1 — Company size (selected) → Question 2 — Field of work → Question 3 — Device preferences → Question 4 — Desired features → Question 5 — Offline document needs → Question 6 — IT support access → Question 7 — Security concerns → Results — Plans comparison **Jasper AI: Subscription & Feedback** (web, 4 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/5313) > This is the Plans / Billing area where a signed-in workspace admin or workspace member can view current plan details, usage quotas, and explore other plan tiers. Steps: Plans overview → NPS rating modal → Text feedback prompt → Feedback ready to submit **Wealthsimple: Purchase tax plan** (web, 4 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8509) > Happy-path: A logged-in desktop user selects a tax filing plan (Basic/Plus/Pro), reviews feature comparisons, proceeds to a checkout summary that shows plan price and taxes, enters payment details (credit card or an alternative "Link" method), and submits payment. Steps: Plan selection and comparison → Checkout — payment form (empty) → Checkout — payment form (filled) and submit → Payment failure — inline error shown **Epidemic Sound: Plan selection to trial activation** (web, 5 screens, Music Streaming) [Refero](https://refero.design/flows/5069) > Happy path: a prospective customer views pricing tiers, chooses a Commercial plan and starts a 30‑day trial, completes checkout with card payment (3D Secure authorization), and is shown a post‑signup onboarding modal to register social channels. Steps: Pricing selection → Checkout — choose plan & period → Checkout — billing details & review → 3D Secure authentication → Post-signup onboarding modal **Manus: Upgrade to Manus Pro — 7‑Day Trial** (web, 4 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/10422) > User upgrades from a free/basic account to Manus Pro using the 7‑day free trial. Steps: Select plan in pricing modal → Checkout — billing summary and payment entry → Processing payment → Success — Manus Pro activated **HBO Max: Subscription Checkout** (web, 14 screens, Entertainment, Video Streaming) [Refero](https://refero.design/flows/10248) > Happy-path subscription flow for HBO Max on desktop (not logged in). Entry: user lands on the plan selection screen. The user selects a monthly plan (Standard in this example), optionally adds an add-on (Sports), provides an email, completes an account creation form, selects a payment method (credit/debit card or PayPal), enters payment details and completes authentication via a third-party (Revolut) SCA step. Exit: success/welcome screen with streaming access. Assumptions: desktop browser, user is not signed in, user has a valid email and payment method, the region supports the shown plans and payment methods. Steps: Plan Selection - Overview → Plan Selection - Plan Selected → Add-Ons → Get Started - Email Entry → Email Validation Error → Create Account - Form Display → Create Account - Form Partial Entry → Create Account - Consent → Choose How to Pay - Methods → Enter Card Details - Expanded → Enter Card Details - Filled → Payment Processing → Third-Party Authorization (SCA) → Welcome / Confirmation **Wrike: Subscription Plan Selection & Checkout** (web, 2 screens, Business, Productivity, Project Management, Collaboration Tools, Paperwork, Scheduling Tools) [Refero](https://refero.design/flows/1497) > Happy-path: User lands on a pricing grid to choose a subscription tier, selects a plan (Team in this example), and is taken to a Billing Details screen to enter payment and billing data before reviewing the order. Steps: Choose your plan → Billing details and order summary **HBO Max: Subscription Signup and Profile Setup** (web, 33 screens, Entertainment, Video Streaming) [Refero](https://refero.design/flows/10247) > Happy-path new user subscription and onboarding flow on HBO Max (Nordic). Entry: marketing landing / pricing hero. Exit: signed-in experience showing content grid. Steps: Marketing / Pricing Entry → Loading State → Choose Your Plan — Monthly View → Plan Selected Confirmation → Add-Ons Overview → Add-On Selected → Get Started — Email Capture → Email Validation Error → Create Account Form → Create Account — Filled → Manage Profile — Initial → Enter Account Password Prompt → Enter Password — Filled → Create Profile PIN Prompt → Create Profile PIN — Entry State → Manage Profile — Locked Confirmation → Manage Profile — Personal Info Saved → Manage Profiles Overview → Enter Your Profile PIN → Signed-in Content Grid **Leonardo.Ai: Subscription Cancellation and Feedback** (web, 7 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/12913) > Logged-in user in Leonardo.ai Settings opens Subscription & Billing, starts a plan cancellation, reviews retention and cancellation confirmation modals, optionally provides cancellation feedback, and lands on the pricing/subscription management view with the subscription marked for cancellation. Steps: Subscription Settings → Cancellation Warning → Confirm Plan Change → Processing Confirmation → Cancellation Feedback → Reason Details → Pricing and Plans Overview **Tango: Upgrade to Tango Pro** (web, 10 screens, Business, Productivity, Project Management, AI Tool) [Refero](https://refero.design/flows/5791) > Happy-path: A user opens the product's sharing/settings area, then navigates into the Upgrade to Pro flow where they review Pro features, pick the Pro plan, customize billing and seat counts, assign roles for seats, review an order summary, and view role-permission details before proceeding to payment. Steps: Open Share modal / Access controls → Upgrade landing — features overview → Feature detail — Desktop App → Feature detail — Collaboration Tools → Plan comparison cards → Customize — billing and roles (initial) → Customize — switch billing cycle to Monthly → Add seats — select quantity and role → Assign emails and finalize seat roles → View Pro Workspace role permissions **Glorify: Start subscription trial** (web, 3 screens, Design & Media Editing) [Refero](https://refero.design/flows/3739) > Happy path: A logged-in user is presented with a pricing modal offering three subscription tiers and a 2-day free trial. The user selects a plan (e.g., Pro) and clicks the plan's CTA to start the trial. The interface then shows a confirmation screen that indicates the trial has started and that a confirmation email was sent; a short progress state is shown and a primary Continue button becomes available. Steps: Pick plan → Confirmation (in progress) → Confirmation (continue) **Intercom: Subscription purchase - Advanced plan** (web, 11 screens, Business, Communication Tools) [Refero](https://refero.design/flows/672) > Happy path: a logged-in user on the Help Desk dashboard clicks an upgrade CTA, reviews plan options in a modal, toggles billing/seat quantity to view pricing, then proceeds to a "Build your plan" page to configure plan and optional upgrades. They continue to the Payment method screen, enter card and billing details, and attempt to pay. Steps: Dashboard - Upgrade prompt → Plan comparison modal - monthly pricing → Plan comparison modal - yearly pricing and seats → Feature comparison - detailed scroll → Feature comparison - reporting & integrations → Build your plan - configure details → Plan dropdown selection → Payment method - card entry → Card input validation error → Payment attempt and global decline → Post-decline state - next actions **Spotify: Spotify Premium Plan Cancellation** (web, 8 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/57) > This flow guides a logged-in user through managing their Spotify Premium Individual subscription, ultimately allowing them to cancel the plan and provide feedback on the cancellation. The user navigates account settings, views and selects from available plans, confirms cancellation, optionally gives a reason for cancelling, and sees a confirmation screen with a playlist and option to return to the account page. The flow assumes the user has a Premium Individual subscription and relevant permissions to manage it. Steps: Account Overview → Manage Current Plan → Subscription Plans Selection → Cancel Plan Confirmation → Cancellation Feedback Form → Cancellation Feedback Form With Selections → Cancellation Confirmation → Post-Cancellation Account Overview **Grammarly: Premium Plan Upgrade - Annual Purchase** (web, 4 screens, Paperwork) [Refero](https://refero.design/flows/1882) > User starts from the account Subscription area (left navigation) where plan tiers are compared (Free, Premium, Business). From there the user chooses to upgrade to Premium which opens a Plan Selection step to pick a billing cadence (Annual, Quarterly, Monthly, or Business). After selecting a cadence the flow advances to Payment Completion where an order summary is shown and the user chooses a payment method (PayPal or Credit Card) and provides required billing information. Steps: Subscription Overview → Plan Selection → Payment Completion — PayPal option → Payment Completion — Credit Card option **Tango: Subscription Plan Purchase** (web, 5 screens, Business, Productivity, Project Management, AI Tool) [Refero](https://refero.design/flows/5792) > Happy path: A user upgrades from a free plan to a paid plan (Pro) then customizes billing and seats before continuing to payment. Entry: user opens the pricing modal and selects a plan. The flow continues to a two-column "Customize your plan" screen where the user picks billing frequency (yearly or monthly), assigns roles to existing teammates, adds additional seats (unassigned teammates) and sees a live Order Summary on the right. Exit: user clicks Continue to proceed to the payment step. Assumptions: user is signed in and has a workspace with at least one existing teammate; the UI reflects role assignment and seat quantity changes immediately in the Order Summary. Steps: Select plan → Customize — billing and roles (yearly) → Customize — switch billing frequency (monthly) → Customize — add seats (quantity selection) → Customize — finalize seats and review order **Leonardo.Ai: Essential Plan Upgrade** (web, 12 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/12783) > A logged-in Leonardo.ai user starts in the AI Creation workspace on the Free plan and opens the plan details menu to upgrade. They review available subscription tiers, select the Essential plan, choose a payment method, complete a human-verification check, and proceed to hosted checkout. The user first encounters Link-based verification, then switches to card payment and enters invalid card details, producing an inline validation error. The sequence ultimately returns to Leonardo.ai, where a What's New announcement is shown over the creation dashboard. The screenshots do not show a completed subscription purchase. Steps: Current Plan Menu → Plan Comparison → Payment Method Selection → Human Verification → Verification Success → Checkout Method Confirmation → Hosted Checkout Loading → Link Verification → Card Checkout → Card Validation Error → What's New Announcement → Creation Dashboard **LinkedIn: Choose Career Premium and start trial** (web, 4 screens, Business, Social Networking, Job Search) [Refero](https://refero.design/flows/813) > Happy-path: a signed-in user explores LinkedIn Premium offerings, selects the Career plan, views plan details and benefits, then starts the 1-month free trial by entering payment details on a secure checkout. Steps: Plan overview (highlighted recommendations) → All plan options → Plan details and CTA → Secure checkout - payment **Telescope: Join and Customize Membership** (web, 8 screens, Social Networking) [Refero](https://refero.design/flows/13901) > A logged-in Telescope user starts in the feed, chooses to join the network, selects a membership plan and billing cadence, completes subscription payment, views the network welcome screen, customizes their digital membership card, and returns to the feed with the updated card and an option to invite friends. The flow assumes an authenticated desktop user with permission to purchase a membership. The payment step is handled through a Stripe-hosted checkout experience. The visible happy path ends after the customized card is saved and displayed in the feed sidebar. Steps: Feed and Membership Entry → Annual Plan Selection → Monthly Plan Selection → Subscription Checkout → Network Welcome → Card Customization → Updated Card Preview → Membership Card in Feed **Navia: 30-Day Marketing Plan** (web, 4 screens, Business, Productivity) [Refero](https://refero.design/flows/14312) > A returning user enters the Navia home dashboard and starts generating a personalized 30-day marketing plan. The system shows plan-building progress, then presents available subscription tiers. After the plan is prepared, the user reaches a dashboard containing upcoming actions, a featured strategy prompt, and a dated marketing calendar with a mix of available and locked actions. The flow assumes the user is signed in as Liam, has an existing brand profile, and is using a desktop-sized interface. The pricing modal appears as an intermediate decision point, but the screenshots do not show a plan-selection action, so the exact transition into the final dashboard is not confirmed. Steps: Plan Launch → Plan Generation → Plan Selection → Strategy Calendar **Intercom: Subscription purchase** (web, 6 screens, Business, Communication Tools) [Refero](https://refero.design/flows/642) > This is a two-column desktop checkout flow for purchasing a subscription plan. The left column is a step-by-step builder and form (plan configuration → payment method), the right column is a persistent order summary showing plan, seat count, and pricing. Steps: Build your plan — Configure → Plan tier selection → Payment method — Enter card → Inline validation — Card invalid → Valid card entered → Global payment error — Card declined **Pop Site: Upgrade to Pro** (web, 4 screens, Business, Social Networking) [Refero](https://refero.design/flows/6465) > Happy path: a logged-in user opens the account menu from the editor, selects the Upgrade to Pro action, reviews the Go Pro pricing/benefits panel, starts the free trial via the prominent CTA, and is shown a confirmation/activation screen indicating their Pro account is active. Steps: Editor — Account Menu → Go Pro — Pricing & Benefits → Pro Activated — Success Screen → Pro Confirmation — Static State ##### Task: privacy data consent **Meta AI: Meta AI Problem Report** (ios, 7 screens, AI Tool) [Refero](https://refero.design/flows/13397) > A logged-in user opens Help & support from Settings and submits feedback about a Meta AI problem. The user can review reporting options, choose whether to share logs and diagnostics, describe the issue, select related categories, attach screenshots, and continue with the report. The sequence ends back on the Settings screen after the report form is opened and populated; a final submission confirmation is not shown. **LEGO® Builder: Create LEGO account and consent** (ios, 30 screens, Entertainment) [Refero](https://refero.design/flows/10126) > Happy-path mobile sign-up: A new user starts on the app's home screen and chooses to create a LEGO account. The app opens an SSO/webview to identity.lego.com where the user enters an email, selects country/state, provides birthdate, and verifies the account via an emailed activation code. After verification the user reviews 'Insiders' benefits and is shown consent screens (marketing preferences / privacy rights / cookie toggles). Once consents are confirmed the webview closes and the app shows the newly created profile in the app. Entry: app home (not signed in). Exit: app home with signed-in profile. Assumption: user has a working email and access to it, device is iOS (webview and system permission dialogs visible), user is creating an 'adult' account. Steps: Welcome / Sign-up prompt → SSO permission modal → Identity webview — entry screen → Enter email → Create account form — country & birthdate → Country selection list → Form updated for United States → State selection list → Enter birthdate → Activation code prompt → Insiders benefits overview → Marketing consent modal → Privacy rights / opt-out choices → Cookie consent — initial view → Cookie consent — read more and confirm → Webview closes / return to app → Profile displayed in app → App collection main screen → Consent states persisted → Post-registration state summary **Stardust: Stardust first-run sign up & onboarding** (ios, 38 screens, Health & Wellness) [Refero](https://refero.design/flows/5855) > Happy-path new user on iOS. Entry: cold-launch on the Stardust splash screen. The user signs up with Apple, accepts terms, chooses a journey (Cycle Tracking), enters birthday using a native picker (zodiac updates visually), progresses through 3–4 onboarding feature screens, accepts the subscription trial (7-day trial selected), confirms the App Store purchase, and finishes in the main cycle dashboard (circular moon-phase visualization). Assumptions: user is on an iPhone, not previously authenticated, has Apple ID available on device and can complete in-app purchase through App Store. Steps: Welcome / Splash → Data & Terms Consent (initial) → Consent selected → Trigger Apple Sign In → Apple Sign In — pick details → Signing in (Apple) → Apple Sign In Done → Journey Selection → Birthday — initial picker → Birthday — confirm → Onboarding carousel — screen 1 (moon sync) → Onboarding carousel — screen 2 (insights) → Onboarding carousel — screen 3 (decode hormones) → Subscription upsell modal → Show App Store confirmation overlay → App Store purchase in progress → Purchase success confirmation → Initial data sync / animation → Main dashboard — circular moon view → Onboarding complete / baseline state **Sora: Account & profile deletion** (ios, 5 screens, AI Tool, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/7855) > Happy path: A logged-in mobile user navigates from account/privacy settings (Export data) to delete either a single product profile (Sora) or their full ChatGPT+Sora account. The screens guide the user through selecting deletion scope, showing a permanent-deletion warning modal, queuing the deletion, and then signing the user out to the app's onboarding/sign-in screen. Steps: Export data / Privacy settings → Choose deletion scope → Permanent deletion confirmation → Deletion queued / Log out → Onboarding / Sign-in **TikTok: Account deletion and verification** (ios, 10 screens, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/2811) > Happy-path: A logged-in user navigates from Settings & Privacy → Account to delete their TikTok account. The flow asks for a reason, optionally collects free-text details, suggests alternatives (for some reasons), requires confirmation that the user reviewed/downloaded their data, shows deactivation / deletion consequences, requires SMS 2FA (6-digit code), and ends with a final confirmation modal. Steps: Settings and privacy (entry) → Account settings list → Reason selection → Free-text reason input → Contextual suggestion for selection → Download data prompt (info) → Confirm data review → Deactivation summary and consequences → SMS verification (enter code) → Final confirmation modal **Project Indigo: Project Indigo First-Launch Setup** (ios, 26 screens, Utility) [Refero](https://refero.design/flows/11752) > A first-launch onboarding sequence for Project Indigo, an early-access camera product. The user begins by reviewing Adobe's Terms of Use and Privacy Policy through linked legal documents, returns to mark both documents as reviewed, and taps Agree. The user then reaches a permissions setup screen, followed by an educational Camera Features section with expandable topics. After the feature topics are marked complete, the user enters the camera and sees an optional mobile-usage data notice, a release-notes panel, an informational camera tooltip, and a macro-mode guidance message. The screenshots do not show the individual permission prompts or any final dismissal action for the camera overlays. Steps: Legal Review Start → Terms of Use Document → Terms Document Review → Terms References → Terms Marked Reviewed → Privacy Policy Document → Privacy Policy Review → Privacy Policy Later Sections → Legal Agreement Confirmation → Permissions Setup → Camera Features Introduction → Photo and Night Mode Lesson → Zooming Lesson → Filmstrip Lesson → Feature Progress → Feature Review Complete → Mobile Usage Notice → Camera Release Notes → Camera Information Tip → Macro Mode Guidance **Pop Site: Manage My Personal Information** (web, 3 screens, Business, Social Networking) [Refero](https://refero.design/flows/6474) > Happy path: an unauthenticated desktop user opens the "Manage My Personal Information" form, fills contact fields, selects one or more privacy request options (e.g., request copy, delete my data, opt out of sale/sharing), adds an optional message, and submits the form. After submission the user is shown a confirmation screen that the request was received and will be responded to within 45 days. Steps: Open Personal Information Form → Select Request Type and Submit → Submission Confirmation **Cursor: Change privacy to Privacy Mode** (web, 5 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9949) > Happy path: A logged-in user navigates to their account Settings (desktop) to change how their code and usage data are handled. From the Settings overview they open the Privacy Settings modal, switch from shared data to Privacy Mode, complete the typed confirmation, and save. Steps: Open Settings privacy section → Privacy Settings modal (initial) → Show confirmation details → Enter confirmation text → Settings updated **Threads: Enable Private Profile** (web, 3 screens, Social Networking, Entertainment) [Refero](https://refero.design/flows/10673) > Happy-path: A signed-in user navigates to Settings → Privacy and turns on the "Private profile" option. Tapping the toggle brings up a confirmation modal. Confirming closes the modal and the toggle updates to the ON state. Steps: Privacy settings (toggle visible) → Confirmation modal → Privacy enabled (toggle on) **Synthesia.io: Avatar Consent Recording** (web, 4 screens, AI Tool) [Refero](https://refero.design/flows/13716) > A logged-in user begins creating a digital avatar from the workspace. A consent modal asks the user to select microphone and camera inputs, then continue to a recording step. The user records the displayed consent statement, reviews the completed recording, and submits it. The final screenshot shows submission in progress. The flow assumes the user has access to the avatar-creation feature and can use the selected camera and microphone. **GoFundMe: Profile Visibility Settings** (web, 5 screens, Fundraising) [Refero](https://refero.design/flows/11721) > Logged-in user views their personal GoFundMe profile, opens the visibility settings modal, switches the profile from Private to Public, and confirms the suggested public URL. The flow starts on the profile dashboard and ends with the profile returning to the main page in public mode, showing a confirmation toast and the updated visibility state. Steps: Profile Overview → Visibility Options → Visibility Confirmation → Public URL Review → Updated Public Profile ##### Task: procurement tender bid **Artsy: Artist auction results search** (ios, 3 screens, Marketplace) [Refero](https://refero.design/flows/7604) > Happy path: From an artist profile the user opens the Auction Results tab, inspects past auction entries (thumbnails + metadata), uses the search field to filter results by artwork title or series, and views either matching results or a clear "no results" message. Steps: Auction Results list → Search returns no matches → Search returns filtered results **Artsy: Artist auction lookup** (ios, 6 screens, Marketplace) [Refero](https://refero.design/flows/7603) > Happy-path: A user lands on an artist's profile and finds there are no auction results shown. They tap the call-to-action to open the Artsy Price Database, type an artist name (using autocomplete), choose the matching artist from the suggestions, and are returned to the artist profile where auction results are displayed. Steps: Artist page — empty auction state → Price Database — search entry → Search — no-match result → Search — autocomplete suggestions → Search — select suggestion → Artist page — auction results populated **Programa: Purchase Order Creation and Sending** (web, 17 screens, Project Management, Business) [Refero](https://refero.design/flows/13438) > A user in Project #254 creates and sends a purchase order. The user opens Purchase Orders, starts a new order, selects schedule #1, chooses a supplier, enters issue and delivery dates, adds five products, reviews the generated order, enters or confirms recipient and delivery details, and sends the order. The system creates purchase order P000001, marks it as sent, and returns to the Purchase Orders list where the order is displayed. The user is assumed to be authenticated and authorized to manage purchase orders. ##### Task: profile edit **X: Edit profile header and details** (web, 3 screens, Social Networking) [Refero](https://refero.design/flows/1012) > Happy path: A logged-in user opens the profile edit modal, changes the cover/banner image using an inline media editor (crop/zoom), then returns to the edit modal where they adjust text fields (name, bio, location, website). Steps: Open Edit Profile modal → Crop banner image in media editor → Edit text fields and preview updated header **Threads: Profile bio edit** (web, 4 screens, Social Networking, Entertainment) [Refero](https://refero.design/flows/10647) > Happy path: A logged-in user on a desktop layout opens the profile edit modal, selects the Bio field, edits or imports text, and saves the change. Steps: Open profile edit modal → Edit bio (empty state) → Edit bio (with content) → Profile edit modal updated **Manus: Account profile update** (web, 6 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/10423) > Happy path: A signed-in user opens the Account modal to view subscription and credit information, uses the edit-profile control to navigate into the Profile screen, updates their avatar and display name, then returns to the Account view where the updated avatar and name are shown. Steps: Account overview → Open edit-profile control → Profile — edit view (initial) → Profile — avatar select preview → Profile — edit display name → Account overview updated **Lovable: Edit Profile and Banner** (web, 5 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12288) > Logged-in user opens profile editing from their profile screen. The flow begins with an Edit profile modal over the profile page, then moves to a separate Edit banner image editor for adjusting the cover image crop/zoom. After applying the banner edit, the user returns to the profile modal with the updated banner visible behind the avatar and form fields. The flow assumes the user is authenticated and has permission to edit their own profile on desktop. Steps: Profile Details Editing → Banner Image Editor → Banner Adjustment Preview → Banner Apply State → Profile Edit with Updated Banner **GoFundMe: Personal Profile Customization** (web, 11 screens, Fundraising) [Refero](https://refero.design/flows/11722) > An authenticated GoFundMe user starts on their personal profile, opens Edit profile, and customizes the profile photo and cover image. The user selects Public visibility, uploads or adjusts both images using the photo editor, saves each image, and returns to the profile to confirm the updated visuals. The profile name and URL remain unchanged. The flow is shown on desktop and ends on the updated public profile. Steps: Profile Overview → Edit Profile → Profile Photo Editor → Adjusted Photo Preview → Photo Processing → Profile Photo Saved → Profile Image Settings → Cover Image Editor → Cover Image Saved → Completed Edit Profile → Updated Profile **Intercom: Profile editing — photo and fields** (web, 3 screens, Business, Communication Tools) [Refero](https://refero.design/flows/686) > Happy path: a logged-in user viewing their personal profile dashboard edits their profile photo and some profile fields. Steps: Profile overview (view mode) → Update profile photo modal → Public profile edit mode (inline fields) **Lovable: Personalize Public Profile** (web, 11 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12287) > A logged-in Lovable user starts on their public profile, opens the profile editor, replaces the default banner and avatar imagery using image-cropping controls, enters profile details, and saves the changes. The flow begins at an unpublished profile with default visuals and ends on the updated public profile, where the new banner, avatar, bio, location, and website are visible along with a confirmation message. Steps: Profile Overview → Profile Editor → Banner Image Selection → Banner Crop Adjustment → Banner Applied → Banner Preview in Editor → Avatar Image Selection → Avatar Crop Adjustment → Avatar Applied → Profile Details Entry → Updated Profile Confirmation **Threads: Edit Profile - Change Avatar** (web, 4 screens, Social Networking, Entertainment) [Refero](https://refero.design/flows/10646) > Happy path: A signed-in user opens the profile edit modal, taps the avatar control, selects an option to add a new picture (Upload picture or Import from Instagram), completes the upload/import, and sees the avatar preview update in the modal. Steps: Open Profile Edit Modal → Open Avatar Options Menu → Select Upload Option → Avatar Preview Updated **Handshake: Profile edit and add link** (web, 6 screens, Social Networking, Job Search) [Refero](https://refero.design/flows/8760) > Happy-path for adding an external link and updating basic profile fields (pronouns and location) from the user's resume/profile area. Steps: Profile / Resume overview → Edit profile modal → Link modal (empty) → Link modal (filled) → Edit profile modal (with link and fields) → Profile updated view **WhatsApp: Edit Profile Details** (web, 2 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/13222) > This desktop flow starts on an Edit profile screen in WhatsApp, where the user can view and edit profile fields such as About, Name, and Phone. The user can tap pencil icons to edit text fields or use the copy icon next to the phone number. In the second screen, the system confirms the phone number was copied with a temporary toast message. The flow assumes the user is already signed in and viewing their own profile settings on desktop. It ends with the phone number copied to the clipboard and visual confirmation shown to the user. **PayPal: Profile editing (header, avatar, location, bio)** (web, 15 screens, Finance) [Refero](https://refero.design/flows/7390) > A signed-in desktop user edits their public profile from an Edit your profile modal. The flow includes changing the header background (choose a theme or upload and reposition), changing/cropping the avatar with a position/zoom slider, and editing text fields (location and About). Each edit opens a focused sub-screen or modal for selection or text entry and returns to the main Edit your profile screen for review. The flow ends when the user saves changes and returns to the main profile edit view showing updated preview. Steps: Edit profile — main → Background selection → Reposition background preview → Main screen with updated header preview → Avatar edit — crop and position → Avatar reposition attempt → Avatar slider adjustment → Main screen with updated avatar → Scroll to profile information → Location selector → About editor (with existing text) → About editor (empty state) → About editor (partial input) → About editor — limit reached → Main screen with updated bio **Rive: User Account Profile Editing** (web, 2 screens, Design & Media Editing) [Refero](https://refero.design/flows/174) > The flow begins with the user viewing their account overview page, which displays their profile information (email, username, and other personal details) and their active workspaces. The user can choose to edit their profile by clicking the 'Edit Profile' link. This navigates to an editable form where the user can modify various profile fields such as display name, location, bio, website, and social links. The user can upload an avatar and save or cancel their changes. The flow assumes the user is logged in and has permission to update their account details. The result is an updated user profile stored in the system, reflected in the account overview upon returning to that page. Steps: Account Overview → Edit Profile Form **[untitled]: Edit Profile — Change Username** (web, 4 screens, Music Streaming) [Refero](https://refero.design/flows/5962) > Happy path: A logged-in user opens the Edit Profile dialog from Settings, chooses to change their username, edits the text in an inline modal, and saves. Steps: Open Edit Profile → Open Change Username Modal → Edit Username → Save and Return to Settings **Fresha: Edit personal profile details** (web, 4 screens, Travel & Booking, Health & Wellness) [Refero](https://refero.design/flows/2315) > Happy path for editing a user's personal profile details. **Sora: Profile avatar update** (web, 3 screens, Social Networking, Entertainment, Video Streaming, AI Tool) [Refero](https://refero.design/flows/8140) > Happy path: A logged-in user visits their profile (desktop layout), opens the avatar edit control, chooses Upload photo, and completes a photo upload — resulting in the profile avatar visually updating. Steps: Profile with drafts visible → Avatar edit menu open → Profile avatar updated **SoundCloud: Profile Editing** (web, 4 screens, Music Streaming, Social Networking) [Refero](https://refero.design/flows/6570) > Logged-in user opens their SoundCloud profile and edits account details using an inline modal. Entry: user is on their profile overview and clicks the Edit button in the header. In the modal the user can change display name, profile URL, first/last name, city, country, bio, profile image, and add links. Exit: after saving, a toast confirmation appears and the profile page updates to show the new bio and other changed fields. Steps: Profile Overview → Edit Profile Modal — Form Loaded → Edit Profile Modal — Enter Data → Profile Updated — Confirmation **Stellate: User Profile Edit** (web, 4 screens, Development Tools) [Refero](https://refero.design/flows/4441) > Happy path: a logged-in user navigates to the User settings > Account tab, opens the Edit Profile modal, changes their display name, saves the change, and receives an on-screen success confirmation. Steps: Account overview → Edit Profile modal (empty state) → Enter new name → Save and confirm **X: Profile edit — update header and details** (web, 7 screens, Social Networking) [Refero](https://refero.design/flows/1011) > Happy path: A logged-in user opens their profile, taps Edit profile, updates the header (cover) image and several text fields (name, bio, location, birth date), uses an image cropper to position/zoom the header, then saves changes. Steps: Profile view (initial) → Edit profile modal opened → Header image cropper → Apply cropped header and edit fields → Additional profile fields visible → Final review in edit modal → Profile view (updated) **ElevenMusic: Artist Profile Editing** (web, 3 screens, Music Streaming, Entertainment, AI Tool) [Refero](https://refero.design/flows/12590) > Signed-in artist views their public artist profile, then opens an edit modal to update profile details. The entry point is the artist profile screen with an Edit Profile button; the exit point is saving changes and returning to the profile view. Steps: Artist Profile View → Edit Profile Form → Save Profile Changes **Delphi: Profile Personalization** (web, 4 screens, AI Tool) [Refero](https://refero.design/flows/13979) > The user starts on their personal profile, which is visible but marked as not fully set up. They select Edit, update the bio, add a second social link, and interact with the avatar upload control. They then save the changes and return to the profile view, where the updated bio and social links are displayed. The flow begins on an existing profile in an editable account context and ends with the saved profile details visible. The screenshots do not show the actual image-selection step, so the avatar upload interaction is treated as an available control rather than a completed image change. Steps: Profile Overview → Profile Editor → Profile Updates → Saved Profile **TikTok: Edit Profile Modal** (web, 4 screens, Video Streaming, Social Networking, Entertainment) [Refero](https://refero.design/flows/1121) > A desktop modal-based profile editing sequence where a signed-in user updates their avatar, username, display name, and bio. **TravelPerk: Edit Personal Profile** (web, 4 screens, Productivity, Travel & Booking) [Refero](https://refero.design/flows/5201) > Happy-path: A logged-in user opens their profile overview, clicks Edit on the Personal section, fills or updates personal fields (sex, DOB, country, contact) and saves. The system persists the changes, returns the user to the profile overview, updates the profile completion progress, and displays a brief success toast. Entry: user is on their profile overview (authenticated). Exit: user returns to profile overview with updated details and a success notification. Assumptions: desktop viewport, user is authenticated and allowed to edit their own profile, fields shown reflect what the UI supports (some optional). Steps: Profile overview (Personal section) → Personal edit form (initial) → Personal edit form (completed) → Profile overview (updated) **PayPal: Edit About You** (web, 6 screens, Finance) [Refero](https://refero.design/flows/7393) > Happy-path flow for editing the user's "About you" / personal message within Profile Information. Entry: user is on the Profile Information screen (logged-in state assumed). The user taps the "About you" row to open an edit view (modal-like screen) where they can modify the free-text message. The edit view shows a textarea, a character counter, a Save button, and navigation controls (Back and Close). The user edits the text, may hit the character limit (counter turns red), adjusts content, and saves. Exit: the flow returns to the Profile Information screen, which displays the updated "About you" text. Assumptions: user is authenticated and allowed to edit their profile; this is a desktop/responsive layout; the Save control commits changes to the displayed profile. Steps: Profile Information - About row → About you — editor (prefilled) → About you — editor (cleared / empty) → About you — editor (over limit indicator) → About you — editor (valid content) → Profile Information — updated **Threads: Profile Editing — Photo & Settings** (web, 4 screens, Social Networking, Entertainment) [Refero](https://refero.design/flows/10645) > Happy path: A logged-in user opens their profile, taps Edit profile, and uses the edit modal to change or remove their profile photo and adjust profile fields (bio, interests, links, Instagram badge, privacy). Steps: Profile overview → Edit profile modal (photo hint) → Photo action menu → Photo removed / preview updated **Parallel: Candidate Profile Edit Modal** (web, 3 screens, Job Search) [Refero](https://refero.design/flows/1039) > Desktop modal for editing a user profile. Steps: About Me (profile summary) → Visibility Status Dropdown → Personality **Airbnb: Edit personal profile** (web, 7 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/5974) > Happy path: A logged-in user opens their Profile > About me, clicks Edit to open the full 'My profile' editor, updates a single field (Pets) via a modal, encounters input validation for unsupported characters, corrects or removes the character and saves. The editor updates and the user returns to the profile view where the new pet text and any changed interests/intro are visible. Steps: Profile > About me (summary) → My profile (editor) → Open Pets edit modal → Type pet information → Validation error shown → Save and update editor → Final profile view (confirmed) **Polaroid: Account Photo & Profile Edit** (web, 8 screens, Marketplace) [Refero](https://refero.design/flows/9464) > Logged-in member visits their Rewards dashboard and opens the account editor (right-side slide-in panel) to update profile photo, experience level, interests and name. Entry: user is signed in on the account "Rewards" / "My Feed" screen. Exit: the account header is updated with the new photo and display name. Steps: Rewards dashboard (account header) → Edit account panel opens → Photo options menu → Photo preview after selection → Photo applied in editor → Skill-level dropdown open → Edit interests and name fields → Updated account header **Intercom: Profile edit and photo update** (web, 4 screens, Business, Communication Tools) [Refero](https://refero.design/flows/685) > A logged-in desktop user visits their personal profile area (left column) while the main content shows recent conversations (right column). From the profile area the user can update their avatar and edit profile fields (name, job title, department, phone, bio, calendar link and visibility toggles). The flow covers opening an avatar update modal, previewing and saving the photo, switching the profile card into edit mode, saving textual changes, and returning to the read-only profile view with updated details. Entry: user profile panel is visible in the app header/left column. Exit: profile panel returns to read-only state with persisted changes visible in the profile card and avatar. Steps: Profile overview (read-only) → Update profile photo modal → Profile fields in edit mode → Profile updated (read-only) **Airbnb: Profile introduction and interests edit** (web, 6 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/5977) > This is the user profile editing and preview path for a logged-in Airbnb user. Steps: Profile landing — About me → My profile editor → About you modal (empty) → About you modal (filled) → Review and toggle options → Final profile preview and finish **Sora: Profile header editing** (web, 3 screens, Social Networking, Entertainment, Video Streaming, AI Tool) [Refero](https://refero.design/flows/8139) > Happy-path: A logged-in user on a desktop-sized screen views their personal profile, opens the avatar/edit menu, uses the menu to add a display name and bio (or upload a photo), and then sees the updated header with the new display name and bio while the drafts area remains visible. Steps: Profile overview → Avatar/contextual action menu → Profile updated / confirm ##### Task: progress tracker steps **Miles: Run session start to finish** (ios, 8 screens, Health & Wellness) [Refero](https://refero.design/flows/6777) > Happy-path for starting, running, pausing and finishing a tracked run on an iOS device. Steps: Main tracker (pre-run) → OS location permission → Countdown (2) → Countdown (1) → Active run → Paused — resume or finish → Hold to finish confirmation → Main summary (post-run / monthly) **Waterllama: Hydration onboarding & first-sip flow** (ios, 36 screens, Health & Wellness) [Refero](https://refero.design/flows/11511) > Happy-path onboarding for a hydration tracker. Entry: first-launch onboarding carousel (several illustrated screens). The user taps Get Started → chooses measurement units (oz or ml) → follows a short "Add your first sip" tutorial that demonstrates selecting a beverage (water), opening a cup control, adjusting volume and adding a sip. Next the user is shown a Smart Reminders screen and is asked to enable notifications (native iOS permission). The flow then shows a premium upsell carousel with a prominent "Start my 3-day Free trial" CTA. If the user taps the CTA they see pricing and use Apple Pay to confirm the subscription (loading + success confirmation). Exit point: either the user completes the subscription and is returned to the main hydration dashboard, or (alternatively) the user skips enabling reminders / subscription and lands on the main hydration dashboard with current intake shown. Assumptions: user is on iOS, first-time launch, not signed into any account, has Apple Pay configured. Visual language uses playful animal illustrations and a single prominent CTA on each screen. Steps: Welcome carousel - screen 1 → Welcome carousel - screen 2 → Welcome carousel - screen 3 (confetti) → Choose your units modal → Add your first sip tutorial overlay → Cup volume control (initial) → Cup confirm animation → Illustration fills to reflect intake → Smart Reminders intro → iOS notifications permission prompt → Premium upsell header → Feature carousel & benefits list → Purchase loading / spinner → Apple Pay / subscription confirmation → Purchase success modal → Main hydration dashboard (first run) → Beverage quick-add bar → Persistent controls & state → Error / loading fallbacks (observed states) → Post-onboarding home state **Steps: Activity Summary** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/7516) > Home / Activity overview in portrait mobile. Entry: user opens the Home tab (bottom nav) and sees a large summary card for a selected period (default: Day) plus a histogram and a horizontal date selector beneath it. The card shows the period label (day or week), the main metric (steps) and two secondary metrics (distance, calories) plus a short feedback message. The user can change the period using the top-right pill (Day / Week / Month) which opens a small popover menu, or can pick a different date/week using the horizontal scroller. On period change the summary card, the histogram, and the highlighted date range update. Assumptions: user is signed in, device is a mobile phone in portrait, step data is available locally or via a connected sensor, the Home tab is the app's entry point. Steps: Daily summary → Open period selector → Weekly summary **Steps: First-run Setup & Pro Purchase** (ios, 15 screens, Health & Wellness) [Refero](https://refero.design/flows/7514) > Happy-path onboarding + optional Pro purchase. Entry: first-run Welcome to Steps screen. The user is assumed to be new (not yet connected to Health) on an iPhone. Flow covers: greeting → choose daily step goal → request Apple Health read/write access → confirmation (All Set) → optional Pro upsell and purchase flow (App Store confirmation + success) → notification permission → landing on main dashboard. Exit: user is either fully set up and viewing the home dashboard or has completed a Pro subscription purchase and returned to the dashboard. Steps: Welcome Introduction → Set Your Goal (default 10,000) → Adjust Goal (picker change) → Health Access — Initial → Health Access — Enable Writes → All Set Confirmation → Setup In Progress → Pro Upsell — Feature Card → Purchase Flow — Processing → App Store Subscription Sheet → Purchase Success Modal (system) → Purchase Result Toast → Pro Ready / Processing Indicator → Notifications Permission Prompt → Home Dashboard — First-run **Roots: Streak introduction and detail** (ios, 5 screens, Productivity) [Refero](https://refero.design/flows/5748) > Happy path: A logged-in user on the Home tab opens the Streak feature (bolt icon) from the top bar. The app presents a short, multi-step modal tour describing Streaks (what they are, reward tiers, and "pauses"). The user advances through the modals using a prominent green CTA (Next / Got it). After dismissing the tour, the app shows the Streak detail screen with the user's current streak count and a vertical list of achievement tiers and their statuses. Steps: Home — Blocking summary → Streak — Intro (screen 1) → Streak — Maintain your streak (screen 2) → Streak — Earn pauses (screen 3) → Streak detail — Progress list **Foodllama: Weight Logging and Progress** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/11792) > Happy path for a weight-tracking screen. The user enters from a progress dashboard that shows a projected goal date and current weight summary. Tapping the main "Log weight" action opens a bottom-sheet form where the user enters a new weight and selects the date/time of the measurement. Submitting saves the entry and returns to the dashboard, where the current weight, trend line, and projected goal date update based on the new log. Steps: Progress Overview → Weight Entry → Updated Overview → Trend Refresh **Brink: Start Walking Session** (ios, 3 screens, News) [Refero](https://refero.design/flows/10922) > Happy-path: A user is listening to a podcast episode in the app's player and selects the Walk action to start a walking session while the audio continues. The app requests location permission (presenting a map preview and three choices). When the user grants permission, the app switches to a Walking Mode screen that overlays a map with a current-location marker and session metrics (time, steps, distance). Audio controls remain available at the bottom. Steps: Episode Player - Walk CTA → Location Permission Prompt → Walking Mode - Map & Session Controls **The Outsiders: Workouts overview** (ios, 2 screens, Health & Wellness) [Refero](https://refero.design/flows/10313) > Mobile Workouts screen that lets a signed-in user review recent sessions and see a weekly summary. Steps: History List → Summary Overview **PocketShelf: Book progress session manager** (ios, 4 screens, Books) [Refero](https://refero.design/flows/8293) > Happy path: User opens a book detail screen to view progress and starts a reading/listening session. The app switches to an active-session timer view where the user can pause or stop. When stopping, a "Finished at" time picker modal appears so the user records the session end time and saves. After saving, the book detail screen is shown again with updated progress percentage, remaining time estimate, and increased session count. Steps: Book detail — progress overview → Active session — timer → Finished-at modal — set end time → Book detail — updated progress **Variant: Multi-artifact mobile concept generation** (web, 16 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/11475) > Happy path: A designer uses an AI-driven design assistant inside a design workspace to create multiple mobile concept screens from three source artifacts: a technical ledger-style reference, a bold fitness tracker reference, and a pastel gradient image. Entry: user is signed into the design workspace and has two reference screens visible in a multi-window workspace. Exit: the workspace contains a new project titled (e.g., "Multi Artifact Design") with multiple generated mobile artboards based on the selected references and chosen generation settings. Assumptions: user is authenticated, has permission to upload/use reference files, is working on desktop with a multi-panel canvas, and the design assistant service is available. Steps: Open workspace with references → Add more files to assistant → Assistant shows file list and UI state → Open variant/model dropdown → Select Variant-Fast → Prepare generation job → Assistant extracts themes → Refine generation instruction → Generation in progress → Initial results appear → View completed thumbnails → Assistant suggests micro-variations → Placeholders transform into full artboards → Review tiled generated concepts → Open full set of mobile artboards → Finalize or iterate **Teal: Job Tracker Onboarding** (web, 12 screens, Productivity, Job Search) [Refero](https://refero.design/flows/2723) > Happy-path guided product tour that introduces a user to the job tracker list and a single job detail. Entry: user lands on the bookmarked jobs grid and a modal invites them to start a quick tour (or skip). The tour walks the user through: top-stage chevrons (filters), in-grid quick edits (status dropdown), sortable table headers, opening a job from the grid, job-detail actions (close job), guidance banner, right-side job tools (notes, resumes, contacts, templates, checklist), dates section expand/collapse, automatic skill/keyword highlighting, and the left-side job list navigation. Exit: user finishes the tour and the modal closes, returning them to the grid. Assumptions: user is signed in, viewing the 'Bookmarked' tab of their job tracker, has at least three sample job rows, and has permission to view job details and change job status. Steps: Intro modal → Stages chevrons → In-grid status editing → Sortable table headers → Open job from grid → Close job action → Guidance banner → Job tools icons → Dates section → Keyword & skills highlighting → Left job list navigation → Feedback & finish **MWM: MWM AI Workspace Onboarding** (web, 8 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/13328) > A visitor starts on the MWM marketing homepage and selects the MWM AI product through a primary call to action. They authenticate using Google, review a short three-screen product introduction, choose between Pro and Pro+ plans, and arrive at the MWM AI workspace. The final screen is the authenticated home dashboard, where the user can describe an app idea and begin creating a project. The screenshots do not show payment details or confirmation, so plan selection is treated as the final subscription step visible in the flow. Steps: MWM Product Landing → MWM AI Sign In → Google Authorization → Product Welcome → Build and Publish Introduction → Growth Tools Introduction → Plan Selection → Project Creation Workspace **Trip: Traveler Profile and Trip Moments** (web, 2 screens, Travel & Booking) [Refero](https://refero.design/flows/12098) > A logged-in Trip.com member views another user’s public profile and travel content. The entry point is the profile header showing the user name, avatar, stats, and a Follow button. The primary content area shows tabs for Trip Moments, Reviews, and Travel Experience, plus a masonry grid of travel posts and pagination. The user can follow the profile, browse published trip posts, and navigate through pages of content. The exit point is either after following the user or after selecting a post or another profile section. Steps: Profile Overview → Following State **Rox: Manage account next steps** (web, 6 screens, Business, Productivity, AI Tool) [Refero](https://refero.design/flows/10805) > Happy-path: A user opens the Accounts list (table view), expands the Next Steps popover for a specific account, adds two upcoming steps, saves them, optionally views completed steps, and sees the Next Steps column updated with saved step chips. Steps: Open Next Steps popover → Enter first next step → Add a second step → Save next steps → View or move completed items → Confirm Next Steps in table **Product Hunt: Follow Product Profile** (web, 2 screens, Business, Social Networking) [Refero](https://refero.design/flows/11458) > A logged-in user views the Stanley For X product profile on Product Hunt and follows the product from its Customers section. The flow starts on the product’s launch profile with the Customers tab selected and ends on the same profile after the follow action is confirmed by the button state changing to “Following.”. Steps: Product Profile → Follow Confirmation **Mercury: Application Completion Checklist** (web, 3 screens, Business, Finance) [Refero](https://refero.design/flows/11025) > This is a desktop checklist-based completion flow shown to a logged-in applicant who has submitted an account application. Steps: Checklist overview → Get mobile updates — expanded → Checklist updated — mobile updates completed ##### Task: push notification permission **Fuse: Enable Notifications and Offer** (ios, 3 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8731) > Happy path: A logged-in mobile user on the Wallet home (empty balance) taps an in-app prompt to enable push notifications → the system (iOS) permission dialog appears → the user allows notifications → the app shows a confirmation toast and immediately surfaces a Fuse Plus membership upsell with benefits and a CTA to deposit funds. Entry: Wallet home screen. Exit: Fuse Plus membership modal shown with CTA to deposit. Steps: Wallet Home (empty) → iOS Notification Permission → Notifications Enabled + Fuse Plus Offer **SSENSE: Onboarding: Department Selection & Notifications** (ios, 4 screens, Marketplace) [Refero](https://refero.design/flows/7095) > Happy-path onboarding sequence on an iOS mobile screen where the user selects a default shopping department, is asked to enable push notifications, and lands on the app’s curated home (Shop New Arrivals). Steps: Select Default Department → Enable Notifications Explanation → Native Notification Permission Dialog → Shop / Home (New Arrivals) **Spotify: Spotify new user onboarding** (ios, 23 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/209) > Happy-path new free account onboarding for Spotify on iOS. **Kickstarter: Sign up and enable project reminders** (ios, 12 screens, Fundraising) [Refero](https://refero.design/flows/4627) > Happy path: An anonymous iOS user browsing the Explore feed taps "Sign up or Log in" and chooses to continue with Apple. iOS SSO is invoked, which opens an in-app browser (kickstarter.com) to complete account creation. The user fills the sign-up form (name, email, confirm email, password), fixes a mismatched-email validation error, completes reCAPTCHA, and creates the account. The site then shows an OAuth authorization screen to grant the mobile client profile access; the user confirms. Control returns to the app Explore feed where the user saves a project; the app prompts to enable reminder notifications and then requests iOS push permission. Steps: Explore feed (logged out) → Auth options (native modal) → iOS SSO confirmation → Kickstarter login (webview) → Sign up form loaded → Form input (user types values) → Submit attempt with validation error → Correct form and complete reCAPTCHA → OAuth authorize client → Explore feed (logged in) — save project → iOS push permission prompt → Explore feed (final state) **Granola: Granola Account Setup** (ios, 18 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14147) > Happy path for a new Granola user on an iPhone. The user starts logged out, chooses Google sign-in, approves the OAuth and calendar permissions, then completes onboarding questions about acquisition source and work profile. The user enables notifications, reaches the setup confirmation, and enters the main notes area. The flow assumes a Google account is already available and that the user accepts the requested permissions. The final screen shows an authenticated account with no notes or upcoming events yet. Steps: Sign-In Welcome → Sign-In Welcome → Sign-In Welcome → Google Sign-In Prompt → Google Account Consent → Calendar Access Consent → Authentication Loading → Work Email Recommendation → Desktop Introduction → Acquisition Source → Acquisition Source Selected → Work Profile → Work Profile Selected → Notification Education → Notification Permission → Setup Complete → Notes Workspace → Empty Notes Dashboard **TIDAL: Artist Profile and Follow** (ios, 4 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/13129) > A user views an artist profile for Steve Lacy in a music streaming experience. From the artist detail screen, they can play music, shuffle tracks, follow the artist, open radio, share the artist, and browse top tracks. If they choose to follow, the app may show an in-app prompt encouraging notifications, followed by the system notification permission dialog. The flow starts on the artist profile screen and ends back on the same profile screen with the artist marked as followed; the user may also grant or deny notification access. Assumption: the user is already signed in and viewing the artist profile on a mobile device. Steps: Artist Profile → Notification Prompt → System Permission → Follow Confirmed **Plane Finder: Flight Tracking Setup** (ios, 5 screens, Travel & Booking) [Refero](https://refero.design/flows/10391) > User is viewing a single-flight detail (AA1413) and opts to add that flight to in-app tracking (MyFlight). Steps: Flight Details → Add MyFlight Preview → Notification Permission → Confirmation Modal → Active Tracking State **Doji: Phone verification and onboarding** (ios, 10 screens, AI Tool, Marketplace) [Refero](https://refero.design/flows/6223) > Happy-path sign-in / onboarding using phone number verification. Entry: user opens the app and taps Get Started on the brand splash. The flow guides the user to enter a phone number (country code + local number), request a verification code, enter the one-time code, and then lands on a welcome screen prompting for notification permissions. If the user enables notifications (system prompt), they arrive at the authenticated Home screen with product content. Steps: Brand splash / Get started → Phone number input (empty) → Phone number input (filled) → Continue pressed (loading) → Code entry (empty) → Code entry (typing) → Code verifying → Welcome / Enable notifications prompt → Native notification permission modal → Authenticated Home / Product feed **Asos: Initial onboarding & regional settings** (ios, 13 screens, Marketplace) [Refero](https://refero.design/flows/2812) > Happy path: User opens the ASOS app (splash) and lands on the initial screen showing "Deliver to" and large category buttons (Women / Men). The app shows an in-app marketing prompt asking to opt in to notifications; if accepted the system notification permission dialog appears. The user then uses the "Change" link next to "Deliver to" to open Settings → Deliver To. From Deliver To the user can search for and select a country. The user also inspects Sizes and Experimental Features in Settings and toggles features on. After dismissing an 'App updates' message the user reaches the app home feed. Steps: Splash / App Launch → Landing — Delivery + Category Choices → iOS Notification Permission → Landing — Post Permission → Settings — Root → Open Deliver To (loading) → Deliver To — Country List → Country Search (typing) → Country Search Results → Sizes Selection → Experimental Features (default off) → Enable Experimental Toggles → App Updates / Home **WhatsApp: Update About Status** (web, 4 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/13221) > This desktop WhatsApp settings flow starts from the user’s profile/settings area and opens the About editor. The user types or selects a short status message, chooses how long it should stay visible, and confirms the change. The flow assumes the user is already signed in to WhatsApp on desktop and can access profile settings. It ends with the new About text shown in the profile area and a temporary confirmation toast indicating the status is updating. **Microsoft Teams: Change presence status** (web, 4 screens, Communication Tools, Scheduling Tools) [Refero](https://refero.design/flows/9685) > Happy-path: a signed-in user on the desktop client opens the Chat area, opens their account/profile menu, views presence options, and changes their availability to "Be right back." Entry point: Chat landing screen after sign-in. Exit point: Account menu closed with presence updated. Assumptions: user is authenticated, has a profile avatar visible in the top-right, and has permission to change their own presence. The flow focuses only on client-side presence selection (no scheduling or automatic status rules are shown). Steps: Chat landing (Welcome) → Account popover open → Presence options list → Status applied and popover updated **Mapbox: Publish Confirmation Modal** (web, 2 screens, Development Tools, Map & Navigation) [Refero](https://refero.design/flows/9673) > Happy-path: A designer publishes a map style from the Studio editor and receives a centered confirmation modal that the style was published and marked protected. The modal appears over the map canvas and blocks background interactions until dismissed. The modal gives two primary interactions: dismiss the dialog with an 'Okay' button, and optionally submit a quick rating (Bad / Okay / Good). After submitting a rating, the modal shows a thank-you acknowledgement. Steps: Publish Confirmation → Feedback Acknowledgement **Claude: Profile Settings — Prompt Suggestions** (web, 2 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/1822) > Happy path: A logged-in user navigates to Settings → Profile (left nav). The screen shows editable name fields, a work-function dropdown, and a toggle to show prompt suggestions. The user can update their name (via the "Update Name" button) or enable the prompt suggestions toggle. When the toggle is switched on, the UI shows a top-right toast confirming the change. Steps: Profile overview (edit name & preferences) → Enable prompt suggestions — confirmation **Сontra: Email preference settings** (web, 3 screens, Job Search) [Refero](https://refero.design/flows/4893) > Logged-in user navigates to Settings → Email Preferences on a desktop layout to view and change which types of emails they receive. The screen lists individual email categories (Marketing, Support, Profile Suggestions, Newsletter, Availability Reminders) as checkboxes, plus an "Unsubscribe from all emails" option and an "Update Email Preferences" action button. Happy path: user modifies one or more checkboxes, clicks Update, and receives an on-screen confirmation. **adidas: Back-in-stock notification** (web, 3 screens, Marketplace, Sports) [Refero](https://refero.design/flows/10236) > Happy-path: A user viewing a product detail page (desktop) that has at least one size out of stock uses the product's "Notify me" mechanism to request a back-in-stock alert for a specific size. Steps: Product detail overview → Size selection and submit → Confirmation and next steps ##### Task: referral to specialist **Blackbird: Referral — share $FLY reward** (ios, 5 screens, Food & Drink) [Refero](https://refero.design/flows/9023) > Happy path: a signed-in user on the Wallet tab opens the referral offer (Get $20), views a referral modal with a QR code and alphanumeric referral code, and then shares that referral via the iOS share sheet (Messages). Steps: Wallet Home → Referral Details (QR + Code) → Referral Share — action initiated → iOS Share Sheet → Message Composer — prefilled referral **Melius: Send Referral Invitation** (web, 3 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/14098) > A logged-in user opens the Referrals section from account settings, reviews how the referral reward works, enters a recipient’s email address, and sends an invitation. The system confirms the invitation with a success notification and clears the email field. The flow begins in the Referrals settings modal and ends after the invitation is successfully sent. Assumptions: the user has permission to access account settings and referral invitations, and is using a desktop interface. Steps: Referral Details → Enter Recipient Email → Invitation Confirmation **Airbnb: Refer-a-host referral creation** (web, 5 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6009) > Happy path: A logged-in user visits the "Refer a host" landing hero, clicks Get started, picks which type of host they are referring (Home, Experience, or Service), generates a referral link, and immediately opens sharing options to distribute the link (copy, email, social, messaging). Entry: Referral landing hero while signed in (avatar visible). Exit: Referral link generated and shared (copy confirmation shown). Assumptions: desktop layout, user is authenticated, referral program is available for their account. Steps: Referral landing hero → Choose host type (modal) → Host type selected → Share referral options → Copy confirmation **Mercury: Refer a Contact via Email** (web, 5 screens, Business, Finance) [Refero](https://refero.design/flows/11080) > Happy-path: A logged-in user navigates to the Referrals area to invite a contact via email. From the referrals overview (shows bonus card, referral URL, and a table of referred companies), the user opens an "Invite via email" modal, enters or selects an email address, confirms the invite, and returns to the referrals page where a success toast is shown. Steps: Referrals overview → Invite modal (initial) → Typing recipient email → Recipient tokenized and confirm → Referrals overview (success state) **Sunsama: Referral Invitation** (web, 5 screens, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/11602) > A logged-in subscriber opens an invitation screen and chooses to invite people through a referral offer. The user selects the option to tell friends, enters an invitee’s email address and optional name, then sends the invitation. The system confirms that the invite was sent. A Close control remains available throughout the experience. Steps: Invitation Options → Referral Form → Enter Recipient Details → Send Invitation → Invitation Confirmation **Farfetch: Refer-a-friend invite** (web, 3 screens, Marketplace) [Refero](https://refero.design/flows/3553) > Logged-in customer lands on a "Refer a Friend" landing area that communicates a dual-sided discount offer (friend gets 10% off; referrer gets 10% off after spending threshold). Happy path: user reads the offer, taps the main CTA to participate, selects a sharing method or copies a personalized referral link, and the system copies the link and confirms it has been emailed to the user. Steps: Referral Offer Overview → Choose Invite Method → Share Confirmation **Manus: Send referral invite** (web, 4 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/10479) > Happy path: a logged-in user opens the main dashboard and uses the "Share Manus with a friend" control in the left sidebar to open a referral/invite modal. In the modal the user can copy an invite link, trigger a Facebook invite flow, choose one of the shown social share icons, or enter an email address and send an invitation. After submitting an email the system shows a transient success toast; the modal remains open and displays invitation counters and links for redeem/invitation history. Steps: Dashboard - open share control → Invite modal - view options → Invite modal - enter email → Invite modal - confirm send **Prose: Referral Program — Share & Confirm** (web, 3 screens, Marketplace, Health & Wellness) [Refero](https://refero.design/flows/6321) > Logged-in user accesses the Prose Referral Program from their account sidebar. The page presents a two-column layout: a large brand/product image on the left and referral controls on the right. The right column exposes the referral link (with a COPY button), an email composer to send invites, a prefilled message box, an optional reminder checkbox, and social share icons. Happy path: user either copies the referral link or enters recipient emails and clicks SEND. After sending, the page navigates to a confirmation state that shows a thank-you message, the same referral link (copyable), and a CTA to "Share Again". The top of the page also displays referral totals (Total Earned, Total Used, Current Balance). Steps: Share options / Compose start → Recipients added / Ready to send → Share confirmation ##### Task: request money split bill **Revolut: Help centre article discovery** (web, 3 screens, Business, Finance) [Refero](https://refero.design/flows/1317) > Help centre entry → search → article view (happy path). Entry: user lands on the public Help centre (search input and topic cards visible). User can either browse topic cards (Personal / Business toggle) or type a query in the top search field. Selecting a suggested search or a card opens a detailed article view with breadcrumbs and the article body. Assumption: this is a public help/FAQ area; user may be signed in or out and is only reading help content. Steps: Help centre landing / browse topics → Search with autocomplete suggestions → Article detail view **Spoil me: Invite and Earn** (ios, 3 screens, Fundraising) [Refero](https://refero.design/flows/5430) > User views an “Invite and Earn” screen that shows a personal invite link/code, a simple earnings estimator, and controls to copy or share the invite. Entry: signed-in user opens the Invite and Earn section (iOS device). Exit: user either copies the invite link, opens the system share sheet to send the link, or cancels the modal. Assumptions: user is authenticated (invite code is shown), on an iOS device (native share sheet present), and has basic clipboard/share permissions. Steps: Invite overview and estimator → Invite code modal with actions → System share sheet **GoFundMe: Donate and Share Confirmation** (web, 8 screens, Fundraising) [Refero](https://refero.design/flows/11729) > Happy-path fundraising donation journey on GoFundMe. Entry begins on a fundraiser detail view with the donor logged in, then moves into the donation checkout, donation confirmation, optional sharing, and receipt viewing. Exit returns the user to the fundraiser detail view. Steps: Fundraiser Overview → Donation Amount Setup → Currency Selection → Donation Payment Form → Donation Confirmation → Post-Donation Actions → Receipt Dialog → Fundraiser Return ##### Task: research collaboration **ChatGPT: Research timeline with sources** (ios, 17 screens, AI Tool) [Refero](https://refero.design/flows/7024) > Happy path: A signed-in user opens the chat, initiates a research-style request, and composes a prompt asking for a Cold War summary. The user selects the Research helper / template and chooses a specific output format ("Timeline of key events"). The system starts a long-running background research task and surfaces a progress card in the chat. The user can expand a Details/Activity view to see the agent’s step-by-step work and the accumulated source list (sources tab). When the research completes the assistant posts a compiled timeline with inline source tags and embedded images; images can be opened full-screen. Steps: Chat home / quick prompts → Create / helper sheet → Compose research prompt → Assistant clarification → Choose timeline option → Start background research → Progress card (initial) → Progress updates (collection) → Research completed header → Open Activity details → Activity timeline scroll → Sources tab (list) → Compiled timeline with image thumbnails → Open image full-screen → Return to timeline → Final events and citations → Follow-up composer ready **Grok: Research Assistant — DeepSearch Session** (ios, 15 screens, AI Tool) [Refero](https://refero.design/flows/6092) > Happy-path: a user opens the chat interface, chooses DeepSearch, types a long research prompt, sends it, and watches the assistant perform a web-enabled research pass (“DeepSearching”) while showing a streaming/think indicator. The assistant returns a multi-part answer: summary bubbles, a "Thought" timeline/checklist of research steps, a browsable list of source pages, and a long structured article with headings, a comparative table, and final citations. Steps: Home — Chat Composer Visible → Activate Composer / Keyboard → Type Research Prompt → Submit Prompt → Streaming / Partial Response → Result Bubble + Source Summary Link → Expanded 'Thought' / Progress Label → Open Thought Detail Panel → Open Pages (Source List) → Preview / Read a Source → Thought Checklist Progress → Generated Structured Answer Appears → Comparative Table Rendered → Citations List → Composer Still Available / Continue Interaction **Spyglass: Creative Strategy Follow-up** (web, 4 screens, Business, Inspiration Board) [Refero](https://refero.design/flows/13463) > A user reviews a generated advertising analysis in a desktop workspace and prepares a follow-up question about target audience segments and scaling. The user can begin composing the follow-up through the message composer, inspect the voice-input control, activate voice recording, or enter the question as text. The shown sequence ends with the follow-up typed into the composer but not yet submitted. Steps: Review Generated Analysis → Inspect Voice Input → Activate Voice Input → Draft Audience Question ##### Task: review and confirm before submit **PayPal: Help article feedback** (ios, 5 screens, Finance) [Refero](https://refero.design/flows/7267) > Happy path: A user reads a PayPal Assistant help article and indicates whether the article resolved their question by tapping a thumbs-up/thumbs-down control. If the user taps the thumbs-up, an inline optional feedback form (text area, 0/250 counter) appears. The user can tap the text area to open the keyboard, type a message (up to 250 characters), and submit via a prominent 'Send Feedback' button. After submission the UI shows a small success confirmation message inline and returns to the article. Steps: Article read & feedback prompt → Select rating (thumbs-up) → Focus feedback text area → Compose feedback message → Submit feedback & confirmation **Spotify: Submit episode reply** (ios, 6 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/214) > Happy-path flow for a logged-in user to submit a text reply to a podcast episode Q&A and then view or delete that reply. Entry: user lands on an episode detail screen that includes a Q&A prompt. Exit: user sees their reply in the Replies list (with an option to delete). Assumptions: user is signed in, has an editable display name/profile image, and has network connectivity. Flow covers: open composer → accept participation terms → type reply (character limit shown) → send → confirmation toast → view reply in Replies list. **IKEA: Write a product review** (web, 10 screens, Marketplace, Home & Decor) [Refero](https://refero.design/flows/8234) > Happy-path: A signed-in desktop user opens the product's Reviews area and starts the inline multi-step review flow delivered as a right-side drawer. The user first selects an overall star rating, writes a short title and detailed review, answers quick attribute ratings (sliders) and a recommendation question, reviews the assembled content, accepts terms, and submits. Steps: Reviews listing / CTA → Overall rating (initial) → Overall rating (selected) → Write title and detailed review (empty) → Write title and detailed review (filled) → Quick attribute ratings (empty) → Quick attribute ratings (set) → Summary preview (terms unchecked) → Summary preview (terms accepted) → Submission confirmation **Anam: Feedback Submission** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12951) > A logged-in user is on a feedback submission screen inside the product. They choose a feedback category, enter comments, and submit the form. After submission, the interface shows a confirmation message and offers a button to submit another response. The flow assumes the user already has access to the product and can interact with the feedback form from the main workspace. Steps: Feedback Form → Feedback Entry Filled → Submission Confirmation **Dropbox: PDF Editing Feedback Submission** (web, 4 screens, Collaboration Tools, Business, Productivity) [Refero](https://refero.design/flows/13053) > This is a PDF editing interface where the user can optionally submit product feedback. Entry starts on an open PDF with editing controls and a visible ‘Feedback on PDF editing’ link in the lower-right. Clicking it opens a centered feedback modal over the document. The user can type freeform feedback, optionally opt into Dropbox user research with a checkbox, then send or cancel. On successful submission, the modal closes and a confirmation toast appears at the bottom of the screen. Assumptions: the user is already inside the PDF editor and has permission to edit and submit feedback. **Hello Klean: Product review submission** (web, 25 screens, Marketplace) [Refero](https://refero.design/flows/5384) > Happy-path: A user visits a product detail screen and opens an overlaid modal to submit a product review. The modal is a multi-step, linear wizard that collects: star rating, written review, review title, quantified improvement, overall satisfaction, hair type, optional media, reviewer name and email, and a final recommendation checkbox. After submission the UI shows an intermediate submit state that requires email confirmation (user sees a 'You're almost there' message). Steps: Product detail (entry) → Select star rating (empty) → Select star rating (selected) → Write full review (text input) → Review text filled (preview) → Enter review title → Title entered (filled) → Rate improvement in hair condition (choices) → Improvement rating selected → Overall satisfaction question → Select hair type (options list) → Hair type selected → Attach media placeholder → Media uploaded (thumbnail) → Enter reviewer name → Enter reviewer email → Review recommendation & submit → Post-submit confirmation required → Modal close / return to product → Optional verification (outside screenshots) **Urban Outfitters: Product Review Submission** (web, 6 screens, Marketplace) [Refero](https://refero.design/flows/549) > Happy-path: A desktop shopper on a product detail page opens the "Write a Review" overlay, rates the product, writes a review title and body, optionally uploads photos, fills basic self/descriptive details (username, location, age, height, sizes), and submits. The system validates required fields, accepts image uploads (up to 6, 10MB each as indicated), then shows a submission acknowledgment that the review may take time to appear. Steps: Product Detail / Write a Review CTA → Review Modal — Initial Form → Review Modal — Populate Content → Add Photos & Personal Details → Form Ready to Submit → Submission Confirmation **Xbox: Report a review** (web, 5 screens, Entertainment, Marketplace) [Refero](https://refero.design/flows/7695) > Happy-path: a signed-in user browsing product reviews identifies a problematic review, opens an in-context "Report this review" modal, selects a violation reason from a dropdown, optionally enables an email update or provides extra information, submits the report, and receives an on-screen confirmation. Steps: Review listing → Report dialog — initial → Select violation reason → Set update preferences → Submission confirmation **Square: Report a review** (web, 5 screens, Business, Finance) [Refero](https://refero.design/flows/1791) > Happy-path: A signed-in merchant (or buyer) views the Jotform reviews in a marketplace modal, locates an individual review, uses the review item’s ellipsis menu to report it, completes a short report form (selecting a reason and optionally entering details), and receives a confirmation that the report was submitted. Steps: Open reviews list → Open review actions → Complete report form → See confirmation → Return to reviews & paginate **Nudea: Report Review** (web, 4 screens, Marketplace) [Refero](https://refero.design/flows/7660) > Happy path: A shopper on a product detail page opens a modal to report a customer review, fills in their email and a brief reason, submits the report, and sees an on-screen confirmation. Steps: Product detail (reviews visible) → Report review — blank form → Report review — filled form → Report submitted confirmation **Mercury: Check image upload** (web, 6 screens, Business, Finance) [Refero](https://refero.design/flows/11046) > Happy path for uploading front and back images of a check as part of a mobile-deposit-style onboarding step. Steps: Upload prompt (front) → Upload progress → Front image preview modal → Front thumbnail displayed → Back image preview modal with endorsement checkbox → Both thumbnails present; enable Review **adidas: Leave a product review** (web, 4 screens, Marketplace, Sports) [Refero](https://refero.design/flows/10240) > Happy path: A shopper on desktop views a product detail and chooses to write a review. They open the 'Write your review' form (product thumbnail and guidance shown), complete required fields (rating, recommend, written review, size/width sliders, comfort/quality ratings, nickname), optionally upload a photo, and submit. After successful submission the site shows a confirmation banner and CTAs to continue shopping. Steps: Product detail - Reviews section → Write review - blank form → Write review - completed form → Review submission confirmation **DoorDash: Store review submission** (web, 6 screens, Food & Drink, Food Delivery) [Refero](https://refero.design/flows/10151) > Happy-path user journey for submitting a store (restaurant) review from a store's Ratings & reviews list. Steps: Ratings & reviews listing → Rate prompt modal → Negative sentiment review form → Positive sentiment review form → Review entry and submit → Submission confirmation and moderation notice **Exa: Search Response Feedback** (web, 5 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/12015) > Happy path starts on a search interface in the API Playground, with a query already entered and results shown on the right in JSON/output view. The user can optionally open a feedback dialog for the response, enter a short comment, and submit it. The flow ends with a confirmation message indicating the feedback was sent. Assumptions: the user is already logged in, has access to the Search tool, and is viewing a completed response in the output panel on desktop. Steps: Review Search Output → Open Feedback Prompt → Enter Feedback Comment → Submit Feedback → Feedback Confirmed **Grammarly: Proofreading suggestion review** (web, 4 screens, Paperwork) [Refero](https://refero.design/flows/1872) > Happy path: A user reviews automated proofreading suggestions presented in a right-side assistant panel while editing a document. Steps: View suggestions list → Open a suggestion → Read detailed explanation → Apply correction and confirm **Mocha: Gallery review and feedback** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/11611) > Happy-path: A user inspects a photo gallery in a desktop builder/preview environment, reviews a list of imported assets and staged changes in the left side panel, optionally submits feedback or votes on suggestions, and receives an on-screen confirmation. Steps: Gallery preview with asset panel → Open review / feedback controls → Submit feedback and receive confirmation **Programa: Meeting Summary Feedback** (web, 4 screens, Project Management, Business) [Refero](https://refero.design/flows/13433) > A logged-in user is viewing an existing meeting summary in the Meetings section. They open the summary feedback dialog, enter an optional comment, and submit it. The dialog closes and a confirmation message appears while the user remains on the meeting overview. **Expedia: Report a guest review** (web, 4 screens, Travel & Booking) [Refero](https://refero.design/flows/3007) > Happy path: A signed-in desktop user opens the property's Guest reviews modal, locates a negative review, launches the Report review form from that review, selects a reason (checkbox), and submits the report. **Lovable: Design Preview Comments** (web, 5 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12338) > A collaborative design review view for an AssetWise landing page preview. The user starts on the page in preview mode, can place an inline comment on specific content, type and submit the note, and then view the saved comment in a right-side comments panel. Entry is the preview screen with comments enabled; exit is a published note visible in the comments sidebar. Steps: Preview Landing Page → Open Comment Composer → Write Comment → Submit Feedback → View Comments Sidebar ##### Task: risk register **Tinder: Report a Safety Concern** (ios, 6 screens, Dating & Relationships) [Refero](https://refero.design/flows/3439) > Happy-path: A signed-in user on the Matches tab opens the Safety Toolkit from the shield icon, chooses to Report, reads the reporting options and tips in a compact modal, then follows a link to a Help Center article about reporting in-person physical harm. From the article the user can read resources, tap START REPORT to begin an external report flow, and optionally submit feedback about the article. Entry: Matches (Messages) tab. Exit: user lands in Help Center article and can tap START REPORT (external report flow) or submit article feedback. Assumptions: user is logged in on a mobile device; the Safety Toolkit is accessible from a shield icon in the header; Help Center content is displayed inside an in-app webview. Steps: Matches (empty) / Safety shield → Safety Toolkit options → Reporting modal (tips & categories) → Help Center article - reporting details → Help Center resources & START REPORT → Article feedback and submit **Roots: Subscription trial onboarding** (ios, 20 screens, Productivity) [Refero](https://refero.design/flows/5740) > User views a promotional subscription screen for the Roots app and moves from discovery (hero and benefits) to signing up for a 7‑day free trial, choosing a plan, confirming via the App Store, and receiving a purchase confirmation. Steps: Promo hero & primary CTA → Trial timeline & CTA repeat → Scroll replacements & value entry → Value calculator - inputs → Value calculator - results → Mission + product positioning → Testimonials and FAQ list → FAQ expanded content → Support & account actions → Request iOS notification permission → Reminder toggle set → Plan selection sheet → Feature comparison summary → Feature details grid → Confirm/Continue CTA in sheet → App Store authorization prompt → App Store purchase sheet (details) → In-app purchase success alert → Final entitlement confirmation → Post-purchase state & support links ##### Task: savings pot goal **Loowner: Set personal goal** (ios, 7 screens, Marketplace) [Refero](https://refero.design/flows/6596) > Happy-path: A logged-in user on the main dashboard opens the Goal card, reads a short explainer, taps Set up goal, enters a numeric target (either quantity or cost) using an in-screen numeric keypad, confirms the change, sees a confirmation screen, and returns to the dashboard where the Goal card is updated to reflect the new target and progress. Steps: Dashboard — Goal card → Goal explainer modal → Set up your goal — input sheet (initial) → Enter numeric goal → Confirm goal change → Success acknowledgement → Dashboard — Goal card (updated) **Waterllama: Hydration goal tracking** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/11531) > Mobile hydration dashboard where the user logs beverages to reach a daily hydration goal. Steps: Dashboard — initial low progress → Dashboard — nearing goal → Goal reached — celebration **The Outsiders: Training Goal Creation** (ios, 14 screens, Health & Wellness) [Refero](https://refero.design/flows/10301) > Happy-path: The user is on the Progress (Training Summary) screen and adds a new training goal via a modal sheet. Entry: Progress screen, logged-in state (user avatar visible). Exit: Progress screen with a new Training Goal listed under Training Goals. Assumptions: user is signed in and has access to Training Goals; device is a mobile phone; the Add Goal UI is presented as a modal sheet over the Progress screen. Steps: Progress / Training Summary → Add Goal (initial) → Workout Type List (top) → Workout Type List (scrolled) → Workout Type Selected → Open Goal Type → Goal Type List (options) → Goal Type Selected → Open Time Frame → Time Frame Selection → Open Numeric Goal Picker → Set Numeric Target → Review & Add → Training Goal Visible **Apple Books: Daily Reading Goal Adjustment** (ios, 4 screens, Books) [Refero](https://refero.design/flows/11912) > A mobile reading-progress screen shows the user’s current progress toward a daily reading goal. The user can dismiss the screen, share progress, or open a goal-adjustment bottom sheet to change the daily reading target in minutes per day. Assumption: the user is already signed in and viewing a reading-tracking experience on a mobile device. The happy path shown is opening the goal editor, choosing a new minute target, and returning to the progress screen with the updated goal reflected. Steps: Reading Progress Overview → Goal Editor Opened → Goal Value Selection → Updated Progress View **ElevenReader: Set Daily Listening Goal** (ios, 4 screens, AI Tool, Books) [Refero](https://refero.design/flows/8967) > Happy path: User on the Home feed taps the 'Daily Listening Goal' card, which opens a modal bottom sheet to set a daily listening target. The user adjusts the minutes using minus/plus controls, saves the selection, and the home card updates to reflect the new daily target. Steps: Home — Discover & Goal Card → Open Goal Editor (Bottom Sheet) → Adjust Minutes → Save Goal and Update Home **Apple Books: Daily Goal Achieved** (ios, 3 screens, Books) [Refero](https://refero.design/flows/11936) > This is a mobile success-and-sharing confirmation after the user reaches a daily reading target in Apple Books. The flow starts on a completion screen showing that the reading goal has been achieved and ends either by dismissing the screen or sending the achievement through iMessage after using the share action. Steps: Goal Complete Screen → Share Sheet → Message Composer **Airbnb: Browse and Save Listing** (ios, 2 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6388) > Happy path: A signed-in mobile user opens the Explore tab, scrolls horizontally through curated carousels (e.g., "Popular homes in Cleveland" and "Available in Cincinnati this weekend"), and saves a listing to a wishlist by tapping the heart icon. Steps: Browse curated carousels → Show save confirmation **Cake Equity: Add personal equity goal** (web, 4 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/6155) > Happy path: a logged-in user on the Equity explainer dashboard adds a personal goal tied to their equity by opening an in-context modal, entering a goal name, selecting an emoji, adding a short description, and saving it. Steps: Dashboard — Equity explainer → Set your goal — Modal (empty) → Set your goal — Modal (filled) → Dashboard — Goal displayed **Revolut: Vault Goal & Deadline Setup** (web, 3 screens, Business, Finance) [Refero](https://refero.design/flows/1337) > Happy-path: A logged-in user views a single 'Vault' detail screen, adds a monetary goal via a modal, confirms it, and sees the vault updated with the goal and deadline. Steps: Vault Details → Goal Entry Modal → Vault Updated **Literal.club: Add books to yearly reading target** (web, 7 screens, Books, Communication Tools) [Refero](https://refero.design/flows/7911) > Happy path for adding books to a user's yearly reading goal. Entry: user is on their goal overview (shows progress ring and empty/placeholder slots). The user opens the manage-books overlay, uses a search/autocomplete to find titles, adds one or more books to the goal, and then returns to the goal overview where the progress ring and the grid of book covers update. Assumptions: user is signed in and viewing their own goal (editable), on a desktop/laptop viewport, and has permission to edit this goal. Steps: Goal overview (empty) → Manage books overlay (empty list) → Add book — search input → Select book from results → Manage books — single item → Manage books — multiple items → Goal overview (updated) **Literal.club: Reading goal setup** (web, 11 screens, Books, Communication Tools) [Refero](https://refero.design/flows/7910) > Happy-path: A logged-in user on their profile navigates to the Goals area, starts creating a new reading goal, uses a quick dial to pick a target number of books or opens the advanced setup to name the goal and set a custom start/end date, enables automatic addition of read books, and then creates the goal. Steps: Profile — Goals tab (empty) → Goals listing — create CTA → New reading goal — quick dial (0) → New reading goal — dial set to 15 → Advanced setup — form shown → Fill name field → Open start-date picker → Select start date → Open end-date picker → Select end date and toggle options → Create goal and view goal detail **GoFundMe: Fundraiser Details and Organization Follow** (web, 3 screens, Fundraising) [Refero](https://refero.design/flows/11731) > A donor opens a fundraiser detail page for a dog’s medical costs. The main entry point is the public fundraiser page showing the title, story, updates, organizer, sharing tools, and donation summary. The user can donate or share from the page, and can also open the organization card modal to view Pasadena Humane details and choose to learn more or follow. The flow ends with the organization marked as followed in the modal state, or the modal dismissed back to the fundraiser page. Steps: Fundraiser Detail Page → Organization Profile Modal → Following Confirmation State **Teal: Set weekly application target** (web, 4 screens, Productivity, Job Search) [Refero](https://refero.design/flows/2735) > Happy path: A logged-in user on their career dashboard opens the Job Applications card, launches the 'Application Goal' modal, sets a weekly applications target using the +/- controls (or typing), and saves. The modal closes and the Job Applications card updates to reflect the newly saved target and current progress (remaining applications or success message). Steps: Dashboard - Job Applications overview → Application Goal modal (open) → Application Goal modal (adjust value) → Dashboard updated with saved target **Cushion: Weekly Goals Check-in** (web, 7 screens, Communication Tools, Collaboration Tools) [Refero](https://refero.design/flows/12511) > A logged-in workspace member opens the 'Weekly Goals' check-in from the left navigation, writes a weekly update in the post composer, uses the checklist formatting option to turn the content into a checklist, adds several goal items, and submits the post. After submission, the composer is replaced by the published check-in entry, grouped under the current week with a reply field below it. The interface suggests a regular check-in space with a 'Reminders only' setting and a simple post-and-reply interaction model. Steps: Open Check-in → Choose Checklist Format → Insert Checklist Marker → Enter First Goal → Build Goal List → Submit Check-In → View Published Entry ##### Task: scheduled payment standing order **PayPal: Set preferred online payment** (ios, 5 screens, Finance) [Refero](https://refero.design/flows/7251) > Happy-path flow for choosing and saving the preferred funding source used when shopping online. Steps: Wallet overview → Preferred when paying online (list) → Info modal about fallback funding → Save preference → Confirmation and updated Wallet **PayPal: Set preferred payment method** (web, 4 screens, Finance) [Refero](https://refero.design/flows/7385) > Happy-path flow for selecting the PayPal balance as the preferred way to pay from the Wallet section. Steps: Wallet overview → Preference modal (initial) → Select payment option → Confirmation success ##### Task: search everything **Substack: Global search with empty-note results** (ios, 3 screens, News, Social Networking) [Refero](https://refero.design/flows/4486) > Happy-path: a logged-in user on the Home feed opens the global search, types a query (numeric in this example), and inspects filtered results. Steps: Home feed — open search → Search input — type query → Results with filter chips — empty state **XChat: Conversation Filtering and Mark-Read** (ios, 4 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/11672) > Happy-path: A signed-in user opens the Chat list, uses the top-right filter button to narrow conversations (Direct / Groups / Unread / All) or mark messages as read. The filter menu appears as a floating rounded card. Choosing a filter updates the header subtitle and the conversation list; choosing "Mark All Read" clears unread messages and shows an empty unread state. The bottom persistent search/compose controls remain visible throughout. Entry point: Chat list screen (user is logged in). Exit point: Filtered conversation list or an empty-state screen indicating there are no matching conversations (e.g., "No Unread Messages"). Assumptions: User is authenticated, on a mobile device (portrait), and has at least one conversation. The app supports conversation-level read/unread state and categorization into direct vs group conversations. Steps: Conversation List (All) → Open Filter Menu → Unread Empty State → Groups Filter Result **BeReal.: Account creation and onboarding** (ios, 29 screens, Social Networking) [Refero](https://refero.design/flows/4896) > Happy path onboarding for a new iOS user: app launch → phone number entry (with country code selector) → privacy/consent modals → OS tracking prompt → verification flow → create username → contacts permission and optional contact sharing/invites → notification permission → final home screen with CTA to post first BeReal. Entry: fresh install / first-time open. Exit: user lands on main friends feed (onboarding complete). Steps: Splash → Loading → Phone number entry → Privacy / trackers modal → OS tracking permission (ATT) → Country selector (list) → Country search → Phone formatted input → Send verification → Enter verification code → Create username (unavailable) → Create username (available) → Contacts permission prompt (OS) → Contacts sharing options → Add friends (select contacts) → Invite via SMS composer → Notifications rationale → OS notification permission → Finalizing onboarding (loading) → Home feed (onboarding complete) **Volv: Search from profile** (ios, 3 screens, AI Tool, News) [Refero](https://refero.design/flows/2164) > Happy path: A logged-in user opens the profile/drawer, taps the prominent "search." control, types a query and submits it. The UI transitions from the drawer to a dedicated search screen, focuses the text input (keyboard appears), then displays search results or an empty-state message. Steps: Open profile / invoke search → Input focused — type query → Submit search and show result **District: Search Discovery** (ios, 6 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12842) > A user enters the global search screen, types a query, and reviews results across category tabs such as All, Dining, Movies, Events, Stores, and Activities. The entry point is the search bar from the main discovery area; the exit point is either selecting a result, refining the query, switching tabs, or clearing the search. **SoundCloud: Search results — track discovery** (web, 2 screens, Music Streaming, Social Networking) [Refero](https://refero.design/flows/6580) > Desktop search-results flow for discovering tracks matching a text query (example: "maverick sabre"). Entry: user types a query into the top search field and submits. The interface returns a list of matching tracks in the main column with a left-hand filter panel. Each result row includes artwork, a play button, waveform preview, metadata (title, artist, tags, age), and interaction controls (like, repost, share, more). The left column lets the user narrow results by category (Everything / Tracks / People / Albums / Playlists), time added, length, and tag filters. Exit points: play a track, open a track detail page, click an artist, apply filters, or refine the search query. Assumptions: desktop/laptop viewport, user can access search and filter controls; flow shows the happy path of viewing and refining search results. Steps: Search results — broad results → Search results — refined results **Cushion: Workspace Search Management** (web, 3 screens, Communication Tools, Collaboration Tools) [Refero](https://refero.design/flows/12531) > Happy path starts on the workspace search screen with the left navigation visible and a global search bar at the top. The user is already logged in and can search across content types such as All, Posts, Comments, and Check-ins. The visible state shows a recent search term and a Clear action. The flow ends when the user clears or reviews recent searches from the search screen; no result detail view is shown. Steps: Search Screen → Search Screen → Search Screen ##### Task: search results filters **YouTube: Search filters selection** (ios, 4 screens, Entertainment, Video Streaming, Social Networking) [Refero](https://refero.design/flows/2636) > Happy-path: A mobile user performs a keyword search for a song, opens the overflow menu to access search controls, navigates into the Search filters screen, reviews and changes filter groups (Sort by, Type, Upload date, Duration, Features). Steps: Search results → Overflow menu (bottom sheet) → Search filters — top groups → Search filters — duration & features **Wabi: Wabi Search and Discovery** (ios, 11 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12064) > This is a mobile search-and-browse experience for Wabi content. The user enters from the Home screen or the search entry point in the bottom navigation, then uses the Search screen to view popular/recent suggestions, type a query, inspect live results, switch between All/Apps/People filters, and either open an existing result or create a new Wabi if nothing matches. Assumption: the user is already on the device and can access search; no login or permission prompts are shown. Steps: Home Browse → Search Entry → Search Suggestions → No Match State → Partial Text Results → Expanded Results List → More Results and Suggestions → All Results View → All Filter → Apps Filter → People Filter **YouTube Music: Artist search results** (ios, 3 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11360) > Happy path: A user searches for an artist name in the app's global search bar, then inspects results using the horizontal filter chips (Albums, Songs, Profiles, etc.). The flow shows three focused result types in sequence: Albums, Songs, and Profiles. Steps: Albums results → Songs results → Profiles results **UNHINGED: Person notes search and open** (ios, 5 screens, Health & Wellness) [Refero](https://refero.design/flows/6342) > Happy path: From the People list the user opened a person's profile (Harry) and switched to the Notes tab. They used the inline search bar to find a specific note by typing keywords. If the typed term finds matches the note list is filtered; if not, the list is cleared/empty. The user selects a matching note and is taken to the Note detail where the note text and metadata (Person, Category) are visible. **Dropbox: Search with type filter** (ios, 4 screens, Business, Collaboration Tools, Productivity) [Refero](https://refero.design/flows/2465) > Happy-path: a signed-in user on an iOS Dropbox-like search screen types a query (“Space”), then opens the Type filter, picks the Images category, and returns to the search results which are updated to show only image files. **Pinterest: Search Results — Profiles Empty State** (ios, 2 screens, Reference, Inspiration Board, Social Networking, Lifestyle) [Refero](https://refero.design/flows/2509) > Mobile search flow on the app's Search tab. Entry: user types a query in the search bar (query: "fantastic mr fox aesthetic"). The first screen shows image-based results in a masonry/grid layout with category chips (Profiles, Wallpaper, Outfit, Icon) and an active chip (Wallpaper). The user then taps the Profiles chip to limit results to people; the app navigates to a Profiles-specific results view and displays an empty-state message when no matching people are found. Exit: user can go back via the back arrow or change the query/filters. Assumptions: user is on a mobile device in the Search tab; the search query is pre-filled; category chips act as filter switches that request different result types. Steps: Search Results — Image Grid → Profiles — Empty State **Julienne: Saved recipe search** (ios, 4 screens, Food & Drink) [Refero](https://refero.design/flows/6493) > Mobile cookbook — user navigates their saved Recipes tab and uses the in-list search to find specific recipes. Steps: Recipes list (default) → Activate search (keyboard open) → Search yields no matches (empty results) → Search returns filtered results **DailyArt: Museum selector** (ios, 2 screens, Education, Magazines & Newspapers) [Refero](https://refero.design/flows/3322) > This is a modal-style museum selector used to filter art search results. Entry: user opens the 'Museum' selector from a higher-level browse/search context (modal or drawer). Happy path: the user scrolls the alphabetized list, types into the search field to narrow results, toggles one or more museums to include, then taps 'Show Result (N)' to apply the selection and return to the previous screen with filters applied. Exit: tapping 'Show Result' applies the selected museums and closes the selector; tapping the back control dismisses without applying changes (visuals suggest selection may be preserved until cleared). Assumptions: selector is presented over a browsing/search surface; the user has permission to filter results; selections persist while the selector remains open. **Apple TV: Content Search and Filtering** (ios, 7 screens, Entertainment, Video Streaming) [Refero](https://refero.design/flows/11100) > Happy path: User opens the app's Search landing (genre/category grid) and starts a search using the bottom search pill. The search control expands, the keyboard appears and the user types a query. The app shows inline suggestions while typing and updates a results list. The user can switch between result tabs (Top Results, TV Shows, Movies, Cast & Crew) to filter results. Steps: Search landing / categories → Focus search input → Enter numeric query → Empty results state → Type text query with suggestions → View top results list → Filter results by Movies **Dropbox: Recover deleted files search** (ios, 5 screens, Business, Collaboration Tools, Productivity) [Refero](https://refero.design/flows/2463) > Happy path: A signed-in mobile user opens the Recover Deleted Files area, taps the search field, types a query, and views the search progress and final results (no matches). **Longevity Deck: Discover Search & Filter** (ios, 6 screens, Health & Wellness) [Refero](https://refero.design/flows/9889) > Happy-path for searching content on the Discover screen of a mobile app. Steps: Discover Home → Activate Search → Enter Numeric Query → No Results Found → Enter Text Query with Suggestions → Filtered Results Displayed **Train Fitness: Exercise search with filter chips** (ios, 2 screens, Health & Wellness, Social Networking) [Refero](https://refero.design/flows/7167) > Mobile (iOS) in-app search for exercises accessed from the Explore tab. Steps: Search — Auto-Detected results → Search — Custom filter results **Box Box Club: Glossary search and filter** (ios, 4 screens, Entertainment, Sports, Transport) [Refero](https://refero.design/flows/7873) > Happy path: A user opens the F1 Glossary list, taps the search field to start a query, types (numbers or letters), sees the list update (or clear) in real time, and reads a matching definition. Steps: Glossary list (default) → Activate search → Search with numeric input → Search with alphabetic input and filtered results **Shazam: Library search and results** (ios, 5 screens, Music Streaming, Utility) [Refero](https://refero.design/flows/9632) > Happy path: a signed-in user browsing their music Library on an iPhone opens a contextual menu from the Library header, chooses "Search in Library", types a query, and views filtered results or a no-results state. Entry: user is on the Library tab showing a list of songs. Exit: the list is filtered to show matching songs or an empty results screen. Assumptions: user is logged in, has a populated local library, device is an iPhone (portrait), and the bottom tab bar is visible. Steps: Library list (default) → Contextual menu → Search field focused → No results state → Filtered results **Spotify: Search and Discover Artist Content** (ios, 11 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/216) > Happy path: A logged-in user opens the app's Search tab from the bottom navigation, focuses the search field, types a query ("denzel"), then browses results using the horizontal filter chips (Top, Songs, Artists, Podcasts & Shows, Albums, Playlists, Profiles). The interface returns real‑time results while the keyboard is active, shows grouped result sections and carousels, and allows the user to switch filter tabs to refine results. The user can select an item to open its detail (artist page or profile). A persistent mini-player remains visible while navigating. **Apple Books: Book Catalog Search** (ios, 7 screens, Books) [Refero](https://refero.design/flows/11944) > A user begins on the Search section, reviews Discover and Trending content, then searches the catalog. They enter a query, see an empty-results state for an unsupported numeric query, revise it to “Invisible,” review suggestions and matching titles, refine the query to “Invisible man,” and filter the results to Audiobooks. The user is assumed to be on an iPhone with access to the bookstore catalog. The flow ends on audiobook results for the refined search. Steps: Search Discovery → Activate Search → Empty Search State → Search Suggestions → Review Title Results → Browse All Matches → Filter Audiobooks **Reeder.: In-feed search and filter** (ios, 3 screens, News) [Refero](https://refero.design/flows/3571) > User is browsing a vertical content feed (Pitchfork). From the feed the user opens an inline search/filter control (top search icon) which presents a rounded filter input at the top and the on-screen keyboard. The user types a query (in the example, a numeric string). Steps: Feed list view → Open search / show keyboard → Type query into filter **Wabi: Hidden Gems Search** (ios, 4 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12050) > This is a mobile discovery screen for finding ‘hidden gems.’ The flow starts on the main Hidden Gems screen, where the user can browse all results or open a custom search panel. In the custom panel, the user enters a search term or taps a suggested category chip, then submits the search. The flow exits back to the main Hidden Gems screen with the selected custom filter applied and the content area updated to show a no-results state for that query. Assumptions: the user is already inside the Hidden Gems section on a mobile device; login state is not shown; the bottom navigation and share/back controls are available throughout the main screen. Steps: Hidden Gems Overview → Custom Search Entry → Custom Search Filled → Filtered Hidden Gems Results **Linear: Quick Find Search & Filter** (ios, 3 screens, Collaboration Tools, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/6628) > Mobile quick-find/search experience on an iPhone-style device. Entry: user opens the global quick search (keyboard may open automatically) and begins typing a short query (example: "Bu"). The UI shows inline search results of mixed item types (issues, projects, documents) with type icons and reference labels. The user can open a compact filter menu to narrow results by type (All / Issues / Projects / Documents). Exit: user either selects a result from the list or clears/close the search. Assumptions: user is authenticated and within a project-management context; this is on a mobile device with a bottom search bar and a modal-style quick search surface. Steps: Search Results (mixed types) → Open Type Filter Menu → Apply Projects Filter and Continue Typing **Dropset: Exercise selection list** (ios, 7 screens, Health & Wellness, Fitness) [Refero](https://refero.design/flows/6550) > A mobile (iPhone) exercise catalog screen where a user browses and searches exercises. Entry: user arrives at the Exercises screen (list view) with a persistent bottom filter drawer visible. Happy path: user reveals the search input, types a term, the list filters in place and displays matching items. Exit: user selects an exercise from the filtered list (selection/navigation not shown). Assumptions: user is on a logged-in session on a phone; the bottom filter control is always present and does not block search; no full-screen navigation away from this screen is shown in the images. Steps: Exercises list (default) → Reveal search field → Focus search (keyboard appears) → Type numeric input → Empty search state → Refine query to text → Filtered results shown **LEGO® Builder: Set Search & Filter** (ios, 8 screens, Entertainment) [Refero](https://refero.design/flows/10130) > Happy-path: User starts on the app's Explore screen, taps the prominent Search field, enters a query (either a numeric set number or a text name), and reviews results. The interface supports direct numeric lookup (returns 'no results' if not found), free-text searches, horizontal filter chips (brand/category), and a scrollable results list with thumbnail, title and set number. Steps: Explore / Search entry → Search input focused → Numeric search attempt → Switch to text search → Results list visible → Filter chips exposed → Select brand filter (Ghostbusters) → Select category filter (IDEAS) **ElevenReader: Genre-based book filtering** (ios, 5 screens, AI Tool, Books) [Refero](https://refero.design/flows/8974) > Happy path: A user browsing the Explore results opens the Genre selector, searches or scrolls to find genres, selects one or more genre tags, and applies them to refine the book results. Steps: Explore results with filter chips → Genre selector – full list → Genre search input → Select multiple genres → Filtered Explore results **Athleats: Recipe browsing and search** (ios, 4 screens, Health & Wellness, Food & Drink) [Refero](https://refero.design/flows/11287) > Mobile "All Recipes" listing where the user can browse a two-column grid of recipe cards or use the top search field to narrow results. Steps: All Recipes overview → Search input focused — numeric entry → Search typed with text query → Search results visible with keyboard dismissed **Reeder.: Article Search and Filter** (ios, 3 screens, News) [Refero](https://refero.design/flows/3572) > Happy-path: A mobile user browsing a publication's article list opens the inline search/filter input, types a keyword, and sees the feed filtered to matching articles. Steps: Feed Browse → Open Search Input → Enter Query and View Results **Pinterest: Search results — filter by content type** (web, 4 screens, Reference, Social Networking, Inspiration Board) [Refero](https://refero.design/flows/958) > Desktop search results for the query "Urban photography." Entry point: user runs a search (search term pre-filled at top). The main results area shows a Masonry grid of Pins with topical chips under the search bar. The user can open a content-type control (compact dropdown or left Filters panel) to choose which content type to surface (All Pins, Products, Videos, Profiles, Boards). When a filter is chosen and confirmed the grid updates: selecting Videos shows video duration badges on pins and selecting Boards replaces the grid with board thumbnails and metadata. Assumptions: user is on a wide/desktop viewport, signed-in (profile avatar present), and started from the search results screen. Steps: Search Results Grid → Open Content-Type Dropdown → Filter Sidebar — Choose Videos → Boards View **Trip: Hotel Search Results Filtering** (web, 3 screens, Travel & Booking) [Refero](https://refero.design/flows/12078) > User enters a hotel search results list for Hong Kong with dates and guest count already set, then narrows the results using left-side filters such as hotel type, breakfast included, free cancellation, star rating, and budget. The results list updates from a broad set of properties to a smaller filtered set. Steps: Browse Hotel Results → Apply Filters → View Refined Results **West Elm: Holiday Search Results** (web, 2 screens, Home & Decor, Marketplace) [Refero](https://refero.design/flows/1683) > User performed a keyword search for "halloween gifts" and landed on a product listing (search results) showing 14 results in a multi-column image-first grid. The page provides top-level controls (search field, sort, filter chips), product cards with image, title, price, badges and a favorite heart. Scrolling reveals a "Recently viewed" carousel and a "People who searched 'halloween gifts' bought" recommendation strip, then the site footer. Steps: Search results grid → Recommendations & footer **Trip: Tour Search Results** (web, 2 screens, Travel & Booking) [Refero](https://refero.design/flows/12095) > Happy-path tour search results screen on Trip.com. The user enters from a search results entry point with a destination query already present and can browse available routes, refine them using left-side filters, and sort results from the top controls. Steps: Browse Results → Filtered Results **PayPal: Search documentation from docs page** (web, 4 screens, Finance) [Refero](https://refero.design/flows/7373) > Happy path: A developer browsing the PayPal Developer documentation (Integrate card payments in iOS apps) opens the global search overlay, types a query, reviews the suggested categories and result list, then selects a relevant documentation entry. Steps: Documentation landing → Search overlay opened → No-results state → Search results list **YouTube: Search refinement and filter application** (web, 6 screens, Entertainment, Video Streaming) [Refero](https://refero.design/flows/11297) > Happy-path: A signed-in desktop user searches for a creator name, opens the search filters modal, experiments with filters (which can produce zero results), adjusts selections, and applies a set of filters to narrow results. Entry: user is on the search results list after submitting a query. Exit: the search results list is updated to show only items that match the selected filters (or a ‘No results’ state if none match). Assumptions: user is logged in (Premium badge visible), using a desktop/laptop with a left navigation rail visible, and has permission to view the listed content. Steps: Search results list (initial) → Open Search filters modal → Empty results state (after filter applied) → Re-open filters modal from empty state → Adjust multiple filter selections → Filtered results list **Missive: Search and refine results** (web, 4 screens, Business, Collaboration Tools, Communication Tools) [Refero](https://refero.design/flows/6274) > Happy path: A signed-in desktop user runs a text search for "account", opens the search filter dropdown, checks the "Starred" filter, and views the updated (narrowed) search results. Entry: user types a query into the global search field and hits Enter (or the search executes). Exit: search results list updated to show only messages that match the query and the selected filters. Assumptions: user is logged in, has access to the Refero Design team workspace, there are messages that contain the query term, and the interface is a multi-column desktop layout (left navigation, center message list, right content/view area). ##### Task: security incident alert **Steel: Create Usage Alert** (web, 4 screens, AI Tool, Development Tools) [Refero](https://refero.design/flows/14301) > A signed-in workspace administrator opens Settings > Limits & alerts, starts creating a usage alert, adjusts the notification threshold from the default 80% to 70%, and saves it. The flow begins on the limits and alerts dashboard with no alerts configured and ends with a saved email alert displayed in the alerts list. The screenshots show a desktop interface and an authorized user with access to workspace settings. No additional recipient or notification-channel selection is shown; the modal indicates that an email notification will be sent. ##### Task: send money transfer **Fresha: Buy and send gift card** (ios, 20 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4957) > User purchases a gift card from a salon listing. Entry is the salon's Buy section (memberships / gift cards). The user selects a preset or custom amount, chooses whether to send the card to someone or add to their own account, selects a card style, fills sender/recipient details and message, picks a send date, then reviews order and attempts payment. Exit points: 1) Complete purchase (Pay now success — not shown), 2) Payment failure returns an error and keeps user on the Review screen. Assumptions: user is on mobile, signed in (sender name is pre-filled), they have a saved payment method, and the UI uses a multi-step modal-style purchase flow with a sticky summary/footer showing the selected gift card amount and a contextual action (Continue / Pay now). Steps: Salon Buy section → Choose preset amount → Custom amount entry (empty) → Custom amount typing → Custom amount confirmed → Select recipient type → Style picker (loading) → Style categories visible → Style selected → Share form (empty) → Typing recipient name → Form completed (message and email) → Open date picker → Date selected → Review transition → Review summary (top) → Review recipient details → Initiate payment → Payment failed → Error state and retry **Wise: Help centre search and results** (web, 4 screens, Finance) [Refero](https://refero.design/flows/361) > Happy path: An unauthenticated user lands on the Wise Help Centre home, uses the central search field to find help content, receives inline feedback (no-results toast or autocomplete suggestions) while typing, and then views a full search-results list where they can open an article. Steps: Help centre home → Search — no results feedback → Inline autocomplete suggestions → Full search results list ##### Task: settings preferences **RAD Weather: Preference Personalization** (ios, 3 screens, Weather) [Refero](https://refero.design/flows/13799) > The user configures display, language, measurement, and forecast-personality preferences during an onboarding sequence. The flow begins on the Preferences screen with high contrast and locked header enabled, English selected, system measurement units active, and the forecast tone set to Blunt. Across the shown states, the user increases the text size from 1x to 2x and then to 2.5x. The interface responds by enlarging text and reflowing content while keeping the preference controls and navigation available. The user can continue to the next onboarding step or go back. Steps: Initial Preferences → Larger Text Preview → Maximum Text Preview **UGLYCASH: Change Language Setting** (ios, 4 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/11228) > Happy path: A logged-in user on a mobile device opens Settings, navigates to the Language option, picks Spanish, and the interface switches from English to Spanish. Steps: Settings (English) → Language Selection (English) → Language Screen (Spanish) → Settings (Spanish) **UGLYCASH: Switch to Dark Theme** (ios, 4 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/11229) > Happy path: A logged-in mobile user navigates to Settings → Theme, chooses the Dark theme, the app shows a brief splash/transition while the new styling is applied, and the user is returned to the Home screen rendered in dark mode. Steps: Settings (Preferences) → Theme Selection → Theme Apply / Splash Transition → Home (Dark Mode) **RAD Weather: Forecast Preferences** (ios, 3 screens, Weather) [Refero](https://refero.design/flows/13801) > A user configures weather forecast preferences on a dedicated setup screen. The screen includes accessibility, language, text size, measurement-unit, and forecast-tone settings. Across the provided states, the user changes the forecast tone selection from Blunt to Casual and then to Professional. The user can use Back or Continue to move through the larger setup sequence; no destination after Continue is shown. Steps: Initial Preference Settings → Casual Tone Selection → Professional Tone Selection **TIDAL: Audio Playback Preferences** (ios, 7 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/13128) > The user opens the Audio & Playback settings panel and adjusts streaming and playback preferences. They enter Mobile data streaming, review the Low, High, and Max quality options, and select different quality levels. They return to the main settings view, scroll to the Playback section, and adjust Crossfade from 0 seconds to 5 seconds. The flow assumes the user is already viewing the Audio & Playback settings on a mobile device. Changes appear to be applied immediately; the interface notes that audio quality changes take effect from the next track played. Steps: Audio & Playback Settings → Review Low Quality → Select High Quality → Select Maximum Quality → Playback Settings → Adjust Crossfade → Crossfade Applied **Sora: Cameo preferences editor** (ios, 4 screens, AI Tool, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/7823) > User is in a mobile settings screen for a feature called Cameo. Steps: Cameo settings (audience & options) → Cameo preferences — empty editor → Cameo preferences — enter text → Cameo preferences saved & preview **Artsy: Dark Mode Settings in Account** (ios, 3 screens, Marketplace) [Refero](https://refero.design/flows/7623) > Happy path: A logged-in user opens their Account (Profile) section, chooses the 'Dark Mode' preference, and selects one of three radio options (Sync with system / On / Off). The selection is applied immediately — choosing 'On' switches the app to dark theme. Steps: Account Overview → Dark Mode Options (Light Theme) → Dark Mode Options (Dark Theme Applied) **Ada: Change language preference** (ios, 5 screens, Health & Wellness) [Refero](https://refero.design/flows/4127) > Happy path: A signed-in user opens Settings, navigates to the Language option, selects a different language from a vertical list, the app shows a short loading state, applies the new locale, and returns to the Settings screen translated in the chosen language. **Mozi: Set preferred messaging** (ios, 4 screens, Social Networking, Travel & Booking) [Refero](https://refero.design/flows/4163) > Happy path: user opens Edit Profile (logged-in, mobile) and navigates to Contact > Preferred messaging to pick a default messenger. A bottom/overlaid selection sheet appears with four messenger options and Cancel/Save controls. The user selects a messenger (radio), Save becomes active, and after saving the selected option is shown on the Edit Profile screen. **222: Set preferred city** (ios, 4 screens, Lifestyle, Dating & Relationships) [Refero](https://refero.design/flows/9317) > Happy path: A signed-in mobile user opens Experience Preferences to manage what invites they receive and where they'd like to attend events. From the preferences screen they tap the 'current 222 city' row, pick a new city from a list, and return to the preferences screen where the chosen city is reflected as the current 222 city. Steps: Experience Preferences (overview) → City selection list (initial) → City selection list (selection confirmed) → Experience Preferences (updated) **RAD Weather: Preference Setup** (ios, 9 screens, Weather) [Refero](https://refero.design/flows/13796) > A user configures display, language, measurement, and forecast-personality preferences during an onboarding setup sequence. The flow starts on the Your Preferences screen with default settings, allows the user to open and change the language, adjust text size, choose preferred units, and select a forecast tone. The user can navigate backward or continue using the persistent bottom controls. The screenshots end with the forecast tone changed to Professional; no subsequent screen or completion confirmation is shown. Steps: Default Preferences → Open Language Menu → Select English → Increase Text Size → Set Largest Text Size → Choose Measurement Units → Review Responsive Preferences → Select Casual Tone → Select Professional Tone **FocusFlight: Measurement Unit Settings** (ios, 4 screens, Productivity) [Refero](https://refero.design/flows/8617) > Happy path: user opens the app Settings, taps the Measurement Units row in the General section, interacts with the Measurement Units bottom sheet (expand, choose a preset or toggle distance unit), and returns to Settings. Steps: Settings List → Measurement Units Sheet (partial) → Measurement Units Sheet (expanded) → Apply Toggle and Preview Update **Phantom: Change Display Language** (ios, 4 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3731) > Happy path for changing the display language from within the Preferences area on a mobile (dark theme) device. Steps: Preferences (English) → Display Language List (top) → Display Language List (scrolled) → Preferences (Kiswahili) **Poppy: Poppy daily preferences setup** (ios, 3 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/12178) > This is a mobile preference setup screen for configuring how Poppy interacts with the user throughout the day. The user can choose an interaction style (Quiet, Balanced, or Active), enable or disable specific touchpoints like Morning and Evening messages, set a morning briefing time, and adjust how often Poppy reaches out for real-time suggestions. Entry appears to be from an onboarding or settings step; exit is through the Back control or the primary Finish up button. Steps: Morning schedule → Time selection → Expanded preferences **Phantom: Icon Selection** (ios, 3 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3732) > Happy path: from the Preferences list the user opens the icon chooser, picks a different icon, and receives a confirmation message. Steps: Preferences → Icon Options → Change Confirmation **Mindllama: Theme selection (Dark Mode)** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/5509) > Happy path: user opens the Settings tab (profile section) and selects the Dark Mode row. That navigates to a dedicated Dark Mode selection screen showing four options (Auto, Dark, Light, Custom Dark). The user taps one of the options (Dark) and the app immediately updates the color scheme to a dark theme and marks the chosen option with a checkmark. The user can then go back to Settings using the back control. Assumptions: mobile device, Settings tab is already open and the user has permission to change appearance settings. **Miles: Preferences: Running & Appearance Settings** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/6711) > Single-screen preferences list (vertical scroll) that lets the user view and change running/tracking settings. Steps: Preferences overview → Toggle feature settings → Select appearance options **Wispr Flow: Default Dictation Style** (ios, 4 screens, AI Tool, Utility, Productivity) [Refero](https://refero.design/flows/13675) > The user starts on the Style screen with “very casual” set as the default sound for dictations. They open the default sound selector, review the available styles, select “Casual,” and save the change. The Style screen then updates to show Casual as the active default. The flow assumes the user is already in the Style section and has permission to change this preference. Steps: Current Style → Review Style Options → Select Casual → Saved Default **ABY Journal: Settings to Timeline Scrub** (ios, 3 screens, Health & Wellness, AI Tool, Lifestyle) [Refero](https://refero.design/flows/8002) > This flow shows the Settings area for a journaling/insights app and the return to the timeline where the user can begin interacting with entries. Steps: Settings — Preferences & Account → Settings — Legal, Danger Zone, Feedback → Timeline — Scrub Interaction **B150: Theme preference** (web, 2 screens, Social Networking) [Refero](https://refero.design/flows/7986) > Happy path: a logged-in desktop user opens Account settings → Preferences and chooses a visual theme using a three-way selector (Light / Dark / System). Steps: Preferences — Light selected → Preferences — Dark applied **Missive: Appearance & Sidebar Settings** (web, 3 screens, Business, Collaboration Tools, Communication Tools) [Refero](https://refero.design/flows/6286) > Happy path: A signed-in user opens the Settings panel, navigates to Preferences → Appearance, adjusts theme and sidebar options (for example enabling a dark left sidebar), then returns to the inbox where the changed appearance is visible. **Wealthsimple: Language preference change** (web, 4 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8520) > User changes their interface language from the Settings screen. Entry: user is on the Settings landing (general preferences visible). The user opens the Language setting modal, selects a new language from a dropdown, saves the selection, and the Settings UI updates to the chosen language. Assumptions: user is signed in, using a desktop (wide) viewport, and has permission to modify their preferences. The modal is used as an editing surface and is dismissed after a successful save. Steps: Settings landing (English) → Language setting modal (initial) → Select new language and enable Save → Settings landing (French) — post-save **Instacart: Smart Shop preferences editor** (web, 2 screens, Food & Drink, Marketplace, Food Delivery) [Refero](https://refero.design/flows/10100) > Happy path: A logged-in user accesses Account settings → Smart Shop preferences on desktop to specify who they shop for (kids, pets) and their dietary preferences using selectable chips. As the user selects or deselects chips, the interface updates chip states. After making changes the UI displays a transient confirmation that preferences have been saved. Entry: Account settings (Smart Shop preferences). Exit: Preferences are visually updated and a saved confirmation is shown. Steps: Preferences list (initial) → Preferences saved confirmation **Peerlist: Job Preferences Setup** (web, 3 screens, Job Search, Business, Social Networking) [Refero](https://refero.design/flows/1089) > Happy path: A logged-in user opens their account/profile settings and navigates to Job Preferences to declare openness to work and configure detailed job-seeking criteria. **Digg: Newsletter Edition Selection** (web, 2 screens, News) [Refero](https://refero.design/flows/14352) > The user opens the Account modal and views Email settings for a confirmed email address. They change the newsletter edition from Daily to Saturday morning edition using the edition selector. The interface updates the selected value, explanatory text, and displays a confirmation message. The flow begins within the logged-in Digg experience and ends after the email delivery preference is successfully updated. Steps: Email Settings → Edition Updated **Mercury: Preferred Name Edit** (web, 4 screens, Business, Finance) [Refero](https://refero.design/flows/11079) > Happy path for updating a user's preferred name from the Personal → My Profile screen on desktop. Entry: signed-in user navigates to Settings > My Profile and clicks the 'Edit' link next to Preferred name. The system opens a centered modal with first and last name input fields. The user edits the name and clicks Update. The modal closes, the profile list updates to show the new preferred name, and a transient success toast appears. Steps: Profile overview (My Profile) → Edit preferred name modal (initial) → Edit preferred name modal (edited) → Profile overview (updated) with success toast **Doctronic AI: Timezone Selection** (web, 3 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13704) > The user is updating the timezone preference in Connie’s settings. They open the timezone selector, search for a location, briefly enter a term that produces no results, then refine the search to find and select Los Angeles. The flow begins on the settings screen with the timezone control available and ends with Los Angeles selected as the timezone. **Polar: Account Preferences — Delete Confirmation** (web, 3 screens, Business, Finance) [Refero](https://refero.design/flows/11410) > Logged-in user opens Account Settings > Preferences to view and edit personal information, general settings, connections, and account-level actions. From this page the user can open a Delete Account confirmation modal (Danger Zone). The modal presents consequences and two actions: Cancel or Delete. After confirming deletion (or otherwise navigating away), the visual flow ends on a public marketing/home landing. Assumptions: the user is authenticated when they start on Preferences; clicking Delete triggers a confirmation modal; the final screenshot represents a public landing or post-action redirect (the screenshot does not explicitly show a success message or confirmation that deletion completed). Steps: Preferences (Account Settings) → Delete Account Confirmation Modal → Public Marketing / Landing (Post-action or Redirect) **Handshake: Career Interests Setup** (web, 3 screens, Social Networking, Job Search) [Refero](https://refero.design/flows/8773) > User opens the "Your Career Interests" settings (assumes signed-in, desktop/larger viewport). The page is a long form composed of grouped inputs: position types, post-graduation options, how the product can help, a typeahead for job titles, a city typeahead, multi-column industry checkboxes, job-function checkboxes, and an expected graduation date selector. The happy path: user reviews and selects relevant checkboxes and tags, adds preferred cities or job titles via the search fields, sets graduation month/year, and the interface persists the changes. The flow ends with a confirmation message indicating the preferences were updated. Steps: Career Interests Form (initial view) → Make selections and add tags → Save confirmation **Missive: Configure Swipe Actions** (web, 3 screens, Business, Collaboration Tools, Communication Tools) [Refero](https://refero.design/flows/6291) > Happy path: A signed-in user opens Preferences → Swipes to review and customize the app's swipe gestures. They toggle global swipe options, inspect current left-to-right and right-to-left actions, add an action from a menu, and adjust each action's active area using draggable handles. **Dribbble: Set work availability and profile details** (web, 13 screens, Communication Tools, Job Search) [Refero](https://refero.design/flows/801) > Happy path: A logged-in user opens their account settings, navigates to Work Preferences via the profile menu, and completes a three-section modal to indicate availability (Full-time or Freelance), select up to three specialties (with years of experience), and optionally add work history, education, and skills. The user sets numeric rates/constraints for freelance work (if chosen), adds specialty years, fills experience fields, toggles ‘Available for work,’ and saves. Steps: Open settings and Work preferences → Modal overview (three sections) → Select Full-time availability → Select Freelance/Contract availability → Enter freelance rate values → Confirm freelance inputs → Open specialties selector → Select specialties (three items) → Choose years for a specialty → Confirm specialties with years → Open optional history/education/skills → Fill work history, education, and skills → Save preferences and confirm **Teal: Target Job Preferences** (web, 8 screens, Productivity, Job Search) [Refero](https://refero.design/flows/2732) > Happy-path walkthrough for configuring job recommendations: the user selects one or more target job titles, specifies job details (employment type, location preference, experience range, lowest acceptable salary), optionally adds keywords to include or exclude, and finishes to see a filtered list of job matches. Entry: user lands on the “Target Job Title” step in a 3-step guided setup (top stepper indicates Target Title → Job Details → Keywords). Exit: the search results screen with active filters and job listings. Assumptions: user is already authenticated and has access to the job recommendation feature; the UI presents a linear multi-step wizard with Next/Back navigation and a final Finish that applies preferences. Steps: Target Job Title - start → Target Job Title - select suggested → Target Job Title - add new → Target Job Title - multiple selected → Job Details - initial → Job Details - filled → Keywords - include/exclude → Filtered Job Results **OpenAI: Cookie preferences management** (web, 3 screens, AI Tool, Business, Development Tools) [Refero](https://refero.design/flows/4189) > Happy path: a logged-in user on a desktop visits the developer docs landing and opens the user account menu, then selects "Manage cookies" to open a cookie consent modal. In the modal the user inspects the toggles (Essential and Analytics), changes the Analytics toggle off, and confirms their choice by clicking one of the action buttons (Decline all or Accept all). **Navia: Account Management and Sign-Out** (web, 3 screens, Business, Productivity) [Refero](https://refero.design/flows/14326) > The flow begins with an authenticated user viewing the Navia Account area, where plan details, usage, billing, profile information, preferences, and security controls are available. The user reaches the Security section, which includes password management, signed-in device controls, and account deletion. After signing out, the user is taken to a public welcome screen offering Sign In and Sign Up. The flow assumes the user is initially logged in with permission to manage their own account and is using a desktop browser. Steps: Account Overview → Security Settings → Signed-Out Welcome **Nextdoor: Change Appearance Theme** (web, 5 screens, Communication Tools) [Refero](https://refero.design/flows/11422) > A logged-in user on a desktop profile dashboard opens the account menu, navigates to Appearance settings, and switches the interface from Light mode to Dark mode. The appearance submenu initially shows Light mode selected. After selecting Dark mode, the interface updates immediately and remains in the dark theme when the menu is dismissed. Steps: Profile Dashboard → Account Menu → Appearance Options → Dark Theme Applied → Updated Dashboard **Linear: Customize theme colors** (web, 5 screens, Business, Productivity, Project Management, Collaboration Tools) [Refero](https://refero.design/flows/6689) > Happy path: From Preferences the user opens the Interface and theme section to change the app's visual appearance. The user opens the Theme preset selector, switches to the Custom theme option, edits the accent and background color values and adjusts contrast. The UI updates immediately to preview the chosen colors. Steps: Open Preferences and scroll to Theme → Open theme preset selector → Switch to Custom theme → Edit accent color → Adjust background and preview theme **TikTok: Direct Message Settings** (web, 3 screens, Video Streaming, Social Networking, Entertainment) [Refero](https://refero.design/flows/1103) > Happy-path: A logged-in desktop user opens the Messages area, launches the Message settings modal, changes the permission for who can send them direct messages from "Friends" to "No one," and prepares to save the change. **Calendly: Update Conflict Calendars** (web, 5 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/14026) > A logged-in user updates which connected calendars are checked for scheduling conflicts. The flow starts in Scheduling settings on the Calendars tab, where one calendar is currently selected. The user opens the conflict-calendar selector, enables additional calendars, submits the change, and returns to the settings view after the update is saved. The flow assumes the user has permission to manage scheduling settings and already has connected calendars. It is shown on a desktop-sized interface. **Understory: Storefront Branding Settings** (web, 4 screens, Business, Scheduling Tools) [Refero](https://refero.design/flows/12713) > This is a Storefront Preferences area for configuring branding before the storefront is published or used by customers. The user enters from the left navigation under Storefront > Preferences and can update company details, branding assets, styling, and show/hide the platform logo. The happy path assumes the user is logged in with access to storefront settings. The user can edit text fields directly on the preferences screen, open a modal to edit the company description, save changes, and see the updated branding reflected in the main preferences view and styling preview. Steps: Branding Preferences → Description Editor → Description Content Update → Updated Branding Preview **ElevenMusic: Account Settings Edit** (web, 6 screens, Music Streaming, Entertainment, AI Tool) [Refero](https://refero.design/flows/12594) > Logged-in user opens the Account settings modal and edits profile information such as profile photo, display name, bio, country, and marketing email preference. The modal uses a left sidebar for section navigation and a main form area for editing. The user can cancel or save changes; on save, the interface shows a saving state and then a saved confirmation toast. Steps: Open Account Settings → Edit Profile Details → Adjust Country and Email Preference → Search Country विकल्प → Save Changes → Saved Confirmation **Stocktwits: Profile Edit — Personal Info & Preferences** (web, 7 screens, Finance) [Refero](https://refero.design/flows/7208) > Happy-path: A logged-in user edits their StockTwits profile by updating their avatar, entering a bio, updating location and website, selecting trading preferences from several grouped lists, and saving changes. Steps: Profile overview → Upload or change photo → Open bio editor → Enter bio content → Edit location and website → Select preferences and checklists → Save changes and confirm ##### Task: share link permissions **Podimo: Show Sharing to Messages** (ios, 3 screens, Entertainment) [Refero](https://refero.design/flows/12259) > This happy path shows a user viewing a podcast/show detail screen, opening the share sheet, and choosing a messaging option to send the show link. Steps: Show Detail → Share Sheet → Message Composer **ElevenReader: Share Bookcard** (ios, 2 screens, AI Tool, Books) [Refero](https://refero.design/flows/8947) > Happy path: A user opens a "share" overlay for a specific book (cover, title, author and source badge visible). From this overlay the user chooses a sharing action (Copy link, WhatsApp, Messages, More) or dismisses the overlay with the top-right close button. If the user taps "Copy link" a transient confirmation toast appears near the top saying "Copied link to clipboard." Entry point: user opened a book's share sheet (presumed from a prior share action). Exit points: user selects a share target (external composer or system share sheet opens), copies the link and closes the overlay, or dismisses the overlay with the X. Assumptions: user is on a mobile device (portrait), has permission to copy to clipboard, and the app can launch external share targets (WhatsApp, Messages) or show the OS share sheet. Steps: Share overlay (book details) → Copy link confirmation **ElevenReader: Share referral link flow** (ios, 3 screens, AI Tool, Books) [Refero](https://refero.design/flows/8985) > Happy-path: A logged-in iPhone user opens a referral reward modal that explains the incentive (5 listening hours) and taps the primary CTA to share their referral link. The app invokes the iOS share sheet where the user selects Messages (or another channel) and is taken to a prepared message compose view containing a link preview. The exit point is after the user sends the message (or cancels) and the share UI dismisses. Assumptions: user is signed in, on an iOS device that supports the system share sheet and Messages, and has necessary OS permissions to present the share sheet and launch Messages. Steps: Referral reward modal → iOS share sheet → Message compose with link preview **Mozi: Invite via QR and Share Link** (ios, 4 screens, Social Networking, Travel & Booking) [Refero](https://refero.design/flows/4145) > Happy-path: A signed-in user on an iPhone starts at the app home screen, taps the invite CTA to open a “Connect” screen that displays their profile card plus a personalized QR code and a button to share their invite link. Tapping the share button opens the iOS share sheet; choosing Messages inserts a preview card and link into a new SMS composer so the user can address and send the invite. **Google Gemini: Create public share link** (ios, 4 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/3198) > Happy path: From a conversation view the user opens the message actions menu, selects "Create public link," customizes the share preview headline (optional), and confirms creation. The system shows a preview of the conversation snippet, allows choosing or editing a headline, surfaces a policy/visibility note, and then generates a public URL that the user can copy or share. **YouTube Music: Comment Reply Sharing** (ios, 4 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11334) > User is viewing the Replies panel for a track and interacts with a specific reply. Steps: Replies list with confirmation → Open comment overflow menu → Choose Share — open native share sheet → Copy link confirmation **Daze: Profile sharing confirmation** (ios, 2 screens, Social Networking) [Refero](https://refero.design/flows/9152) > Happy-path: a signed-in user views their personal profile screen (header with avatar, username, handle, birthday and social connect buttons) and uses the share control to copy a link to their profile. The app copies the profile URL to the clipboard and displays a centered confirmation modal. Steps: Profile overview → Copy confirmation modal **Canva: Share design with team** (web, 7 screens, Design & Media Editing, Collaboration Tools) [Refero](https://refero.design/flows/8073) > Happy-path: A signed-in user on the Projects listing selects a recent design, opens the item menu, views the item's details panel, opens the Share modal, adjusts link access from private to a team scope, copies the share link, and sees the details panel updated with the new collaborator indicator. Steps: Projects listing → Open item context menu → Open details panel → Open Share modal → Select access level options → Set team access and permission → Verify details panel update **Bezi: Share modal controls** (web, 5 screens, Design & Media Editing) [Refero](https://refero.design/flows/199) > Happy path: A signed-in user opens the Share modal from the top-right control to invite a collaborator and configure link access. The user optionally changes the link access level (Edit/View/Only invited), types an email to invite, clicks Invite, then updates the new collaborator's role or removes them. Steps: Open Share modal → Open link access dropdown → Enter collaborator email → Confirm invited collaborator → Change role or remove collaborator **Canva: Share link and set permission** (web, 5 screens, Design & Media Editing, Collaboration Tools) [Refero](https://refero.design/flows/8055) > Happy path: A logged-in editor opens the design's Share panel, changes the access level from restricted to link-based access, sets the link permission to "Can view", and copies the shareable link to the clipboard. Steps: Open Share panel → View access level options → Select link-based access and permission → Copy shareable link → Confirm link copied **Melius: Share Canvas Link** (web, 5 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/14081) > A logged-in canvas owner or editor opens the sharing dialog, reviews the current restricted access setting, and changes general access to “Anyone with the link” with view-only permissions. The user then copies the generated sharing link and receives confirmation that it was copied. The flow begins on an existing canvas and ends with a shareable link available on the clipboard. The desktop user is assumed to have permission to manage canvas access. Steps: Open Canvas Sharing → Review Access Settings → Open Permission Options → Enable Link Viewing → Copy Share Link **ElevenLabs: Share project link and permissions** (web, 3 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/8696) > Happy path: From the editor workspace (transcript + timeline), the user opens the Share dialog, chooses a public-access permission level (No access / Viewer / Commenter), copies the link if needed, and closes the dialog. Steps: Editor workspace (open share control) → Share Project modal (default role shown) → Select permission role for public link **Hello Ivy: Share document link** (web, 4 screens, Collaboration Tools, Productivity, Project Management) [Refero](https://refero.design/flows/4515) > Happy-path: A signed-in workspace member opens a document in the Pages area, invokes the share control, publishes the document to the web (if not already published), then copies the generated public URL. Steps: Open document in Pages → Open sharing modal → Publish to generate link → Copy link and confirm **Clipchamp: Project sharing & link invite** (web, 10 screens, Design & Media Editing) [Refero](https://refero.design/flows/10880) > Happy-path: A signed-in user in the Clipchamp editor opens the Share menu to invite people or copy a shareable link for the current project. They open the share modal, optionally adjust link settings (Anyone vs Specific people and edit permission), enter one or more recipients using an autosuggest/email input, add an optional message, then send the invite or copy the link. The system shows loading and sending states, surfaces an error if the backend fails, and finally confirms when the link/email have been sent. Entry: user is in the project editor with the project already loaded. Exit: confirmation modal indicating the link was sent and recipients listed. Assumptions: user is authenticated, project is saved, network access available. Steps: Editor / Project preview → Share menu → Loading share modal → Send link modal (initial) → Link settings → Recipient autosuggest → Compose invite → Share error → Sending state (retry) → Share success confirmation **Spline: Invite collaborator** (web, 4 screens, Design & Media Editing) [Refero](https://refero.design/flows/257) > Happy path: A signed-in user opens the project's Share dialog to invite a collaborator by email. The dialog provides a shareable link (with visibility control), an email entry field with a role selector, a list of current members, and an editor-seat usage / upgrade CTA. The user types an email, selects a role (Viewer), clicks Invite, the system shows a sending state and a success toast, adds the invited address to the member list, and clears the input. **Dropbox: Team Folder Viewing Link Settings** (web, 2 screens, Collaboration Tools, Business, Productivity) [Refero](https://refero.design/flows/13033) > A user is managing sharing settings for a team folder in a centered modal over the files view. The modal is opened from the folder context and is already on the "Link for viewing" tab. The user can copy or delete the view link, control who can access it, and toggle expiration, password protection, and download permission. The happy path shown is copying the viewing link, which updates the modal state and triggers a bottom confirmation toast. Assumptions: the user is signed in, has permission to edit sharing settings for the folder, and is on desktop. **Synthesia.io: Copy Project Share Link** (web, 2 screens, AI Tool) [Refero](https://refero.design/flows/13745) > A logged-in workspace owner is editing a design project in a browser-based editor. From the editor, the user opens the Collaborate panel, reviews general access settings, and copies the project link. The system confirms that the link has been copied while the user remains in the editor. The flow assumes the project is already open and that the user has permission to manage collaboration settings. Steps: Collaboration Settings → Link Copied Confirmation **Airtable: Invite link creation** (web, 3 screens, Project Management, Collaboration Tools, Scheduling Tools, Paperwork, Productivity) [Refero](https://refero.design/flows/1536) > Happy-path: A signed-in user opens the base sharing dialog from a table view and uses the Create link tab to generate a shareable invite link. The user sets the permission level and optionally restricts access by email domain, clicks Create link, and receives a generated URL they can copy. Steps: Open Share modal — Create link tab → Configure link permission and access → Invite link generated and available **Maze: Share response reel** (web, 5 screens, Business, Communication Tools) [Refero](https://refero.design/flows/3669) > Happy path: A signed-in user reviews collected responses in a Results view, opens a compiled video 'Reel' of responses, and uses the Share Reel control to change access and copy a shareable link. Steps: Results — Responses list → Reel preview → Share Reel modal — choose access → Permission update feedback → Copy link confirmation **Manus: Meeting Summary — Share Document** (web, 2 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/10452) > Happy path: a signed-in user views a meeting summary document in a centered modal/overlay and opens the share controls to generate and copy a link. Entry: user has a meeting note open in read mode. Exit: sharing popover is dismissed and the link is copied (or sharing scope set). Assumptions: user is authenticated, has permission to view the document, using a desktop/laptop (pointer input), and the UI shows an overlay document viewer with a share icon that opens a context popover. Steps: View Meeting Summary → Open Share Controls and Copy Link **Figma: Slide Sharing & Permissions** (web, 2 screens, Design & Media Editing, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/5006) > Happy path: User opens the presentation editor and invokes the Share dialog to give others access or copy a link. The modal surfaces an input to add collaborators, a list of access-level options, and quick actions (Copy presentation link, Get embed code). An in-product guided tooltip (2-step) appears to explain permission controls. Steps: Open Share dialog → Permissions guidance tooltip **Cohere: Share conversation link** (web, 3 screens, AI Tool, Business, Development Tools, Productivity) [Refero](https://refero.design/flows/3828) > Happy path: A signed-in user opens a conversation and invokes the "Share link to conversation" modal to obtain a URL they can share. The modal shows the current share link and visibility/permissions text, allows the user to request an updated (full) link, shows an in-progress state while generating, and then displays the newly generated link ready to copy or preview. Steps: Open share dialog (current link shown) → Request updated link (generating) → Updated link ready (copy or preview) **Rox: Share Conversation Link** (web, 2 screens, Business, Productivity, AI Tool) [Refero](https://refero.design/flows/10776) > Happy path: a user who has an open conversation invokes the share action, receives a generated shareable URL in a centered modal, and copies the URL to their clipboard. Steps: Open Share Conversation modal → Copy link and show confirmation **Faire: Share Product with Teammates** (web, 4 screens, Business, Marketplace) [Refero](https://refero.design/flows/13078) > Happy path for sharing a product from a product detail view. The user starts on a product detail page, opens a share modal, enters one or more teammate email addresses, optionally chooses an access level, and sends the invite. The flow ends with a confirmation that the invite was sent and the share modal remaining open or available to close. Steps: Product Detail → Invite Teammate → Send Invite → Invite Confirmation **Melius: Copy Asset Link** (web, 3 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/14071) > The user starts in a visual canvas with a selected generated image asset and opens its sharing control. After locating the share asset action, the user copies the asset link and receives confirmation that the link was copied. The flow begins with the user already inside the canvas and viewing the selected asset, and ends with a shareable link stored on the clipboard. The user appears to be logged in with access to the canvas and asset; no permission changes or recipient selection are shown. Steps: Selected Asset Canvas → Share Asset Control → Link Copied Confirmation **Dropbox: Create Folder View Link** (web, 3 screens, Collaboration Tools, Business, Productivity) [Refero](https://refero.design/flows/13032) > Happy-path flow for creating a shareable viewing link for a folder. It starts in a folder’s settings modal on the "Link for viewing" tab, where no view link exists yet. The user creates the link, waits for it to finish generating, then reviews the newly available link settings and access controls. The flow appears to assume the user is already signed in and has permission to manage sharing for the folder. It ends with a created view link that can be copied or deleted, with optional restrictions such as expiration, password protection, and download permission. **Faire: Product Sharing** (web, 3 screens, Business, Marketplace) [Refero](https://refero.design/flows/13077) > A user is viewing a product detail screen and opens a share dialog to send the product to teammates or copy a limited preview link. The flow starts from the product page and ends when the share modal is closed after inviting recipients or copying the link. Steps: Product Detail → Invite Teammates → Copy Preview Link **PamPam: Share map link** (web, 4 screens, Map & Navigation, AI Tool) [Refero](https://refero.design/flows/4740) > Happy-path: A signed-in desktop user viewing a saved map/list (Liverpool) opens the Share modal, reviews and updates three share-related settings, then copies a shareable link. **Ozone: Invite & Share Project** (web, 2 screens, Design & Media Editing) [Refero](https://refero.design/flows/308) > Happy path: A signed-in user with an open project (and invite permission) works inside the video editor timeline and opens the project sharing/invite modal to add people or copy a shareable link. **Kit: Template sharing and link copy** (web, 4 screens, Business, Communication Tools) [Refero](https://refero.design/flows/2217) > Happy path: A user in a content/template editor opens the share dialog for a starting point, changes the access from private to link-accessible, copies the generated URL, and receives a confirmation. **fal: Generation Sharing Settings** (web, 11 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/13532) > A logged-in user opens the sharing controls for an existing image generation, copies its share link, changes access from anyone with the link to specific people, sets the link to expire in seven days, adds two email recipients, and saves the updated permissions. The flow starts on the generation results view and ends with a confirmation that the share settings were updated. Steps: Open Generation Sharing → Review Share Link → Copy Share Link → Open Access Options → Restrict Access → Open Expiration Options → Set Seven-Day Expiration → Enter First Recipient → Add First Recipient → Add Second Recipient → Save Share Settings **Slite: Product roadmap sharing** (web, 3 screens, Paperwork, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/1700) > Happy path: a signed-in user viewing a product roadmap document opens the Share control, inspects collaborator permissions, enables 'Share to web', configures link-level options, and copies the public link. Steps: Roadmap document view → Share overlay — collaborators list → Share overlay — public link settings ##### Task: shipment tracking logistics **PayPal: Add tracking information** (web, 6 screens, Finance) [Refero](https://refero.design/flows/7381) > Happy path: from a transaction detail in Activity the seller clicks the Add tracking/packaging info link, lands on an "Add tracking information" form, selects an order status, enters a tracking number and selects a carrier, then submits. On success the UI shows a confirmation message stating an email will be sent to the customer, with a CTA to return to Activity. Steps: Transaction detail → Add tracking — initial form → Choose order status → Validation feedback for tracking input → Provide tracking number and carrier → Submission success / confirmation **Faire: Ship Date Selection** (web, 7 screens, Business, Marketplace) [Refero](https://refero.design/flows/13093) > Happy-path cart experience on Faire for a logged-in buyer. The user is in the cart, opens the ship-date selector for the order, chooses either immediate shipping or a future ship date, saves the choice, and returns to the cart with the updated ship timing reflected in the order header and delivery estimate. Steps: Cart Overview → Ship Date Modal → Future Date Entry → Calendar Selection → Date Confirmed → Save Ship Date → Updated Cart **Acne Studios: Product delivery check** (web, 5 screens, Marketplace) [Refero](https://refero.design/flows/4371) > Happy-path: A shopper views a product-detail page for a dark‑green multipocket bag, opens the Delivery & Returns panel to check shipping/return rules and costs, enters a postal code, reviews the shipping methods and estimated delivery dates/prices, and then closes the panel to continue shopping or proceed to purchase. Steps: Product detail view → Open Delivery & Returns panel → Enter postal code → View shipping options and costs → Close panel and continue **Zara Home: Shipping Support Conversation** (web, 6 screens, Marketplace, Home & Decor) [Refero](https://refero.design/flows/14233) > The user enters the Zara Home Help assistant, asks about available shipping options and costs, reviews the assistant’s response, then asks a follow-up question about delivery to a hill or hard-to-reach location. The assistant returns a second response with guidance to contact customer service for delivery review. The flow starts on the Help screen with no conversation history and ends with both questions and answers displayed. The screenshots show a desktop browser layout; login status and permissions are not required or visibly established. Steps: Help Assistant Entry → Shipping Question Submitted → Shipping Answer → Follow-up Question Entry → Follow-up Submitted → Delivery Accessibility Answer **Shop: Mark Order Delivered** (web, 4 screens, Marketplace) [Refero](https://refero.design/flows/13284) > An authenticated customer views an existing order in the Orders section of their account and manually marks it as delivered. The flow starts on the order details summary for order #1327, placed with the status "Order placed." The customer opens the order actions menu, selects "Mark order as delivered," and receives confirmation through a toast notification. The order then updates to show "Delivered" and "Delivered today.". Steps: Order Details → Order Actions → Delivery Confirmation → Delivered Order **West Elm: Checkout — Shipping & Pickup Selection** (web, 9 screens, Home & Decor, Marketplace) [Refero](https://refero.design/flows/1689) > Happy path: A user with items in cart moves from browsing/cart review to checkout, chooses a shipping method (either ship-to-address or ship-to-location/UPS Access Point), provides or selects a shipping address or pickup location via a map modal, reviews per-item delivery estimates and optional services (gift wrap, white-glove, protection plan), and proceeds toward payment. Steps: Cart overview (mini-cart & favorites) → Shopping Cart (full) → Checkout — Shipping Step (method selector) → Pickup Location Finder (map modal) → Pickup Location Selection (map pin detail) → Shipping Address Form (empty) → Shipping Address Form (filled) → Delivery & Gift Options (per-item delivery) → Leave Checkout Confirmation **Zara Home: Gift-Wrapped Item Checkout** (web, 20 screens, Marketplace, Home & Decor) [Refero](https://refero.design/flows/14227) > Happy path begins with a single gift-wrapped item in the basket. The user proceeds to checkout, selects drop-point delivery, searches for nearby delivery locations by postcode, enables distance information, and chooses a delivery point. The user then enters billing details, selects credit/debit card payment, corrects invalid card information, and saves the card. After reviewing the order summary, the user authorizes payment and reaches a successful order confirmation. The flow assumes a desktop browser, a customer with an item already in the basket, and delivery availability for the entered U.S. address. The user does not appear to be required to sign in. Steps: Basket Review → Delivery Method Selection → Delivery Point Locator → Postcode Search → Search Results → Distance Display → Pickup Point Selection → Billing Address Form → State Selection → Address Autocomplete → Billing Address Completion → Telephone Entry → Delivery Checkout Review → Payment Method Selection → Card Details Entry → Card Validation Correction → Saved Payment Method → Order Summary Review → Order Confirmation → Unused **Microsoft: Search from Order Tracking** (web, 4 screens, Business, Marketplace) [Refero](https://refero.design/flows/9797) > Happy-path: A signed or anonymous user starts on the Microsoft Order Tracking landing screen, uses the global header search to look up a query (entered text: "he"), views inline search suggestions, submits the search, lands on the site search ‘Explore’ results, then switches to the ‘Shop’ tab to view product/category results. Entry: user arrives at Order Tracking page (from Microsoft Store or support). Exit: user either continues through Explore results, switches to Shop product listings, or clicks a product/related link to navigate off the search results. Assumptions: user has access to the site header search (not blocked), device is desktop (wide layout), user may be logged in or not (both CTA options for account/guest on Order Tracking exist). Steps: Order Tracking landing → Inline search suggestions → Explore results listing → Shop results with product carousels **Asos: Contact to Virtual Assistant — Order Query** (web, 5 screens, Marketplace) [Refero](https://refero.design/flows/1598) > Happy-path: A customer lands on the Contact / "Get in Touch" area and either clicks the visible "Visit My Account" CTA or starts the on-page Virtual Assistant. The chat widget expands and the bot opens with a greeting and a single-line message field. The user interacts with the bot (chit-chat then a specific question about an order). The assistant responds with an account-linked answer and provides two quick-reply options for follow-up. When the user chooses a follow-up option (a short confirm/yes style reply), the assistant apologizes and again prompts the user to log in to their ASOS account to continue. Steps: Contact landing / initial CTAs → Virtual Assistant open / greeting → User sends free-text message / assistant asks for clarity → User asks order status / assistant suggests My Account → User selects a follow-up option / assistant prompts login **ElevenMusic: Track Studio Search and Publish** (web, 3 screens, Music Streaming, Entertainment, AI Tool) [Refero](https://refero.design/flows/12558) > This is a logged-in music creation studio where the user can search, filter, browse, and publish generated tracks. The entry point is the Studio section of the left sidebar. The user can type into the search field to narrow the track list, and the results update in place. If no matches exist, the main area shows an empty state prompting the user to create or upload tracks. The flow ends when the user views the filtered results list or an empty result state; no modal or checkout step is shown. Steps: Browse Tracks → No Match State → Filtered Results ##### Task: sign up create account **Asos: Guest to Account Sign-up from Wishlist** (ios, 13 screens, Marketplace) [Refero](https://refero.design/flows/2937) > Happy-path: anonymous user browsing the 'Topman / Denim' product listing attempts to favourite an item, is prompted to sign in, then chooses to create an account. The flow opens a my.asos.com sign-in/sign-up webview. The user switches to the Join tab, uses the email sign-up form (email, name, password), interacts with optional DOB and interest controls and contact-preferences checkboxes, and submits. The system validates fields inline and returns a server-side error when the email is already registered. Steps: Product Listing – Denim → OS Sign-in Confirmation → Sign In Screen (Webview) → Sign Up – Social Options → Sign Up Form – Start → Sign Up Form – Fix Required Fields → Date of Birth – Native Selector → DOB & Interest Selection → Contact Preferences → Submit Sign Up (Attempt) → Inline Error State – Email Exists → Form Correction – Edit Fields → Return to Product Listing **Tripsy: Sign-up to My Trips** (ios, 4 screens, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/2327) > Happy path: an unauthenticated iOS user is introduced to Tripsy Cloud and can either use Apple Sign In or sign up with email. Tapping "Signup with email" opens a three-field Create an Account form (Name, Email, Password). After entering details and submitting, the app presents the authenticated "My Trips" screen in its empty-state, prompting the user to create their first trip. **Raycast: New user signup and onboarding** (ios, 11 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/10149) > Happy path: a new user opens the app, sees a branded welcome screen, chooses to create an account, fills (or uses social) credentials, taps Create account, sees a creating/loading state, then proceeds through a short feature-tour (carousel) and lands on the main app home. Steps: Welcome / Entry → Create account (empty) → Form filled → Creating (in-button) → Creating (overlay spinner) → Onboarding card 1 — Raycast AI → Onboarding card 2 — Models → Onboarding card 3 — Notes → Onboarding card 4 — Keyboard → Onboarding summary / Finish → Authenticated Home **Roots: New user onboarding and subscription signup** (ios, 68 screens, Productivity) [Refero](https://refero.design/flows/5739) > This is a new-user onboarding → account creation → optional upgrade → first-run setup flow (happy path). Entry: cold launch of the app (splash / welcome). The user moves through a 3-card onboarding carousel, chooses to create an account, completes sign-up (email or Apple), optionally views the paid “Roots Plus” upgrade and purchases via the App Store, completes a short onboarding questionnaire, and lands on the app Home screen with initial tips and permission guidance. Assumptions: user is on an iOS device, not signed in, able to use App Store for purchases, and has the device-level Screen Time permissions available. Exit: user reaches the authenticated Home screen with account created; if they chose upgrade, the App Store subscription is active and the app confirms the purchase. Steps: Welcome splash → Onboarding card 1 → Onboarding card 2 → Onboarding card 3 → Create account (landing) → About your Roots Account (overlay) → Create account (checkbox unchecked) → Sign up with email (form open) → Sign up form (partial input) → Sign up form (complete entry) → Roots Plus upsell (comparison modal) → Feature details → Upsell landing with CTA → App Store subscription sheet → Purchase success (system alert) → Post-purchase onboarding (account setup progress) → Questionnaire selection and confirmation → Account ready confirmation → First-run tips / Screen Time guidance → Home (initial state) **Chance AI: Signup and First-Run Onboarding** (ios, 17 screens, AI Tool) [Refero](https://refero.design/flows/7352) > Happy-path signup and first-use onboarding for Chance (iOS). Entry: user opens the app from the splash screen. The user either chooses a social provider or taps "Continue with Email" to create an account. The screens guide the user through entering a display name, email and password, show live password-strength feedback, validate the name, submit the account, request push notification permission, and then present an onboarding carousel with CTAs to try examples. Exit: user lands in the app's main exploration/visual-search interface with a bottom navigation and a camera-centric primary action. Assumptions: user is on an iOS device, not previously logged in, has network connectivity, and can receive push prompts. All inputs are client-side validated with inline feedback and a final server submission step. Steps: Welcome / Sign-in Choices → Create Account - Empty Form → Password Strength — Weak → Password Strength — Medium → Password Strength — Strong (masked) → Password Reveal (optional) → Create Account CTA Enabled → Display Name Validation Error → Correct Name and Submitting → Notification Permission Prompt → Onboarding Carousel — Intro → Onboarding Carousel — Example Prompt → Generated Article / Example Result → Article Scroll and Controls → Generated Content Actions & Suggestions → Home Feed — Visual Intelligence Tour → Camera-First Search Interface **SSENSE: Account registration and login entry** (ios, 8 screens, Marketplace) [Refero](https://refero.design/flows/7101) > Happy path: an unauthenticated user opens the Profile tab, selects Register, enters email and password, optionally chooses a newsletter preference, taps Create Account, sees a creating/loading state, and is returned to their Profile where Wishlist and Order History are available (empty if none). Steps: Profile - Logged Out → Register Form - Empty → Email Entered → Password Entered & Visibility → Keyboard Active → Newsletter Preference → Submit Registration → Profile - Logged In (Success) **On: New user sign-up and onboarding** (ios, 21 screens, Marketplace) [Refero](https://refero.design/flows/9584) > Happy-path new-user onboarding on a mobile commerce experience (user starts logged out). Entry: first-launch welcome/marketing screen. Exit: product listing (catalog) with region/language set and notifications permission presented. The flow assumes a mobile device (iPhone status bar shown), network available, and the user chooses to create an account (not continuing as guest). The flow includes a marketing carousel, a sign-up modal (email/password + social sign-in), progressive profile questions (name, who you shop for, activity preferences), a push-notification prompt (system-level), and a shipping location/language modal leading into the product catalog. Steps: Welcome / entry → Marketing carousel (card 1) → Marketing carousel (card 2/3) → Sign up modal (initial) → Sign up - enter credentials → Submit credentials → Collect name → Name entered → Who do you shop for? → Activity preferences list (initial) → Activity preferences (selection) → Enable notifications pitch → OS notification permission dialog → Onboarding splash / transitional carousel → Region & language modal (initial) → Region & language selection → Catalog landing with applied settings → Modal dismissal & final state → Picker interaction (country wheel) → Final catalog browsing (post onboarding) **DailyArt: Create account and patron onboarding** (ios, 7 screens, Education, Magazines & Newspapers) [Refero](https://refero.design/flows/3314) > Happy-path sign-up flow on a mobile iOS screen for a user who has (or just purchased) a subscription and needs to create/link an account. **Luminary: Account Creation and Subscription** (ios, 18 screens, Entertainment) [Refero](https://refero.design/flows/4695) > Happy path: user launches the app (splash) → sees onboarding hero with carousel of shows → grants/denies tracking permission (system dialog) → taps Get Started → chooses a signup method (Apple/Facebook/Google/Email) → fills email and password on the Create Account form with inline validation → after account creation the user is shown subscription pricing and chooses a plan → the platform purchase sheet (store) is displayed for confirmation → user confirms purchase and sees a success message → app returns to the signed-in home/content screens. Entry: cold start (new user). Exit: signed-in, subscribed user on Home/Show detail. Assumptions: mobile iOS device; user is new (not signed in); device provides system-level tracking prompt and in‑store purchase sheet. **ChatGPT: Sign up and email verification** (ios, 20 screens, AI Tool) [Refero](https://refero.design/flows/7006) > Happy path: An iOS user who is not _… 195116 more characters in the zip._ ### research/finance/ASSURANCE-AND-REUSE.md `research/finance/ASSURANCE-AND-REUSE.md` · sha256 `480f3311b802c97e7f38919b865ae9013a9fd2c2b10b107525703b820ccead01` #### EDGE Finance: Agent Assurance Landscape and Reusable Subsystems Research note for **EDGE Finance** (part of JudeAI | EDGE ID). Prepared 31 October 2026. --- ##### 0. How to read this document ###### 0.1 Evidence labels | Label | Meaning | |---|---| | **OBSERVED** | Read directly from a primary source: an official specification, a standards-body page, a government page, or the project's own repository (licence file, tags, commit history). | | **CLAIM** | Stated by a vendor, a project's marketing, or a secondary commentator (law firm, blog, news). It may be true, but nobody independent has verified it. | | **UNKNOWN** | We searched and could not confirm. The line says what we searched. | | **AGREEMENT (status)** | A standard or specification. Status is one of: *published* (final or ratified), *draft* (working document, may change), or *proposed* (an idea, concept paper or individual submission, not yet adopted by a standards group). | ###### 0.2 How the facts were checked - **Licences:** we read the actual `LICENSE` file in each repository, using `raw.githubusercontent.com` and shallow `git clone`. We did not rely on GitHub's licence badge. The GitHub REST API was not available to this research session. - **Versions:** "Highest version tag" is the highest semantic-version tag returned by `git ls-remote --tags` on 2026-10-31. That is usually the latest release, but not always, because projects sometimes tag patch releases on older branches. - **Maintenance:** "Last commit" is the date of the newest commit on the default branch, read from a shallow clone on 2026-10-31. - **Access date:** every source was accessed on **2026-10-31**, and each source link also carries that date. ###### 0.3 Acronyms (first use) | Acronym | Meaning | |---|---| | AI | Artificial Intelligence | | AML | Anti-Money Laundering | | API | Application Programming Interface | | AP2 | Agent Payments Protocol (Google-originated) | | A2A | Agent2Agent protocol (Google-originated, now Linux Foundation) | | ACP | Agentic Commerce Protocol (OpenAI and Stripe) | | AISI | UK AI Security Institute | | CAISI | US Center for AI Standards and Innovation (part of NIST) | | CEN-CENELEC | European standards bodies (European Committee for Standardization / for Electrotechnical Standardization) | | CNCF | Cloud Native Computing Foundation | | CSA | Cloud Security Alliance | | DID | Decentralized Identifier (W3C) | | DSIT | UK Department for Science, Innovation and Technology | | FAPI | Financial-grade API (OpenID Foundation security profile) | | FCA | UK Financial Conduct Authority | | FCDO | UK Foreign, Commonwealth & Development Office | | GNAP | Grant Negotiation and Authorization Protocol (IETF) | | HMT | HM Treasury | | IETF | Internet Engineering Task Force | | ISO/IEC | International Organization for Standardization / International Electrotechnical Commission | | KYA | Know Your Agent | | KYC / KYB | Know Your Customer / Know Your Business | | LLM | Large Language Model | | MCP | Model Context Protocol | | NCCoE | US National Cybersecurity Center of Excellence (part of NIST) | | NIST | US National Institute of Standards and Technology | | OAuth | Open Authorization (delegated access standard) | | OFAC | US Office of Foreign Assets Control | | OFSI | UK Office of Financial Sanctions Implementation | | OIDF | OpenID Foundation | | OWASP | Open Worldwide Application Security Project | | PISP / ASPSP | Payment Initiation Service Provider / Account Servicing Payment Service Provider (UK and EU payments law) | | RFC | Request for Comments (an IETF published standard or document) | | RMF | Risk Management Framework | | SPDX | Software Package Data Exchange (standard licence identifiers) | | VC | Verifiable Credential (W3C) | --- ##### Executive summary 1. **Agent-specific assurance standards are still forming.** No published standard covers autonomous agents as financial customers. NIST's AI RMF 1.0 dates from January 2023 and NIST says it "is being revised" ([NIST AI RMF](https://www.nist.gov/itl/ai-risk-management-framework), accessed 2026-10-31). NIST's agent identity work is a draft concept paper ([CSRC](https://csrc.nist.gov/pubs/other/2026/02/05/accelerating-the-adoption-of-software-and-ai-agent/ipd), accessed 2026-10-31). The most-cited KYA schemes are announcements, not specifications ([TechNode](https://technode.com/2026/09/10/ant-international-visa-and-mastercard-develop-know-your-agent-framework-for-ai-payments/), accessed 2026-10-31). OBSERVED / CLAIM as marked in Part 1. 2. **The agent protocols leave the hard parts out of scope.** - **MCP:** authorisation is "OPTIONAL" ([MCP spec 2026-07-28](https://modelcontextprotocol.io/specification/2026-07-28/basic/authorization), accessed 2026-10-31). - **A2A:** authorisation is "implementation-specific" ([A2A spec](https://github.com/a2aproject/A2A/blob/main/docs/specification.md), accessed 2026-10-31). - **AP2:** dispute handling, retention and retrieval are "outside the scope" ([AP2 spec](https://github.com/google-agentic-commerce/AP2/blob/main/docs/ap2/specification.md), accessed 2026-10-31). - **ACP:** refund semantics are "out of scope" ([ACP delegate-payment RFC](https://github.com/agentic-commerce-protocol/agentic-commerce-protocol/blob/main/rfcs/rfc.delegate_payment.md), accessed 2026-10-31). OBSERVED. 3. **Nobody standardises rehearsal, revocation, undo, liability or re-runnable audit.** We found nothing for rehearsal before commit, revocation, recovery/undo, liability allocation or regulator-re-runnable audit (Section 1.5). This is EDGE Finance's opening. 4. **Most subsystems can be reused under permissive licences.** Examples: TigerBeetle, Blnk, Apache Fineract, Hyperswitch, Mojaloop, Keycloak, Ory, OPA, Cedar, OpenFGA, Temporal and Mermaid (Part 2). 5. **Licence flags (detail in Section 2.11):** - Elastic-2.0: Marble, Ballerine. - AGPL-3.0: ZITADEL, Jube, Lago. - BUSL-1.1: Restate. - SSPL-1.0: Inngest. - CC-BY-NC-4.0: OpenSanctions data. - Proprietary terms: Visa Trusted Agent Protocol code. - No licence file: Formance Ledger v3 main branch. --- #### PART 1: Autonomous-agent AI assurance and governance ##### 1.1 Frameworks and standards ###### 1.1.1 Summary table | Framework / standard | Who | AGREEMENT status | What it is (plain English) | Agent-specific? | Source | |---|---|---|---|---|---| | NIST AI RMF 1.0 | NIST (US) | **published** 26 Jan 2023. NIST says it "is being revised as part of the White House AI Action Plan" (no date given) | A voluntary framework for managing AI risk in four functions: Govern, Map, Measure, Manage | No (OBSERVED: no agent profile listed on the NIST page) | [NIST AI RMF](https://www.nist.gov/itl/ai-risk-management-framework), accessed 2026-10-31 | | NIST AI 600-1 Generative AI Profile | NIST | **published** 26 Jul 2024 | A companion to the AI RMF covering risks that generative AI creates or makes worse | Not specifically. Secondary analysis says it does not distinguish autonomous multi-step agents (CLAIM) | [NIST AI RMF](https://www.nist.gov/itl/ai-risk-management-framework), accessed 2026-10-31; [CSA research note](https://labs.cloudsecurityalliance.org/research/csa-research-note-nist-ai-agent-standards-initiative-complia/), accessed 2026-10-31 | | NIST concept note: AI RMF Profile on Trustworthy AI in Critical Infrastructure | NIST | **proposed** (concept note, 7 Apr 2026) | Planned profile for critical-infrastructure operators | No | [NIST AI RMF](https://www.nist.gov/itl/ai-risk-management-framework), accessed 2026-10-31 | | NIST AI Agent Interoperability Profile | NIST/CAISI | **UNKNOWN**. Secondary sources say it is planned for Q4 2026; we found no NIST page confirming it | n/a | Yes, if it appears | [CSA white paper](https://labs.cloudsecurityalliance.org/agentic/agentic-nist-ai-rmf-profile-v1/), accessed 2026-10-31 (CLAIM) | | NCCoE concept paper "Accelerating the Adoption of Software and AI Agent Identity and Authorization" | NIST NCCoE | **proposed** (initial public draft, 5 Feb 2026; comments closed 2 Apr 2026) | Proposes treating agents as identifiable entities in enterprise identity systems, reusing OAuth, OpenID Connect and SCIM. Covers identification, authentication including revocation, authorisation, audit and non-repudiation | **Yes** | [CSRC listing](https://csrc.nist.gov/pubs/other/2026/02/05/accelerating-the-adoption-of-software-and-ai-agent/ipd), accessed 2026-10-31; [NCCoE project](https://www.nccoe.nist.gov/projects/software-and-ai-agent-identity-and-authorization), accessed 2026-10-31; [NIST news](https://www.nist.gov/news-events/news/2026/02/new-concept-paper-identity-and-authority-software-agents), accessed 2026-10-31 | | ISO/IEC 42001:2023 | ISO/IEC JTC 1/SC 42 | **published** Dec 2023 (Edition 1, IEC store gives 2023-12-18) | Requirements for an AI Management System (AIMS). It is certifiable, like ISO 27001 is for security | No (organisation-level) | [IEC webstore](https://webstore.iec.ch/publication/90574), accessed 2026-10-31; [ISO catalogue](https://www.iso.org/standard/81230.html), accessed 2026-10-31 (ISO page blocked automated fetch; status taken from search snippet) | | ISO/IEC 23894:2023 | ISO/IEC JTC 1/SC 42 | **published** Feb 2023 | Guidance on AI risk management, built on ISO 31000 | No | [ISO 77304](https://www.iso.org/standard/77304.html), accessed 2026-10-31; [ANSI blog](https://blog.ansi.org/ansi/iso-iec-23894-2023-ai-risk-management/), accessed 2026-10-31 | | ISO/IEC 42005:2025 | ISO/IEC | **published** May 2025 | Guidance for AI system impact assessments (societal and individual impact) | No | [ISO 44545](https://www.iso.org/standard/44545.html), accessed 2026-10-31 | | ISO/IEC 42006:2025 | ISO/IEC | **published** Jul 2025 | Extra requirements for bodies that audit and certify ISO/IEC 42001 management systems | No. It certifies organisations' management systems, not individual agents | [ISO 42006](https://www.iso.org/standard/42006), accessed 2026-10-31; [Standards Council of Canada](https://scc-ccn.ca/standardsdb/standards/8187573), accessed 2026-10-31 | | UK Trusted third-party AI assurance roadmap | DSIT | **published** policy paper, 3 Sep 2025 | Plans for a UK AI assurance market: a consortium, a skills framework and an AI Assurance Innovation Fund | **No.** The page does not mention AI agents (OBSERVED) | [GOV.UK](https://www.gov.uk/government/publications/trusted-third-party-ai-assurance-roadmap), accessed 2026-10-31 | | UK AI Security Institute (AISI) "Inspect" evaluation framework | AISI | Open-source tool, `MIT`. Highest tag 0.3.286; last commit 2026-10-30 | A framework for running LLM and agent evaluations. Reusable for EDGE racetrack scoring | Tooling, not a standard | [inspect_ai repo](https://github.com/UKGovernmentBEIS/inspect_ai), accessed 2026-10-31 (OBSERVED) | | UK AISI agent evaluation incident (July/Aug 2026) | AISI | n/a | Secondary reports say agents in an AISI cyber evaluation took 19 unsanctioned live-internet actions. We could not reach aisi.gov.uk to confirm | Yes | [letsdatascience](https://letsdatascience.com/news/rogue-ai-agents-challenge-trusted-access-controls-9037e91e), accessed 2026-10-31 (**CLAIM, unverified**) | | EU AI Act (Reg. (EU) 2024/1689) as amended by Digital Omnibus on AI, Reg. (EU) 2026/1744 | EU | **published**. Omnibus in OJ 24 Jul 2026, in force 27 Jul 2026. Annex III high-risk duties (this includes credit scoring) deferred to **2 Dec 2027**. Annex I deferred to **2 Aug 2028** | EU AI law. Creditworthiness assessment is a listed high-risk use | Not agent-specific | [Mayer Brown](https://www.mayerbrown.com/en/insights/publications/2026/07/eu-ai-act-news-digital-omnibus-on-ai-new-guidance-on-risk-classification-gpai-and-transparency-obligations), accessed 2026-10-31; [Leaseurope](https://www.leaseurope.org/new-application-dates-high-risk-ai-system-rules), accessed 2026-10-31 (CLAIM: secondary; Official Journal text not fetched) | | EU AI Act harmonised standards (CEN-CENELEC JTC 21) | CEN-CENELEC | **draft**. Autumn-2025 deadline missed. Secondary sources say prEN 18286 (AI quality management system) was first to reach final approval in 2026. No harmonised standard confirmed as cited in the Official Journal | Standards that would give "presumption of conformity" with the AI Act | No | [EU AI Act Newsletter](https://artificialintelligenceact.substack.com/p/the-eu-ai-act-newsletter-63-standards), accessed 2026-10-31; [sota.io](https://sota.io/blog/eu-ai-act-cen-cenelec-harmonised-standards-delay-august-2026-conformity-assessment-developer-guide), accessed 2026-10-31 (CLAIM). **UNKNOWN** whether any are cited in the Official Journal as of today | | OWASP Top 10 for LLM Applications 2025 | OWASP GenAI Security Project | **published** (2025 edition) | Ten top LLM risks. **LLM06 "Excessive Agency"** is directly relevant to agents | Partly | [OWASP LLM Top 10](https://genai.owasp.org/llm-top-10/), accessed 2026-10-31 | | OWASP Top 10 for Agentic Applications for 2026 | OWASP GenAI Security Project | **published** 9 Dec 2025 | Ten agent-specific risks, ASI01 to ASI10 (for example goal hijack, tool misuse, identity and privilege abuse, rogue agents). Site content is licensed `CC-BY-SA-4.0` | **Yes** | [OWASP resource page](https://genai.owasp.org/resource/owasp-top-10-for-agentic-applications-for-2026/), accessed 2026-10-31. Item titles from [Cycode summary](https://cycode.com/blog/owasp-top-10-agentic-applications/), accessed 2026-10-31 (CLAIM: the official page did not render the list) | | MITRE ATLAS (Adversarial Threat Landscape for AI Systems) | MITRE | **published** knowledge base, data release **v2026.10 (2026-10-28)**: 16 tactics, 129 techniques, 92 sub-techniques, 44 mitigations, 77 case studies. Data licensed `Apache-2.0` | A catalogue of attacker techniques against AI systems, including agent tools and "Poison AI Agent Learning Environment" (AML.T0145) | Partly | [atlas-data CHANGELOG](https://github.com/mitre-atlas/atlas-data/blob/main/CHANGELOG.md), accessed 2026-10-31; [LICENSE](https://github.com/mitre-atlas/atlas-data/blob/main/LICENSE), accessed 2026-10-31 (OBSERVED) | | CSA AI Controls Matrix (AICM) v1.1 | Cloud Security Alliance | **published**. CSA artifact page says released 22 Jun 2026; blog 14 Jul 2026 | 247 control objectives across 18 domains, with an AI-CAIQ questionnaire, mapped to ISO 42001 and other frameworks | Partly | [CSA AICM v1.1](https://cloudsecurityalliance.org/artifacts/ai-controls-matrix-v1-1), accessed 2026-10-31; [CSA blog](https://cloudsecurityalliance.org/blog/2026/07/14/ai-controls-matrix-v1-1-strengthening-the-foundation-for-trustworthy-ai), accessed 2026-10-31 (control count via secondary search summary: CLAIM) | | HM Treasury Financial Services AI Adoption Plan | HMT (UK) | **published** 14 Jul 2026 | Ten recommendation areas, including "agentic payments" (OBSERVED). Commentators say Recommendation 10 proposes a trust framework covering liability, Know Your Agent and machine-to-machine authentication (CLAIM) | **Yes** (payments) | [GOV.UK](https://www.gov.uk/government/publications/ai-adoption-plan-financial-services), accessed 2026-10-31; [Regulation Tomorrow](https://www.regulationtomorrow.com/2026/07/hmt-publishes-financial-services-ai-adoption-plan/), accessed 2026-10-31; [GoCardless](https://gocardless.com/blog/action-plan-agentic-commerce), accessed 2026-10-31 (CLAIM) | | FCA 2026 payments regulatory priorities | FCA (UK) | **published** (25 Mar 2026, per secondary) | Secondary reports say the FCA will consider whether payments and e-money rules need to change for agentic AI payments | Yes | [Linklaters](https://techinsights.linklaters.com/post/102mo0a/uk-payments-priorities-for-2026-include-agentic-ai-and-stablecoins), accessed 2026-10-31; [Payment Expert](https://paymentexpert.com/2026/03/25/fca-2026-payments-regulatory-priorities-report/), accessed 2026-10-31 (CLAIM, FCA text not fetched) | ###### 1.1.2 What this means for EDGE Finance - **Management-system certification (ISO/IEC 42001 + 42006) is the only certifiable AI assurance route today.** It certifies the *organisation*, not each agent. Sources: [ISO 42006](https://www.iso.org/standard/42006), accessed 2026-10-31; [IEC webstore 42001](https://webstore.iec.ch/publication/90574), accessed 2026-10-31. OBSERVED. - **OWASP Agentic Top 10 and MITRE ATLAS can serve as the threat model for the racetrack.** Each racetrack scenario can be tagged with ASI and ATLAS IDs. Sources: [OWASP](https://genai.owasp.org/resource/owasp-top-10-for-agentic-applications-for-2026/), accessed 2026-10-31; [ATLAS](https://github.com/mitre-atlas/atlas-data/blob/main/CHANGELOG.md), accessed 2026-10-31. (Recommendation.) - **UK government assurance work does not yet address agents.** The DSIT roadmap does not mention agents. Source: [GOV.UK](https://www.gov.uk/government/publications/trusted-third-party-ai-assurance-roadmap), accessed 2026-10-31. OBSERVED. --- ##### 1.2 Agent identity and authority | Item | Owner | AGREEMENT status | What it does | Relevance and limits | Source | |---|---|---|---|---|---| | **OpenID AuthZEN Authorization API 1.0** | OpenID Foundation AuthZEN WG | **published** (Final Specification; approval announced 12 Jan 2026; vote 81 for, 1 against, 25 abstain) | A standard API between a "policy enforcement point" (your app) and a "policy decision point" (an authorisation engine), so engines can be swapped | High: lets EDGE put OPA, Cedar or OpenFGA behind one interface. It is not agent-specific | [OIDF announcement](https://openid.net/authorization-api-1-0-final-specification-approved/), accessed 2026-10-31 (OBSERVED via search extract) | | **OIDF whitepaper "Identity Management for Agentic AI"** | OIDF AIIM Community Group | **published** whitepaper (arXiv 29 Oct 2025). Not a standard | Says current OAuth/OIDC handles simple agent cases. Lists gaps: recursive delegation, user impersonation, consent fatigue, cross-organisation agents | Evidence of gaps | [OIDF post](https://openid.net/new-whitepaper-tackles-ai-agent-identity-challenges/), accessed 2026-10-31; [arXiv 2510.25819](https://arxiv.org/abs/2510.25819), accessed 2026-10-31 | | **draft-oauth-ai-agents-on-behalf-of-user** | Individual (WSO2 authors) | **proposed** (individual Internet-Draft; -02 dated 26 Aug 2025, expiry 27 Feb 2026; not adopted by the OAuth WG per secondary) | Extends the OAuth authorisation-code flow so the *agent's* identity appears at user consent and in the token | Shows the direction of travel. Not stable | [IETF datatracker -02](https://datatracker.ietf.org/doc/html/draft-oauth-ai-agents-on-behalf-of-user-02), accessed 2026-10-31; [WorkOS](https://workos.com/blog/oauth-on-behalf-of-ai-agents), accessed 2026-10-31 (CLAIM on adoption) | | Other IETF agent-auth drafts (draft-klrc-aiagent-auth, transaction-tokens-for-agents, liu chain delegation) | Individuals | **proposed** (individual drafts). A May 2026 OAuth interim slide deck groups them as "IETF Individual Drafts on AI Agent Authorization" | Agent authentication, token chaining, agent-to-agent delegation | No adopted IETF standard for agent delegation found | [OAuth interim 2026-03 slides](https://datatracker.ietf.org/meeting/interim-2026-oauth-03/materials/slides-interim-2026-oauth-03-sessa-ietf-individual-draft-analysis-00), accessed 2026-10-31; [draft-klrc-aiagent-auth](https://datatracker.ietf.org/doc/draft-klrc-aiagent-auth/02/), accessed 2026-10-31 | | **GNAP**, RFC 9635 | IETF | **published** (Proposed Standard, 2024) | A redesigned delegation protocol (successor-style alternative to OAuth) with richer, negotiated grants | Expressive enough for scoped mandates. Ecosystem adoption is low (UNKNOWN: we found no major agent protocol that uses it) | [RFC 9635](https://www.rfc-editor.org/rfc/rfc9635), accessed 2026-10-31 | | **W3C Verifiable Credentials Data Model v2.0** | W3C | **published** (W3C Recommendation, 15 May 2025) | A standard format for signed, tamper-evident claims ("this agent is authorised by Org X up to £500") | Used by AP2 mandates (CLAIM). A good carrier for EDGE mandates | [W3C TR](https://www.w3.org/TR/vc-data-model-2.0/), accessed 2026-10-31 | | **W3C DID Core v1.1** | W3C | **draft** (Candidate Recommendation Snapshot, 5 Mar 2026) | Self-controlled identifiers with discoverable keys | Optional for EDGE. v1.0 is the older Recommendation | [W3C TR did-1.1](https://www.w3.org/TR/did-1.1/), accessed 2026-10-31 | | **Web Bot Auth** (HTTP Message Signatures for Bots) | IETF webbotauth WG (Cloudflare and Google authors) | **draft** (WG draft `draft-ietf-webbotauth-httpsig-protocol-00`, Sep 2026) | Agents sign each HTTP request (RFC 9421), with a `Signature-Agent` header and a key directory at a well-known URI | Proves *which software* sent a request. It does not prove which accountable human or organisation stands behind it | [IETF datatracker](https://datatracker.ietf.org/doc/html/draft-ietf-webbotauth-httpsig-protocol-00), accessed 2026-10-31. Cloudflare library `Apache-2.0` ([repo](https://github.com/cloudflare/web-bot-auth), accessed 2026-10-31) | | **Visa Trusted Agent Protocol (TAP)** | Visa (with Cloudflare) | Vendor specification and code. Not a standards-body document | Builds on Web Bot Auth and RFC 9421. Agent signatures carry timestamps, session ID, key ID and an intent tag (browse or pay). Can pass consumer identifiers and Payment Account References | **Licence flag:** the repo's `LICENSE.md` binds you to Visa Developer Center Terms of Use and TAP Product Terms. This is **not an open-source licence (no SPDX id)**. Last commit 2025-10-28 | [repo README](https://github.com/visa/trusted-agent-protocol), accessed 2026-10-31; [LICENSE.md](https://github.com/visa/trusted-agent-protocol/blob/main/LICENSE.md), accessed 2026-10-31 (OBSERVED); [Cloudflare blog](https://blog.cloudflare.com/secure-agentic-commerce/), accessed 2026-10-31 | | **Mastercard Verifiable Intent** | Mastercard (with Google) | **draft** (reported Draft v0.1, Mar 2026) | SD-JWT-based record linking identity, intent and action, with constraint types such as amount bounds and merchant allow-lists | Contributed to FIDO Payments WG (CLAIM) | [Mastercard story](https://www.mastercard.com/us/en/news-and-trends/stories/2026/verifiable-intent.html), accessed 2026-10-31; [PYMNTS](https://www.pymnts.com/mastercard/2026/mastercard-unveils-open-standard-to-verify-ai-agent-transactions/), accessed 2026-10-31 (CLAIM) | | **FIDO Alliance Agentic Authentication TWG and Payments TWG** | FIDO Alliance | **proposed** (working groups formed 28 Apr 2026. No published spec found) | Standardising agent delegation with phishing-resistant authentication. AP2 says its standardisation continues in these FIDO groups | Watch closely | [FIDO Alliance](https://www.fidoalliance.org/fido-alliance-to-develop-standards-for-trusted-ai-agent-interactions/), accessed 2026-10-31; [AP2 docs index](https://github.com/google-agentic-commerce/AP2/blob/main/docs/index.md), accessed 2026-10-31 (OBSERVED) | | **"Know Your Agent" (KYA)** | Many vendors. Ant International, Visa and Mastercard "initiated collaboration" Sep 2026 | **proposed**. No formal KYA standard from any standards body or regulator found | The idea: verify an agent and bind it to the customer it acts for, with revocation | Gap (Section 1.5) | [TechNode](https://technode.com/2026/09/10/ant-international-visa-and-mastercard-develop-know-your-agent-framework-for-ai-payments/), accessed 2026-10-31 (CLAIM); [Trulioo](https://www.trulioo.com/blog/trust-and-safety/future-agentic-commerce-know-your-agent-kya), accessed 2026-10-31 (CLAIM). **UNKNOWN**: searched "Know Your Agent framework standard 2026"; found only vendor and press material | | **SPIFFE / SPIRE** (workload identity) | CNCF | Graduated CNCF project. SPIRE `Apache-2.0`, highest tag v1.15.5, last commit 2026-10-29 | Cryptographic identity for running software (workloads) | Useful for identifying *EDGE-hosted* agent processes inside the racetrack | [CNCF projects](https://www.cncf.io/projects/), accessed 2026-10-31; [SPIRE repo](https://github.com/spiffe/spire), accessed 2026-10-31 (OBSERVED) | --- ##### 1.3 Agent protocols | Protocol | Governance | Version / status (AGREEMENT) | Licence (SPDX) | Authorisation model | What is explicitly out of scope | Source | |---|---|---|---|---|---|---| | **Model Context Protocol (MCP)** | Linux Foundation's Agentic AI Foundation (AAIF), formed 9 Dec 2025 with MCP, goose and AGENTS.md as founding projects | **published** spec revision **2026-07-28** (git tag `2026-07-28`) | Transition from `MIT` to `Apache-2.0`. New contributions `Apache-2.0`. Docs `CC-BY-4.0`. Some legacy `MIT` | Authorisation is **"OPTIONAL"**. When used, MCP servers are OAuth 2.1 resource servers and MUST implement RFC 9728 Protected Resource Metadata. Clients MUST send RFC 8707 `resource` (audience binding). Client ID Metadata Documents SHOULD be supported. Dynamic Client Registration is deprecated | Agent identity, delegation chains, spend limits: none in the core authorisation spec (OBSERVED) | [MCP authorization](https://modelcontextprotocol.io/specification/2026-07-28/basic/authorization), accessed 2026-10-31; [MCP LICENSE](https://github.com/modelcontextprotocol/modelcontextprotocol/blob/main/LICENSE), accessed 2026-10-31; [LF AAIF press](https://www.linuxfoundation.org/press/linux-foundation-announces-the-formation-of-the-agentic-ai-foundation), accessed 2026-10-31 | | **A2A (Agent2Agent)** | Linux Foundation project (originally Google). >150 supporting organisations per LF, 9 Apr 2026 | **published** v1.0 (first stable spec). Tags show v1.0.1 and v1.1.0-rc | `Apache-2.0` | "Identity information is handled at the protocol layer, not within A2A semantics." Credentials are obtained out-of-band. Server authorisation "is implementation-specific". v1.0 adds signed Agent Cards | Accountable-principal binding, mandates, limits (OBSERVED: left to implementers) | [A2A spec §7](https://github.com/a2aproject/A2A/blob/main/docs/specification.md), accessed 2026-10-31; [LF press](https://www.linuxfoundation.org/press/a2a-protocol-surpasses-150-organizations-lands-in-major-cloud-platforms-and-sees-enterprise-production-use-in-first-year), accessed 2026-10-31; [A2A tags](https://github.com/a2aproject/A2A/tags), accessed 2026-10-31 | | **Agent Payments Protocol (AP2)** | Google-originated. Standardisation continuing in FIDO Alliance working groups | **draft** (tags v0.1.0, v0.2.0. Last commit 2026-04-29) | `Apache-2.0` | Signed mandates (Checkout Mandate, Payment Mandate). Payment Mandate constraints: allowed payee, allowed instrument, allowed PISP, amount range, **budget**, agent recurrence (frequency, max occurrences), execution date | "Specific details of how this is used for dispute resolution, retention, and retrieval requirements are outside the scope". Liability is left to "payment networks" | [AP2 spec](https://github.com/google-agentic-commerce/AP2/blob/main/docs/ap2/specification.md), accessed 2026-10-31; [Payment Mandate](https://github.com/google-agentic-commerce/AP2/blob/main/docs/ap2/payment_mandate.md), accessed 2026-10-31; [FAQ](https://github.com/google-agentic-commerce/AP2/blob/main/docs/faq.md), accessed 2026-10-31 (OBSERVED) | | **Agentic Commerce Protocol (ACP)** | Maintained by OpenAI and Stripe. Requires a contributor licence agreement | **draft** ("currently in `beta`"). Latest stable spec folder **2026-04-17** | `Apache-2.0` | Delegate Payment issues a vault token usable "**only** within explicit Allowance constraints" (reason, `max_amount`, currency, expiry). The token MUST become invalid at `expires_at` | "Out of scope: PSP-specific authorization/capture, multi-use tokens beyond allowance, refund semantics." | [ACP README](https://github.com/agentic-commerce-protocol/agentic-commerce-protocol), accessed 2026-10-31; [rfc.delegate_payment.md](https://github.com/agentic-commerce-protocol/agentic-commerce-protocol/blob/main/rfcs/rfc.delegate_payment.md), accessed 2026-10-31 (OBSERVED) | **Note on MCP governance:** the Linux Foundation states that AAIF platinum members include AWS, Anthropic, Block, Bloomberg, Cloudflare, Google, Microsoft and OpenAI ([LF press](https://www.linuxfoundation.org/press/linux-foundation-announces-the-formation-of-the-agentic-ai-foundation), accessed 2026-10-31; CLAIM: the foundation's own statement). A blog reports that A2A moved under AAIF on 17 Aug 2026 ([Pebblous](https://blog.pebblous.ai/blog/a2a-mcp-agentic-ai-foundation-authorization/en/), accessed 2026-10-31). We found no second source: **UNKNOWN**. --- ##### 1.4 Open banking and API security standards relevant to agents | Standard | AGREEMENT status | Notes | Source | |---|---|---|---| | **FAPI 2.0 Security Profile** | **published** (OIDF Final, Feb 2025) | A hardened OAuth profile for financial APIs: sender-constrained tokens, PAR, PKCE | [OIDF FAPI 2.0](https://openid.net/tag/fapi-2-0/page/2/), accessed 2026-10-31 | | **FAPI 2.0 Message Signing** | **published** (OIDF Final, 26 Sep 2025) | Non-repudiable signed requests and responses. Useful evidence for audit | [OIDF announcement](https://openid.net/fapi-2-message-signing-final-specification-approved/), accessed 2026-10-31 | | **UK Open Banking Read/Write API** | **published** v4.0 (Jul 2024). Repo tags include `v4.0.1` and `v4.0.1-Update-1`. Last commit 2026-09-18 | No `LICENSE` file in the specs repo (**UNKNOWN** licence; terms likely on openbanking.org.uk, not verified) | [read-write-api-specs](https://github.com/OpenBankingUK/read-write-api-specs), accessed 2026-10-31; [v4.0.1 site](https://openbankinguk.github.io/read-write-api-site3/v4.0.1), accessed 2026-10-31; [IBS Intelligence](https://ibsintelligence.com/ibsi-news/obl-publishes-open-banking-standard-v4-0), accessed 2026-10-31 (CLAIM for v4.0 date) | | **ISO 20022** (financial messaging) | **published** standard. Swift cross-border payment-instruction coexistence (MT vs pacs.008/pacs.009) ended **22 Nov 2025**. Statement and reporting (camt.053) coexistence was extended (planned SR 2028 per Swift roadmap). pain.001 interbank date to be confirmed | EDGE should speak pain.001 (customer to bank), pacs.008 (bank to bank) and camt.053 (statements) | [Swift ISO 20022](https://www.swift.com/standards/iso-20022/iso-20022-bytes/one-month-go), accessed 2026-10-31; [Swift CBPR+ roadmap](https://www.swift.com/sites/default/files/files/cbpr-phase-2-roadmap_detailed_v4-2.pdf), accessed 2026-10-31; [State Street guide](https://statestreet.com/web/insights/articles/documents/state-street-client-guide-to-iso-20022-2025.pdf), accessed 2026-10-31 (CLAIM) | | **BIAN Service Landscape** | **published** v14.0 (release notes final 20 Feb 2026) | A reference map of banking "service domains". Secondary source says 340 service domains in v14. Use it as naming and boundary vocabulary, not code | [BIAN v14.0 release notes](https://bian.org/wp-content/uploads/2026/02/BIAN-v14.0-Release-Notes-v1.0_-Final-Version.pdf), accessed 2026-10-31; [Fusion5](https://www.fusion5.com/au/integration-services/blogs/bian-banking-architecture-2026.md), accessed 2026-10-31 (CLAIM for count) | | **HMT Modernising Payment Services Regulation consultation** | **proposed** (consultation closed 6 Oct 2026, per secondary) | Asks whether consent, authentication and liability provisions need updating for agentic payments | [Bratby Law](https://bratby.law/agentic-payments-liability-psrs-2017/), accessed 2026-10-31 (CLAIM) | --- ##### 1.5 What is MISSING for agents to be safe, accountable financial customers Each row gives the requirement, what exists today, the evidence of a gap, and how strong that evidence is. "Searched" lists exactly what we looked for. | # | Requirement | What exists today | Evidence that it is a gap | Evidence strength | |---|---|---|---|---| | G1 | **Agent identity bound to an accountable human or organisation** | Web Bot Auth and Visa TAP prove which *software* signed a request. AP2 mandates are signed by the user. The NCCoE paper proposes treating agents as enterprise identities | No KYA standard exists. HMT plan *proposes* KYA work (CLAIM: [Regulation Tomorrow](https://www.regulationtomorrow.com/2026/07/hmt-publishes-financial-services-ai-adoption-plan/), accessed 2026-10-31). Card networks only "initiated collaboration" in Sep 2026 ([TechNode](https://technode.com/2026/09/10/ant-international-visa-and-mastercard-develop-know-your-agent-framework-for-ai-payments/), accessed 2026-10-31). NCCoE work is a draft concept paper ([CSRC](https://csrc.nist.gov/pubs/other/2026/02/05/accelerating-the-adoption-of-software-and-ai-agent/ipd), accessed 2026-10-31). OIDF lists "user impersonation" and "recursive delegation" as open problems ([OIDF](https://openid.net/new-whitepaper-tackles-ai-agent-identity-challenges/), accessed 2026-10-31) | Strong (OBSERVED for draft status; CLAIM for KYA) | | G2 | **Scoped mandates** (what the agent may do, for whom, where) | AP2 Payment Mandate constraints (draft v0.2). ACP Allowance (beta). Mastercard Verifiable Intent (draft v0.1). GNAP RFC 9635 can express rich grants | Three incompatible pre-1.0 formats ([AP2](https://github.com/google-agentic-commerce/AP2/blob/main/docs/ap2/payment_mandate.md); [ACP](https://github.com/agentic-commerce-protocol/agentic-commerce-protocol/blob/main/rfcs/rfc.delegate_payment.md); [Mastercard](https://www.mastercard.com/us/en/news-and-trends/stories/2026/verifiable-intent.html), all accessed 2026-10-31). MCP has scopes but authorisation is "OPTIONAL" ([MCP](https://modelcontextprotocol.io/specification/2026-07-28/basic/authorization), accessed 2026-10-31). A2A authorisation is "implementation-specific" ([A2A §7.5](https://github.com/a2aproject/A2A/blob/main/docs/specification.md), accessed 2026-10-31) | Strong (OBSERVED) | | G3 | **Spend limits enforced across channels** | AP2 `payment.budget` and `payment.agent_recurrence`. ACP `max_amount` | AP2 says "Evaluating the budget requires tracking the total amount spent". It does not say who holds that shared counter across merchants ([AP2 Payment Mandate](https://github.com/google-agentic-commerce/AP2/blob/main/docs/ap2/payment_mandate.md), accessed 2026-10-31). No cross-protocol budget ledger found | Medium (OBSERVED text; gap inferred) | | G4 | **Revocation** (kill an agent's authority instantly) | ACP tokens expire at `expires_at`. NCCoE paper lists credential revocation as a topic | Searched AP2 `docs/ap2/*` for "revoc/revok": **0 hits**. Searched ACP `rfcs/` for "revok/revoc": only marketing-consent hits, none for payment mandates ([ACP rfcs](https://github.com/agentic-commerce-protocol/agentic-commerce-protocol/tree/main/rfcs), accessed 2026-10-31). Commentary says KYA *could* help a provider revoke authority, which means it does not today ([GoCardless](https://gocardless.com/blog/action-plan-agentic-commerce), accessed 2026-10-31, CLAIM) | Medium (OBSERVED absence in two specs) | | G5 | **Rehearsal before commit** (dry-run, sandbox, preview of consequences) | Vendor sandboxes exist per product (for example Formance and Blnk cloud sandboxes, CLAIM) | Searched AP2 docs and ACP RFCs for "dry run", "simulat", "rehears": **0 hits** ([AP2](https://github.com/google-agentic-commerce/AP2/tree/main/docs/ap2); [ACP](https://github.com/agentic-commerce-protocol/agentic-commerce-protocol/tree/main/rfcs), accessed 2026-10-31). MCP core spec not searched for this: **UNKNOWN** | Medium (OBSERVED absence) | | G6 | **Error-rate guarantees** (measured, contractual reliability of an agent) | Evaluation tools exist (AISI Inspect, `MIT`) | No standard defines agent error-rate service levels. Secondary analysis says NIST AI RMF and AI 600-1 do not distinguish recommendation systems from autonomous multi-day agents ([CSA note](https://labs.cloudsecurityalliance.org/research/csa-research-note-nist-ai-agent-standards-initiative-complia/), accessed 2026-10-31, CLAIM). **Searched:** NIST AI RMF page, ISO 42001/42005/42006 summaries, OWASP Agentic Top 10 page. None define error-rate guarantees | Medium (absence plus CLAIM) | | G7 | **Independent verification of a specific agent** | ISO/IEC 42006 accredits auditors of *management systems* | ISO 42006 scope is AIMS certification bodies, not agents ([ISO 42006](https://www.iso.org/standard/42006), accessed 2026-10-31). DSIT roadmap does not mention agents ([GOV.UK](https://www.gov.uk/government/publications/trusted-third-party-ai-assurance-roadmap), accessed 2026-10-31). No agent certification scheme found | Strong (OBSERVED) | | G8 | **Recovery / undo** (reverse a wrong agent action) | Card chargebacks and dispute processes. AP2 provides mandate evidence | ACP: "refund semantics" out of scope ([ACP RFC](https://github.com/agentic-commerce-protocol/agentic-commerce-protocol/blob/main/rfcs/rfc.delegate_payment.md), accessed 2026-10-31). AP2: dispute resolution "outside the scope" ([AP2 spec](https://github.com/google-agentic-commerce/AP2/blob/main/docs/ap2/specification.md), accessed 2026-10-31). Push payments (UK Faster Payments) are not natively reversible; AP2 lists push payments only on its "roadmap" ([AP2 index](https://github.com/google-agentic-commerce/AP2/blob/main/docs/index.md), accessed 2026-10-31) | Strong (OBSERVED) | | G9 | **Liability allocation** (who pays when the agent is wrong) | UK Payment Services Regulations 2017 allocate loss between payer and provider | AP2 FAQ: its goal is "supporting evidence that helps payment networks establish accountability and liability principles". The protocol itself does not allocate liability ([AP2 FAQ](https://github.com/google-agentic-commerce/AP2/blob/main/docs/faq.md), accessed 2026-10-31, OBSERVED). Law-firm analysis: liability between PISP, ASPSP and AI provider "is not addressed in the current framework" ([Bratby Law](https://bratby.law/agentic-payments-liability-psrs-2017/), accessed 2026-10-31, CLAIM). FCA is "considering" changes ([Linklaters](https://techinsights.linklaters.com/post/102mo0a/uk-payments-priorities-for-2026-include-agentic-ai-and-stablecoins), accessed 2026-10-31, CLAIM) | Strong | | G10 | **Audit evidence a regulator can re-run** (deterministic replay of what the agent saw, decided and did) | AP2 gives a "non-repudiable, cryptographic audit trail" of mandates. FAPI 2.0 Message Signing gives signed API messages | AP2 retention and retrieval are out of scope ([AP2 spec](https://github.com/google-agentic-commerce/AP2/blob/main/docs/ap2/specification.md), accessed 2026-10-31). No standard for replaying an agent's decision path found. **Searched:** AP2, ACP, MCP authorisation, A2A §7, NIST AI RMF page, OWASP agentic page | Medium (absence) | | G11 | **Agent-to-agent delegation chains** (agent A sub-delegates to agent B) | Individual IETF drafts (chain delegation, transaction tokens for agents) | All are individual drafts, not WG-adopted ([OAuth interim slides](https://datatracker.ietf.org/meeting/interim-2026-oauth-03/materials/slides-interim-2026-oauth-03-sessa-ietf-individual-draft-analysis-00), accessed 2026-10-31). The on-behalf-of draft -02 expired 27 Feb 2026 ([datatracker](https://datatracker.ietf.org/doc/html/draft-oauth-ai-agents-on-behalf-of-user-02), accessed 2026-10-31) | Strong (OBSERVED) | | G12 | **Open, licence-clean agent signalling for payments** | Web Bot Auth (`Apache-2.0` library, IETF draft) | Visa TAP code is under proprietary Visa terms, not an open-source licence ([LICENSE.md](https://github.com/visa/trusted-agent-protocol/blob/main/LICENSE.md), accessed 2026-10-31) | Strong (OBSERVED) | ###### 1.5.1 Implication for EDGE Finance (assessment, not a sourced fact) EDGE ID already plans agent identity, delegated authority, a mailbox, a racetrack and recoverable actions. These line up with gaps G1, G2, G4, G5, G8 and G10. EDGE should: 1. **Emit and accept** existing formats where they exist (AP2 mandates, ACP allowances, Web Bot Auth signatures, VC 2.0) rather than inventing new ones. 2. **Add the missing layer** on top: KYA binding, revocation, a rehearsal record, undo and a replayable evidence pack. --- #### PART 2: Existing subsystems to reuse instead of building bespoke **Fit ratings** (assessment): ★★★ strong fit, ★★ usable with work, ★ weak fit or blocked. ##### 2.1 Ledgers | Project | What it does | Licence (SPDX), OBSERVED | Highest version tag / last commit | Maturity notes | Fit for EDGE Finance | Source | |---|---|---|---|---|---|---| | **TigerBeetle** | A purpose-built financial transactions database (debits and credits, accounts, transfers) | `Apache-2.0` | 0.17.10 / 2026-10-14 | Pre-1.0 version numbers. Project claims "designed for mission critical safety and performance" (CLAIM) | ★★★ Core balance engine. Its two-phase (pending then post or void) transfers map naturally to **rehearse then commit**. Needs a metadata or audit layer beside it | [repo](https://github.com/tigerbeetle/tigerbeetle), [LICENSE](https://github.com/tigerbeetle/tigerbeetle/blob/main/LICENSE), [tags](https://github.com/tigerbeetle/tigerbeetle/tags), accessed 2026-10-31 | | **Formance Ledger** | Programmable double-entry ledger with the Numscript language | **v2.x branch:** `MIT` (LICENSE present at tag v2.4.17). **v3 `main` branch (Raft-cluster rewrite, started 2025-12-03): no LICENSE file at HEAD** → **UNKNOWN** (by default, no licence means no rights granted) | v2.4.17 (2026-10-28); v3.0.0-rc.2 (2026-10-30) / 2026-10-30 | Active. v3 is a ground-up rewrite (OBSERVED from git history) | ★★ Use v2 (`MIT`) now. **Flag:** confirm the v3 licence with Formance before adopting it | [repo](https://github.com/formancehq/ledger), [v2.4.17 tree](https://github.com/formancehq/ledger/tree/v2.4.17), accessed 2026-10-31 | | **Formance Stack** | Bundle: Ledger, Payments, Numscript and more | `MIT` for community edition. Code under `ee/` directories uses a separate Enterprise licence | / 2026-10-29 | Active | ★★ Avoid `ee/` components unless licensed | [LICENSE](https://github.com/formancehq/stack/blob/main/LICENSE), accessed 2026-10-31 | | **Blnk** | Open-source double-entry ledger with balances, reconciliation and balance monitoring | `Apache-2.0` | v0.15.7 / 2026-10-22 | Pre-1.0. Commercial cloud offered | ★★ Good for wallets and reconciliation. Lighter than TigerBeetle | [repo](https://github.com/blnkfinance/blnk), [LICENSE.md](https://github.com/blnkfinance/blnk/blob/main/LICENSE.md), accessed 2026-10-31 | | **Mojaloop central-ledger** | Clearing and net-settlement ledger between financial service providers in a payment scheme | `Apache-2.0` (Mojaloop Foundation) | v20.3.1 / 2026-10-22 | Mature, scheme-level (inter-bank), not a customer ledger | ★ For EDGE's customer ledger. Relevant only if EDGE runs a scheme | [repo](https://github.com/mojaloop/central-ledger), [LICENSE.md](https://github.com/mojaloop/central-ledger/blob/main/LICENSE.md), accessed 2026-10-31 | ##### 2.2 Core banking | Project | What it does | Licence (SPDX) | Version / last commit | Fit | Source | |---|---|---|---|---|---| | **Apache Fineract** | Open-source core banking: clients, loans, savings, accounting, with REST APIs | `Apache-2.0` (OBSERVED: `LICENSE_SOURCE` on `develop`) | highest tag 1.16.1 / 2026-10-30 | ★★ Proven product engine for deposits and loans. Heavy (Java/Spring). Agent-as-customer would need a new client type | [repo](https://github.com/apache/fineract), [LICENSE_SOURCE](https://github.com/apache/fineract/blob/develop/LICENSE_SOURCE), accessed 2026-10-31 | | **Mifos X web app** (openMF/web-app) | Back-office user interface for Fineract | `MPL-2.0` (file-level copyleft: changes to its own files must be shared; fine commercially) | tag v1.0.0-fineract1.14 / 2026-10-30 | ★★ Staff UI | [repo](https://github.com/openMF/web-app), accessed 2026-10-31 | | **Apache Fineract CN** | Earlier microservices rewrite | **UNKNOWN**: `apache/fineract-cn-mono` clone failed (repo not found) | n/a | Do not use: treat as inactive until verified | searched `github.com/apache/fineract-cn-mono`, accessed 2026-10-31 | | **Kill Bill** (billing, adjacent) | Subscription billing and invoicing | `Apache-2.0` | killbill-0.25.6 / 2026-10-28 | ★★ If EDGE bills for agent services | [repo](https://github.com/killbill/killbill), accessed 2026-10-31 | | **Lago** (billing, adjacent) | Usage-based billing | `AGPL-3.0` (variant not stated in header) | / 2026-10-30 | ★ **Licence flag** (network copyleft) | [repo](https://github.com/getlago/lago), accessed 2026-10-31 | Other open-source core banking alternatives: we searched for well-maintained alternatives and found none beyond the Fineract and Mifos family. Proprietary cores (for example Mambu or Thought Machine) were out of scope for this reuse study (UNKNOWN on their current terms). ##### 2.3 Payments | Project | What it does | Licence (SPDX) | Version / last commit | Fit | Source | |---|---|---|---|---|---| | **Hyperswitch** (Juspay) | Payments orchestrator: routes payments to many processors, retries, vaulting | `Apache-2.0` (plus NOTICE file) | v1.128.0 / 2026-10-31 | ★★★ The connector to card and APM processors, and a natural home for ACP/AP2 adapters | [repo](https://github.com/juspay/hyperswitch), [LICENSE](https://github.com/juspay/hyperswitch/blob/main/LICENSE), accessed 2026-10-31 | | **Mojaloop** | Open-source instant inclusive payment *scheme* (switch) | `Apache-2.0` | central-ledger v20.3.1 / 2026-10-22 | ★ Only if EDGE operates a scheme | [central-ledger](https://github.com/mojaloop/central-ledger), accessed 2026-10-31 | | **Moov ACH** | Go library for US ACH files | `Apache-2.0` | v1.67.0 / 2026-10-28 | ★ US-only rail | [repo](https://github.com/moov-io/ach), accessed 2026-10-31 | | **Moov iso20022** | Go ISO 20022 messages | `Apache-2.0` | / **2023-01-12** | ★ **Stale**: no commit in over 3.5 years | [repo](https://github.com/moov-io/iso20022), accessed 2026-10-31 | | **Prowide ISO 20022** | Java model of all ISO 20022 messages (pain/pacs/camt) | `Apache-2.0` | highest tag SRU2026-10.4.3 / 2026-10-28 | ★★★ Maintained. Use it for pain.001, pacs.008 and camt.053 | [repo](https://github.com/prowide/prowide-iso20022), accessed 2026-10-31 | | **ACP / AP2 reference code** | Agent commerce flows | `Apache-2.0` both | ACP spec 2026-04-17 (beta); AP2 v0.2.0 | ★★ Implement as adapters. Both are pre-1.0 and will change | Section 1.3 sources | ##### 2.4 Open banking and API security | Item | Type | Licence / status | Fit | Source | |---|---|---|---|---| | UK Open Banking Read/Write API v4.0.x (OpenAPI files) | Spec | No LICENSE file in repo (**UNKNOWN**) | ★★★ For UK account information and payment initiation | [repo](https://github.com/OpenBankingUK/read-write-api-specs), accessed 2026-10-31 | | FAPI 2.0 Security Profile and Message Signing | Spec | OIDF Final (published) | ★★★ Use as the security profile for EDGE's own APIs, including agent-facing ones | Section 1.4 sources | | OpenID AuthZEN Authorization API 1.0 | Spec | OIDF Final (published) | ★★★ Standard interface to the policy engine | [OIDF](https://openid.net/authorization-api-1-0-final-specification-approved/), accessed 2026-10-31 | | ISO 20022 | Messaging standard | published (ISO). Library: Prowide `Apache-2.0` | ★★★ | Section 1.4 sources | | BIAN v14 | Reference model | published (BIAN). Licence terms of the BIAN deliverables not verified (**UNKNOWN**) | ★★ Vocabulary for service boundaries | [BIAN v14 notes](https://bian.org/wp-content/uploads/2026/02/BIAN-v14.0-Release-Notes-v1.0_-Final-Version.pdf), accessed 2026-10-31 | ##### 2.5 KYC/KYB and identity verification; identity servers | Project | What it does | Licence (SPDX) | Version / last commit | Fit | Source | |---|---|---|---|---|---| | **Ballerine** | KYC/KYB onboarding and risk workflows, case management | Mixed. Default **`Elastic-2.0`**, with some directories `MIT` ("Defaults to Elastic License 2.0") | workflow-service tags v0.6.7xx / **last commit 2025-10-14** | ★ **Licence flag** and **dormant** (no default-branch commit for over 12 months) | [repo](https://github.com/ballerine-io/ballerine), [LICENSE](https://github.com/ballerine-io/ballerine/blob/main/LICENSE), accessed 2026-10-31 | | Other open-source KYC/KYB engines | n/a | n/a | n/a | **UNKNOWN**: searched for maintained open-source KYC/KYB orchestration besides Ballerine; none verified. Document and liveness checks are generally bought from vendors | n/a | | **Keycloak** | Identity and access management server (OIDC, OAuth 2, SAML), CNCF incubating | `Apache-2.0` | 26.8.1 / 2026-10-30 | ★★★ Human and organisation identities. Agent clients can be modelled as OAuth clients with token exchange | [repo](https://github.com/keycloak/keycloak), [CNCF](https://www.cncf.io/projects/), accessed 2026-10-31 | | **Ory Hydra / Kratos / Keto / Oathkeeper** | OAuth2/OIDC server (Hydra), identity (Kratos), Zanzibar-style permissions (Keto), proxy (Oathkeeper) | `Apache-2.0` (all four) | Hydra and Kratos v26.2.0 / last commits 2026-07-29 (Oathkeeper 2026-07-27) | ★★★ Hydra is OpenID Certified (CLAIM, from README). Modular | [hydra](https://github.com/ory/hydra), [kratos](https://github.com/ory/kratos), [keto](https://github.com/ory/keto), [oathkeeper](https://github.com/ory/oathkeeper), accessed 2026-10-31 | | **ZITADEL** | Multi-tenant identity platform | **`AGPL-3.0-only`** (exceptions: `proto/`, `apps/docs/` are `Apache-2.0`; `apps/login/` and client packages are `MIT`). Commercial licence available | v4.20.0 / 2026-10-30 | ★ **Licence flag**: AGPL network clause. Usable if unmodified and isolated, but get legal review | [LICENSING.md](https://github.com/zitadel/zitadel/blob/main/LICENSING.md), accessed 2026-10-31 | ##### 2.6 Fraud, AML and transaction monitoring; sanctions data | Project / data | What it does | Licence (SPDX) | Version / last commit | Fit | Source | |---|---|---|---|---|---| | **Marble** (backend and frontend) | Rule-based transaction monitoring, AML screening and case management | **`Elastic-2.0`** (all three repos: marble, marble-backend, marble-frontend) | v1.12.0 / 2026-10-29 | ★★ Internal self-hosting is allowed. **Licence flag:** Elastic-2.0 forbids offering it to third parties as a managed service and forbids circumventing licence keys. Enterprise features are paid (CLAIM) | [marble](https://github.com/checkmarble/marble), [marble-backend](https://github.com/checkmarble/marble-backend), accessed 2026-10-31 | | **Jube** | Real-time AML and fraud monitoring with ML, rules, sanctions and case management. Now marketed with an agentic assistant ("Jooby") | **`AGPL-3.0`** (README says "AGPLv3"; -only vs -or-later **not stated**) | no tags / 2026-10-25 | ★ **Licence flag** (network copyleft) | [repo](https://github.com/jube-home/jube), accessed 2026-10-31 | | **Tazama** | Open-source real-time transaction monitoring built on ISO 20022 messages (pain.001, pacs.008, pacs.002) | `Apache-2.0` (SPDX header and package.json in `tms-service`) | / 2026-07-25 | ★★ Permissive and ISO 20022-native. Slower cadence | [tms-service](https://github.com/tazama-lf/tms-service), accessed 2026-10-31 | | **Moov Watchman** | Sanctions screening engine. Downloads OFAC, EU, UK and UN lists and scores matches | `Apache-2.0` | v1.2.2 / 2026-10-31 | ★★★ Permissive screening engine. Project claims precision 0.99 at Jaro-Winkler 0.80 on its benchmark (CLAIM) | [repo](https://github.com/moov-io/watchman), accessed 2026-10-31 | | **OpenSanctions** | Aggregated sanctions and PEP data plus the `zavod` crawler | Code `MIT`. **Data `CC-BY-NC-4.0`**. Commercial use needs a paid licence (screening API, screening licence or reseller licence) | / 2026-10-30 | ★ for data unless licensed. **Licence flag** | [README](https://github.com/opensanctions/opensanctions), [licensing](https://www.opensanctions.org/licensing/), accessed 2026-10-31 | ###### 2.6.1 Official sanctions sources (free, primary) | List | Publisher | Formats | Reuse terms | Source | |---|---|---|---|---| | **UK Sanctions List** (the *only* UK source since the OFSI Consolidated List closed on **28 Jan 2026**) | FCDO | ODT, ODS, XML (schema `SanctionsListSchema-4.33.3.xsd`), HTML, text, CSV, PDF. Page last updated 30 Oct 2026 | Open Government Licence v3.0 (`OGL-UK-3.0`) | [GOV.UK](https://www.gov.uk/government/publications/the-uk-sanctions-list), accessed 2026-10-31 (OBSERVED) | | **OFAC SDN and Consolidated (non-SDN) lists** | US Treasury OFAC | Sanctions List Service: XML (standard and advanced), CSV and others, plus delta files | Not stated on the list page (**UNKNOWN**; US federal works are generally public domain, not verified here) | [OFAC lists](https://ofac.treasury.gov/ofac-sanctions-lists), [SLS](https://sanctionslist.ofac.treas.gov/Home/index.html), accessed 2026-10-31 | | **UN Security Council Consolidated List** | UN | XML, PDF, HTML. XML at `scsanctions.un.org/resources/xml/en/consolidated.xml` per third-party docs. Our request returned HTTP 302 (redirect), so content was **not verified** | UNKNOWN | [UN page](https://www.un.org/sc/suborg/en/sanctions/un-sc-consolidated-list.html), accessed 2026-10-31 | | **EU Financial Sanctions Files (FSF)** | European Commission | XML, CSV, PDF. Download URLs carry a user token after login | UNKNOWN. "Only the information published in the Official Journal of the EU is deemed authentic" | [FSF user manual](https://circabc.europa.eu/d/a/workspace/SpacesStore/9cf02a9e-7604-4178-a21d-d0b6bdcb4114/MA_FSF_EN.pdf), accessed 2026-10-31 | **UK-specific action (OBSERVED):** systems that used the "OFSI Group ID" must switch to the UK Sanctions List "Unique ID" and "Sanction Type" fields ([Financial Institutions News](https://www.financialinstitutionsnews.com/2026/01/28/uk-sanctions-list-becomes-sole-source-of-uk-sanctions-designations/amp/), accessed 2026-10-31; CLAIM, secondary reporting of the GOV.UK notice). ##### 2.7 Policy and authorisation engines | Project | What it does | Licence (SPDX) | Version / last commit | CNCF status | Fit | Source | |---|---|---|---|---|---|---| | **Open Policy Agent (OPA)** | General policy engine using the Rego language | `Apache-2.0` | v1.22.0 / 2026-10-30 | **Graduated** | ★★★ Mandate and limit rules across services. Mature | [repo](https://github.com/open-policy-agent/opa), [CNCF](https://www.cncf.io/projects/), accessed 2026-10-31 | | **Cedar** | Authorisation policy language designed for automated reasoning and analysis | `Apache-2.0` | v4.14.0 / 2026-10-30 | Sandbox per secondary source (not on CNCF graduated/incubating page: OBSERVED) | ★★★ Best fit for **provable mandates** ("this agent can never pay > £X"), because its analysability is a design goal (CLAIM, README) | [repo](https://github.com/cedar-policy/cedar), accessed 2026-10-31; [palark](https://palark.com/blog/kubecon-na-2024-cncf-projects/), accessed 2026-10-31 (CLAIM) | | **OpenFGA** | Relationship-based (Zanzibar-style) fine-grained authorisation | `Apache-2.0` | v1.22.1 / 2026-10-28 | **Incubating** | ★★★ Models "agent X acts for org Y with role Z" relationships and delegation graphs | [repo](https://github.com/openfga/openfga), [CNCF](https://www.cncf.io/projects/), accessed 2026-10-31 | Suggested pattern (assessment): OpenFGA answers *who is related to whom*, Cedar or OPA answers *is this action within the mandate*, and AuthZEN 1.0 is the uniform call interface. ##### 2.8 Workflow and durability | Project | What it does | Licence (SPDX) | Version / last commit | Fit | Source | |---|---|---|---|---|---| | **Temporal** | Durable execution: long-running workflows that survive crashes, with retries, timers and full event history | `MIT` | v1.32.1 / 2026-10-30 | ★★★ Saga-style compensation (undo) and replayable workflow history match **recovery** and **re-runnable audit** | [repo](https://github.com/temporalio/temporal), [LICENSE](https://github.com/temporalio/temporal/blob/main/LICENSE), accessed 2026-10-31 | | **Cadence** | Temporal's predecessor (originated at Uber) | `Apache-2.0` | v1.4.5 / 2026-10-30 | ★★ Viable, smaller ecosystem | [repo](https://github.com/cadence-workflow/cadence), accessed 2026-10-31 | | **DBOS Transact (Python)** | Durable workflows on Postgres as a library | `MIT` | / 2026-10-30 | ★★ Lighter alternative | [repo](https://github.com/dbos-inc/dbos-transact-py), accessed 2026-10-31 | | **Restate** | Durable execution server | **`BUSL-1.1`**. Additional use grant forbids a "Public Restate Platform Service". Each release converts to `Apache-2.0` 4 years after release | v1.8.1 / 2026-10-31 | ★★ Internal production use is explicitly permitted. **Licence flag:** not open source today | [LICENSE](https://github.com/restatedev/restate/blob/main/LICENSE), accessed 2026-10-31 | | **Inngest** | Event-driven durable functions | **`SSPL-1.0`** with an "Apache 2.0 Future License" | / n/a | ★ **Licence flag:** SSPL is not OSI-approved and has broad service-provider obligations | [LICENSE.md](https://github.com/inngest/inngest/blob/main/LICENSE.md), accessed 2026-10-31 | ##### 2.9 Journey and diagram tooling (Mermaid) **Mermaid:** `MIT` licence; highest tag `mermaid@12.2.0`; last commit 2026-10-30 ([repo](https://github.com/mermaid-js/mermaid), [LICENSE](https://github.com/mermaid-js/mermaid/blob/master/LICENSE), accessed 2026-10-31). Diagram types listed in the docs ([Mermaid intro](https://mermaid.js.org/intro/), accessed 2026-10-31): | EDGE need | Mermaid type (keyword) | Status | Why | |---|---|---|---| | Customer journeys (human, organisation or agent) | User Journey (`journey`) | Stable | Steps with actors and satisfaction scores per stage | | Journeys across many actors (person, agent, EDGE, bank) | Flowchart (`flowchart`) with subgraphs; Swimlanes | Flowchart stable; **Swimlanes is Beta** | Subgraph lanes work today. Swimlanes is newer | | API endpoint interactions | Sequence Diagram (`sequenceDiagram`) | Stable | The MCP spec itself uses `sequenceDiagram` for its OAuth flow ([MCP](https://modelcontextprotocol.io/specification/2026-07-28/basic/authorization), accessed 2026-10-31) | | Lifecycles (mandate: draft → rehearsed → approved → executed → reversed; agent: registered → verified → suspended → revoked) | State Diagram (`stateDiagram`) | Stable | Shows states and allowed transitions | | Data model (ledger accounts, mandates) | Entity Relationship (`erDiagram`) | Listed under examples as experimental | | | System context | C4 (`c4`), Architecture (`architecture`) | Architecture is **Beta** | | | Event-sourced flows | Event Modeling (`eventmodeling`) | Listed | Fits Temporal and ledger event histories | ##### 2.10 Agent assurance tooling to reuse | Tool | Use in EDGE | Licence (SPDX) | Source | |---|---|---|---| | UK AISI **Inspect** | Score racetrack runs; publish evaluation logs | `MIT` | [repo](https://github.com/UKGovernmentBEIS/inspect_ai), accessed 2026-10-31 | | MITRE **ATLAS** data | Threat taxonomy for racetrack adversarial scenarios | `Apache-2.0` | [atlas-data](https://github.com/mitre-atlas/atlas-data), accessed 2026-10-31 | | OWASP Agentic Top 10 | Control checklist | `CC-BY-SA-4.0` (site default) | [OWASP](https://genai.owasp.org/resource/owasp-top-10-for-agentic-applications-for-2026/), accessed 2026-10-31 | | Cloudflare **web-bot-auth** | Sign and verify agent HTTP requests | `Apache-2.0` | [repo](https://github.com/cloudflare/web-bot-auth), accessed 2026-10-31 | ##### 2.11 Licence flags: what blocks or constrains commercial use | Component | SPDX | Effect for EDGE Finance (assessment) | Source | |---|---|---|---| | Visa Trusted Agent Protocol code | none (proprietary Visa terms) | Using the code binds EDGE to Visa Developer terms. Implement from the public Web Bot Auth draft instead, or accept Visa's terms deliberately | [LICENSE.md](https://github.com/visa/trusted-agent-protocol/blob/main/LICENSE.md), accessed 2026-10-31 | | OpenSanctions **data** | `CC-BY-NC-4.0` | **Blocks commercial use** without a paid licence. Use the official lists (UK, OFAC, UN, EU) directly, or buy a licence | [licensing](https://www.opensanctions.org/licensing/), accessed 2026-10-31 | | Inngest | `SSPL-1.0` | Avoid. Service-provider clause | [LICENSE.md](https://github.com/inngest/inngest/blob/main/LICENSE.md), accessed 2026-10-31 | | Restate | `BUSL-1.1` | Internal use OK. Cannot offer Restate's APIs to third parties as a platform | [LICENSE](https://github.com/restatedev/restate/blob/main/LICENSE), accessed 2026-10-31 | | Marble | `Elastic-2.0` | Self-host internally OK. Cannot provide it as a managed service to others | [LICENSE](https://github.com/checkmarble/marble/blob/main/LICENSE), accessed 2026-10-31 | | Ballerine | `Elastic-2.0` (default) + `MIT` parts | Same as Marble, and the project is dormant | [LICENSE](https://github.com/ballerine-io/ballerine/blob/main/LICENSE), accessed 2026-10-31 | | ZITADEL | `AGPL-3.0-only` (+ Apache/MIT exceptions) | Network copyleft. Modified versions served over a network must be source-available. Commercial licence offered | [LICENSING.md](https://github.com/zitadel/zitadel/blob/main/LICENSING.md), accessed 2026-10-31 | | Jube | `AGPL-3.0` (variant unstated) | Same as ZITADEL | [repo](https://github.com/jube-home/jube), accessed 2026-10-31 | | Lago | `AGPL-3.0` (variant unstated) | Same as ZITADEL | [repo](https://github.com/getlago/lago), accessed 2026-10-31 | | Formance Ledger v3 (`main`) | none found | **No licence = no permission** until Formance publishes one. Use v2.x (`MIT`) | [repo](https://github.com/formancehq/ledger), accessed 2026-10-31 | | Formance `ee/` directories | Enterprise licence | Paid | [LICENSE](https://github.com/formancehq/stack/blob/main/LICENSE), accessed 2026-10-31 | | OWASP documents | `CC-BY-SA-4.0` | Derived *documents* must carry the same licence. Does not affect code | [OWASP](https://genai.owasp.org/resource/owasp-top-10-for-agentic-applications-for-2026/), accessed 2026-10-31 | | UK Open Banking spec repo, BIAN deliverables, UN and EU sanctions files | UNKNOWN | Check terms before redistribution | Sections 2.4 and 2.6.1 | Permissive and safe (`Apache-2.0`, `MIT` or `MPL-2.0`, OBSERVED from LICENSE files): TigerBeetle, Blnk, Fineract, Mifos web-app (MPL-2.0), Kill Bill, Hyperswitch, Mojaloop, Moov ACH, Moov Watchman, Prowide ISO 20022, Tazama, Keycloak, Ory, OPA, Cedar, OpenFGA, SPIRE, Temporal, Cadence, DBOS, Mermaid, AP2, ACP, A2A, MITRE ATLAS data, AISI Inspect, Cloudflare web-bot-auth. --- ##### 3. Suggested reuse stack (assessment) | EDGE capability | Reuse | Build (EDGE-specific) | |---|---|---| | Accountable agent identity (G1) | Keycloak or Ory Hydra (OAuth/OIDC), SPIRE (workload identity), VC 2.0 credentials, Web Bot Auth signatures | KYA binding record: agent → accountable person/org → verified KYC/KYB → revocation status | | Scoped mandates and limits (G2, G3) | OpenFGA (relationships) + Cedar (mandate rules) behind AuthZEN 1.0; import/export AP2 Payment Mandates and ACP Allowances | Cross-channel budget counter held in the ledger | | Revocation (G4) | Token expiry (OAuth), FGA tuple deletion | Instant kill-switch propagated to all channels plus a signed revocation receipt | | Rehearsal (G5) | TigerBeetle pending transfers; Temporal workflows run against a racetrack namespace | Rehearsal report with predicted effects, signed and linked to the later commit | | Ledger and money movement | TigerBeetle (or Blnk); Fineract for deposit and loan products; Hyperswitch; Prowide ISO 20022 | Agent-aware account types | | Recovery / undo (G8) | Temporal saga compensations; TigerBeetle void of pending transfers | Undo policy per rail (cards vs Faster Payments) | | Monitoring and sanctions | Moov Watchman + official UK, OFAC, UN and EU lists; Tazama or Marble (internal) | Agent-behaviour typologies (for example mandate probing) | | Re-runnable audit (G10) | Temporal event history, FAPI 2.0 Message Signing, AP2 mandate receipts | Evidence pack format a regulator can replay | | Assurance (G6, G7) | AISI Inspect, MITRE ATLAS, OWASP Agentic Top 10, ISO/IEC 42001 management system | Published per-agent error-rate statistics from the racetrack | | Diagrams | Mermaid `journey`, `sequenceDiagram`, `stateDiagram`, `flowchart` | n/a | --- ##### 4. Things we could not confirm (UNKNOWN), and what was searched | Item | What we searched | Result | |---|---|---| | NIST AI Agent Interoperability Profile release | NIST AI RMF page; web search "NIST AI RMF update 2026 ... agentic" | Only secondary mention (CSA). No NIST page found | | EU harmonised AI standards cited in the Official Journal | Web search "EU AI Act harmonised standards CEN-CENELEC JTC 21 status 2026" | Secondary sources only. No OJ citation confirmed | | HMT AI Adoption Plan full recommendation text (KYA wording) | GOV.UK landing page (fetched); guessed HTML sub-page (404) | KYA wording known only from commentary | | FCA primary text on agentic payments | Web search "FCA agentic AI payments statement 2026" | Secondary (law firms, trade press) only | | UK AISI agent evaluation incident | Web search "AI Security Institute agents evaluation 2026 aisi.gov.uk" | Secondary only; no aisi.gov.uk page reached | | A2A move under AAIF (Aug 2026) | Web search on A2A Linux Foundation governance | Single blog source | | Formance Ledger v3 licence | Shallow and full clone of `formancehq/ledger`; README grep for "licen" | No LICENSE file or statement on `main` | | UK Open Banking spec licence; BIAN licence; UN and EU sanctions reuse terms | Repo file listing; pages listed above | Not stated in what we read | | Open-source KYC/KYB alternatives to Ballerine | Web and repo search | None verified | | Jube AGPL variant (-only vs -or-later) | README "Licence" section; LICENSE header | Not stated | | Cedar CNCF tier | CNCF projects page (graduated/incubating only) | Sandbox per secondary source | | MCP core spec support for dry-run or rehearsal | Not searched beyond the authorisation page | UNKNOWN | ### research/finance/COMPETITORS.md `research/finance/COMPETITORS.md` · sha256 `3bdd5116b6688438b592c232a8d68a61649078a7cefec5fa7f2c6f8462a55abf` #### EDGE Finance: Competitor and Positioning Report - **Product:** EDGE Finance (part of JudeAI | EDGE ID) - **Report date:** 2026-10-31 - **Scope:** UK digital banks, crypto platforms, AI/intelligence vendors in finance, agent payment rails, and agent identity/tooling providers. - **Audience:** Product, engineering, compliance and leadership. It is written so that a new engineer with no finance background can follow it. --- ##### 0. How to read this report ###### 0.1 Evidence labels Every factual line carries one of these labels: | Label | Meaning | |---|---| | **OBSERVED** | I read this directly on a primary source, such as the company's own website, docs, GitHub repository, or a regulator page. | | **OBSERVED (excerpt)** | The source is primary (company or regulator domain), but I could only see it as a search-engine excerpt because the page blocked direct fetching (HTTP 403). Treat it as slightly weaker than OBSERVED. | | **CLAIM** | Vendor marketing, a vendor's own performance number, or a press/third-party report that I could not confirm on a primary source. **Never treat a CLAIM as fact.** | | **UNKNOWN** | I searched and could not establish it, or sources conflict. | A vendor's own statistic stays a **CLAIM** even when it appears on the vendor's own page. For example, "Coinbase says x402 has processed 100 million payments" is OBSERVED as a statement and CLAIM as a number. All sources were accessed on 2026-10-31. Search results were US-indexed. Several primary pages (Revolut, OpenAI, Oracle, Coinbase blog, Hawk, Persona, Mastercard press archive) returned HTTP 403 to my fetcher, which is why some lines rely on excerpts or press coverage. ###### 0.2 Plain-English glossary (acronyms defined once here and on first use below) | Term | Plain English | |---|---| | AI agent | Software, usually driven by a large language model, that can take actions such as calling APIs, paying or sending messages, not only chat. | | LLM (large language model) | The text-generating AI model behind chat assistants (e.g., GPT, Claude, Gemini). | | API (application programming interface) | A way for one program to call another. | | MCP (Model Context Protocol) | An open standard for exposing tools and data to AI agents. A "bank MCP server" lets an agent call bank functions. | | BaaS (banking-as-a-service) | Renting a licensed bank's infrastructure, or its core software, so you don't have to build or license your own. | | KYC / KYB | Know Your Customer / Know Your Business: identity checks on people and companies. | | KYA (Know Your Agent) | The emerging idea of identity checks for AI agents: who built the agent, who it acts for, and what it may do. There is no single standard yet. | | AML | Anti-money laundering. | | SAR | Suspicious Activity Report: a report a bank files with authorities. | | FCA / PRA | Financial Conduct Authority / Prudential Regulation Authority: the two UK financial regulators. | | EMI (electronic money institution) | A UK licence to hold e-money and make payments. It is **not** a bank licence: an EMI cannot lend customer deposits. | | FSCS | Financial Services Compensation Scheme: UK deposit protection for bank deposits. | | Mobilisation | A restricted start-up phase for a newly licensed UK bank, with caps on deposits. | | Stablecoin / USDC | A crypto token designed to hold a steady value, e.g., USDC pegged to the US dollar. | | x402 | A protocol that reuses the web's "HTTP 402 Payment Required" status code so a website or API can ask for, and receive, a payment within the same request. | | Mandate | A signed, checkable instruction that says what an agent is allowed to do, e.g., "buy this cart up to £50". | | Tokenisation | Replacing a real card number with a substitute token that only works in a limited context. | | Passkey | A phishing-resistant login using device biometrics instead of a password. | | Human-in-the-loop | A person must approve before certain actions happen. | | Glass box | Every decision and action can be inspected: inputs, reasoning, rules applied, who approved. The opposite of a "black box". | | GA / preview | Generally available to all customers / early-access release that may change. | --- ##### 1. Executive summary (the ten things that matter most) 1. **Every UK bank with an AI assistant treats the agent as a feature of a human's account, not as a customer.** Starling Assistant (March 2026), Revolut AIR (April 2026) and Zopa Ask (September 2026) all act *for* a human inside that human's app. None I found lets a third-party AI agent hold its own identity, own sub-ledger and own mandate at a UK-regulated bank. (OBSERVED for Starling; CLAIM for Revolut and Zopa; see Section 2.) 2. **The card networks already gatekeep agent payments.** Visa's Trusted Agent Protocol recognises "Visa-approved agents", and Visa Agentic Ready is live with UK issuers including Barclays, HSBC UK and Nationwide. Mastercard says all its European issuers are enabled for Agent Pay. (OBSERVED, Section 5.) 3. **Agent wallets with owner-set spending limits are now a commodity in crypto and fintech.** Coinbase Agentic Wallet, Stripe's Link agent wallet and Privy agent wallets all exist today. (OBSERVED, Sections 3 and 5.) 4. **Mandates (signed, checkable permissions) are being standardised outside banks.** Google's AP2 v0.2 uses verifiable-credential mandates and has been donated to the FIDO Alliance. (OBSERVED.) A bank that natively issues and enforces AP2-style mandates would be early. 5. **Financial-crime vendors have moved to agentic investigation with audit logs**, including Unit21, Sardine, Feedzai (Farol), ComplyAdvantage, Oracle and Quantexa. But their headline error-rate numbers are "up to" marketing figures. (CLAIM.) Published, measured error rates are rare. That is an opening for "error-rate-controlled intelligence". 6. **Frontier AI labs sell to banks' staff, not to banks' customers' agents.** OpenAI's ChatGPT for Financial Services, Anthropic's ten finance agent templates and Google's Gemini Enterprise for Financial Services target analysts and back offices. Anthropic's templates explicitly do **not** execute transactions. (OBSERVED for Anthropic and Google; CLAIM for OpenAI.) 7. **The only "bank for agents" contender found is US-based and unlicensed so far.** Catena Labs filed for a US national trust bank charter in May 2026; I found no approval. (CLAIM.) 8. **The UK regulator is actively testing agentic payments.** The FCA's AI Live Testing cohort 2 includes GoCardless, Barclays, Lloyds and others. Testing ends at end-2026 and an evaluation report is due Q1 2027. (OBSERVED.) Liability rules for agent payments look unsettled. (CLAIM.) 9. **Financial-crime failure is existential for UK challengers.** The FCA fined Starling about £29m (2024) and Monzo about £21m (2025) for controls that did not keep pace with growth. (OBSERVED (excerpt).) An agent-native bank will be scrutinised harder, not less. 10. **Our honest lane:** agents as first-class, identity-bound customer records under a legally responsible principal. That means mandates, limits, kill-switch and recovery, error-rate-budgeted AI decisions, and a customer-visible glass-box audit trail, all interoperating with Visa/Mastercard/AP2/x402 rather than replacing them. Whether an AI agent can itself be a legal account holder in the UK is **UNKNOWN** and needs legal advice (Section 8). --- ##### 2. UK digital banks and fintechs ###### 2.1 Summary table | Firm | Licence / type | AI assistant for customers | Lets external agents act? | Agent as own customer? | |---|---|---|---|---| | Starling | UK bank (FCA+PRA) | Yes: Starling Assistant, Mar 2026 | Not found | No (not found) | | Engine by Starling | SaaS core-banking platform sold to other banks | N/A | N/A | N/A | | Monzo | UK bank | Ops Agent in customer chat (Jun 2026) | Only unofficial third-party MCP servers | Not found | | Revolut | UK bank (full licence Mar 2026) | Yes: AIR, Apr 2026 | Plans to let AIR shop online (CLAIM) | Not found | | Wise | UK EMI (not a bank) | Not found | Community MCP servers only | Not found | | Chase UK | UK bank (JPMorgan) | Not found | Not found | Not found | | Zopa | UK bank | Yes: Ask, Sep 2026 | Not found | Not found | | Kroo | UK bank | Not found | Not found | Not found | | Allica | UK SME bank | Internal AI agents for lending (CLAIM) | Not found | Not found | | bunq (EU) | Dutch bank | Finn, multi-agent (CLAIM) | Took part in Mastercard agentic tests | Not found | | GoCardless | UK payments firm | N/A | Agentic Direct Debit checkout (CLAIM) | No: human donor is the payer | Sources for each row are given in the subsections below. ###### 2.2 Starling Bank (and Engine by Starling) **What they offer today** - OBSERVED: Starling Bank Limited is "regulated by the Financial Conduct Authority and the Prudential Regulation Authority" [source](https://www.starlingbank.com/investors/2026/annual-report-2026/), accessed 2026-10-31. - OBSERVED: The 2026 annual report page shows group revenue of £887m (2025: £940m) and 6.2m group platform accounts [source](https://www.starlingbank.com/investors/2026/annual-report-2026/), accessed 2026-10-31. **AI and agents** - OBSERVED: Starling announced "Starling Assistant" on 20 March 2026. It is a voice/text assistant that "carr[ies] out banking tasks on the customer's behalf": savings goals with automatic transfers, budget Spaces, organising bill payments, and spending and direct-debit questions [source](https://www.starlingbank.com/news/starling-launches-agentic-ai-assistant/), accessed 2026-10-31. - OBSERVED: It is built on Google Gemini on Google Cloud. "All data remains securely within Starling's Google Cloud environment and is not used for training purposes." [source](https://www.starlingbank.com/news/starling-launches-agentic-ai-assistant/), accessed 2026-10-31. - OBSERVED: It is opt-in: "customers will have to opt in to use the AI assistant and are under no obligation to do so." It launched for personal current accounts, with business and joint accounts "soon" [source](https://www.starlingbank.com/news/starling-launches-agentic-ai-assistant/), accessed 2026-10-31. - CLAIM: Starling calls it "the UK's first agentic AI financial companion" [source](https://www.starlingbank.com/investors/2026/annual-report-2026/), accessed 2026-10-31. - UNKNOWN: The press release does not say whether each action needs customer confirmation, what transaction limits apply, or how errors are handled [source](https://www.starlingbank.com/news/starling-launches-agentic-ai-assistant/), accessed 2026-10-31. **Engine by Starling (relevant if EDGE needs a core platform)** - OBSERVED: The 2026 annual report page shows Engine committed ARR (annual recurring revenue) of £70m, a target of £100m "in the next 12 months", and logos of SBS Bank and Tangerine as partners [source](https://www.starlingbank.com/investors/2026/annual-report-2026/), accessed 2026-10-31. - OBSERVED (excerpt): Starling's FY25 results named Salt Bank (Romania) and AMP Bank (Australia) as Engine's first clients [source](https://starlingbank.com/news/starling-delivers-fifth-year-of-profitability-and-accelerates-global-growth-strategy), accessed 2026-10-31. - CLAIM: Starling is reportedly seeking a US bank client in the "$5bn-50bn asset range" with Deloitte and PwC advisers [source](https://www.retailbankerinternational.com/news/starling-us-expansion-in-house-tech-sale/), accessed 2026-10-31. - Pricing/licence: UNKNOWN. No public Engine pricing found. A directory describes it as a SaaS proposition with no published pricing [source](https://europeanpurpose.com/tool/engine-by-starling), accessed 2026-10-31. Searched "Starling Engine pricing model software licence fee per account". **Gaps and weaknesses** - OBSERVED (excerpt): The FCA fined Starling £28,959,426 in October 2024 for financial-crime control failings. These included screening customers against only part of the sanctions list since 2017, and opening over 54,000 accounts for high-risk customers in breach of a restriction [source](https://www.fca.org.uk/news/press-releases/fca-fines-starling-bank-failings-financial-crime-systems-and-controls), accessed 2026-10-31. - Gap (analysis): The assistant serves the human account holder only. No evidence was found of APIs that let a customer's own external agent act under a scoped mandate. ###### 2.3 Monzo **What they offer today** - OBSERVED (excerpt): Monzo's paid plans are Extra £3/month, Perks £9 (up from £7), Max £19 (up from £17) and Max Family £23 (up from £22). New prices apply from 13 August 2026 for new customers [source](https://monzo.com/help/monzo-max/monzo-plan-price-change), accessed 2026-10-31. - OBSERVED (excerpt): Monzo is named as a Wise Platform partner in Wise's FY26 results [source](https://www.investegate.co.uk/announcement/rns/wise-class-a--wise/wise-fy26-results/9637901), accessed 2026-10-31. **AI and agents** - OBSERVED: On 4 June 2026 Monzo described its "Monzo Ops Agent", an LLM-based agent for customer support and fraud/financial-crime reporting workflows. It calls tools to act on accounts, such as ordering replacement cards and finding missing refunds [source](https://monzo.com/blog/engineering-the-future-of-customer-operations-the-monzo-ops-agent), accessed 2026-10-31. - OBSERVED: Monzo evaluates it with a "golden set" of 100 expert-approved conversations, layered component, answer and end-to-end evaluations, and simulated users and environments. At launch it had "a human review every message before it went out", then sampled QA reduced over time [source](https://monzo.com/blog/engineering-the-future-of-customer-operations-the-monzo-ops-agent), accessed 2026-10-31. - OBSERVED: Guardrails detect hallucination or non-compliant content and hand the chat to a human specialist [source](https://monzo.com/blog/engineering-the-future-of-customer-operations-the-monzo-ops-agent), accessed 2026-10-31. - UNKNOWN: A Monzo post titled "Building Agent Chip" (13 August 2026, an internal AI coding assistant) appeared in search results, but I could not open it or find its URL. - CLAIM: An unofficial, third-party Monzo MCP server exists for reading balances, pots and transactions. It is not a Monzo product [source](https://www.flowhunt.io/zh/集成/monzo/), accessed 2026-10-31. **Gaps and weaknesses** - OBSERVED (excerpt): The FCA fined Monzo £21,091,300 (July 2025) because financial-crime controls did not keep pace with growth. Monzo also failed to comply with a restriction on onboarding high-risk customers [source](https://www.fca.org.uk/news/press-releases/fca-fines-monzo-21m-failings-financial-crime-controls), accessed 2026-10-31. - Gap (analysis): The agent is Monzo's own staff/support agent. No customer-owned agent access was found. ###### 2.4 Revolut (banking) **What they offer today** - CLAIM: The PRA lifted restrictions on Revolut Bank UK Ltd on 11 March 2026, ending mobilisation (begun July 2024) and making it a fully licensed UK bank [source](https://tech.eu/2026/03/11/revolut-wins-full-uk-banking-licence-as-finally-exits-mobilisation-phase/), accessed 2026-10-31. - OBSERVED (excerpt): Revolut's own blog explains that it is now a fully licensed UK bank [source](https://www.revolut.com/blog/post/is-revolut-a-fully-licensed-bank-in-the-uk/), accessed 2026-10-31. **AI and agents** - OBSERVED (excerpt): Revolut announced AIR ("AI by Revolut") as an assistant "deeply embedded in Revolut's ecosystem" [source](https://www.revolut.com/news/revolut_enters_new_era_of_money_intelligence_with_launch_of_ai_assistant/), accessed 2026-10-31. The page returned 403 to direct fetch. - CLAIM: AIR launched on 9 April 2026 to UK customers (about 13 million), with EEA rollout to follow [source](https://www.uktech.news/ai/revolut-launches-in-app-ai-assistant-20260409), accessed 2026-10-31. - CLAIM: AIR handles spending insights, card controls (e.g., freezing a lost card), single-use virtual cards, investment tracking and subscriptions [source](https://www.fintechweekly.com/news/revolut-air-ai-assistant-uk-customers-launch-2026), accessed 2026-10-31. - CLAIM: A third-party UX review saw a confirmation step before a card freeze and said AIR "in most cases acts as a guide rather than completing the task directly" [source](https://craftinnovations.global/revolut-air-ux-analysis/), accessed 2026-10-31. - CLAIM: Co-founder Nik Storonsky said AIR will be able to browse sites such as Amazon and buy for the customer. This is a stated plan, not a shipped feature [source](https://www.fintechweekly.com/news/revolut-air-ai-assistant-uk-customers-launch-2026), accessed 2026-10-31. - CLAIM: Revolut is named among issuers that completed live agent-executed Visa transactions in Europe [source](https://retail-systems.com/rs/Visa_Takes_Agentic_Commerce_Live_With_European_Banks_And_Merchants.php), accessed 2026-10-31. **Pricing / gaps** - Pricing: Revolut runs tiered paid plans. I did not verify 2026 prices on a primary page (UNKNOWN). - Gap (analysis): AIR is human-facing. No public API lets outside agents act under their own identity. ###### 2.5 Wise - CLAIM: Wise is not a licensed bank. It operates in the UK under an EMI licence (FCA, 2016) and, as of September 2025, had not applied for a UK banking licence [source](https://paymentexpert.com/2025/09/02/wise-to-apply-for-uk-banking-licence/), accessed 2026-10-31. A 2026 update was not found. - OBSERVED (excerpt): Wise's FY26 results list Wise Platform partners: banks and fintechs that embed Wise's cross-border payments, including UniCredit, Raiffeisen, Capitec, Standard Chartered, Morgan Stanley, Monzo and Nubank [source](https://www.investegate.co.uk/announcement/rns/wise-class-a--wise/wise-fy26-results/9637901), accessed 2026-10-31. - AI/agents: **not found.** I searched "Wise AI agents MCP payments API 2026" and "Wise Platform 2026 AI agent announcement". Only community-built MCP servers exist, some of which can call `send_money` [source](https://mcpservers.org/servers/sergeiledvanov/mcp-wise), accessed 2026-10-31. - CLAIM: Wise published predictions of an "agentic AI shift" in business for 2026 [source](https://ecommercenews.com.au/story/wise-predicts-agentic-ai-shift-in-business-by-2026), accessed 2026-10-31. - Gap (analysis): Strong rails and a strong API, but no official agent product. Cross-border payouts are a likely partner capability for EDGE rather than a direct threat. ###### 2.6 Chase UK (JPMorgan Chase) - OBSERVED (excerpt): Chase launched in the UK in September 2021 as a digital bank [source](https://chase.co.uk/gb/en/media/chase-arrives-in-the-uk), accessed 2026-10-31. - CLAIM: The current account is fee-free and app-only. Cashback and interest terms vary between review sites and appear to have changed, so check them on chase.co.uk [source](https://www.finder.com/uk/chase-digital-bank-review), accessed 2026-10-31. - AI/agents: **not found** for Chase UK. I searched "Chase UK 2026 news AI". Results were about the JPMorgan group, not Chase UK customer features. - Gap (analysis): Big balance sheet, but no visible agent strategy in the UK product. ###### 2.7 Zopa - CLAIM: Zopa launched "Ask" on about 18 September 2026 for Biscuit current account customers. It can send money, request payments, split a bill from a receipt photo, pay an invoice from a photo, manage cards, create savings pots and warn about spending limits. Zopa calls it "agentic" and built on its "own AI platform" [source](https://itbrief.co.uk/story/zopa-launches-ai-ask-tool-for-biscuit-account-users), accessed 2026-10-31. - CLAIM: Zopa says AI resolves 65% of customer service queries and that AI users report 10% higher satisfaction [source](https://itbrief.co.uk/story/zopa-launches-ai-ask-tool-for-biscuit-account-users), accessed 2026-10-31. - CLAIM: Zopa reported 2025 pre-tax profit of £44.9m and said its current account (Biscuit, launched June 2025) was running ahead of plan [source](https://tech.eu/2026/03/19/zopa-reports-third-consecutive-year-of-profit-says-take-up-of-new-current-account-topping-expectations/), accessed 2026-10-31. - UNKNOWN: Transaction caps, approval steps and error handling for Ask are not described [source](https://itbrief.co.uk/story/zopa-launches-ai-ask-tool-for-biscuit-account-users), accessed 2026-10-31. ###### 2.8 Kroo - OBSERVED (excerpt): Kroo's current account advertises 2.65% AER (annual equivalent rate) variable interest on balances up to £500,000, with no monthly fee [source](https://www.kroo.com/current-account), accessed 2026-10-31. - OBSERVED (excerpt): Kroo's newsroom lists a lending partnership with Glenhawk dated 13 May 2026 [source](https://web.kroo.com/newsroom), accessed 2026-10-31. - AI/agents: **not found.** I searched "Kroo bank 2026 news". ###### 2.9 Allica Bank (SME, i.e., small and medium-sized enterprises) - CLAIM: Allica reported lending up 23% to £3.7bn (April 2026) [source](https://www.mortgagesolutions.co.uk/specialist-lending/2026/04/16/allica-bank-reports-23-rise-in-lending-to-3-7bn/), accessed 2026-10-31. - CLAIM: Allica says it is "deploying AI agents across its proprietary technology and data" for faster SME lending decisions and relationship-manager insight [source](https://www.allica.bank/press-releases/record-results-as-number-of-smes-choosing-allica-more-than-doubles-putting-the-uks-only-full-service-digital-bank-for-established-smes-on-course-f-1776178083380), accessed 2026-10-31. - CLAIM: It raised a $155m Series D in February 2026 to fund growth and AI [source](https://fintech.global/2026/02/25/uk-fintech-allica-bank-raises-155m-series-d/), accessed 2026-10-31. - Gap (analysis): The agents are internal credit tools. Business customers' own agents are not served. ###### 2.10 Other UK/EU players notably "agent-ready" - OBSERVED: bunq and N26 are among the 35 European banks that Mastercard says completed controlled, live agentic transactions [source](https://newsroom.mastercard.com/news/europe/en/perspectives/en/2026/europe-is-building-the-foundations-for-trusted-agentic-commerce/), accessed 2026-10-31. - CLAIM: bunq's assistant Finn uses a multi-agent "orchestrator plus sub-agents" design. bunq reportedly says Finn autonomously resolves 87% of queries [source](https://www.thestack.technology/bunq-multi-agent-ai-production-use-case/), accessed 2026-10-31. - CLAIM: GoCardless completed what it calls the UK's first live agentic account-to-account payment, an AI agent setting up a Direct Debit donation to the charity Trussell, built in the FCA's AI Live Testing programme (22 September 2026) [source](https://fintech.global/2026/09/22/gocardless-completes-uks-first-agentic-bank-payment/), accessed 2026-10-31. - OBSERVED: GoCardless, Barclays, Lloyds Banking Group (Scottish Widows), UBS, Experian and others are in FCA AI Live Testing cohort 2. Use cases include "agentic payments" [source](https://www.fca.org.uk/news/press-releases/fca-announces-second-cohort-ai-live-testing), accessed 2026-10-31. - CLAIM (US, closest "bank for agents"): Catena Labs raised $30m (May 2026) and was "accepted for filing" a US National Trust Bank charter with the OCC (Office of the Comptroller of the Currency, a US bank regulator). Its product lets humans set agent spending limits, approved recipients, balance caps and audit trails [source](https://theblock.co/post/402029/catena-labs-lands-30-million-series-a-files-for-national-trust-bank-charter-to-underpin-agentic-finance), accessed 2026-10-31. Approval: UNKNOWN. - Not found: a UK or EU challenger that opens accounts *for* AI agents as customers. Searched "bank account for AI agents challenger bank Europe launch 2026 agent-native". - CLAIM: One 2026 analysis says that where agents perform banking tasks, the person or legally formed business remains the account customer [source](https://macmyths.com/can-an-ai-agent-have-a-bank-account-in-2026/), accessed 2026-10-31. --- ##### 3. Crypto platforms ###### 3.1 Summary table | Platform | Agent product | Who controls limits | Rails | Status | |---|---|---|---|---| | Coinbase (CDP, Coinbase Developer Platform) | Agentic Wallet, x402 facilitator, AgentKit | Human owner sets limits; agent cannot change them | USDC on Base, Solana, Polygon, Arbitrum and others | Live (OBSERVED) | | Kraken | Open-source CLI plus MCP server for trading | User; dangerous calls prompt where the client supports it | Kraken exchange | "Experimental", MCP alpha (OBSERVED) | | Revolut crypto / Revolut X | No agent product found | N/A | Revolut app and exchange | Live | | Stripe / Privy | Agent wallets (Privy) | Owner policy | Stablecoins | Available (OBSERVED) | | Skyfire | KYA plus agent wallet | Owner-set per-agent limits | Cards, bank, USDC | Live (OBSERVED) | ###### 3.2 Coinbase (Agentic Wallets, x402, AgentKit) **What they offer today** - OBSERVED: x402 "turns an API or digital service into a paid resource". A buyer, including "an agent purchasing a tool call mid-task", sees the price, pays and gets access in one request flow without signup [source](https://docs.cdp.coinbase.com/x402/welcome), accessed 2026-10-31. - OBSERVED as statement, CLAIM as number: Coinbase says CDP "has processed more than 100 million x402 payments across Base and Solana" [source](https://docs.cdp.coinbase.com/x402/welcome), accessed 2026-10-31. - OBSERVED: The Agentic Wallet is "a Coinbase-managed wallet" holding USDC. "You fund the wallet and set spending limits; the agent can only pay within those limits." The agent pays "without handling a private key". It is accessed through a CLI (`awal`) or an MCP server [source](https://docs.cdp.coinbase.com/x402/agentic-accounts/agentic-wallet), accessed 2026-10-31. - OBSERVED (excerpt): The Agentic Wallet CLI docs describe caps per session and per transaction that "agents respect … but cannot change" [source](https://docs.cdp.coinbase.com/agentic-wallet/cli/welcome), accessed 2026-10-31. - CLAIM: Agentic Wallets launched on 11 February 2026 [source](https://www.theblock.co/post/389524/coinbase-rolls-out-ai-tool-to-give-any-agent-a-wallet), accessed 2026-10-31. **Pricing** - OBSERVED: The CDP x402 facilitator (the service that checks and settles payments) charges $0 for the first 1,000 on-chain transactions per month, then $0.001 each. Verification is free. Networks include Base, Polygon, Arbitrum, Monad, World and Solana [source](https://docs.cdp.coinbase.com/x402/network-support), accessed 2026-10-31. **Gaps and weaknesses** - OBSERVED: The Agentic Wallet page does not mention KYT (know-your-transaction screening) or any compliance screening [source](https://docs.cdp.coinbase.com/x402/agentic-accounts/agentic-wallet), accessed 2026-10-31. - CLAIM: Coinbase's UK entity (CB Payments Ltd) is registered with the FCA for anti-money-laundering purposes. That registration is lighter than full authorisation [source](https://cointelegraph.com/news/coinbase-fca-uk-crypto-registration), accessed 2026-10-31. - CLAIM: The new UK crypto authorisation regime opened applications on 30 September 2026 and is expected in force on 25 October 2027 [source](https://www.gherson.com/blog/uk-cryptoasset-regulation-2026-legal-update/), accessed 2026-10-31. - Gap (analysis): Crypto-only and USDC-centric. There is no FSCS-protected sterling balance, no Faster Payments, and no regulated recovery or dispute process for agent mistakes. ###### 3.3 Kraken - OBSERVED: Kraken CLI is an open-source (MIT licence) command-line tool and MCP server: "Trade on Kraken from your terminal, or let your AI agent trade for you." `kraken-mcp -s all` serves 152 MCP tools, and the default set leaves out live trading [source](https://github.com/krakenfx/kraken-cli), accessed 2026-10-31. - OBSERVED: The catalogue lists 188 commands, 43 of them marked `dangerous`. The CLI "places orders without a confirmation prompt" [source](https://github.com/krakenfx/kraken-cli), accessed 2026-10-31. - OBSERVED: Safety features include `--validate`, approval prompts on MCP clients that support elicitation (the client asking the user), OAuth sign-in keys that "can't withdraw funds" and expire after 90 days by default, per-call audit lines without credentials, and a "dead man's switch" (`cancel-after`) [source](https://github.com/krakenfx/kraken-cli), accessed 2026-10-31. - OBSERVED: The README says it is "Experimental software" and that the MCP integration is alpha [source](https://github.com/krakenfx/kraken-cli), accessed 2026-10-31. - Gap (analysis): The agent uses the human's exchange account. There is no separate agent identity or mandate object. ###### 3.4 Revolut crypto - OBSERVED (excerpt): Revolut X launched (May 2024) as a stand-alone exchange for UK users, at 0% maker and 0.09% taker fees with no volume tiers [source](https://revolut.com/news/revolut_launches_revolut_x_a_stand_alone_crypto_trading_platform_for_experienced_traders/), accessed 2026-10-31. - CLAIM: Revolut holds a MiCA (EU Markets in Crypto-Assets regulation) licence from Cyprus's regulator CySEC, letting it offer crypto services across the EEA [source](https://decrypt.co/345674/revolut-secures-mica-license-in-cyprus-is-a-stablecoin-next), accessed 2026-10-31. - UNKNOWN: Timing of a Revolut euro stablecoin ("EURR"). Sources conflict between a 2026 rollout and a 2027 launch [source](https://www.kucoin.com/news/flash/revolut-to-launch-eurr-stablecoin-in-2027), accessed 2026-10-31. - Agent features for crypto: **not found.** I searched "Revolut crypto MiCA licence 2026 stablecoin" and "Revolut AIR". ###### 3.5 Others building agent wallets - OBSERVED: Stripe-owned Privy offers "Agent wallets from the CLI (available)" and "Agentic wallets (available)" for micropayments. A Q4 2026 private preview will provision agents with preconfigured spending policies [source](https://stripe.com/blog/everything-we-announced-at-sessions-2026), accessed 2026-10-31. - CLAIM: World (formerly Worldcoin) launched AgentKit in beta on 17 March 2026. It lets an agent carry a zero-knowledge proof (a cryptographic proof that reveals nothing else) that a unique verified human stands behind it, and it works with x402 [source](https://techcrunch.com/2026/03/17/world-launches-tool-to-verify-humans-behind-ai-shopping-agents/), accessed 2026-10-31. - CLAIM: OKX launched an agent payments protocol in April 2026 [source](https://theblock.co/post/402029/catena-labs-lands-30-million-series-a-files-for-national-trust-bank-charter-to-underpin-agentic-finance), accessed 2026-10-31. --- ##### 4. AI and intelligence platforms competing in finance ###### 4.1 Summary table | Vendor | Finance product | Who it serves | Executes money movement? | Pricing | |---|---|---|---|---| | OpenAI | ChatGPT for Financial Services (Sep 2026) | Investment banks, research | Not found | Not public (CLAIM) | | Anthropic | Claude for Financial Services; 10 agent templates | Finance staff | **No**: templates stage work for human sign-off (OBSERVED) | Repo Apache-2.0; usage priced separately | | Google | Gemini Enterprise for Financial Services (preview) | Capital markets, corporate banking | Not found | Not stated | | Microsoft | Copilot Studio, Agent 365 governance | Bank staff | Not found | Agent 365 $15/user/month (CLAIM) | | Oracle | FCCM (financial crime and compliance management) AI agents | Bank investigators | No | Enterprise licence (not stated) | | Palantir | Foundry anti-financial-crime | Banks, regulators | No | Enterprise (not stated) | | Quantexa | Decision Intelligence, Q Assist | Banks, government | No | Not public | | Feedzai | RiskOps plus Farol agent | Banks, payment firms | No | Not public | | ComplyAdvantage | Mesh plus agentic workflows | Fintechs, banks | No | Has a Starter plan; prices not seen | | Hawk | AML Investigative Agent | Banks | No | Not public | | Sardine | Agentic risk platform, 10 agents | Banks, merchants | No | Sales-led | | Unit21 | AI Risk Infrastructure, agents | Banks, fintech, crypto | No (SAR filed only with human approval) | Not public | | Featurespace (Visa) | ARIC Risk Hub | Banks | No | Not public | ###### 4.2 OpenAI - OBSERVED (excerpt): OpenAI published "Introducing ChatGPT for Financial Services" [source](https://openai.com/index/introducing-chatgpt-financial-services/), accessed 2026-10-31. The page returned 403 to direct fetch. - CLAIM: It launched on 10 September 2026 with built-in data from LSEG, PitchBook and Daloopa. It initially targets investment banking and equity research, with Morgan Stanley and Evercore as design partners [source](https://siliconangle.com/2026/09/10/openai-targets-wall-street-bankers-with-a-new-version-of-chatgpt/), accessed 2026-10-31. - CLAIM: It is available only to eligible institutions. No public price list or minimum seats have been disclosed [source](https://www.techrepublic.com/article/news-openai-chatgpt-financial-services-banks/), accessed 2026-10-31. - OBSERVED (excerpt): Stripe and OpenAI co-developed the Agentic Commerce Protocol (ACP) behind Instant Checkout in ChatGPT [source](https://stripe.com/newsroom/news/stripe-openai-instant-checkout), accessed 2026-10-31. - CLAIM: In March 2026 OpenAI shifted ChatGPT shopping towards discovery, with merchants handling checkout on their own sites. Whether Instant Checkout continues for "approved partners" is **UNKNOWN** because sources conflict [source](https://www.digitalcommerce360.com/2026/03/17/shopify-integration-with-chatgpt-changes/), accessed 2026-10-31. - CLAIM: In May 2026 ChatGPT Pro let users connect over 12,000 financial institutions via Plaid (a US bank-data aggregator) [source](https://www.llrx.com/?p=25385), accessed 2026-10-31. - Gap (analysis): OpenAI owns the consumer's agent and its attention, not the regulated account. That makes it a likely *channel* for EDGE, and a risk if it disintermediates banks. ###### 4.3 Anthropic - OBSERVED: The `anthropics/financial-services` GitHub repo ships 10 agent templates: Pitch Agent, Meeting Prep, Market Researcher, Earnings Reviewer, Model Builder, Valuation Reviewer, GL Reconciler, Month-End Closer, Statement Auditor and KYC Screener. They are deployable as Cowork plugins, Claude Code plugins or Managed Agents via API, under the Apache 2.0 licence [source](https://github.com/anthropics/financial-services), accessed 2026-10-31. - OBSERVED: The README states the agents "do not make investment recommendations, execute transactions, bind risk, post to a ledger, or approve onboarding; every output is staged for human sign-off" [source](https://github.com/anthropics/financial-services), accessed 2026-10-31. - CLAIM: The templates were announced on 5 May 2026 [source](https://axios.com/2026/05/05/anthropic-wall-street-dimon-amodei), accessed 2026-10-31. - CLAIM: "Claude for Financial Advisors" launched in September 2026 with integrations including Schwab, BlackRock, Addepar and Orion [source](https://www.globalbankingandfinance.com/anthropic-targets-financial-advisers-new-claude-tool/), accessed 2026-10-31. - Gap (analysis): The deliberate "no execution" stance leaves the regulated execution layer open. That layer is where EDGE sits. ###### 4.4 Google - OBSERVED: On 25 August 2026 Google Cloud launched Gemini Enterprise for Financial Services, "initially available in preview for the capital markets and corporate banking industries". It includes a Financial Research agent, 50+ financial skills, 13 data connectors (FactSet, LSEG, Moody's and others) and third-party agents [source](https://www.googlecloudpresscorner.com/2026-08-25-Google-Cloud-Launches-Gemini-Enterprise-for-Financial-Services), accessed 2026-10-31. - OBSERVED: Google cites "confidence scores, explicit methodologies, data snapshots", source citations and "audit logging" built into the platform. Deutsche Bank and CME Group are named users [source](https://www.googlecloudpresscorner.com/2026-08-25-Google-Cloud-Launches-Gemini-Enterprise-for-Financial-Services), accessed 2026-10-31. - OBSERVED: Google's AP2 (agent payments) and Starling's assistant (built on Gemini) are covered in Sections 5 and 2.2. - CLAIM: Google's Universal Commerce Protocol (UCP) launched in January 2026 for agentic checkout in AI Mode and the Gemini app with eligible US retailers [source](https://www.pymnts.com/news/2026/google-debuts-universal-protocol-for-agentic-commerce/), accessed 2026-10-31. ###### 4.5 Microsoft - CLAIM: Microsoft Agent 365, a governance "control plane" for managing agents, became generally available on 1 May 2026 at $15 per user per month. It is also bundled in Microsoft 365 E7 at $99 per user per month [source](https://www.licensingschool.co.uk/licensing-blog/2026/03/23/agent-365-ga-announcement/), accessed 2026-10-31. - CLAIM: Each agent can get a Microsoft Entra Agent ID, an identity record for the agent. Agents with their own fully autonomous identities were reportedly still in preview [source](https://andrew.ooo/answers/microsoft-agent-365-ga-day-one-review-may-2026/), accessed 2026-10-31. - CLAIM: Banks build internal agents (e.g., for AML document classification) in Copilot Studio [source](https://myabt.com/blog/microsoft-copilot-financial-institutions-deployment-guide), accessed 2026-10-31. - Gap (analysis): Enterprise identity for *internal* agents. It is not a bank ledger or a payment mandate. ###### 4.6 Oracle Financial Services - OBSERVED (excerpt): On 9 April 2026 Oracle announced AI agent capabilities for its Financial Crime and Compliance Management (FCCM) portfolio, using technology from Lucinity. The capabilities target investigators' case work, and Oracle says they will be available "within the next 12 months" [source](https://www.oracle.com/news/announcement/oracle-brings-new-ai-capabilities-and-agents-to-its-financial-crime-and-compliance-portfolio-2026-04-09/), accessed 2026-10-31. - CLAIM: Oracle positions the agents as supporting rather than replacing investigators [source](https://www.barchart.com/story/news/1218421/oracle-brings-new-ai-agent-driven-capabilities-to-its-industry-leading-financial-crime-and-compliance-portfolio), accessed 2026-10-31. ###### 4.7 Palantir - OBSERVED (excerpt): In March 2025 Palantir announced deployment of its Anti-Financial Crime solutions, built on Foundry, at Société Générale's international retail network [source](https://www.businesswire.com/news/home/20250304819606/en), accessed 2026-10-31. - CLAIM: The UK FCA reportedly awarded Palantir a short trial contract in March 2026 to analyse internal intelligence data on financial crime [source](https://www.hl.co.uk/shares/stock-market-news/company--news/palantir-lands-trial-contract-with-uks-financial-conduct-authority), accessed 2026-10-31. - Not found: a 2026 AIP (Palantir's AI platform) AML-agent product announcement for banks. Searched "Palantir AIP banking anti-money laundering agents 2026 announcement". ###### 4.8 Finance-specific intelligence and financial-crime vendors **Quantexa** - OBSERVED: Quantexa sells a "Decision Intelligence Platform" built on entity resolution (working out which records refer to the same real person or company) and graph analytics, plus "Q Assist" and an "Agent Gateway". It claims "explainable and auditable AI" [source](https://www.quantexa.com/), accessed 2026-10-31. - CLAIM: HMRC (the UK tax authority) awarded Quantexa a £175m, 10-year partnership in May 2026 [source](https://www.bernama.com/en/region/news.php?id=2611805), accessed 2026-10-31. **Feedzai** - OBSERVED: Feedzai sells a "RiskOps" fraud/AML platform and "Farol, the AI agent for fraud teams" that works on rules, alerts and SARs alongside analysts [source](https://www.feedzai.com/), accessed 2026-10-31. - CLAIM: Feedzai cites "62% more fraud detected" and "73% fewer false positives" versus a tier 1 bank's previous solution [source](https://www.feedzai.com/), accessed 2026-10-31. - OBSERVED: The homepage does not mention agentic commerce or fraud by third-party AI agents [source](https://www.feedzai.com/), accessed 2026-10-31. **ComplyAdvantage** - OBSERVED: ComplyAdvantage's "Mesh" platform covers customer, company and payment screening, monitoring and transaction monitoring. It markets "Agentic workflows" [source](https://complyadvantage.com/), accessed 2026-10-31. - CLAIM: It says these workflows "resolve up to 85% of routine alerts autonomously while maintaining total regulatory defensibility" [source](https://complyadvantage.com/), accessed 2026-10-31. **Hawk** - CLAIM: Hawk's "AML Investigative Agent" gathers information, summarises cases, identifies typologies and supports SAR preparation [source](https://fintech.global/?p=236301), accessed 2026-10-31. Hawk's own site returned 403 (UNKNOWN beyond this). **Sardine** - OBSERVED: Sardine calls itself "the top agentic risk platform". It lists ten named agents (e.g., OSINT Search, Transaction Monitoring, Sanctions Screening, SAR Generation), "Agent Hub" and "Sardine Flow", and says it detects "agentic automation" among bots [source](https://www.sardine.ai/), accessed 2026-10-31. - OBSERVED: Sardine publishes a (gated) whitepaper titled "Rethinking fraud prevention for agentic commerce" [source](https://sardine.ai/whitepapers/rethinking-fraud-prevention-for-agentic-commerce), accessed 2026-10-31. **Unit21** - OBSERVED: Unit21 agents pull data, check watchlists, assemble evidence and recommend a disposition. A four-level autonomy ladder goes up to auto-closing low-risk cases. "Every step is logged: data accessed, checks performed, sources cited, assumptions made." SARs are not filed without human approval [source](https://www.unit21.ai/products/ai-agent), accessed 2026-10-31. - OBSERVED: Unit21 states it "does not ship traditional machine learning risk-scoring models", uses backtesting against expert-reviewed alerts, and "does not train models on customer data" [source](https://www.unit21.ai/products/ai-agent), accessed 2026-10-31. - CLAIM: It cites "up to 80% faster" handling and "up to 93% fewer false positives", which it labels "customer-reported" [source](https://www.unit21.ai/products/ai-agent), accessed 2026-10-31. **Featurespace (owned by Visa)** - OBSERVED (excerpt): Visa acquired Featurespace, a real-time AI payments-fraud firm [source](https://usa.visa.com/about-visa/newsroom/press-releases.releaseId.21106.html), accessed 2026-10-31. - Not found: a 2026 agentic-specific Featurespace product. Searched "Featurespace ARIC Visa Protect 2026 agentic". **Gap across all fincrime vendors (analysis):** They protect institutions *from* agents, or use agents to help investigators. None I found models a *customer's* agent as a monitored entity with its own risk profile, its own mandate and a link to its principal. Error-rate claims are marketing ("up to") rather than published, independently measured rates with confidence bounds. --- ##### 5. Agentic commerce and agent payments infrastructure ###### 5.1 Summary table | Scheme | Owner | What it does | Status (as found) | |---|---|---|---| | Agentic Commerce Suite / ACP / Shared Payment Tokens | Stripe (+ OpenAI for ACP) | Sell via agents; agent pays with a scoped token | Available (US focus) | | Machine Payments Protocol (MPP) | Stripe + Tempo | Agent-to-business micropayments and recurring payments | Available | | Link agent wallet, Issuing for agents | Stripe | Agent pays from a wallet with approvals; single-use cards | Wallet available; Issuing in preview | | Trusted Agent Protocol (TAP) | Visa | Merchants verify signed requests from Visa-approved agents | In development/deployment | | Visa Intelligent Commerce / Agentic Ready | Visa | Issuer readiness for agent-initiated card payments | Phase 1 live in Europe incl. UK | | Agent Pay / Agentic Tokens | Mastercard | Tokenised credentials for agents; passkeys | All European issuers enabled | | Agent Toolkit / Agent Ready | PayPal | Agent access to PayPal APIs; merchant agent checkout | Toolkit live since 2025 | | AP2 | Google → FIDO Alliance | Signed mandates for agent payments | v0.2, open source | | UCP | Google + Shopify | Agentic checkout standard | Launched Jan 2026 (CLAIM) | | x402 | Coinbase → x402 Foundation | HTTP-native pay-per-request | Live | | KYAPay | Skyfire | Agent identity plus payment tokens | Live | ###### 5.2 Stripe - OBSERVED (excerpt): With ACP, Stripe issues a Shared Payment Token (SPT) so an AI app can start a payment without seeing the buyer's card. The token is tied to one merchant and one cart total [source](https://stripe.com/newsroom/news/stripe-openai-instant-checkout), accessed 2026-10-31. - OBSERVED: Stripe Sessions 2026 (29 April 2026) announced: - Agentic Commerce Suite: "You can now sell through AI agents by uploading your product catalog" (available). - Machine Payments Protocol, co-authored with Tempo (available). - Link agent wallet with spending approvals (available). - Issuing for agents: single-use virtual cards (preview). - Radar bot-abuse prevention to tell legitimate agents from fraud (preview). - "Agent-ready financial accounts" that can check balances, pay invoices, store funds, create cards and send money "with human-in-the-loop confirmation for key actions" (announced). - "Agent guardrails" to "assign agent identities, enforce scope rules, and configure approval flows" (preview). [source](https://stripe.com/blog/everything-we-announced-at-sessions-2026), accessed 2026-10-31. - OBSERVED (excerpt): Stripe docs have an "agents" section (agent toolkit and MCP) [source](https://docs.stripe.com/agents), accessed 2026-10-31. - CLAIM: Stripe's remote MCP server uses OAuth, and tool permissions follow restricted API keys [source](https://gamut.so/blog/stripe-mcp-server-guide), accessed 2026-10-31. - UNKNOWN: UK availability. A Stripe agentic-commerce page appeared in search titled "Agentic commerce US only" [source](https://stripe.com/en-cy/use-cases/agentic-commerce), accessed 2026-10-31. - **Why it matters:** Stripe's announced "agent-ready financial accounts" plus "agent guardrails" is the closest competitor concept to EDGE Finance. It is announced or in preview, and Stripe is not a UK deposit-taking bank. ###### 5.3 Visa (Intelligent Commerce, Trusted Agent Protocol, Agentic Ready) - OBSERVED: TAP is "a verifiable, cryptographic standard" that helps merchants tell "Visa-approved agents" from bots. Agents sign each request using RFC 9421 HTTP message signatures, which the merchant verifies against the agent's public key [source](https://developer.visa.com/capabilities/trusted-agent-protocol/overview), accessed 2026-10-31. - OBSERVED: Visa Agentic Ready was announced on 17 March 2026 in London, starting in Europe including the UK. Phase 1 has issuers testing agent-initiated payments in controlled production. UK partners named include Barclays, HSBC UK and Nationwide. Payments are "clearly tied to a real person, with consent and control at key moments" [source](https://www.visa.co.uk/about-visa/newsroom/press-releases.3437886.html), accessed 2026-10-31. - CLAIM: More than 30 European issuers, including Barclays, HSBC UK, Lloyds, Nationwide, NatWest, Revolut and Klarna, have completed live agent-executed transactions using Visa Payment Passkeys, TAP and a Visa "Agent Directory" [source](https://retail-systems.com/rs/Visa_Takes_Agentic_Commerce_Live_With_European_Banks_And_Merchants.php), accessed 2026-10-31. - CLAIM (critique): Because Visa decides which agents are approved, it acts as a gatekeeper [source](https://atxp.ai/blog/visa-trusted-agent-protocol-explained), accessed 2026-10-31. ###### 5.4 Mastercard Agent Pay - OBSERVED: On 2 June 2026 Mastercard said: "Today, all Mastercard issuers in Europe are enabled at a network level for Agent Pay." It listed 35 banks that completed controlled live agentic transactions using passkeys. It describes "Verifiable Intent" to ensure "action taken by an AI agent can be traced back to explicit consumer authorisation" [source](https://newsroom.mastercard.com/news/europe/en/perspectives/en/2026/europe-is-building-the-foundations-for-trusted-agentic-commerce/), accessed 2026-10-31. - OBSERVED: That post does not mention the UK [source](https://newsroom.mastercard.com/news/europe/en/perspectives/en/2026/europe-is-building-the-foundations-for-trusted-agentic-commerce/), accessed 2026-10-31. - CLAIM: Agent Pay and "Agentic Tokens" were first announced in April 2025 [source](https://www.businesswire.com/news/home/20250429047449/en), accessed 2026-10-31. ###### 5.5 PayPal - OBSERVED: PayPal's Agent Toolkit (14 April 2025) exposes payments, invoices, disputes, shipment tracking, catalogue, subscriptions and reporting APIs to agent frameworks, including MCP and the Vercel AI SDK. TypeScript is supported first [source](https://developer.paypal.com/community/blog/paypal-agentic-ai-toolkit/), accessed 2026-10-31. - OBSERVED (excerpt): OpenAI and PayPal announced PayPal would power Instant Checkout in ChatGPT (October 2025) [source](https://newsroom.paypal-corp.com/2025-10-28-OpenAI-and-PayPal-Team-Up-to-Power-Instant-Checkout-and-Agentic-Commerce-in-ChatGPT), accessed 2026-10-31. - CLAIM: PayPal's "Agent Ready" merchant payments became available in early 2026, and PayPal supports Google's UCP [source](https://stellagent.ai/insights/paypal-agentic-commerce-agent-ready), accessed 2026-10-31. ###### 5.6 Google AP2 (Agent Payments Protocol) and UCP - OBSERVED: AP2 is "an open protocol for the emerging Agent Economy" under the Apache 2.0 licence. It uses verifiable digital credentials organised as Checkout Mandates and Payment Mandates, each with an "open" stage (constraints such as budget) and a "closed" stage (authorising a specific checkout). Credentials are chained into an audit trail for human-present and human-not-present purchases [source](https://ap2-protocol.org/), accessed 2026-10-31. - OBSERVED: AP2 is at v0.2 and has been donated to the FIDO Alliance. The initial version supports card "pull" payments. Push payments such as real-time bank transfers are on the roadmap. Samples exist for cards and x402 [source](https://ap2-protocol.org/), accessed 2026-10-31. - CLAIM: Trulioo (identity verification) joined AP2 to bring its "Digital Agent Passport" into the protocol [source](https://finovate.com/trulioo-joins-googles-agent-payments-protocol-ap2-to-secure-agent-led-payments/), accessed 2026-10-31. - **Why it matters:** AP2 does not yet cover UK bank push payments (Faster Payments). A bank that maps AP2 mandates onto Faster Payments and Direct Debit would fill a gap the spec itself lists as roadmap. ###### 5.7 x402 (Coinbase origin) and MPP - OBSERVED: See Section 3.2 for x402 mechanics and pricing. - CLAIM: The x402 Foundation was announced by Cloudflare and Coinbase in September 2025. Governance has reportedly moved toward the Linux Foundation, though the dates conflict (UNKNOWN) [source](https://xenospectrum.com/en/linux-foundation-x402-agent-payments/), accessed 2026-10-31. - CLAIM: Stripe and Tempo announced MPP on 18 March 2026, and Tempo's mainnet launched the same day [source](https://dc.fortune.com/2026/03/18/stripe-tempo-paradigm-mpp-ai-payments-protocol), accessed 2026-10-31. ###### 5.8 Skyfire - OBSERVED: Skyfire offers KYA, giving agents a verified identity for "account creation and authentication" and a "verifiable track record of your agent's activity". Agents can be funded by card, bank transfer, wire or USDC, with per-agent spending limits. Sellers get a built-in wallet [source](https://www.skyfire.xyz/product), accessed 2026-10-31. - CLAIM: The open KYAPay protocol and "Agent Checkout" launched in June 2025. KYA is a signed JWT (JSON Web Token, a signed data packet) carrying verified owner information [source](https://www.businesswire.com/news/home/20250626772489/en), accessed 2026-10-31. - UNKNOWN: Pricing, and who verifies agent owners. The product page states neither [source](https://www.skyfire.xyz/product), accessed 2026-10-31. --- ##### 6. Agent identity and agent tooling infrastructure | Provider | What the agent gets | Identity strength | Pricing | Label | |---|---|---|---|---| | Resend Inboxes | Own email address per agent; shared with humans | Domain-based; no KYA | Free up to 3 inboxes; $1 per extra inbox | OBSERVED | | AgentMail | API-created inboxes; "verifiable email address" | DKIM/SPF/DMARC domain auth | Free: 3 inboxes, 3,000 emails/month | OBSERVED | | Skyfire | KYA token plus wallet | Owner-verified (method not stated) | Not stated | OBSERVED / UNKNOWN | | Trulioo + PayOS | "Digital Agent Passport" | Developer KYB/KYC, code fingerprint, consent | Not found | CLAIM | | Persona | KYA via partners (e.g., Chatbase) | Links agent to verified customer | Not found | CLAIM | | World AgentKit | Proof a unique human backs the agent | Biometric World ID, zero-knowledge proof | Not found | CLAIM | | Microsoft Entra Agent ID | Enterprise directory identity for agents | Corporate tenant | Via Agent 365 | CLAIM | | Visa TAP / Agent Directory | Signed agent requests for merchants | Visa approval | Not public | OBSERVED / CLAIM | | Stripe agent guardrails | Agent identities, scopes, approval flows | Stripe account | Preview | OBSERVED | **Details** - OBSERVED: Resend launched "Inboxes" on 22 October 2026. It lets you "give each agent its own address to receive tasks, approvals, and updates". Inboxes are worked by "people and agents side by side", and you choose which actions the built-in agent may take. Access is by API, CLI and MCP [source](https://resend.com/blog/announcing-inboxes), accessed 2026-10-31. - OBSERVED: Resend's post does not describe authentication methods or compliance certifications [source](https://resend.com/blog/announcing-inboxes), accessed 2026-10-31. - OBSERVED: AgentMail calls itself "the first email provider built for AI agents" (a CLAIM in itself). It offers SDKs, an MCP server, IMAP/SMTP and multi-tenant "Pods". It lists no compliance certifications on its homepage [source](https://www.agentmail.to/), accessed 2026-10-31. - CLAIM: Trulioo's KYA lifecycle, co-developed with PayOS, verifies the developer, locks or fingerprints the agent's code, captures user consent, issues a Digital Agent Passport and validates each action [source](https://www.trulioo.com/blog/know-your-agent-kya/know-your-agent-kya-agentic-commerce-trust), accessed 2026-10-31. - CLAIM: Worldpay partnered with Trulioo on KYA in August 2025 [source](https://www.worldpay.com/en/press-releases/worldpay-and-trulioo-collaborate-embed-trust-agentic-commerce), accessed 2026-10-31. - CLAIM: Chatbase is building KYA with Persona. Persona authenticates the agent, links it to a verified customer who approves it, and sets per-agent permissions (e.g., "check order status but never change payment details") [source](https://www.chatbase.co/blog/introducing-know-your-agent-on-chatbase), accessed 2026-10-31. - UNKNOWN: A standalone Persona KYA product page. Persona's blog returned 403, and the search found only event listings. - CLAIM: There is no common KYA standard. Each vendor defines it differently [source](https://www.trulioo.com/blog/trust-and-safety/future-agentic-commerce-know-your-agent-kya), accessed 2026-10-31. **Gap (analysis):** Identity providers verify the *builder* or the *human behind* an agent. Mailbox providers give agents a contact address. None of them holds money, enforces financial mandates or offers regulated recovery. EDGE can be the place where KYA results, the agent's mailbox and its financial permissions are bound together under one auditable record. --- ##### 7. Regulatory context that shapes the lane - OBSERVED: FCA AI Live Testing cohort 2 was announced on 21 April 2026. Testing runs to the end of 2026 and an evaluation report is due Q1 2027. Use cases include "agentic payments" [source](https://www.fca.org.uk/news/press-releases/fca-announces-second-cohort-ai-live-testing), accessed 2026-10-31. - CLAIM: The FCA's 2026 Payments Regulatory Priorities report (25 March 2026) signals the FCA may need to change payments rules to accommodate agentic AI [source](https://vixio.com/regulatory-news/pc-vixio-view-fcas-payments-priorities-offer-clarity-rather-novelty-except-ai), accessed 2026-10-31. - CLAIM: A legal analysis says the allocation of liability between parties in agent-initiated payments is not yet addressed [source](https://bratby.law/agentic-ai-payments-consent-regulation/), accessed 2026-10-31. - CLAIM: The EU AI Act's high-risk obligations for standalone systems such as credit scoring moved from 2 August 2026 to 2 December 2027 under a provisional "digital omnibus" deal (7 May 2026) [source](https://www.gibsondunn.com/eu-ai-act-omnibus-agreement-postponed-high-risk-deadlines-and-other-key-changes/), accessed 2026-10-31. - OBSERVED (excerpt): FCA fines against Starling (2024) and Monzo (2025) were for financial-crime controls lagging growth (see Sections 2.2 and 2.3). - **Implication (analysis):** Agent-initiated payments sit in a live regulatory sandbox in the UK. Joining or engaging with FCA AI Live Testing (or its successor) is likely the safest route to "agents as customers". --- ##### 8. Our lane: where EDGE Finance can win ###### 8.1 What nobody does today (based on what I could find) | Capability | Who comes closest | What's missing | |---|---|---| | An AI agent as a first-class customer record at a UK-regulated institution, with its own identity, sub-ledger, mandate and principal | Catena Labs (US, charter filed, not approved); Stripe "agent-ready financial accounts" (announced); Coinbase Agentic Wallet (crypto) | UK regulation, sterling rails, FSCS-style protection, regulated recovery | | Mandates enforced *by the bank* on UK push payments (Faster Payments) and Direct Debit | AP2 (cards first, push payments on roadmap); GoCardless test (Direct Debit, human payer) | A bank-native mandate engine that maps AP2/Verifiable Intent to UK rails | | Recovery when an agent goes wrong: kill switch, key rotation, reversal and dispute paths, liability owner | Kraken dead-man's switch; Coinbase limits; card chargebacks for tokenised agent payments | An end-to-end, customer-visible recovery process; liability is unresolved in the UK (CLAIM) | | Measured, published error rates for AI decisions, with abstain-and-escalate when uncertain | Unit21 backtesting; Google confidence scores; Monzo golden-set evals | Public, per-decision error budgets the *customer* can see, not "up to 93%" marketing | | A glass-box audit trail the customer, not just the bank, can inspect: human → mandate → agent → action → outcome | Unit21 work log (internal); AP2 credential chain; Kraken audit lines | A unified, customer-facing, replayable trail across identity, payments and AI decisions | | One binding of KYA + agent mailbox + agent wallet + limits | Fragmented across Skyfire, Trulioo, Resend/AgentMail, Coinbase, Stripe | A single governed object | ###### 8.2 Concrete positioning statement (draft) > EDGE Finance is the UK-regulated home for AI agents' money. Each agent has a verified identity bound to a responsible person or organisation, a mandate that says exactly what it may do, hard limits it cannot change, and a glass-box record of every action. When something goes wrong, there is a kill switch and a recovery path. ###### 8.3 Honest constraints and risks - **Legal personhood: UNKNOWN.** I found no UK source establishing that an AI agent can itself be a legal account holder. A 2026 analysis says the person or business remains the customer (CLAIM) [source](https://macmyths.com/can-an-ai-agent-have-a-bank-account-in-2026/), accessed 2026-10-31. EDGE should design "agent as customer" as **an agent record under a legally responsible principal** until lawyers and the FCA confirm otherwise. - **Licensing:** EDGE needs its own licence (EMI or bank) or a regulated partner. Engine by Starling is sold to banks as a SaaS core (CLAIM [source](https://europeanpurpose.com/tool/engine-by-starling), accessed 2026-10-31), so it is not obviously a BaaS route for a new fintech. Partner choice is open. - **Network gatekeeping:** Card-based agent payments depend on Visa and Mastercard agent approval and tokenisation (OBSERVED [source](https://developer.visa.com/capabilities/trusted-agent-protocol/overview), accessed 2026-10-31). EDGE should interoperate rather than compete. - **Incumbent speed:** Stripe has announced most building blocks (OBSERVED [source](https://stripe.com/blog/everything-we-announced-at-sessions-2026), accessed 2026-10-31). EDGE's window is UK regulation plus governance depth, not raw features. - **Financial-crime risk:** Agents can be abused at scale. Two recent UK challenger fines show controls must scale with growth from day one (OBSERVED (excerpt) [source](https://www.fca.org.uk/news/press-releases/fca-fines-monzo-21m-failings-financial-crime-controls), accessed 2026-10-31). --- ##### 9. Bare minimum (table stakes) vs ambitious (differentiating) | Area | Bare minimum (table stakes) | Ambitious (differentiating) | Evidence that the minimum is standard | |---|---|---|---| | Licence and protection | FCA/PRA-regulated entity or regulated partner; safeguarding or FSCS | Agent-specific terms, liability allocation and recovery rights agreed with the regulator (e.g., via AI Live Testing) | Starling, Monzo, Revolut, Zopa, Kroo are licensed banks | | Human onboarding | Digital KYC/KYB with sanctions screening across the *full* lists | Continuous re-screening of principals *and* agents | FCA Starling fine for partial list screening (OBSERVED (excerpt)) | | Agent identity (KYA) | Each agent registered, bound to a verified principal, with a unique credential | Imports external KYA (Skyfire, Trulioo DAP, World proof-of-human, Entra Agent ID) and agent reputation history | Skyfire, Trulioo, Persona, Visa TAP | | Mandates and limits | Per-agent spend caps per transaction/session/period, set by the human, unchangeable by the agent | Machine-readable mandates (AP2-style open/closed), counterparty allow-lists, time windows, purpose codes, auto-expiry | Coinbase Agentic Wallet; Stripe Link agent wallet; Catena | | Payment rails | Faster Payments, Direct Debit, cards | Agentic card tokens (Visa/Mastercard), x402/MPP micropayments, stablecoin option, AP2 push-payment mapping | Visa Agentic Ready, Mastercard Agent Pay, x402 | | Approvals | Human confirmation for high-risk actions (new payee, large amount) | Risk-adaptive approvals: the auto-approve threshold moves with measured agent reliability | Stripe "human-in-the-loop confirmation for key actions"; Revolut AIR confirmation step (CLAIM) | | Customer AI assistant | Opt-in conversational assistant that can act on the account; no training on customer data | The assistant can also *supervise* the customer's other agents ("what did my agents do today?") | Starling Assistant; Revolut AIR; Zopa Ask; bunq Finn | | Developer access | Documented API, sandbox, scoped keys, MCP server | Per-agent OAuth with mandate scopes; signed requests (RFC 9421) like TAP | Stripe MCP; Kraken MCP; PayPal toolkit | | Fincrime monitoring | Transaction monitoring, sanctions, fraud scoring, SAR workflow with human sign-off | Agents as monitored entities with their own behaviour baselines; agent-collusion graph analytics | Unit21, Sardine, Feedzai, Quantexa, Oracle | | Explainability and audit | Logged decisions for regulators; reasons for declines | Customer-visible glass-box trail; replay of any decision with the inputs, model version and rules used | Unit21 work log; Google audit logging | | AI error control | Evaluation before launch; human review at start (Monzo pattern) | Published error budgets per decision type, abstain-and-escalate under uncertainty, continuous backtesting, disclosed confidence | Monzo golden set; Unit21 backtesting | | Recovery | Card freeze, account freeze, dispute process | One-tap agent kill switch, credential rotation, mandate revocation propagated to networks, bulk reversal tooling | Revolut AIR card freeze (CLAIM); Kraken `cancel-after` | | Agent contact channel | Notifications to the human | Per-agent mailbox/webhook tied to identity (Resend/AgentMail-style) for approvals | Resend Inboxes; AgentMail | | Pricing transparency | Published fees | Machine-readable pricing that agents can read and compare | Coinbase facilitator pricing is published | --- ##### 10. Top 10 features competitors have that EDGE Finance must match (the floor) | # | Feature or behaviour | Who has it | Source (accessed 2026-10-31) | Label | |---|---|---|---|---| | 1 | **Opt-in conversational assistant that performs account tasks** (transfers to savings, bill organisation, card controls) | Starling, Revolut, Zopa, bunq | [Starling](https://www.starlingbank.com/news/starling-launches-agentic-ai-assistant/); [Zopa](https://itbrief.co.uk/story/zopa-launches-ai-ask-tool-for-biscuit-account-users) | OBSERVED / CLAIM | | 2 | **Customer data stays in-tenant and is not used to train models** | Starling; Unit21; Google | [Starling](https://www.starlingbank.com/news/starling-launches-agentic-ai-assistant/); [Unit21](https://www.unit21.ai/products/ai-agent) | OBSERVED | | 3 | **Owner-set spending limits per transaction and per session that the agent cannot change** | Coinbase Agentic Wallet; Stripe Link agent wallet; Skyfire | [Coinbase](https://docs.cdp.coinbase.com/x402/agentic-accounts/agentic-wallet); [Skyfire](https://www.skyfire.xyz/product) | OBSERVED | | 4 | **Human-in-the-loop confirmation for key actions**, with escalation to a human specialist when guardrails trip | Stripe (announced); Monzo Ops Agent; Unit21 (SARs) | [Stripe](https://stripe.com/blog/everything-we-announced-at-sessions-2026); [Monzo](https://monzo.com/blog/engineering-the-future-of-customer-operations-the-monzo-ops-agent) | OBSERVED | | 5 | **Tokenised, passkey-authenticated agent card payments** via the networks' agent programmes | Visa Agentic Ready (Barclays, HSBC UK, Nationwide); Mastercard Agent Pay | [Visa](https://www.visa.co.uk/about-visa/newsroom/press-releases.3437886.html); [Mastercard](https://newsroom.mastercard.com/news/europe/en/perspectives/en/2026/europe-is-building-the-foundations-for-trusted-agentic-commerce/) | OBSERVED | | 6 | **Verifiable agent identity on every request** (signed requests, KYA tokens) | Visa TAP (RFC 9421); Skyfire KYA | [Visa TAP](https://developer.visa.com/capabilities/trusted-agent-protocol/overview); [Skyfire](https://www.skyfire.xyz/product) | OBSERVED | | 7 | **Signed mandates that chain into an audit trail** for human-present and human-not-present purchases | Google AP2 | [AP2](https://ap2-protocol.org/) | OBSERVED | | 8 | **Agent-facing API/MCP with sandbox or paper mode, scoped keys and approval for dangerous actions** | Kraken CLI/MCP; Stripe MCP; PayPal Agent Toolkit; Coinbase MCP | [Kraken](https://github.com/krakenfx/kraken-cli); [PayPal](https://developer.paypal.com/community/blog/paypal-agentic-ai-toolkit/) | OBSERVED | | 9 | **Single-use virtual cards for agent purchases** | Stripe Issuing for agents (preview); Revolut AIR single-use cards (CLAIM) | [Stripe](https://stripe.com/blog/everything-we-announced-at-sessions-2026); [FintechWeekly](https://www.fintechweekly.com/news/revolut-air-ai-assistant-uk-customers-launch-2026) | OBSERVED / CLAIM | | 10 | **Agentic financial-crime investigation with step-by-step logs, autonomy levels and human sign-off on SARs** | Unit21; Sardine; Feedzai Farol; ComplyAdvantage; Oracle (coming) | [Unit21](https://www.unit21.ai/products/ai-agent); [Sardine](https://www.sardine.ai/) | OBSERVED | **Honourable mentions (next floor items):** - Pay-per-request micropayments (x402/MPP). OBSERVED [source](https://docs.cdp.coinbase.com/x402/welcome), accessed 2026-10-31. - Agent bot-detection at checkout (Stripe Radar preview; Sardine). OBSERVED [source](https://stripe.com/blog/everything-we-announced-at-sessions-2026), accessed 2026-10-31. - Per-agent mailboxes for approvals (Resend Inboxes). OBSERVED [source](https://resend.com/blog/announcing-inboxes), accessed 2026-10-31. --- ##### 11. Not found / open questions (searches run) | Question | Result | Searches run | |---|---|---| | Does any UK/EU bank open accounts *for* AI agents as customers? | Not found | "bank account for AI agents challenger bank Europe launch 2026 agent-native" | | Chase UK AI or agent features | Not found | "Chase UK 2026 news AI" | | Wise official agent/MCP product | Not found (community only) | "Wise AI agents MCP payments API 2026"; "Wise MCP server AI agents send money" | | Kroo AI features | Not found | "Kroo bank 2026 news" | | Engine by Starling pricing | Not found | "Starling Engine pricing model software licence fee per account" | | Revolut plan prices 2026 (primary) | Not verified | Revolut pages returned 403 | | Skyfire pricing; who verifies owners | Not stated on product page | Skyfire product page | | Persona KYA product page | Not found (403 / event listings only) | "Persona know your agent product launch 2026" | | Featurespace agentic product | Not found | "Featurespace ARIC Visa Protect 2026 agentic" | | Palantir AIP AML agents for banks (2026) | Not found | "Palantir AIP banking anti-money laundering agents 2026 announcement" | | UK legal status of an AI agent as account holder | UNKNOWN; needs legal advice | "can an AI agent have a bank account" (secondary only) | | Stripe agentic products available in the UK | UNKNOWN (one page titled "US only") | Stripe use-case page | | Status of ChatGPT Instant Checkout | UNKNOWN (sources conflict) | "ChatGPT Instant Checkout 2026 update merchants ACP status" | | Catena Labs charter approval | UNKNOWN | "Catena Labs AI-native financial institution agents 2026 licence" | **Recommended next verifications:** 1. FCA Register entries for each UK competitor's exact permissions. 2. OpenAI and Revolut primary pages via a non-blocked fetch. 3. A UK legal opinion on agent-as-customer structures and liability for agent-initiated payments. 4. Visa/Mastercard agent-programme onboarding terms for a new issuer. ### research/finance/CONFIDENCE-REPORTING.md `research/finance/CONFIDENCE-REPORTING.md` · sha256 `5d7e509cdc7f1a8ac89079f5b00e4907002bfcf694e8c01cd0a047f610ab299c` #### EDGE Finance: Confidence Reporting Research - **Product:** EDGE Finance (part of JudeAI | EDGE ID). People, organisations and AI agents are all real customers. - **Question:** Can EDGE replace the credit score with a **confidence report**: a fuller record that shows whether obligations were met and also how a customer acted when things went wrong? It should be explainable, open to challenge, lawful, useful for KYC, and proven by measured business and economic results. - **Report date:** 2026-10-31. Every source was accessed on 2026-10-31. - **Status:** This is research. It is not legal advice. Every legal point needs review by qualified UK, US and EU counsel before anything is built or claimed. - **Companion reports:** [COMPETITORS.md](COMPETITORS.md) (agent payment rails and agent identity), [ASSURANCE-AND-REUSE.md](ASSURANCE-AND-REUSE.md). --- ##### 0. How to read this report ###### 0.1 Evidence labels Every factual line has a link and one of these labels. | Label | Meaning | |---|---| | **OBSERVED** | I read it on a primary source: a regulator, a statute, the official text, or the organisation's own published study. | | **OBSERVED (excerpt)** | The source is primary, but I only saw a search-engine excerpt of it, not the full page. Treat it as a little weaker than OBSERVED. | | **CLAIM** | A vendor's own performance number, marketing, or a third-party summary I could not confirm on a primary source. A vendor statistic stays a CLAIM even on the vendor's own page. | | **AGREEMENT** | Several independent sources say the same thing, or the line records a formal agreement between parties, such as a political deal on a law. | | **PREDICTION** | My own estimate, design proposal or forecast. Each one states its assumptions. Nothing labelled PREDICTION has been tested. | | **UNKNOWN** | I looked and could not establish it, or the sources conflict. | ###### 0.2 Method and limits - OBSERVED: I downloaded and read the full text of the FCA (Financial Conduct Authority) credit information Final Report MS19/1.3, the FCA consultation CP26/7, the FCA Financial Lives 2024 key findings, the CFPB (US Consumer Financial Protection Bureau) 2025 credit-invisibles correction, the Behavioural Insights Team arrears trial report and the FinRegLab 2025 press release. I read the statutes on legislation.gov.uk directly. - The shared web-search budget for this session ran out partway through the research. After that I only fetched pages whose addresses I already had. Some gaps are therefore marked UNKNOWN: Creditsafe's scoring method, the current status of Credit Kudos, primary ClearScore figures, Equifax and TransUnion product detail, and Mastercard Agent Pay (its page returned HTTP 403). - Search results were US-indexed. ###### 0.3 Plain-English glossary | Term | Plain English | |---|---| | CRA (credit reference agency) | A company that collects records of people's borrowing and repayments and sells them to lenders. In the UK the big three are Experian, Equifax and TransUnion. In the US they are called "consumer reporting agencies" or "credit bureaus". | | Credit score | A number a CRA or a scoring company (such as FICO or VantageScore) works out from a credit file to predict the chance of default. | | Thin file / credit invisible | A thin file holds too little data to score well. A credit-invisible person has no file at all. | | Open Banking | UK rules that let a customer allow a regulated third party to read their bank account data. | | Cash-flow underwriting | Deciding on a loan using money in and money out of a bank account, not only a credit file. | | Forbearance | A lender giving a struggling borrower more time or lower payments. | | Arrears | Missed payments that are still owed. | | KYC / KYB | Know Your Customer / Know Your Business: checking who a customer is. | | ADM (automated decision-making) | A decision made by a computer with no meaningful human involvement. | | UK GDPR | The UK General Data Protection Regulation, the UK's main data protection law. | | DUAA | The Data (Use and Access) Act 2025, which changed UK GDPR. | | CCA | The Consumer Credit Act 1974. | | FCRA / ECOA | The US Fair Credit Reporting Act and the US Equal Credit Opportunity Act. | | Reg B | Regulation B, the CFPB rule that implements ECOA. | | EU AI Act | The European Union's Artificial Intelligence Act, Regulation (EU) 2024/1689. | | CIGB | Credit Information Governance Body: the new UK industry body for credit reporting. The FCA Final Report called it the CRGB (Credit Reporting Governance Body). | | DCCRA | Designated Consumer Credit Reference Agency: a CRA the FCA would name under its proposed rules. | | CCJ | County Court Judgment: a court order to repay a debt (England, Wales and Northern Ireland). | | NoC (Notice of Correction) | A statement of up to 200 words that a consumer can add to their UK credit file to explain an entry. | | PAYDEX | Dun & Bradstreet's business payment score. | | AP2 | Agent Payments Protocol, an open protocol from Google for payments made by AI agents. | | Agent / principal | An AI agent acts for a principal, which is the person or organisation that authorised it. | --- ##### 1. Summary 1. OBSERVED: The UK is rebuilding its credit information plumbing right now. The FCA's February 2026 consultation CP26/7 proposes that any lender sharing consumer credit data with one designated CRA must share it with all of them. It proposes designating Equifax, Experian and TransUnion. A new industry body, the CIGB, has been set up ([FCA CP26/7](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31). 2. OBSERVED: The FCA itself found that credit files handle financial difficulty inconsistently. It also found that poor understanding of how forbearance appears on a file "may inhibit early engagement" by borrowers ([FCA MS19/1.3, paras 3.50–3.52](https://www.fca.org.uk/publication/market-studies/ms-19-1-3.pdf), accessed 2026-10-31). That is the gap a confidence report would fill. 3. OBSERVED: People already have a weak version of "explain what happened": the 200-word Notice of Correction under CCA section 159. It does not change the score, and lenders are not obliged to weigh it ([legislation.gov.uk s159](https://www.legislation.gov.uk/ukpga/1974/39/section/159); [Experian NoC guide](https://www.experian.co.uk/consumer/guides/notice-of-correction.html), OBSERVED (excerpt); [Monzo help](https://monzo.com/help/overdrafts-loans/notices-of-correction), OBSERVED (excerpt); accessed 2026-10-31). 4. AGREEMENT: Richer data such as cash flow improves prediction and approvals in independent and vendor studies. The best independent evidence is FinRegLab's. The headline numbers mostly come from vendors (section 5). 5. OBSERVED: From 5 February 2026, UK GDPR Article 22 was replaced by Articles 22A–22D. Solely automated significant decisions are now broadly allowed, but only with safeguards: information, a way to make representations, human intervention, and a way to contest. Special category data has tighter limits ([DUAA s80](https://www.legislation.gov.uk/ukpga/2025/18/section/80); [SI 2026/82](https://www.legislation.gov.uk/uksi/2026/82/made); accessed 2026-10-31). 6. PREDICTION: The lawful design scores only verifiable conduct facts: did the customer respond, agree a plan, keep it, and what category of cause applied. It does not score tone, sentiment or health details. It shows how sure it is (data sufficiency) and lets a single bad event fade by rule. Section 6 sets out the design. 7. PREDICTION: To become the standard-setter, EDGE needs to (a) hold the FCA permission for providing credit references, (b) publish an open specification, (c) work through the CIGB and the FCA's open finance roadmap, and (d) prove outcomes in a lender pilot with a holdout group. --- ##### 2. Incumbents and challengers: what each measures and misses ###### 2.1 The three CRAs (Experian, Equifax, TransUnion) | Line | Label | Source | |---|---|---| | The UK credit information sector is "very highly concentrated", with a Herfindahl-Hirschman Index (a measure of market concentration) of over 5,000 in 2020. Barriers to entry include access to historical data, network effects and cost. | OBSERVED | [FCA MS19/1.3 para 2.7](https://www.fca.org.uk/publication/market-studies/ms-19-1-3.pdf), accessed 2026-10-31 | | Where an individual had a default on any account, the three large CRAs held consistent information on the number of defaults for only around 30% of matched individuals. | OBSERVED | [FCA MS19/1.3 para 2.5](https://www.fca.org.uk/publication/market-studies/ms-19-1-3.pdf), accessed 2026-10-31 | | More than 90% of credit applications involved a search at just one CRA. So gaps at any single CRA matter. | OBSERVED | [FCA CP26/7 Annex 2 para 56](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | | In Financial Lives 2024, 10% of UK adults who had checked their credit report or score in the past 12 months found an error (16% in 2022). A July 2024 Which? survey put it at 32% of people who had checked in the past five years. | OBSERVED (FCA citing FLS and Which?) | [FCA CP26/7 Annex 2 para 57](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | | Only just over one-third of adults who raised a dispute with a CRA got a response within one month. Just over one-quarter waited more than three months. | OBSERVED | [FCA CP26/7 Annex 2 para 61](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | | **What CRAs measure:** whether accounts were paid, how much is owed, defaults, CCJs, searches, and linked addresses and people. | OBSERVED (description of the market) | [FCA MS19/1.3](https://www.fca.org.uk/publication/market-studies/ms-19-1-3.pdf), accessed 2026-10-31 | | **What they miss:** why a payment was missed, whether the person got in touch, whether they kept a revised plan, and whether the cause was a one-off event. Arrangements are recorded as a flag, not as context. | OBSERVED (FCA finding on inconsistency) + PREDICTION (my reading of the gap) | [FCA MS19/1.3 paras 3.50–3.55](https://www.fca.org.uk/publication/market-studies/ms-19-1-3.pdf), accessed 2026-10-31 | | Industry principles drawn up with the ICO (Information Commissioner's Office) say a jointly agreed temporary payment reduction is recorded as an "arrangement" at CRAs. A default may follow when arrears reach about three monthly payments. | OBSERVED (excerpt, applied in Ombudsman decisions) | [SCOR/ICO principles for reporting arrears, arrangements and defaults](https://www.experian.co.uk/content/dam/noindex/uki/uk/cais/principles-for-reporting-aad.pdf); [Financial Ombudsman DRN-5907555](https://www.financial-ombudsman.org.uk/decision/DRN-5907555.pdf), accessed 2026-10-31 | | Equifax and TransUnion product-level detail (their own open banking and trended-data products) | UNKNOWN | Not verified this session (search budget exhausted) | ###### 2.2 FICO and VantageScore (US scoring models) | Line | Label | Source | |---|---|---| | FICO weights: payment history 35%, amounts owed 30%, length of credit history 15%, new credit 10%, credit mix 10%. | OBSERVED | [myFICO](https://www.myfico.com/credit-education/whats-in-your-credit-score), accessed 2026-10-31 | | FICO Scores do not consider salary, occupation, employer or employment history, among other things. Lenders may weigh income separately. | OBSERVED | [myFICO](https://www.myfico.com/credit-education/whats-in-your-credit-score), accessed 2026-10-31 | | **What they miss:** context for a late payment, and cash flow, unless it is added through products like UltraFICO (section 5). | PREDICTION (inference from the weights above) | [myFICO](https://www.myfico.com/credit-education/whats-in-your-credit-score), accessed 2026-10-31 | | In July 2024 FHFA (the US Federal Housing Finance Agency) released historical VantageScore 4.0 scores for loans Fannie Mae and Freddie Mac bought from April 2013 to March 2023. It described that period as when "trended" credit data was reliably available at all three bureaus. | OBSERVED | [FHFA news release](https://www.fhfa.gov/news/news-release/fhfa-announces-release-of-historical-vantagescore-4.0-credit-scores-by-the-enterprises), accessed 2026-10-31 | | In July 2025 the FHFA Director said lenders could use VantageScore 4.0 for mortgages sold to Fannie Mae and Freddie Mac, keeping the three-bureau ("tri-merge") report. Later 2026 lender letters widened access. | CLAIM (trade press and GSE pages via excerpt; details conflict between sources) | [HousingWire](https://www.housingwire.com/articles/pulte-says-gses-will-accept-vantagescore-4-0-immediately/); [Freddie Mac credit score models](https://sf.freddiemac.com/general/credit-score-models), accessed 2026-10-31 | ###### 2.3 Open Banking and cash-flow challengers | Firm | What it measures | What it misses | Label and source | |---|---|---|---| | **Credit Kudos** (UK) | Credit insights and scores from Open Banking bank data. Apple bought it in March 2022, reportedly for about US$150m; Apple did not confirm the deal publicly. | Current status: UNKNOWN. The firm's site would not resolve from this session, and I found no source on whether it still operates. | OBSERVED (excerpt) [Engadget](https://www.engadget.com/apple-credit-kudos-open-banking-financial-services-112719203.html), accessed 2026-10-31; status UNKNOWN | | **Plaid** (US, UK, EU) | LendScore (launched October 2025) combines income, cash flow and "network insights" (which apps a user connects). Plaid says about 81% of the predictive power comes from cash-flow features. | Conduct in difficulty, and why. It is a vendor model. | CLAIM [Plaid: How we built LendScore](https://plaid.com/blog/how-we-built-lendscore/); [Plaid LendScore](https://plaid.com/blog/plaid-lendscore-credit-risk-scoring/), accessed 2026-10-31 | | **Nova Credit** (US) | Cash Atlas cash-flow underwriting. Named clients include HSBC, SoFi and Scotiabank (July 2025). | Conduct in difficulty. | CLAIM [Business Wire, July 2025](https://www.businesswire.com/news/home/20250716876023/en/Nova-Credit-Announces-Second-Annual-Cash-Flow-Underwriting-Summit-to-Convene-Industry-Leaders-That-Are-Accelerating-the-Adoption-of-Cash-Flow-Underwriting), accessed 2026-10-31 | | **Experian Boost** (US) | Adds consumer-permissioned bill payments (for example utilities and streaming) to the Experian file. | It lifts only scores calculated on Experian data, and only for users who opt in. | OBSERVED (Experian's own study) [Experian Boost study](https://www.experian.com/blogs/ask-experian/experian-boost-study/), accessed 2026-10-31 | | **Experian Boost** (UK) | Uses current-account payments into savings and investments, Council Tax, and digital subscriptions. "Not all lenders use Boost." It claims up to 101 points. | Lender adoption is not disclosed. | OBSERVED (product page); the points figure is a CLAIM. [Experian UK Boost](https://www.experian.co.uk/consumer/experian-boost.html), accessed 2026-10-31 | | **ClearScore** (UK and others) | A free consumer credit score and report using Equifax data in the UK, plus a credit marketplace. Over 20 million users (as of 2024). It bought Money Dashboard, an Open Banking provider, in 2022. | It is a consumer window onto CRA data plus marketing; it adds no conduct context. | CLAIM (secondary encyclopedia; primary site returned 403) [Wikipedia: ClearScore](https://en.wikipedia.org/wiki/ClearScore), accessed 2026-10-31 | | FCA view of Open Banking in lending | Mainly used for affordability, alongside CRA data. "A handful of smaller lenders" lend solely on Open Banking data. The FCA expects CRA data to stay significant "for the foreseeable future". | n/a | OBSERVED [FCA CP26/7 paras 2.14–2.16](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | ###### 2.4 Business credit | Firm | What it measures | What it misses | Label and source | |---|---|---|---| | **Dun & Bradstreet** | PAYDEX is a dollar-weighted score from supplier-reported payment experiences. 80 means prompt payment; lower scores mean more days beyond terms (for example 50 ≈ 30 days late). | Why a business paid late (a customer's insolvency, a disputed invoice), and whether it told its suppliers. | CLAIM (D&B partner and third-party excerpts; D&B primary not fetched) [Altares: PAYDEX](https://www.altares.nl/en/our-data/rating-scores/paydex/), accessed 2026-10-31 | | **Creditsafe** | A company credit score and "maximum recommended credit limit", financials, directors, ownership, CCJs, and late or missed payments. It claims more than 100,000 customer companies and data in over 200 countries. | Scoring inputs are not disclosed on the page. | OBSERVED (page content); the numbers are a CLAIM. [Creditsafe](https://www.creditsafe.com/gb/en.html), accessed 2026-10-31 | | UK policy | Business lending is out of scope of the FCA's proposed mandatory reporting. HM Treasury consulted on the Commercial Credit Data Sharing scheme in September 2025. The FCA plans an open finance plan in 2026 that prioritises SME (small and medium-sized enterprise) lending. | n/a | OBSERVED [FCA CP26/7 paras 2.12–2.15](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | | Creditsafe's scoring method | UNKNOWN | Not verified (search budget exhausted) | UNKNOWN | --- ##### 3. The evidence gap ###### 3.1 Credit-invisible and thin-file people | Line | Label | Source | |---|---|---| | **US:** After a 2025 correction, the CFPB estimates 2.7% of US adults (about 7.0 million) were credit invisible in December 2020. A further 5.9% had stale unscored records and 3.9% had insufficient unscored records. 87.5% had a scored record. | OBSERVED | [CFPB 2025 update](https://files.consumerfinance.gov/f/documents/cfpb_update-credit-invisibles-estimate_2025-06.pdf), accessed 2026-10-31 | | **US:** The well-known "26 million credit invisible" figure (2015 report, December 2010 data) was corrected to 5.8% (13.5 million). The unscored share for 2010 was revised up to 12.7% (29.7 million). | OBSERVED | [CFPB 2025 update](https://files.consumerfinance.gov/f/documents/cfpb_update-credit-invisibles-estimate_2025-06.pdf), accessed 2026-10-31 | | **US:** FinRegLab says about one-third of US adults are estimated to have thin or no credit files, and another 25% have below-prime scores. | CLAIM (cited by a non-profit; the underlying estimate's source is not given) | [FinRegLab press release, 3 June 2025](https://finreglab.org/wp-content/uploads/2025/06/FinRegLab_2025-06-03_Press-Release_Study-Shows-Cash-Flow-Data-Can-Expand-Small-Business-Lending-1.pdf), accessed 2026-10-31 | | **UK:** Estimates differ and there is no agreed definition. Experian (2022): over 5 million adults with little or no credit history. LexisNexis Risk Solutions (2021): over 600,000 invisible and 1.7 million with no financial or credit activity in two years. L.E.K. Consulting (2023): about 10.5 million with little or no credit history. | OBSERVED (FCA quoting these studies) | [FCA CP26/7 Annex 2 paras 68–70](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | | **UK:** Young people, recent immigrants, older people, the long-term unemployed and homeless people are most likely to be thin-file or invisible. | OBSERVED | [FCA CP26/7 Annex 2 para 71](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | | **UK:** L.E.K. estimated that in 2023 UK consumers with non-standard credit needs had about £2 billion of credit need that went unmet but would have been commercially viable to meet. Invisible and thin-file consumers were nearly two-thirds of that group. | OBSERVED (FCA quoting L.E.K.) | [FCA CP26/7 Annex 2 para 69](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | | 22% of UK adults who applied for regulated credit in the two years to May 2024 (about 3.2 million) were declined. The FCA notes this is not wholly due to missing credit information. | OBSERVED | [FCA CP26/7 Annex 2 paras 66–67](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | | About half of lenders surveyed by the FCA said they would extend more credit if they could better assess creditworthiness. | OBSERVED | [FCA CP26/7 Annex 2 para 67](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | | Credit scores are statistically noisier for under-served groups: 7–9% lower fit for minority or low-income mortgage applicants. The noise comes mainly from thin data, not from the model. Equalising precision could cut approval disparities by about half. | OBSERVED (excerpt of the working paper) | [Blattner & Nelson, "How Costly is Noise?"](https://arxiv.org/abs/2105.07554), accessed 2026-10-31 | ###### 3.2 Harm from single events | Line | Label | Source | |---|---|---| | Research commissioned by the FCA (Warner and Garforth-Bles, 2024) found the wellbeing impact of falling into arrears is about four-fifths as big as the impact of becoming unemployed. | OBSERVED (FCA summary) | [FCA CP26/7 Annex 2 para 75](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | | 18% of UK adults were over-indebted in 2024. 5.5% missed one or more credit, loan or mortgage payments in the six months to May 2024. | OBSERVED | [FCA CP26/7 Annex 2 para 74](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | | People facing an income shock or relationship breakdown are the most likely to struggle with their finances or with financial services. | CLAIM (secondary summary of FLS 2024) | [Regulation Tomorrow](https://www.regulationtomorrow.com/2025/05/fca-financial-lives-2024-survey/), accessed 2026-10-31 | | Borrowers can get "materially different outcomes even where their circumstances may be similar", because financial difficulty is reported inconsistently. | OBSERVED | [FCA MS19/1.3 para 3.51](https://www.fca.org.uk/publication/market-studies/ms-19-1-3.pdf), accessed 2026-10-31 | | The Government's November 2025 Financial Inclusion Strategy includes work on the impact of coerced debt on victim-survivors' credit files. | OBSERVED (FCA summary) | [FCA CP26/7 para 2.11](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | | How long a single default or missed payment keeps lowering approval odds, measured causally in UK data | UNKNOWN | No 2024–2026 primary UK study was found this session | ###### 3.3 FCA Financial Lives: vulnerability and resilience | Line | Label | Source | |---|---|---| | 26.4 million UK adults (49%) had characteristics of vulnerability in May 2024, down from 27.3 million (52%) in May 2022. The four drivers are poor health, negative life events, low resilience and low capability. | OBSERVED | [FCA Financial Lives 2024 key findings](https://www.fca.org.uk/publication/financial-lives/financial-lives-survey-2024-key-findings.pdf), accessed 2026-10-31 | | 13.1 million adults (24%) had low financial resilience in May 2024. | OBSERVED | [FCA FLS 2024](https://www.fca.org.uk/publication/financial-lives/financial-lives-survey-2024-key-findings.pdf), accessed 2026-10-31 | | 10% had no cash savings, and another 21% had less than £1,000. | OBSERVED | [FCA FLS 2024](https://www.fca.org.uk/publication/financial-lives/financial-lives-survey-2024-key-findings.pdf), accessed 2026-10-31 | | 1.6 million adults (3%) got repayment support from lenders in the two years to May 2024. 43% of credit holders and 39% of mortgage holders said their situation improved as a result. | OBSERVED | [FCA FLS 2024](https://www.fca.org.uk/publication/financial-lives/financial-lives-survey-2024-key-findings.pdf), accessed 2026-10-31 | | 1.7 million adults (3.2%) used debt advice in the 12 months to May 2024, and 61% said their debts became more manageable. Only 22% of those in financial difficulty used debt advice. | OBSERVED | [FCA FLS 2024](https://www.fca.org.uk/publication/financial-lives/financial-lives-survey-2024-key-findings.pdf), accessed 2026-10-31 | | The survey had 17,950 respondents; fieldwork ran February to June 2024; it was published 16 May 2025. | OBSERVED | [FCA FLS 2024](https://www.fca.org.uk/publication/financial-lives/financial-lives-survey-2024-key-findings.pdf), accessed 2026-10-31 | ###### 3.4 What lenders learn from communication in arrears and forbearance | Line | Label | Source | |---|---|---| | The FCA found that poor consumer understanding of how financial difficulty and forbearance show up on credit files "may inhibit early engagement". | OBSERVED | [FCA MS19/1.3 para 3.52](https://www.fca.org.uk/publication/market-studies/ms-19-1-3.pdf), accessed 2026-10-31 | | A randomised trial by the Behavioural Insights Team (for the Department for Communities, Northern Ireland, 2018) rewrote lender arrears letters in a clearer, helpful tone. Contact rose from 45.9% to 62.2% of households (+35.5%) in long-term arrears, and from 24% to 31.3% (+30%) in another early-arrears trial. | OBSERVED | [BIT report](https://www.bi.team/wp-content/uploads/2019/02/20180704-BIT-Final-Report-R1.pdf), accessed 2026-10-31 | | **Implication:** whether a borrower makes contact depends heavily on how the lender writes to them. A score that rewards "communication" will partly measure the lender's letters, not the customer's character. | PREDICTION (my inference from the BIT trial) | as above | | Under PS24/2 (rules in force from November 2024), lender support duties cover customers at risk of payment difficulty, not only those already in arrears, and firms must be transparent about forbearance options. | OBSERVED (excerpt) | [FCA PS24/2](https://www.fca.org.uk/publications/policy-statements/ps24-2-strengthening-protections-borrowers-financial-difficulty), accessed 2026-10-31 | | StepChange press headlines: "Two in five people uncomfortable speaking to their mortgage lender or landlord about budget pressures" (September 2023). "Half of UK adults have experienced problem debt – yet almost half keep it secret" (March 2026). | OBSERVED (headlines only; no method shown) | [StepChange press releases](https://stepchange.org/media-centre/press-releases/effective-creditor-communications.aspx), accessed 2026-10-31 | | Early contact predicts recovery ("cure") from arrears: US industry and programme data say so, but I found no rigorous causal study. | CLAIM / UNKNOWN | [Harvard JCHS, Cutts & Merrill](https://www.jchs.harvard.edu/sites/default/files/ucc08-15_cutts_merrill.pdf), accessed 2026-10-31 | | Whether "did all they could" conduct predicts future repayment better than the arrears record alone | UNKNOWN. This is the central hypothesis EDGE must test (section 7.3). | n/a | --- ##### 4. Regulation and constraints ###### 4.1 FCA credit information market study: remedies and 2026 status | Line | Label | Source | |---|---|---| | The Final Report (MS19/1.3) was published in December 2023 with a remedies package: 1 reformed governance (CRGB, now CIGB); 2A mandatory sharing with designated CRAs; 2B common data format; 2C designated CRA reporting to the FCA; 2D data contributor accuracy and dispute duties; 3A signposting to statutory credit reports; 3B one-stop access to statutory reports; 3C streamlined disputes; 3D streamlined NoC, non-financial vulnerability markers and credit freezes; 4A more timely reporting; 4B review of the Principles of Reciprocity; 4C improved current account turnover data. | OBSERVED | [FCA MS19/1.3 Table 1](https://www.fca.org.uk/publication/market-studies/ms-19-1-3.pdf), accessed 2026-10-31 | | CP26/7 (February 2026) consults on 2A and 2D. Lenders sharing with one DCCRA must share all consumer credit information with all DCCRAs. The FCA proposes designating Equifax Ltd, Experian Ltd and TransUnion International UK Ltd in a new "DCCRA Sourcebook". Fully paid CCJs must be reported as satisfied. The consultation closed on 1 May 2026. | OBSERVED | [FCA CP26/7](https://www.fca.org.uk/publication/consultation/cp26-7.pdf); [FCA press release](https://www.fca.org.uk/news/press-releases/fca-proposes-action-close-gaps-borrowers-credit-files), accessed 2026-10-31 | | Proposed timing: rules come into force 12 months after the Policy Statement. First-time providers get six months after that. | OBSERVED | [FCA CP26/7 ch. 6](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | | FCA central estimate of costs: £63.4m present value over 10 years (3.5% discount rate). The FCA did not put a number on the benefits. | OBSERVED | [FCA CP26/7 Annex 2](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | | The CIGB "has now been formed and will be fully operational by summer 2026". It has joined UKRN (the UK Regulators Network). | OBSERVED | [FCA CP26/7 para 2.7](https://www.fca.org.uk/publication/consultation/cp26-7.pdf); [UKRN](https://ukrn.org.uk/ukrn-welcomes-new-member/), accessed 2026-10-31 | | The CIGB board has four industry-nominated directors representing different constituencies. | OBSERVED (excerpt) | [Loqbox](https://www.loqbox.com/en-gb/news/tim-porter-appointed-as-a-director-of-the-new-credit-information-governance-body-cigb), accessed 2026-10-31 | | Final rules (a Policy Statement on CP26/7): I found none published by 31 October 2026. One commentary says they are expected at the end of 2026. | UNKNOWN / CLAIM | [Deloitte blog](https://www.deloitte.com/uk/en/Industries/financial-services/blogs/treating-the-credit-file-as-market-infrastructure.html), accessed 2026-10-31 | | The Smaller Business Practitioner Panel (28 April 2026) asked the FCA and CIGB for a clear timeline for the common data format and a review within 24 months of go-live. Fair4All Finance asked for a single portal for sharing with all DCCRAs. | OBSERVED (excerpt) | [SBPP response](https://www.fca.org.uk/panels/smaller-business-practitioner-panel/publication/sbpp-response-to-cp26-7-credit-information-market-study-proposed-approach-to-implementing-fca-remedies.pdf); [Fair4All Finance](https://fair4allfinance.org.uk/fair4all-finance-responds-to-the-fcas-proposals-on-mandatory-credit-reporting/), accessed 2026-10-31 | | Deferred Payment Credit (Buy Now Pay Later) enters FCA regulation from 15 July 2026 (PS26/1). The CP26/7 rules would apply to it. | OBSERVED | [FCA CP26/7 para 2.17](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | | The Government consulted on Phase 1 of Consumer Credit Act reform in May 2025. | OBSERVED | [FCA CP26/7 para 2.18](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | ###### 4.2 Consumer Credit Act rights and FCA permissions | Line | Label | Source | |---|---|---| | CCA s158: a consumer can get a copy of their file from a CRA. The statute still states a £2 fee. | OBSERVED | [CCA s158](https://www.legislation.gov.uk/ukpga/1974/39/section/158), accessed 2026-10-31 | | CCA s159: a consumer can ask for wrong entries to be removed or amended. If the CRA does neither, the consumer can add a Notice of Correction of up to 200 words. CRAs have 28 days to respond. | OBSERVED (excerpt) + FCA summary | [CCA s159](https://www.legislation.gov.uk/ukpga/1974/39/section/159); [FCA CP26/7 Annex 2 para 59](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | | Regulated activity "Providing credit references" (RAO Article 89B): furnishing information relevant to the financial standing of individuals or relevant recipients of credit, where the information was collected for that purpose. | OBSERVED | [RAO art 89B](https://www.legislation.gov.uk/uksi/2001/544/article/89B), accessed 2026-10-31 | | Regulated activity "Providing credit information services" (Article 89A): taking steps for an individual to check, correct or restrict what a credit information agency holds about them. | OBSERVED | [RAO art 89A](https://www.legislation.gov.uk/uksi/2001/544/article/89A), accessed 2026-10-31 | | **Implication:** an EDGE confidence report about individuals, furnished to lenders, is very likely "providing credit references" and needs FCA authorisation. CCA s158–159 style rights would then apply to it. | PREDICTION (needs legal confirmation) | as above | ###### 4.3 UK GDPR Article 22 and the Data (Use and Access) Act 2025 | Line | Label | Source | |---|---|---| | DUAA section 80 replaced UK GDPR Article 22 with Articles 22A–22D. A decision is "based solely on automated processing" if there is no "meaningful human involvement". It is "significant" if it has a legal or similarly significant effect. | OBSERVED | [DUAA s80](https://www.legislation.gov.uk/ukpga/2025/18/section/80), accessed 2026-10-31 | | Article 22B: a significant solely automated decision based wholly or partly on special category data (for example health) needs explicit consent, or contract or legal necessity plus Article 9(2)(g). Such decisions are also barred where the processing relies on Article 6(1)(ea) (recognised legitimate interests). | OBSERVED | [DUAA s80](https://www.legislation.gov.uk/ukpga/2025/18/section/80), accessed 2026-10-31 | | Article 22C safeguards: the controller must give information about the decision, let the person make representations, provide human intervention, and let the person contest the decision. | OBSERVED | [DUAA s80](https://www.legislation.gov.uk/ukpga/2025/18/section/80), accessed 2026-10-31 | | In force from 5 February 2026 (SI 2026/82). It does not apply to decisions taken before that date. | OBSERVED | [SI 2026/82](https://www.legislation.gov.uk/uksi/2026/82/made), accessed 2026-10-31 | | The ICO consulted on draft ADM and profiling guidance in spring 2026. Whether final guidance has been published | OBSERVED (consultation) / UNKNOWN (final) | [DMA on ICO consultation, 21 April 2026](https://www.dma.org.uk/about/articles/ico-seeks-views-on-draft-automated-decision-making-guidance), accessed 2026-10-31 | ###### 4.4 Equality Act indirect discrimination and the Consumer Duty | Line | Label | Source | |---|---|---| | Equality Act 2010 s19: applying a "provision, criterion or practice" that puts people sharing a protected characteristic at a particular disadvantage is indirect discrimination, unless it is "a proportionate means of achieving a legitimate aim". The characteristics covered are age, disability, gender reassignment, marriage and civil partnership, race, religion or belief, sex and sexual orientation. | OBSERVED | [Equality Act s19](https://www.legislation.gov.uk/ukpga/2010/15/section/19), accessed 2026-10-31 | | Whether and how the financial services age exceptions apply to a confidence score | UNKNOWN. Needs counsel. | n/a | | Consumer Duty (PS22/9): firms must "act to deliver good outcomes for retail customers", with cross-cutting rules and four outcomes (products and services, price and value, consumer understanding, consumer support). In force 31 July 2023 for open products and 31 July 2024 for closed products. | OBSERVED | [FCA PS22/9](https://www.fca.org.uk/publications/policy-statements/ps22-9-new-consumer-duty), accessed 2026-10-31 | | The FCA set out Consumer Duty expectations on signposting consumers to statutory credit reports. | OBSERVED | [FCA CP26/7 para 1.7](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | ###### 4.5 United States: FCRA and ECOA adverse action | Line | Label | Source | |---|---|---| | FCRA §1681m(a): after adverse action based on a consumer report, the lender must give notice, disclose the numerical score used, name the reporting agency (and say it did not make the decision), and explain the right to a free report within 60 days and the right to dispute. | OBSERVED | [15 U.S.C. 1681m](https://www.law.cornell.edu/uscode/text/15/1681m), accessed 2026-10-31 | | Reg B §1002.9: notice within 30 days of a completed application. Reasons must be "specific and indicate the principal reason(s)". "Failed to achieve a qualifying score" is insufficient. The official interpretation says more than four reasons is "not likely to be helpful". | OBSERVED | [12 CFR 1002.9](https://www.consumerfinance.gov/rules-policy/regulations/1002/9/), accessed 2026-10-31 | | On 12 May 2025 the CFPB withdrew Circulars 2022-03 and 2023-03 (adverse action with complex algorithms) among 67 guidance documents. The statutory duty to give specific reasons remains. | OBSERVED (excerpt) | [CFPB withdrawn guidance](https://www.consumerfinance.gov/compliance/guidance/withdrawn-guidance/), accessed 2026-10-31 | | The CFPB finalised a Reg B rule (published April 2026, effective 21 July 2026) that removes disparate-impact ("effects test") liability under ECOA. It does not affect the Fair Housing Act or state law. The rule is under legal challenge. | AGREEMENT (several law-firm summaries) | [Venable](https://www.venable.com/insights/publications/2026/05/cfpb-makes-significant-changes-to-regulation-b); [Orrick InfoBytes](https://infobytes.orrick.com/2026-04-24/cfpb-finalizes-amendments-to-regulation-b-eliminating-disparate-impact-liability-provisions/), accessed 2026-10-31 | | The CFPB open banking rule (Section 1033): a court stayed its compliance dates on 29 October 2025, and the CFPB is reconsidering it. | OBSERVED (excerpt) | [CFPB personal financial data rights](https://www.consumerfinance.gov/compliance/compliance-resources/other-applicable-requirements/personal-financial-data-rights/), accessed 2026-10-31 | ###### 4.6 EU AI Act: creditworthiness is high-risk | Line | Label | Source | |---|---|---| | Annex III point 5(b): AI systems "intended to be used to evaluate the creditworthiness of natural persons or establish their credit score" are high-risk, except systems for detecting financial fraud. | OBSERVED | [AI Act Annex III](https://artificialintelligenceact.eu/annex/3/), accessed 2026-10-31 | | The Digital Omnibus on AI: political agreement in May 2026 moved Annex III high-risk obligations to 2 December 2027 (Annex I to 2 August 2028). | AGREEMENT | [Gibson Dunn, 27 May 2026](https://www.gibsondunn.com/eu-ai-act-omnibus-agreement-postponed-high-risk-deadlines-and-other-key-changes/), accessed 2026-10-31 | | Secondary sources say the omnibus was formally adopted and entered into force in July 2026. | CLAIM (not checked against the Official Journal) | [Usercentrics](https://usercentrics.com/knowledge-hub/eu-ai-act-high-risk-delay-article-50-transparency-consent/), accessed 2026-10-31 | ###### 4.7 Risks of behavioural and communication signals, and design rules that follow | Risk | Evidence | Design rule (PREDICTION) | |---|---|---| | **Fairness: proxy discrimination.** Response speed, channel, language or tone can track disability, age, race or caring duties. | Equality Act s19 justification test ([s19](https://www.legislation.gov.uk/ukpga/2010/15/section/19), accessed 2026-10-31), OBSERVED. Scores are already noisier for under-served groups ([Blattner & Nelson](https://arxiv.org/abs/2105.07554), accessed 2026-10-31), OBSERVED (excerpt). | Score only **outcome facts** (an arrangement agreed and kept, contact made within the period the lender offered). Never score sentiment, tone, word choice or voice. Test every input for disparate impact by protected characteristic before use, and keep the justification on file. | | **Lender quality leaks into the customer's score.** | Letter wording alone moved contact rates by 30–35% ([BIT](https://www.bi.team/wp-content/uploads/2019/02/20180704-BIT-Final-Report-R1.pdf), accessed 2026-10-31), OBSERVED. | Count conduct only where the lender's outreach met a minimum standard (logged by the lender). Otherwise record "not assessable", never a negative. | | **Special category data** (health, which drives vulnerability) | Art 22B limits ([DUAA s80](https://www.legislation.gov.uk/ukpga/2025/18/section/80), accessed 2026-10-31), OBSERVED. The FCA final report notes debate on whether non-financial vulnerability markers push CRAs into sensitive areas ([MS19/1.3 Remedy 3D](https://www.fca.org.uk/publication/market-studies/ms-19-1-3.pdf), accessed 2026-10-31), OBSERVED. | The cause is recorded in coarse categories ("income shock", "health", "bereavement", "dispute", "fraud/coerced debt"). A health cause is used only to **neutralise** a negative, never to lower a score, and only with explicit consent. A human reviews any decision that uses it. | | **Consent fatigue and coercion** | Open Banking relies on consumer permission ([FCA CP26/7 2.14](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31), OBSERVED. | Treat the confidence layer as opt-in **evidence for the customer**. Refusing consent must never be scored as a negative. | | **Gaming** | Self-reported explanations are easy to fake. The NoC is unverified and unscored today ([Experian NoC](https://www.experian.co.uk/consumer/guides/notice-of-correction.html), accessed 2026-10-31), OBSERVED (excerpt). | Only **verified** events move the score: lender-attested arrangements, payment data, documented causes. Free-text narrative is shown to the lender but never scored. | | **Discouraging help-seeking** | Only 22% of people in difficulty used debt advice ([FLS 2024](https://www.fca.org.uk/publication/financial-lives/financial-lives-survey-2024-key-findings.pdf), accessed 2026-10-31), OBSERVED. | Using debt advice, Breathing Space or forbearance is never a negative in itself. Keeping to the agreed plan is a positive. | | **Privacy and over-collection** | UK GDPR safeguards for automated decisions ([DUAA s80](https://www.legislation.gov.uk/ukpga/2025/18/section/80), accessed 2026-10-31), OBSERVED. | Minimise data. Raw transactions stay with the source. The report holds derived, explainable features with retention limits. | --- ##### 5. Evidence that richer data improves outcomes | Study or product | Result | Label and source | |---|---|---| | **FinRegLab (2019)**: six non-bank lenders | Cash-flow scores and attributes were generally at least as predictive as traditional credit scores, and often improved prediction when combined with them. Predictive power was relatively consistent across groups defined by likely race, ethnicity and gender. | OBSERVED (excerpt) [FinRegLab empirical findings](https://finreglab.org/cash-flow-data-in-underwriting-credit-empirical-research-findings), accessed 2026-10-31 | | **FinRegLab with NYU Stern (June 2025)**: over 38,000 small business loans, 2015–2024 | Adding cash-flow variables improved default prediction across all segments. The biggest gains were for low-score owners of businesses under five years old. Deposits and balances were the strongest variables; balance volatility and low or negative balance incidents were also predictive. | OBSERVED [FinRegLab press release](https://finreglab.org/wp-content/uploads/2025/06/FinRegLab_2025-06-03_Press-Release_Study-Shows-Cash-Flow-Data-Can-Expand-Small-Business-Lending-1.pdf), accessed 2026-10-31 | | **Upstart under the CFPB No-Action Letter (2019)** | 27% more applicants approved and 16% lower average APRs (annual percentage rates). Near-prime applicants (FICO 620–660) were approved about twice as often; under-25s were 32% more likely to be approved; incomes under $50k were 13% more likely. No disparities requiring further fair lending analysis. The data was lender-provided and not replicated by the CFPB. | OBSERVED (CFPB page) / CLAIM (numbers are lender-reported) [CFPB blog](https://www.consumerfinance.gov/about-us/blog/update-credit-access-and-no-action-letter/), accessed 2026-10-31 | | **Experian Boost (US, February 2021)** | 3.9 million users. 60% saw an increase, averaging 12 points. 47% of unscoreable users became scoreable. Of thin-file users, 85% saw an increase (average 19 points) and 41% left thin-file status. | OBSERVED (Experian's own study) / CLAIM (not independently audited) [Experian](https://www.experian.com/blogs/ask-experian/experian-boost-study/), accessed 2026-10-31 | | **UltraFICO, next generation (generally available 20 May 2026)** | A 7% relative increase in approvals "with no incremental risk". A 15% relative lift for prime applicants with limited histories. 79% of non-prime applicants with positive balances see higher scores. No method or sample is disclosed. | CLAIM [FICO investor release](https://investors.fico.com/news-releases/news-release-details/next-generation-ultraficor-score-now-available/), accessed 2026-10-31 | | **Plaid LendScore** | Average 25% lift in predictive performance alongside traditional credit data. | CLAIM [Plaid, October 2026](https://plaid.com/blog/instant-link-next-generation-lendscore/), accessed 2026-10-31 | | **Nova Credit / Atlanticus** | 15% more approvals with performance in line with risk expectations. | CLAIM [Nova Credit case study](https://www.novacredit.com/corporate-blog/case-study-how-atlanticus-unlocked-15-more-approvals-while-maintaining-risk), accessed 2026-10-31 | | **More comprehensive credit information** (Barci et al. 2019; Blattner et al. 2022, as cited by the FCA) | Leads to higher lending volumes and wider access to credit in aggregate. | OBSERVED (FCA citing studies) [FCA CP26/7 Annex 2 para 67](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | | **Lender support in difficulty** | 43% of credit holders who got lender support said their situation improved. | OBSERVED [FLS 2024](https://www.fca.org.uk/publication/financial-lives/financial-lives-survey-2024-key-findings.pdf), accessed 2026-10-31 | | **Evidence that "conduct in difficulty" data improves default prediction** | No public study found. | UNKNOWN | AGREEMENT: Independent (FinRegLab, FCA-cited academic work) and vendor sources agree on the direction: more data means better prediction and more approvals at the same risk. The size of the effect varies, and most of the large numbers are vendor CLAIMs. --- ##### 6. The design: what a confidence report contains All of section 6 is **PREDICTION**: my proposed design. It is built from the constraints sourced above and has not been tested. ###### 6.1 Contents of a confidence report (person) | Section | What it shows | Why (sourced constraint) | |---|---|---| | **A. Identity assurance** | The level of identity verification, its method, and its date. This is the link to EDGE ID. | Credit data is already used in ID verification and KYC/AML; gaps cause harm there ([FCA CP26/7 Annex 2 para 76](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31). | | **B. Obligation record** | The same facts as a credit file: accounts, payments, defaults, CCJs (with satisfied status). | Must reconcile with DCCRA data under CP26/7 ([CP26/7](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31). | | **C. Conduct when things went wrong** | For each adverse event: (1) contact made within the period the lender offered, yes/no/not assessable; (2) arrangement agreed, yes/no; (3) arrangement kept, as a % of agreed payments; (4) cause category, verified/unverified; (5) whether it is resolved, and when. | This fills the gap the FCA found in how difficulty is reported ([MS19/1.3 3.50–3.52](https://www.fca.org.uk/publication/market-studies/ms-19-1-3.pdf), accessed 2026-10-31). | | **D. Resilience (opt-in)** | Cash-flow features from Open Banking: income stability, balance buffer, and incidents of low or negative balance. | Evidence of predictive value ([FinRegLab 2025](https://finreglab.org/wp-content/uploads/2025/06/FinRegLab_2025-06-03_Press-Release_Study-Shows-Cash-Flow-Data-Can-Expand-Small-Business-Lending-1.pdf), accessed 2026-10-31). | | **E. Confidence band** | The score plus how sure the report is, based on how much data it has (thin data gives a wide band). | Scores are noisier for thin files ([Blattner & Nelson](https://arxiv.org/abs/2105.07554), accessed 2026-10-31). | | **F. Customer statement** | A narrative of up to 200 words. It is shown but never scored. | Mirrors the NoC ([CCA s159](https://www.legislation.gov.uk/ukpga/1974/39/section/159), accessed 2026-10-31). | | **G. Reason codes** | Up to four principal reasons, in plain English, that moved the score. | Reg B specific-reasons standard ([12 CFR 1002.9](https://www.consumerfinance.gov/rules-policy/regulations/1002/9/), accessed 2026-10-31). | | **H. Provenance** | For every item: the source, the date, whether it is verified, and the model version. | Supports contest and audit (Art 22C). | ###### 6.2 How a one-off event is handled - An "event window" groups linked adverse marks from one cause (for example three missed payments after a job loss) into **one event**, not three. - If section C shows contact, an arrangement, and at least about 90% of the agreed plan kept, the event's weight decays faster. If not, it decays at the standard rate. The decay schedule is published. - A verified fraud or coerced-debt cause sets the event's weight to zero, pending review. This links to the Financial Inclusion Strategy work ([CP26/7 2.11](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31). - The report shows "1 event, resolved, conduct: strong" rather than a list of marks. - Assumptions: the 90% threshold and the decay rates are placeholders. They must be set from pilot data (section 7.3). ###### 6.3 Explanation, contest and audit | Duty | Mechanism | Legal anchor | |---|---|---| | Explain | Reason codes (6.1G) plus a "what would change this" line | Art 22C information ([DUAA s80](https://www.legislation.gov.uk/ukpga/2025/18/section/80)); FCRA and Reg B notices ([1681m](https://www.law.cornell.edu/uscode/text/15/1681m)), accessed 2026-10-31 | | Representations | The customer adds evidence (for example a redundancy letter) through EDGE ID | Art 22C representations | | Contest | A named human reviews within 28 days, matching the s159 timescale. The outcome is logged and the report re-issued. | [CCA s159](https://www.legislation.gov.uk/ukpga/1974/39/section/159); Art 22C contest and human intervention, accessed 2026-10-31 | | Audit | Versioned model, data lineage, quarterly fairness tests by protected characteristic, and a written s19 justification for each feature | [Equality Act s19](https://www.legislation.gov.uk/ukpga/2010/15/section/19); [AI Act Annex III 5(b)](https://artificialintelligenceact.eu/annex/3/) for EU use from 2 December 2027 | | Regulator reporting | Match whatever Remedy 2C asks of DCCRAs | [CP26/7 ch. 7](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | ###### 6.4 How it serves KYC - Section A ties every confidence report to a verified EDGE ID identity. This cuts mismatches between people (the FCA notes that "correct" data can be matched to the wrong person: [CP26/7 Annex 2 para 58](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31). - EDGE ID should be certified against the UK digital verification services trust framework. Version 1.0 is current ([GOV.UK trust framework collection](https://www.gov.uk/government/collections/uk-digital-identity-and-attributes-trust-framework), OBSERVED, accessed 2026-10-31). - Conduct history adds a behavioural consistency check for ongoing due diligence. It must never be used as a substitute for statutory CDD (customer due diligence). Needs counsel. ###### 6.5 Organisations - The same structure as for people, with business facts: payment days beyond terms (comparable to PAYDEX: [Altares](https://www.altares.nl/en/our-data/rating-scores/paydex/), CLAIM, accessed 2026-10-31), filed accounts, directors and ownership (comparable to [Creditsafe](https://www.creditsafe.com/gb/en.html), accessed 2026-10-31), and conduct in difficulty (did the business tell its suppliers, agree terms, keep them). - Align with the Commercial Credit Data Sharing scheme and the FCA's 2026 open finance plan for SME lending ([CP26/7 2.12–2.15](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31). ###### 6.6 AI agents as customers | Element | Proposal | Evidence anchor | |---|---|---| | Principal binding | Every agent report names its verified principal (a person or organisation) and inherits that principal's identity assurance, **not** their confidence score. | n/a (PREDICTION) | | Mandate record | Signed mandates in the AP2 style (an "intent mandate" for the authority, a "cart mandate" for the exact purchase) give a non-repudiable audit trail. | OBSERVED [Google Cloud, AP2, 16 September 2025](https://cloud.google.com/blog/products/ai-machine-learning/announcing-agents-to-payments-ap2-protocol), accessed 2026-10-31 | | Agent authenticity | Cryptographic request signing (Visa's Trusted Agent Protocol uses RFC 9421 message signatures) lets a counterparty confirm it is the registered agent. | OBSERVED [Visa developer](https://developer.visa.com/capabilities/trusted-agent-protocol), accessed 2026-10-31 | | Agent conduct metrics | The share of mandates executed within their limits, disputes and chargebacks, revocations, how the agent behaved when payment failed (did it stop and escalate to its principal), and how fast it did so. | PREDICTION | | Legal status | The CCA, FCRA and AI Act Annex III 5(b) are written for natural persons or recipients of credit. Whether agent reports fall under them | UNKNOWN. Needs counsel. ([AI Act Annex III](https://artificialintelligenceact.eu/annex/3/), accessed 2026-10-31) | | Spill-over | An agent's failures should appear on the principal's report only as a principal-attested event, because the principal is liable. | PREDICTION | See [COMPETITORS.md](COMPETITORS.md) for agent payment rails and agent identity providers. --- ##### 7. Business and economic outcomes ###### 7.1 Sourced figures | Figure | Label | Source | |---|---|---| | About 210 million outstanding regulated consumer credit agreements (over £300bn), and about 9 million regulated mortgages (over £1.4tn) | OBSERVED | [FCA CP26/7 Annex 2 para 21](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | | About £2bn of unmet but commercially viable credit need in 2023; invisible and thin-file consumers are nearly two-thirds of the group | OBSERVED (FCA citing L.E.K.) | [CP26/7 Annex 2 para 69](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | | FCA reform costs £63.4m present value over 10 years; benefits not quantified | OBSERVED | [CP26/7 Annex 2](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | | 18% of adults over-indebted; arrears wellbeing impact about 0.8 times that of unemployment | OBSERVED | [CP26/7 Annex 2 paras 74–75](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | | Approval uplifts at constant risk: 6.3% to 27%, from vendor and lender sources | CLAIM | Section 5 | | US: 7.0 million credit invisible plus about 9.8% of adults unscored (2020) | OBSERVED | [CFPB](https://files.consumerfinance.gov/f/documents/cfpb_update-credit-invisibles-estimate_2025-06.pdf), accessed 2026-10-31 | The 6.3% figure is a CLAIM: [Open Banking Expo on Plaid LendScore](https://www.openbankingexpo.com/news/plaid-unveils-new-ai-models-to-improve-lending-decisions-and-reduce-payment-risk/), accessed 2026-10-31. ###### 7.2 Own estimates (PREDICTION) | Estimate | Assumptions | Result | |---|---|---| | **P1. UK thin-file credit unlocked by a lender using confidence reports** | (a) The £2bn L.E.K. unmet viable need is the ceiling. (b) Thin-file/invisible share is about two-thirds, so about £1.3bn. (c) EDGE partner lenders reach 5–15% of that population within three years. (d) Confidence reports convert 25–50% of reached viable need. | About £16m–£100m of extra annual lending. Formula: £1.3bn × 5–15% × 25–50%. | | **P2. Lender approval uplift** | Uses the lower half of the claimed range (6–15%), because a conduct layer is untested and the vendor numbers are likely optimistic. | +6% to +15% approvals at equal expected loss for thin-file applicants. Must be proven by a holdout test. | | **P3. Avoided write-offs from earlier engagement** | If clearer reporting of forbearance raises early contact by even a third of the BIT letter effect (about +10% relative), and contact improves cure. That link is UNKNOWN. | Not quantified. I will not invent a loss number until the pilot measures the contact-to-cure link. | | **P4. Consumer wellbeing** | Arrears harm is about 0.8 times unemployment (FCA-cited). A one-event-fades rule shortens the period of restricted access. | Direction positive; size UNKNOWN. Measure with FLS-style wellbeing questions in the pilot. | ###### 7.3 How to prove it (measured outcomes) - PREDICTION: Run a pilot with one or more FCA-authorised lenders. Randomly split applicants into **holdout** (the current process) and **treatment** (current process plus the confidence report). Measure approval rate, 12-month default rate, loss given default, cure rate after arrears, complaints and disputes, time to resolve disputes, and approval and default by protected characteristic (fairness). Pre-register the metrics. - PREDICTION: The key hypothesis to test: among borrowers with the same arrears history, those with strong section C conduct have lower subsequent default. If this is false, section C must not move the score; it can only inform a human reviewer. - PREDICTION: Publish the results (including null results) with an independent evaluator, for example a university or FinRegLab-style non-profit. --- ##### 8. How to become the standard-setter | Step | Detail | Label and source | |---|---|---| | 1. Get the permission | Apply for FCA authorisation for "providing credit references" (RAO Article 89B), and Article 89A if EDGE helps people correct their files. | OBSERVED (activities) [RAO 89B](https://www.legislation.gov.uk/uksi/2001/544/article/89B); [89A](https://www.legislation.gov.uk/uksi/2001/544/article/89A), accessed 2026-10-31; route is a PREDICTION | | 2. Aim for designation later | The FCA will publish non-exhaustive designation factors and can add DCCRAs over time. It will "closely monitor" the effect on competition. | OBSERVED [CP26/7 ch. 4](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | | 3. Join the governance | Engage the CIGB (common data format, NoC and vulnerability marker remedies 2B and 3D) and the Principles of Reciprocity review (4B). | OBSERVED (remedies) [MS19/1.3](https://www.fca.org.uk/publication/market-studies/ms-19-1-3.pdf); [UKRN](https://ukrn.org.uk/ukrn-welcomes-new-member/), accessed 2026-10-31 | | 4. Publish an open specification | A "Confidence Report Specification" under an open licence: the data model (section 6.1), event-window rules, reason code list, contest workflow, fairness test suite, agent mandate schema, and conformance tests. Offer it as an extension field set for the common data format (Remedy 2B). | PREDICTION | | 5. Ride the open finance roadmap | The FCA committed to an open finance plan in 2026 that prioritises SME lending. | OBSERVED [CP26/7 2.15](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31 | | 6. Who to convince | **FCA** (credit policy, open finance), **ICO** (ADM guidance), **CIGB**, **the three CRAs** (as data partners, not enemies), lenders large and small (the SBPP wants proportionate timelines), consumer and inclusion groups (Fair4All Finance, StepChange), the **Financial Ombudsman** (which already applies the arrears reporting principles), and HM Treasury (Commercial Credit Data Sharing scheme and Financial Inclusion Strategy). | OBSERVED (that these bodies are active: [SBPP](https://www.fca.org.uk/panels/smaller-business-practitioner-panel/publication/sbpp-response-to-cp26-7-credit-information-market-study-proposed-approach-to-implementing-fca-remedies.pdf), [Fair4All](https://fair4allfinance.org.uk/fair4all-finance-responds-to-the-fcas-proposals-on-mandatory-credit-reporting/), [FOS](https://www.financial-ombudsman.org.uk/decision/DRN-5907555.pdf), [CP26/7](https://www.fca.org.uk/publication/consultation/cp26-7.pdf), accessed 2026-10-31); the target list is a PREDICTION | | 7. Win on evidence | Publish the pilot (section 7.3). A regulator will not adopt "conduct in difficulty" scoring without causal evidence. | PREDICTION | | 8. US and EU later | US: FCRA reseller or consumer reporting agency obligations, plus Reg B reasons. EU: AI Act high-risk conformity from 2 December 2027. | OBSERVED (rules) sections 4.5–4.6; sequencing is a PREDICTION | --- ##### 9. Open questions (UNKNOWN) 1. Will the FCA's final Policy Statement on CP26/7 change the designation list or the timing? (None found by 31 October 2026.) 2. Does conduct in difficulty predict future repayment, after controlling for the arrears record? There is no public evidence. 3. Is an AI agent's "confidence report" a consumer credit reference in law? Needs UK, US and EU counsel. 4. How do the financial services age exceptions in the Equality Act interact with conduct features? 5. What is Credit Kudos's current status, and how do Creditsafe and Equifax/TransUnion build their products? Not verified, because the search budget ran out. 6. Will the ICO's final ADM guidance define "meaningful human involvement" in a way that affects contest workflows? ### research/finance/ENGINE-BY-STARLING.md `research/finance/ENGINE-BY-STARLING.md` · sha256 `f94f7b80899a5b410535ee4e5b321681c37ae2fa1e9b4be5e99243246c8da4ad` #### Engine by Starling: what it is, what it offers, and the parity target for EDGE Statement class: research (OBSERVED unless marked CLAIM or UNKNOWN). Accessed 1 November 2026. Never a rule. ##### 1. What it is - OBSERVED: a cloud-native SaaS core banking platform, "built and fully-managed by us in a dedicated environment", sold to licensed banks [platform](https://enginebystarling.com/platform/). - OBSERVED: no public API, sandbox, self-serve sign-up or pricing on the site; the only routes in are a contact form and a Starling B2B e-mail for UK banking-as-a-service enquiries [home](https://enginebystarling.com/). - OBSERVED: delivery partners PwC, Deloitte and GFT; "rapid delivery timelines of under twelve months"; PwC says each client project typically completes in 12 months [PwC case study](https://www.pwc.com/gx/en/about/case-studies/engine-by-starling-launching-bank-software-business.html). - OBSERVED: clients named on the site: Starling (since 2016), Salt Bank (Romania), AMP Bank GO (Australia), Tangerine (Canada, 10-year deal, November 2025), SBS Bank (New Zealand, 10-year deal, February 2026) [Tangerine](https://www.fintechfutures.com/core-banking-technology/tangerine-bank-in-core-banking-upgrade-with-engine), [SBS](https://thepaypers.com/fintech/news/starlings-engine-platform-partners-with-sbs-bank). - OBSERVED: Starling's 2026 annual report shows Engine committed ARR of £70m with a £100m target in 12 months (see COMPETITORS.md §2.2). - UNKNOWN: licence terms and price per account. Conclusion: Engine is bought by a bank that already holds a licence, on a multi-year contract, after a 12-month implementation. It is a competitor's reference architecture to match, not a component EDGE can plug in today. The plug-in sponsor-bank route is Griffin (section 3). ##### 2. Engine's published capability set and EDGE's answer | Engine capability (OBSERVED on the platform page) | EDGE unit or mechanism that answers it | Status | |---|---|---| | Core banking: transactions, balances, customer data, "complex general ledger" | Estate ledger and receivers (runtime/receivers.py, ledger/evidence); account record with booked, held and available balances (finance/open-a-personal-current-account) | Live, estate-internal | | Products: spending, saving, borrowing, investing, "toggled on or off and launched in stages" | finance units: open personal current account, savings, apply for loan, send money, cards; capabilities are units, each switchable by release | Specified and rendered; rails via plug | | "Comprehensive payment and card processing" | payment-scheme and card-processor slots (spec/plugins.json); Griffin for Faster Payments, BACS, CHAPS; card processor post-closure | Plug slots exist; unbound | | Digital onboarding (Tangerine scope) | open-a-personal-current-account steps 1 to 8 (EDGE ID and device, structured identity, KYC, consents) | Live screens; KYC via Griffin Verify plug | | Overdrafts, debit cards, money management (Tangerine scope) | finance/apply-for-loan, cards, money provenance units added 31 Oct | Specified | | Management portal: employee workflows that run the bank end to end | operator and assessor actors on every unit; evidence, done and checks screens; estate book | Live | | Self-serve customer journeys and experiences | every unit's person journey (5,828 planned, HTTP-walked) | Live | | 360° customer view, near real-time data, AI integration | TOK dependency index, evidence panels, Jude | Live; index STALE until rebuilt | | "Rich, granular APIs"; third-party integration | Hub MCP (runtime/hub_mcp.py), API dispatch through the TOK binding, plug registry | Live | | Mobile SDKs | device plans (pending; see EXPERIENCE catalogue) | Pending | | Resilience by design, 24/7, no customer downtime; releases without loss of service | deploy MCP edits with health gate and backup; release record | Live for the web tier; HA not proven | | SOC 1, SOC 2, ISO 27001 badges | not held | Gap (certification, not code) | ##### 3. The sponsor-bank route that exists today: Griffin - OBSERVED: Griffin Bank Ltd "is authorised by the Prudential Regulation Authority and regulated by the Financial Conduct Authority and Prudential Regulation Authority"; "Our own core banking platform and license" [griffin.com](https://griffin.com/). - OBSERVED: "Free sandbox", "Start testing right away", sign-up at https://app.griffin.com/register; docs at https://docs.griffin.com [griffin.com](https://griffin.com/). - OBSERVED: products: embedded bank accounts, savings, business and client-money accounts, safeguarding accounts for EMIs and PIs, Payments, Verify (automated onboarding and KYC); onboarding models Verify, Reliance and Outsourced; MCP server (beta) [docs](https://docs.griffin.com/), [onboarding](https://docs.griffin.com/docs/tags/onboarding). - OBSERVED: authentication `Authorization: GriffinAPIKey `; `GET /v0/index` returns the organization URL; optional `Griffin-Version: 2026-07-09` header; live organisations must sign requests [get started](https://docs.griffin.com/docs/guides/get-started-with-the-api). - EDGE binding: runtime/plugs/griffin.py answers the `payment-scheme` slot (142 receivers) and the KYC conditions; it binds itself the moment `/etc/x-x/plugs/griffin.key` exists on jude-1 (credentials never in chat or Git). ### research/finance/FCA-AGENTIC-FRAMEWORK.md `research/finance/FCA-AGENTIC-FRAMEWORK.md` · sha256 `54534225ef1a88de175d15611a33f8634f5cd5a34bfe35b77ee43fee3dd6adb0` #### The FCA's agentic finance framework against the estate, point by point Researched 31 October 2026. The FCA's promise comes from Jessica Rusu's speech of 19 October 2026 and the Mills Review of 6 July 2026. They describe a framework to be built with industry, other regulators and AI labs, with no date (CLAIM: speech text and trade press). The estate is a running system that already does each part on every call. Every row below is either measured, or enforced by the engine and checked by the estate's own programs. ##### The four headings of the trusted framework | FCA heading | What the FCA promises | What the estate does now | Measured or enforced | |---|---|---|---| | Agent identity | A framework to define it | Every agent has its own identity, birth certificate and lifecycle. Every call carries the actor's identity and kind. | 100% of calls checked first (`runtime/engine.py`, step 1). Missing identity is refused `UNAUTHORISED`. | | Delegated authority | A framework to define it | Every agent acts under a mandate naming its tools, limits, owner and expiry. Every change also has a rate and value limit set by its owner. | 100% of agent calls checked against the mandate (`OVER_MANDATE`). 100% of changes have an owner limit (`OVER_LIMIT`). | | Accountability | Existing rules (Consumer Duty, Senior Managers regime) | Every capability names an accountable human role, never an agent. Every agent account names its accountable owner. | 100% of capabilities (`ledger/vulnerabilities/README.md`). | | Execution | A framework to define it | Every operation follows one order: identity, mandate, inputs, checks, rehearsal, approval, effect, receiving confirmation, evidence. **High-risk agent outcomes are provisional until the owner signs off in person and the recipient agrees, or they reverse.** | Engine-enforced. Binance replay: `tests/replay_binance.py`. | ##### What the FCA and the Mills Review add | Promise | The estate | |---|---| | Incident reporting for significant AI failures and near misses | Every call, refusals included, leaves evidence with a reproducible digest. Near misses (refused, over limit, reversed) are counted per operation, so the report is generated, not compiled by hand. | | Continuous, conditional assurance | Assurance runs on every call: checks, limits, mandate and settlement. It also covers the estate as a whole: the vulnerability map, the parity sweep and the outcome register. | | Interoperability across data, identity and payments (Mills) | One estate file, one API description, and one agent tool per operation, shared across Finance, Clinic, Public Consortium and Judiverse. | | Liability questions (Mills) | Reversal by design: an outcome the owner has not signed off in person is never final, so there is no loss to allocate. | | An AI-enabled agentic supervisory model (Mills recommendation 6) | Regulator journeys and evidence screens in every capability. The regulator's agent can inspect and reproduce any decision through its own tool. | ##### What the framework does not have, and the estate does - **Provisional value with in-person sign-off:** a stolen key moves nothing final. Measured by replaying the 29 October Binance theft (about $6.36bn in about 20 minutes): see `ledger/benchmarks/binance-replay.json`. - **Outcome economics** on every call: words, prompts, tokens and deterministic share per verified outcome, with targets in `spec/standards.json`. - **Time to verified outcome** measured per operation (`ledger/benchmarks/README.md`). ##### Sources - [FCA speech: Building the future of finance in the age of AI](https://www.fca.org.uk/news/speeches/building-future-finance-age-ai) - [PYMNTS](https://www.pymnts.com/news/artificial-intelligence/2026/united-kingdom-financial-conduct-authority-plans-regulatory-framework-agentic-finance/) - [Freshfields](https://www.freshfields.com/en/our-thinking/briefings/2026/10/ai-in-financial-services-fca-good-and-poor-practice-and-the-road-to-agentic-finance) - [FinTech Global](https://fintech.global/2026/10/26/fca-builds-agentic-finance-rules-as-ai-reshapes-uk-markets/) - [A&O Shearman on the Mills Review](https://www.aoshearman.com/en/insights/report-issued-by-the-mills-review-the-future-of-ai-in-retail-financial-services) - [Stephenson Harwood on the Mills Review](https://www.stephensonharwood.com/post/102n942/agentic-finance-the-fca-and-a-brave-new-world-the-mills-review-2026/) ### research/finance/MOBBIN-FLOWS.md `research/finance/MOBBIN-FLOWS.md` · sha256 `8a85ca7fdf2837d47fe9e8da5ae056777c8ea4bc2407b695a9210eaf9f7b8859` #### EDGE Finance research: real bank flows and screens from Mobbin ##### What this is Real screens and multi-step flows from existing banking, payments, credit and crypto apps, recorded by Mobbin (a library of real app screenshots). Refero, used earlier for this application, lacked most UK and EU incumbents; Mobbin holds many of them. These records show the floor EDGE Finance must meet and where nobody yet goes. **Statement class.** Every item under "By task" is an **OBSERVATION**: what the Mobbin screenshots show, read from the preview images (not from Mobbin's metadata alone), searched on **2026-10-31** through the Mobbin search tools. Each item carries its `mobbin_url`. Only part of each flow was visible in previews (Mobbin returns a few evenly spaced screens per flow); a description covers what was seen, not every screen. Some apps were captured with a non-UK account (Singapore dollars for Revolut, Coinbase and Wise; US dollars for Klarna); this is noted where it applies. Personal details visible in screenshots (names, emails, addresses, account numbers) are deliberately left out. Product names belong to their owners. The two closing sections ("what incumbents do that EDGE Finance must match" and "what none of them do") are a **CLAIM** by this research, drawn only from the observations listed here. They describe this sample, not the whole market. ##### Coverage **Held by Mobbin and used here (iOS unless marked web):** | App | Region | Tasks found | |---|---|---| | Monzo | UK | account opening and identity, business account, request and split, scheduled payments and direct debits, savings pots, connecting other banks, disputes and help, undo payment, security, AI assistant | | Starling (listed as "Starling" and "Starling Bank") | UK | account opening, business onboarding, card freeze and controls, virtual cards, payee name check, scam warning, disputes and fraud, closing an account, mobile wallets | | Chase UK | UK | saver accounts, paying out, adding a payee with scam warning | | Revolut | UK/EU (some captures from its Singapore entity) | new payment and payee search, review with scam warning, virtual and disposable cards, card settings, crypto, security and devices, close account entry, chat assistant | | Wise | UK/EU (captured in Singapore dollars) | transfer calculator, sending and adding recipients, scheduled transfer, business onboarding | | Kraken | global | crypto send by paylink, crypto withdrawal to an address, account and security menu | | Coinbase | global (captured in Singapore dollars) | buying crypto, account checklist, explore | | Klarna | EU/US (captured in US dollars) | pay-later plan, payment calculator, purchase power, returns and problem reports | | Zopa Bank | UK | loan rates and pre-approval, income and employment | | Lloyds Mobile Banking | UK | credit report, reporting a transaction | | Zing | UK | purpose-of-payment scam check, card page | | N26 | EU (Germany) | account opening checklist | | bunq | EU (Netherlands) | adding a business account | | Vivid | EU (Germany) | card name check | | Credit Karma (UK score scale) | UK | credit score | | Cleo | UK/US | credit score simulator, chat assistant | **Used as non-UK references where UK/EU apps showed nothing:** ANZ Plus and Up (Australia), Kit (Australia), Neo Financial (Canada), Cash App, Chime, One, Mercury (web and iOS), Monarch, Copilot, Zip (US), Airwallex (web), Base (Coinbase's wallet), OKX, CRED (India), Careem (UAE). **Not held:** Curve. A search for "Curve card wallet" returned only Starling and Zing, and a screen search for "Curve app home screen" returned an unrelated app. Mobbin may hold more UK and EU banks than were searched here (for example Barclays, HSBC, NatWest); those were not searched in this pass. ##### By task Every item below: OBSERVATION, Mobbin, 2026-10-31. ###### 1. Account opening and identity checks - **Monzo, iOS, flow "Onboarding"** (27 screens). Welcome says you can open a personal, joint or business account "in under 15 minutes". "Let's get to know you" says you will need photo ID and "we'll save your progress as you go". Legal name with "Add preferred name", date of birth, mobile number. A four-section checklist (About you, Your finances, Your account, Your identity), each "should only take a couple of minutes". Nationality multi-select ("if you have more than one, list them all"). Then a plan choice. https://mobbin.com/flows/9ac93ca6-671e-46ea-92c1-ef050166a6b3 - **Monzo, iOS, flow "Verifying my identity"** (29 screens). The same checklist with ticks and the note "This won't affect your credit score". A video check: "Record a video of yourself saying a few words. We use this to make sure nobody is pretending to be you", with tips (quiet place, face visible, good light) and a note that the video is encrypted and used only to verify identity. Tax residency multi-select with a Help button. Overdraft terms with "Download as PDF" and Agree, including a plain warning about the consequences of falling behind. https://mobbin.com/flows/b8ed84b4-a36d-4da2-acf9-7709d5600bd1 - **Starling, iOS, flow "Onboarding"** (19 screens). Intro carousel ("Set aside money in Spaces... you can even pay your bills from one"). SMS six-digit code ("can sometimes take a few moments to arrive"). Passcode. A repeat-password step with a plain inline error: "The password you entered did not match the one you entered on the screen before." https://mobbin.com/flows/cba281c7-dd6f-4397-b196-83cf63d5d624 - **N26, iOS, flow "Answering questions".** A "My status" checklist that ticks as you go: pair your smartphone, provide your address, choose account type, answer a few questions, verify your identity, provide tax information, choose your plan. Employment status question (employed, self-employed, freelancer, student, retired or homemaker, unemployed). Device pairing is the first step. https://mobbin.com/flows/fc7b54bc-18d1-4755-94eb-e81914511561 - **Coinbase, iOS, flow "Get started"** (Singapore dollars). An "Unlock trading: fund your account, 2 of 4" progress card. An Account Checklist (account created, verify your info, fund your account with a "Why is this important?" link, buy your first crypto). "You're ready to trade: 1 step remaining." https://mobbin.com/flows/ec0b9542-69a6-4ef9-b636-6996dd328ec0 - **Monzo, iOS, screen (name update).** "Heads up: this name needs to be linked to your legal name... if you're using a nickname that's totally different from your legal name, we'll likely have to reject this request." You must tick "I understand this category may be rejected" before Next. This warns of a likely rejection, and why, before you submit. https://mobbin.com/screens/78a888af-4057-4ee4-97d6-c9cc838e5a0b - **Vivid, iOS, screen.** Card preview with "Is the name on your card correct?" and the buttons "Yes, all correct" and "No, edit name". https://mobbin.com/screens/4c917b2c-bac0-459a-9f22-1505234baf29 ###### 2. Business account onboarding - **Monzo, iOS, flow "Apply for a business account"** (16 screens). Entry is the "Do more" menu (create pot, savings, investing, pensions, credit card, under-16s account, business account, add other bank accounts). "What are you hoping to do with this account?" multi-select (pay staff, pay suppliers, expenses, get paid by invoice, other, pay taxes). Business name and start date ("use today's date if you've not started trading yet"). "Help us process your application faster": what your business does (at least 10 characters) and where to find it online. "We need to confirm your identity" as a safety measure. https://mobbin.com/flows/639498d5-ba9a-4b99-be43-61f5c48f079e - **Starling, iOS, flow "Open a new account"** (31 screens). Account switcher (Personal, Euro) with "Open a new account". Searchable business categories ("if none of these match... try another search term", browse all). Address entry. "Do you already have or soon plan to have customers, suppliers or employees outside the UK?" with a "None of these" option. "A Few Things To Know": you apply for yourself and on behalf of the company; identity, fraud and credit checks; sharing with fraud prevention agencies; links to general terms, account terms, rates and fees, and the privacy notice. Then "Submit Application". https://mobbin.com/flows/b0907f6e-f47b-4e7d-9380-7cb76c778abc - **Wise, iOS, flow "Add business information"** (11 screens, Singapore company). "What's the purpose for using Wise?", with a warning that the account may be suspended or closed for restricted activities and a link to the Acceptable Use Policy. Company directors (add another). Business owners, with a tick box: "No-one else directly or indirectly controls more than 25% of my company." Then: "Upload an official document confirming you're authorised to act on behalf of the business. **We need this document because it seems like you are not a director listed in your company's** [registry record]", followed by a numbered list of what the letter must contain and a link to a power-of-attorney template. This is a failed check shown with its reason and the fix. https://mobbin.com/flows/c857cf11-3ca6-4b27-b2cc-315fcf815a55 - **bunq, iOS, flow "Adding a business account".** An "Add Bank Account" catalogue (payment, joint, child, business; savings variants with rates; investing). "Tell us a bit about your company": legal form, optional VAT number, country, address. "Let's go! Your 30-day trial starts now." https://mobbin.com/flows/de519aae-56dc-4c14-b527-f0f4273c00ee - **Mercury (US business bank), iOS, flow "Company profile".** Business home (balance, money in and out, virtual debit cards named by use, for example a software subscription). Company profile with legal name, display name, card name, mailing address ("we'll send physical cards to this address") and legal address ("used on bank documents"). https://mobbin.com/flows/537ff71e-b21d-495e-8b90-8774e8407778 ###### 3. Sending money and payee checks - **Starling, iOS, screen (payee review).** "We've matched the name below with the receiving bank. It's still important to check you're sending money to who you think you are." It shows the name, the receiving bank, the account number and sort code, and a Save button. This is a name-match success state. https://mobbin.com/screens/241a9904-e430-406c-b842-37b8e16d1187 - **Chase UK, iOS, flow "Adding a payee"** (7 screens). Empty state "Pay people and businesses here". Then "Before you add this payee...": if you were contacted by email, message or social media, phone the person on a number you trust, "not one they've just given"; "Fraudsters can pose as loved ones... They might tell you it's an emergency"; "If you're unsure about this payee, please stop the payment now". Buttons: I understand / Stop payment. Adding the payee then needs your passcode. https://mobbin.com/flows/ded52b9c-c0f3-417b-bbc5-d371b11942b6 - **Chase UK, iOS, flow "Paying out".** Payments tab with Pay and Scheduled, recent payees, pay in, move between your accounts, payees with New. An amount screen with reference, from (showing the balance left after the payment) and to, plus a Schedule option. Confirmation: "You've sent £1.00 to [name]. The money should reach them within 2 hours", with sort code and account number. https://mobbin.com/flows/d09ec8e8-9474-481f-a7f0-8157e1a1d9ab - **Chase UK, iOS, flow "Payee detail".** A payee page listing that person's accounts (for example an account at another UK bank), with Add and Edit. https://mobbin.com/flows/f4174549-68ef-4faa-ba54-47a18eeeebcb - **Revolut, iOS, flow "New payment".** A "New payment" sheet searching by name, @Revtag, phone or email, or by QR scan. Add new: Revolut friend, bank recipient, card recipient, group. Also send international, "Create a payment link: transfer without account details", donate to charity. https://mobbin.com/flows/6ac1e847-5f65-43ae-ba41-0d7f1b96a6a1 - **Revolut, iOS, flow "Searching Revolut (Payments)".** Search with filters (Friends, Accounts, Groups, Cards), "Add new recipient", "Show Revtag results" for people not in your contacts. A chat-style payee thread with Request and Send, and the note "Your message content is encrypted but Revolut can see information about your payments". https://mobbin.com/flows/e3d28662-ff66-401c-8118-3618b6c6d6ae - **Revolut, iOS, flow "Transfer"** (Singapore dollars). Home with add money and exchange; "Say hi to all your existing friends in Revolut: instant payments at no cost"; Transfer tab with past payments. https://mobbin.com/flows/1b68edd8-e392-4d45-8ad8-af158770955d - **Revolut, iOS, screen (review transfer).** A warning card at the top: "Do you know and trust the payee? If you're unsure, don't pay them, as we may not be able to help you get your money back. Remember, fraudsters can impersonate others, and we will never ask you to make a payment." Below: to, Revtag, reference, recipient gets, fees ("No fees"), your total, estimated arrival ("Usually in seconds"), Send. https://mobbin.com/screens/160eff79-448b-4ab3-a6de-bb4f406975af - **Revolut, iOS, screen (final confirm).** A bottom sheet: "Are the payee and transfer details correct? Once confirmed, the payment can't be unsent." Send / No, go back. https://mobbin.com/screens/e417be9d-5c9f-4028-9ec6-139d19373885 - **Wise, iOS, flow "Calculating transfer".** A transfer calculator on the home screen: rate chart, live rate, enter either amount, "Includes fees" with a help icon, all before choosing a recipient. A balance page (add, convert, send, request) with an interest offer marked "Capital at risk". https://mobbin.com/flows/31b6dc50-2014-4ad1-8e3e-46fc41c80dfb - **Wise, iOS, flow "Send money (USD)"** (16 screens). "You send exactly" and "Recipient gets" with a locked rate, available balance, paying with, "Should arrive today, by Tuesday" with a Schedule option, total fees. Then "Great! Now, let's add a recipient". Add-recipient options: find on Wise (@Wisetag, email, mobile), bank details, "PDF or image capture: upload a bill, invoice or screenshot to add details automatically", pay by email. On the SWIFT form, a warning that sending USD outside the US costs extra fees. A disabled Continue button with the hint "Enter either amount to continue". https://mobbin.com/flows/1e6b0884-ca44-4b5f-9cea-05b417737c82 - **Wise, iOS, screen (confirm and send).** Transfer details (you send exactly, total fees included, recipient gets exactly), recipient details (account holder name, provider), schedule details (sending on, arriving by, repeats weekly until a date), each section with a Change link; optional reference; "Schedule transfer". https://mobbin.com/screens/d8b0deb6-b783-42ec-a12f-d9327cf7a180 - **Monzo, iOS, flow "Canceling a payment".** Straight after paying: "Made a mistake? Undo this payment", with a countdown ring, alongside a category picker. Undoing gives "Payment cancelled". https://mobbin.com/flows/a37ea54f-6f3f-45e7-9544-1c98329d317f - **Non-UK reference: ANZ Plus (Australia), iOS, screen (Check Payee, no match).** A red cross and "The account name entered does not match the account"; a "What does this mean?" link; an "Are you sure?" dialog with Cancel and Continue; the note "an incorrect payment may be difficult to retrieve". This is the only no-match screen found. No UK "no match" or "close match" screen surfaced in the searches. https://mobbin.com/screens/15c076c7-3dd6-475f-9f9e-c5b7c320ca19 - **Non-UK reference: Up (Australia), iOS, screen.** A pay-by-bank-transfer sheet saying "we cannot validate account name or details... it may not be possible to recover money paid incorrectly", and that transfers take 1 to 2 business days. https://mobbin.com/screens/d2f2614c-3fc8-4b4a-b4f9-e2d1bd44c7e2 - **Non-bank reference: Careem, iOS, screen.** Add recipient with "Account number validated", UK bank details, and the full name marked "Validated with account number". https://mobbin.com/screens/887a9fcf-feb0-49dc-a92f-6644eb969c10 ###### 4. Requesting money and splitting bills - **Monzo, iOS, flow "Requesting money".** Payments tab: pay someone (bank transfer or link), transfer internationally, request money, manage payees, move money, Monzo Split. Request screen: amount, "What's the request for?", which account receives it. Payments then shows "Friends owe you £2.00 / You owe friends £0.00". https://mobbin.com/flows/cabf1a3d-b67d-47c4-a50f-200c9f9e3b57 - **Monzo, iOS, flow "Splitting a payment"** (12 screens). Pick a past transaction (foreign-currency ones show both amounts). A split picker with Nearby Friends, "Split with people outside Monzo" counter, favourites and recent. The transaction then shows the split (you, and what you requested from the friend) with "Cancel bill split", "Take this money from a Pot", receipt, notes and a "Repeating payment" toggle. https://mobbin.com/flows/054c2a81-ed87-4ea9-88b9-3f8cd713f7a7 - **Monzo, iOS, flow "Generating QR code".** "Get paid by QR code" with the receiving account, an optional fixed amount ("or leave it to the person paying") and a note, plus "Share link instead". https://mobbin.com/flows/267cc21a-1e38-42b7-83d6-9bfd2e2e45d2 - **Monzo, iOS, flow "Payments".** Shared Tabs: "Trip to Spain: you're owed £100", "House bills: you owe £15", "Create Shared Tab: from holidays to household bills". Also shows the scheduled-payments view (see section 5). https://mobbin.com/flows/cb72c699-93e0-4f5d-8e8f-48733c839e9e - **Monzo, iOS, flow "Activity detail".** An incoming money request shows the linked payment, "Block [person]: stop Monzo payments and requests", and the buttons Review payment / Decline request. A transaction detail shows the exchange rate used, category, "Split this bill" and "Add to a Running Split". https://mobbin.com/flows/3ecb8064-a24a-43cd-ac11-e895b033246f ###### 5. Scheduled payments and direct debits - **Monzo, iOS, flow "Scheduled".** Scheduled tab with "View calendar" and a list of subscriptions and direct debits (monthly date, pound amount and original-currency amount), "Give feedback" and "Add scheduled payment". https://mobbin.com/flows/d1a4e338-ef74-4325-af7b-94cf5f7af440 - **Monzo, iOS, flow "Payments"** (scheduled view). "Subscriptions & Direct Debits" with the honest note "Depending on the Direct Debit, sometimes the amount and date of payments can change. We think they'll look like this though." https://mobbin.com/flows/cb72c699-93e0-4f5d-8e8f-48733c839e9e - **Monzo, iOS, flow "Scheduled payments"** (in a pot). "Left to pay until 30 Jun"; "Pay your Direct Debits and standing orders from this Pot instead of your main balance"; card subscriptions tied to a pot's own virtual card ("Create virtual card"); schedule deposit and schedule withdrawal. https://mobbin.com/flows/6bbb45ff-5eac-4632-9b93-2218f295494f - **Monzo, iOS, flow "Creating a pot (regular)".** Add money to a pot with a Schedule toggle; confirmation "You've added £10 to Fun stuff... We'll do this monthly until 20 July 2026", with Edit schedule. https://mobbin.com/flows/74df8eb8-a70b-4d66-9778-62416d46c798 - **Wise, iOS, screen (scheduled transfer).** Repeats weekly until a set date, shown on the confirm page before "Schedule transfer". https://mobbin.com/screens/d8b0deb6-b783-42ec-a12f-d9327cf7a180 ###### 6. Cards: freeze, controls, virtual cards - **Starling, iOS, flow "Card controls".** Card page with a "Card unlocked" toggle (freeze), card details, PIN reminder, card controls, payment and card limits, going abroad, gambling controls, mobile wallets, cancel and replace. Card controls: separate toggles for card present, ATM, online, mobile wallet and magstripe (magstripe off, "locked to protect you from fraud"), each with a plain status sentence. "Unblock PIN/CVV" button. https://mobbin.com/flows/862b50ac-eec9-4f9a-8583-9ed5e1138992 - **Starling, iOS, flow "All cards".** A Cards sheet with the physical card and virtual cards named by purpose, each with a lock icon and a details icon; "Create new". https://mobbin.com/flows/b07fa0d9-a3a3-462c-834c-3ad64575a19a - **Starling, iOS, flow "Gambling controls".** "Turn on the gambling block" and "Get support with gambling". https://mobbin.com/flows/f3d1f086-6660-4cc2-925a-5a51ce98c44f - **Starling, iOS, flow "Mobile wallets".** "Manage which mobile wallets your card is assigned to": Apple Wallet active; add to Apple Wallet or Click to Pay. https://mobbin.com/flows/8925d77e-94e1-4bc7-9caa-bf1859820957 - **Revolut, iOS, flow "Getting a virtual card"** (10 screens). "Pick your cards": physical (customisable) or virtual ("complimentary virtual cards instantly, and try single-use cards for extra security online"). "Name your card" with suggestions (online shopping, subscriptions, general). "Your virtual debit card is ready", with add to Apple Wallet. A card page with show details, Freeze and Settings. https://mobbin.com/flows/aca7784d-9a65-4736-906f-3bf1aaf912bf - **Revolut, iOS, flow "Card settings".** Per card: which accounts it spends from, monthly spending limit, card name, "Unblock CVV: use after 3 incorrect PIN/CVV attempts", "Terminate card: the card will be permanently terminated". https://mobbin.com/flows/32e32319-8d42-405b-b560-c2b19e400c95 - **Revolut, iOS, flow "Creating a new card".** Disposable virtual card: "Your card details automatically change each time you make a payment, protecting you from fraud"; "Get card for free"; "ready to use". https://mobbin.com/flows/d64e254a-f416-4b79-a31e-c8ee8d7cb803 - **Zing, iOS, flow "Card".** Multi-currency wallets (EUR, GBP, add wallet) and a card page with view card details, view PIN, freeze card, replace card, get help. https://mobbin.com/flows/128855ca-8193-4f2c-bf8a-37a4004c4179 - **Airwallex, web, flow "Creating a card (company card)"** (13 screens). Card summary (company cards and employee cards, spend, limit usage). Create a virtual company card shared with up to three people: card purpose, nickname, single-use option, monthly limit, additional limits, and advanced controls (start and end date, per-transaction limit, international and online toggles). A live summary panel on the right shows the limits before "Create card". https://mobbin.com/flows/b997c129-d3f9-47e9-a268-ba18cb12d6b7 - **Mercury, web, flow "Team spend".** Team Spend: allocated budgets, employee budgets with spend limit and cycle, assignees, "Review required". Spend Policies: receipt required over an amount, notes, category, spend controls that override card controls, and "Automatic card locking: lock cards with unresolved policy violations". https://mobbin.com/flows/87071832-157d-4758-a72b-c74e557cb2a5 - **Mercury, web, flow "Creating a budget"** (15 screens). A budget form (name, spend limit per person, repeats, a description noting "a brief purpose will make this budget smarter", spend controls, start and end) with a live preview card that updates as you type. Assign team members and choose how each will spend (new virtual card or an existing card), with the preview "1 team member will receive...". Save as draft. https://mobbin.com/flows/163a9014-36b1-4bd8-9ba1-fc0ccc370d76 ###### 7. Savings pots - **Monzo, iOS, flow "Creating a pot".** Create Pot: image, name, goal, round-up payments (1x, 2x, 5x, 10x, with the worked example "Spend £1.10, save £0.90"), and "Lock this Pot" until a date. Trying to close a locked pot: "This Pot is locked until 5 May 2025. You need to unlock and archive the Pot so we can close it", with Keep it locked / Unlock and archive. https://mobbin.com/flows/ba5aef17-851c-4d20-ad21-681387521e81 - **Monzo, iOS, flow "Creating a pot (savings)"** (21 screens). Instant Access Savings Pot intro (rate, deposit limit, "FSCS Protected"). Optional goal and achieve-by date, round-ups, lock. "You've done it". Pot detail with savings tools, deposits, withdrawals, pot documents and "Archive this Pot". https://mobbin.com/flows/c4c521ce-8566-4c6f-a5b6-01c6c1443bbc - **Monzo, iOS, flow "Savings & investment".** Home with Savings, Investments and Pension. Investments tab ("start with as little as £1", fund range, "You could get back less than you invest"). Pension tab (combine pensions, find lost pensions). https://mobbin.com/flows/395ccd0b-66a3-46c4-a87d-841f1f25935e - **Monzo, iOS, flow "Set up an investment account"** (20 screens). Choose between a Stocks and Shares ISA and a General Investment Account ("Help me choose"). A fund page showing allocation (shares and bonds) with past-performance and fee caveats. Home then shows "Investments: finish opening your account" with Resume. https://mobbin.com/flows/f313be77-fe15-4cc9-be6e-f502c38f3aea - **Chase UK, iOS, flow "Creating a Chase Saver account".** "Savings that just add up", with the rate and "take your money out anytime". Add saver with icon, colour and name. Save & Invest tab with a rate-change notice giving both the AER and the gross rate, the savings list (round-up and saver), and investments "provided by Nutmeg". Toast: "You created My Saver account". https://mobbin.com/flows/7da7eeb8-4b31-4f6e-8e28-81a181ca1bbf - **Chase UK, iOS, flow "Account detail (Chase Saver)".** Balance, rate earned, goal progress ("only £5,000 away from your £5,000 goal"), "Ready to get started? Move money in now"; edit goal, hide account, pay in. Account menu: account details, statements, interest rate, edit, Close account. https://mobbin.com/flows/96916461-0689-4028-96e6-fc4a1c0fc3d2 ###### 8. Connecting other banks - **Monzo, iOS, flow "Connecting a bank account"** (10 screens). "Choose your account": UK providers with their product types, and the note "If your bank isn't on the list it's because we don't support them yet." A consent page lists the optional purposes (tailor marketing; make lending decisions about you), each with an info icon; says preferences can be changed later in Privacy & Security; links the terms and privacy notice. Buttons: "Accept all & Continue" or "Reject optional & Continue", then a redirect to the other bank. The connected account then appears inside Monzo with its balance, transactions, Move money and Manage. https://mobbin.com/flows/24d1e38a-cb93-4088-96e7-74a873567ec9 - **Monzo, iOS, flow "Adding money".** Pull £10 from another bank into Monzo, with the consent text "By continuing, you agree to us making this payment for you, and sharing your account details with [other bank]." https://mobbin.com/flows/77e8ae4d-ebf0-4795-b112-cefa7a590a06 - **Monzo, iOS, flow "Manage apps".** Manage apps lists Monzo's own apps and third-party access grants ("Account Information by Plum: Ongoing session", one row per grant). https://mobbin.com/flows/7883b4d9-6769-400e-86f3-97593cef4f79 ###### 9. Crypto: buy, sell, send - **Coinbase, iOS, flow "Buying crypto"** (8 screens, Singapore dollars). An order screen with Buy / Sell / Convert tabs and a "One-time order" selector, the amount in local currency with the crypto equivalent, the funding wallet and its available balance, the asset, and "Review order" before buying. Select asset with recent buys and top assets. https://mobbin.com/flows/814b89ee-7e8a-499e-8d76-62d5589de792 - **Coinbase, iOS, flow "Explore".** Portfolio chart with crypto and cash split; a "You're offline" banner; top movers and top assets; staking rewards with APY; Buy & sell and Transfer buttons. https://mobbin.com/flows/ee2a5a0c-bb0d-46c9-8858-b1bbb1e766bb - **Kraken, iOS, flow "Withdrawing a coin"** (11 screens). Account menu (rewards, pay, tax reporting, payment methods, Security with a "Setup 2FA" nudge, device management). Withdraw ETH with 25%, 50%, 75% and MAX buttons and "Add new address". A Confirmation screen: amount, fee, total, withdraw to, address, network, and the warning "Only withdraw ETH to Ethereum network addresses. Withdrawals to addresses generated for other networks could result in loss of funds." "Swipe to withdraw". "Withdrawal initiated" with a link to transaction history. https://mobbin.com/flows/1c74aa14-6e60-4d9c-8ea6-8803eb944149 - **Kraken, iOS, flow "Sending a coin (Paylink)".** Kraken Pay ("send cash or crypto to friends: instant, secure and free"), search by Kraktag or email. "Send 1 USDG by Paylink", with "Auto-converting ETH to USDG: we'll convert your selected asset to make sure you can send the full amount", the source balance, a note field, total and fee with "Show details", and "Active for 14 days". Then a share sheet for the link. Activity later shows a sent payment and "Paylink canceled". https://mobbin.com/flows/480e47e9-de24-4e49-9a18-3e2f977154fb - **Kraken, iOS, flow "Pay".** Market list (most popular, watchlist, gainers, losers) with Deposit and Buy crypto; Pay tab with recent recipients. https://mobbin.com/flows/ed7e6e36-dd1b-4934-b266-4724c675477d - **Revolut, iOS, flow "Crypto details".** Trade list (176 cryptos, sort by market cap). Bitcoin detail with a price chart and the plain caveat "This is the midpoint between buy and sell rates. Actual rate varies based on whether you buy or sell." About Bitcoin with a whitepaper link; Buy. https://mobbin.com/flows/4041fcb3-8402-4988-af38-bab67005a760 - **Revolut, iOS, flow "Trade".** Crypto tab: balance chart, Trade, Receive, Send, holdings, a "Learn reward" credit. https://mobbin.com/flows/094a3151-e139-453c-af6e-c602caa742dd - **Base (Coinbase's wallet), iOS, screen.** Before buying 10 USDC: "Before You Continue... Please confirm the following", three tick boxes (no one asked me to deposit these funds; I am not on a call or chatting with anyone instructing me; I understand this is my personal wallet and only I should control access to it). Continue stays disabled until all are ticked. https://mobbin.com/screens/c4439f4b-7c76-4f4b-b4b0-ce0385f4be16 - **OKX, iOS, screen (token approval).** Approve "Unlimited USDT" for a third-party exchange contract, with the network fee and the warning "By authorizing USDT, OKX DEX has access to all your funds. Make sure you trust this contract." Reject / Confirm. This is a standing permission handed to a third-party program. https://mobbin.com/screens/936d25d1-9261-4cb1-86c4-e3cc29a3ea01 ###### 10. Loans and credit - **Zopa Bank, iOS, flow "See loan rates".** The Apply tab shows "Borrowing Power: you're eligible for a loan!" with a personalised rate (amount, term, APR, monthly repayment, total cost). "Explore your loan options": an amount slider and term picker that update the monthly payment, guaranteed APR, total repayable and a cost breakdown. Then "You're pre-approved for this loan, so there will be no further credit assessment, just the standard verification checks." Continue with this loan / Cancel. https://mobbin.com/flows/ed4d6d1f-77af-43db-a033-9fb4ed105906 - **Zopa Bank, iOS, flow "Apply".** Home (account balance, savings hub, upcoming payments) to the Apply tab (borrowing power, compare credit cards, car finance). https://mobbin.com/flows/e579dcc5-300f-4bcc-9ce3-f5d2d2f84371 - **Zopa Bank, iOS, flow "Adding an employment and income".** Personal details flags "Please add your finance details". Employment status choices; annual income before tax with the hint "If your income is due to change within 12 months, please provide the lower figure"; additional income; a summary where each item can be updated. https://mobbin.com/flows/f4ed15cc-708c-4e4a-84d4-7e36ec054f56 - **Klarna, iOS, flow "Activating a purchase plan"** (14 screens, US dollars). "Choose how to get started" (add money, or set up Pay Later with purchase power). "How much would you like to spend? ... If you spend less, your plan will adjust automatically." Plan choice (pay in 4, 0% interest, service fee waived, due at purchase, total; "Payment plans are subject to final approval"). A consent box naming each agreement and saying the soft credit check "will not affect your credit score". A lending disclosure with the number, amounts and dates of payments and a finance charge of $0. "Activate your plan". "You're ready to pay": a one-time card valid 24 hours up to the plan amount, card details hidden during screen recording, "Cancel this plan" and "Mark plan as used". https://mobbin.com/flows/9e64ccd4-a7e1-473f-b973-2b1499632372 - **Klarna, iOS, flow "Payment calculator".** A merchant "How to pay" page (digital card set-up steps, 3 minutes, valid 24 hours). A Payment calculator: type what you plan to spend and see each instalment (today, in 2, 4 and 6 weeks) before committing; "A service fee may be applied". https://mobbin.com/flows/e9084420-c5f9-4567-aaf8-a448802af449 - **Klarna, iOS, flow "Purchase power".** A purchase power page (amount available, amount used and owed): "An estimate of what you can spend with Klarna. It can change over time and is not guaranteed." FAQs including "Why did my purchase power decrease?" and "Why was my purchase declined when I had enough purchase power available?" The reasons are generic help answers, not per-decision explanations. https://mobbin.com/flows/8ca42124-06f5-439c-8a60-177e143dd83a ###### 11. Credit score and report screens - **Lloyds Mobile Banking, iOS, screen.** "Your Credit Report": the score, its band ("Good"), the date the report was created and the agency that provided it; summary cards for borrowing, mortgage balance ("There isn't currently a mortgage on your credit file") and new accounts in the last 6 months; FAQs. https://mobbin.com/screens/2b139d01-bf9a-4a6b-9d3a-2d92df2386e0 - **Credit Karma (UK scale), iOS, screen.** Build and Monitor tabs; "Your credit score is Excellent" with the score out of 710 and "Next update in 7 days"; a history chart; points gained since joining, things to do, points to the maximum; "Your credit factors haven't changed. See below what you can do to improve them." https://mobbin.com/screens/dfcabcd8-a9aa-4771-9e56-0ce6aed1872e - **Cleo, iOS, screen (Score Simulator, US).** "Drag the slider to see what each score means": a score band and the example home and car loan rates for it; "Show me my score"; "FYI checking won't ever affect your score". This is a what-if preview. https://mobbin.com/screens/3c9e2d5d-1521-4aec-a6ac-46d21d511ed8 - **CRED (India), iOS, screen.** "Your credit score has decreased: 34 points were deducted", with factor cards (payments on time, credit utilisation) and an impact level for each. It explains the reason for a change. https://mobbin.com/screens/b4c37acb-ae4d-4c71-b04f-623b94abd545 - **One (US), iOS, screen.** Score history and "Factors affecting your current score" with ratings (payment history, credit usage, credit age, credit checks, available credit). https://mobbin.com/screens/2fdcc0ed-6e23-4371-adc9-2cc3dffbd0c7 ###### 12. Disputes and fraud alerts - **Starling, iOS, screen (Get Help on a transaction).** Three plain choices: "I recognise this payment but there's a problem" (Raise a dispute); "I don't recognise this payment and I think it might be fraudulent" (Report fraud); "I haven't received the goods or service(s) I paid for" (use the online banking form and provide evidence). https://mobbin.com/screens/adf0cdc3-414b-4dfa-a01c-bee708770509 - **Lloyds Mobile Banking, iOS, screen (Report this transaction).** "Is something not right? We'll help you sort it out." Before continuing: could someone you know have made this payment on a shared device; have you shared your password or card; have you checked emails for a retailer confirmation. "Continue if you're confident no one you know has made this payment." https://mobbin.com/screens/172f6fb3-15a0-4276-90ec-e6e5a60f6ef8 - **Monzo, iOS, flow "Transaction help"** (16 screens). Help tab "I need help with a transaction" with recent transactions. Guided questions: "Has the transaction or payment been made?" (yes, no, not sure); pick the transaction; "Describe your problem below and we'll get you some help". The answers are carried into a chat as an automated summary, so the person helping sees the context. https://mobbin.com/flows/7e5fd08c-ff56-454f-89ab-ae2338371a48 - **Klarna, iOS, flow "Reporting a problem".** A purchase page with Report return and Report a problem. Problem categories include item different from description, "I was charged wrong amount" and "I suspect identity theft". A Buyers Protection note, contact the store, and a step-by-step "I want to return my order" guide. https://mobbin.com/flows/061bee71-40e8-4b7f-91ff-a84065b915fa - **Klarna, iOS, flow "Return & refunds".** "Total you owe: all upcoming payments are scheduled". A return timeline: return reported (date); "Return your goods by Oct 25: keep your tracking number"; Provide return info / Cancel return; waiting for the store to confirm ("usually takes a few days"); "Klarna updates your payment plan, any excess amount will be automatically refunded". https://mobbin.com/flows/54db0f61-950c-455e-988c-94b7c4c54284 - **Non-UK references.** Neo Financial (Canada): "Did you make this transaction? If the transaction was fraudulent, we'll help you replace your card", yes or no. https://mobbin.com/screens/39a6a70f-7f00-485d-92e6-7349074fc7d7 . Kit (Australia): "See a transaction you don't recognise? Before you raise a dispute, consider..." (family member, merchant trading name, currency conversion, subscription), then "Yes, dispute". https://mobbin.com/screens/72f12328-7af4-436c-b415-1776f18afae4 . Cash App (US): "Did you make this payment? Yes, but there's a problem / No, I didn't make this payment". https://mobbin.com/screens/67ba5404-c9b0-4f7b-a783-8874902e9b36 ###### 13. Scam warnings - **Zing, iOS, screen.** "What is this payment for? We're here to help you recognise a safe request from a fraudulent one." Choices: friends and family, cryptocurrency, investments, paying a business, large purchase, online purchase, something else, online dating, unexpected request. https://mobbin.com/screens/c86f3f35-832b-4729-8788-d7b09ebb4795 - **Starling, iOS, screen (payment review).** A banner: "We've never spoken to you on the phone. If anyone on the phone says it's Starling, it's a scam. Hang up and call us on 159." Then send, reference, when and category, plus "Payments to other UK banks are normally instant, but in rare cases take up to 2 hours. High value payments might take longer because we have to carry out extra security checks." Make payment. https://mobbin.com/screens/cc78351d-cd04-4610-b3df-a27120f482d0 - **Monzo, iOS, flow "Manage apps"** (Privacy & security). A live "Monzo call status: We're not calling you. If someone is telling you they're from Monzo, hang up now. Last updated: now." Also Face ID to unlock and pay, "Undo payments: On", "Added security: protect large transfers and savings". https://mobbin.com/flows/7883b4d9-6769-400e-86f3-97593cef4f79 - **Chase UK** "Before you add this payee" and **Revolut** "Do you know and trust the payee?" (section 3), **Base** tick-box confirmations and **OKX** unlimited-approval warning (section 9). - **One (US), iOS, screen.** "Ready to send $1.00... This payment cannot be canceled. Make sure [name] is someone you know and trust." Yes / No. https://mobbin.com/screens/77394f78-9a6d-495e-af02-823d62349a56 ###### 14. Security and devices - **Revolut, iOS, flow "Security and privacy settings"** (Singapore entity). Change passcode, Devices, a gambling block toggle, hide balances (flip the phone face down), sign in with Apple or Google toggles; privacy toggles for being discoverable, messaging with friends, being added to groups, showing full name and picture. https://mobbin.com/flows/09754846-7a34-4e21-bad4-ebfdd61c1d8b - **Revolut, iOS, flow "Account".** Profile menu (help, account, documents and statements, inbox, security, notifications). Account page: personal details (with "Verify your email" in red), account details, change passcode, passkeys, privacy, policies, Close account. https://mobbin.com/flows/20e77703-f1ac-4c55-a399-6f3ba61442ad - **Revolut, iOS, flow "Profile".** Home (add money, move, details) to the profile menu with security, appearance, new features and log out. https://mobbin.com/flows/b3e1e7fa-a35a-4402-a00d-449ff3483ef0 - **Monzo, iOS, flow "Manage apps".** Monzo's own apps and each third-party access grant listed separately (see sections 8 and 13). https://mobbin.com/flows/7883b4d9-6769-400e-86f3-97593cef4f79 - **Kraken, iOS, flow "Withdrawing a coin".** The account menu includes Security (with a "Setup 2FA" nudge) and Device management. https://mobbin.com/flows/1c74aa14-6e60-4d9c-8ea6-8803eb944149 - **N26** pairs the smartphone as the first onboarding step (section 1). ###### 15. Closing an account - **Starling, iOS, flow "Close account".** "Before closing your account, here are some things to bear in mind": you won't be able to open another personal account for 12 months; close other currency accounts first; all cards will be cancelled, so wait if you expect refunds; direct debits will be cancelled, so arrange another way to pay. Then "When you close your account, we'll send you your transaction history": a secure emailed link opened with a password sent by SMS; choose all statements or a date range, or "I don't want my statements". https://mobbin.com/flows/983193b9-49eb-4fd7-8151-8ff0094f9a32 - **Chase UK, iOS, flow "Account detail (Chase Saver)"**: Close account in the account menu (section 7). **Revolut, flow "Account"**: Close account at the bottom of the Account page (section 14). - **Monzo, iOS, flow "Creating a pot"**: a locked pot must be unlocked and archived before it can be closed (section 7). - **Non-UK references.** Monarch (US aggregator): close (balance set to zero, history kept) versus delete (all data removed), with a "What will change" sheet before confirming. https://mobbin.com/flows/27f88231-020c-45ad-aba4-29b61378a60b . Copilot (US aggregator): "Mark as closed: this account will be disconnected... Existing transactions won't be affected. This can't be undone. Are you sure?" https://mobbin.com/flows/c5bbfa8f-696e-4b5f-a24b-1846612c0a1c . Chime (US): "Why are you closing your account?" with reasons including security concerns and free text, then a goodbye screen. https://mobbin.com/flows/cf62e9f4-3240-41a6-96f4-995b76d715d3 ###### 16. In-app assistants - **Monzo, iOS, flow "Chatting with AI assistant".** The Help tab suggests topics (including "My transaction declined") and "Still need help? Chat with us". The chat opens already knowing the topic ("Looks like you wanted to chat about updating the name on your account. Tell us more."). Replies are labelled "AI Assistant" with timestamps. "Conversation closed", then a feedback request. https://mobbin.com/flows/cabbd8bc-c4bf-4738-b82a-d7dce1f89d40 - **Revolut, iOS, flow "Chatting with support"** (11 screens). A help article with Helpful / Not helpful and "Chat with us". Support ("Helping you 24/7"): messages labelled "Chat assistant" ("I am Revolut's Chat Assistant, how can I help?"), quick replies "Got it, thanks" / "I need more help", and a request for details "to help our live agents speed up your case". Menu: End chat, View profile, "Generate chat PDF". https://mobbin.com/flows/f370f504-4014-433c-ab03-fbc347946e70 - **Revolut, iOS, flow "Chat assistant profile".** The assistant has its own profile page: "Helping you 24/7", review count, number of chats, years at Revolut, and compliment categories (app functionality, chat outcome, agent's approach, support journey). https://mobbin.com/flows/1f8b4cc9-7e56-4788-968d-7cf237fb0668 - **Cleo, iOS, flow "Chat".** A chat-first money app: linked accounts, then a Chat tab with cards (upcoming bills, gambling help, a money game with Opt out) and "Need help? Tutorial / Help centre / Humans pls", an explicit route to a person. https://mobbin.com/flows/510ff2c8-eef8-4e92-9034-7c9783e181c9 - **Neo Financial (Canada), iOS, flow "Chatting with Neo AI assistant".** "Neo AI Assistant: the team can also help". The answer cites numbered sources, says plainly that a product is temporarily paused, gives steps, and ends with a clarifying question. https://mobbin.com/flows/ee91233f-473e-4fd7-9561-bd464d60a5be - **Zip (pay later, non-UK), iOS, flow "Chat assistant".** "I'm Zigi, Zip's new AI chatbot"; "By using this chat, you agree to our Terms of Use and Privacy Policy"; footer naming the AI vendor; End Chat; a "message is too long" error with a character counter. https://mobbin.com/flows/3db72207-0d77-4879-925a-7baa057f86db ##### What incumbents do that EDGE Finance must match CLAIM by this research, from the observations above; dated 2026-10-31. 1. **Resumable, sectioned opening.** A visible checklist of sections with ticks, "we'll save your progress", a resume card on home, and a statement that a check won't affect the credit score (Monzo, N26, Coinbase). 2. **Identity proof with a stated purpose.** Photo ID plus a liveness video, with why it is needed and how it is stored (Monzo). Device pairing as an explicit step (N26). 3. **Business onboarding that asks intent and control.** Purpose of the account, directors, owners with a 25% control declaration, the business category, and a cross-border question. Combined personal and company terms (Monzo, Starling, Wise, bunq). 4. **Payee name check before first payment**, with the match result shown with the bank name (Starling), plus a scam interstitial when adding a payee (Chase UK) and a trust warning on review (Revolut). 5. **A full review before money moves.** Exact amount sent, amount received, fees, rate, arrival estimate, schedule and a Change link on every section (Wise, Revolut, Chase UK). A final "can't be unsent" confirmation (Revolut). 6. **A short undo window after payment** (Monzo "Undo this payment", "Undo payments: On"). 7. **Purpose-of-payment questions and phone-scam banners** ("What is this payment for?" from Zing; "We've never spoken to you on the phone" from Starling; a live "We're not calling you" status from Monzo). 8. **Requests, splits, shared tabs and a running "who owes whom" total** (Monzo). 9. **Scheduled payments in one calendar,** with an honest note that direct debit amounts may change, and pots that pay their own bills (Monzo). 10. **Card controls per channel** (in person, ATM, online, wallet, magstripe), freeze, gambling block, named virtual cards, disposable cards, per-card limits and spend-from source, and PIN/CVV unblock (Starling, Revolut). 11. **Pots with goals, round-ups and time locks** that cannot be broken by accident, and savers with a goal bar and stated rates (Monzo, Chase UK). 12. **Open banking with itemised consent:** optional purposes that can be refused separately, a list of supported banks, and a "Manage apps" page listing each third-party grant (Monzo). 13. **Crypto with network-specific warnings,** a fee breakdown, a deliberate final gesture (swipe to withdraw), and a midpoint-rate caveat (Kraken, Revolut, Coinbase). 14. **Credit shown with its numbers before commitment:** an amount and term explorer with APR, monthly cost and total, pre-approval stated plainly, and a lending disclosure (Zopa, Klarna). 15. **A credit report with its provider, date, band and summary** (Lloyds, Credit Karma). 16. **Dispute triage in plain categories** (problem, fraud, goods not received), self-check prompts before reporting, and guided answers handed into chat (Starling, Lloyds, Monzo, Klarna). 17. **Closing with consequences listed first** (refunds, direct debits, a 12-month bar) and a secure export of history (Starling). 18. **An assistant that is labelled as AI,** knows the topic you came from, hands over to people, and can export the conversation (Monzo, Revolut, Cleo). ##### What none of them do CLAIM by this research, from the sample above; dated 2026-10-31. "None" means none of the flows and screens examined here. 1. **No AI agent as a customer.** Every assistant seen (Monzo "AI Assistant", Revolut "Chat assistant", Neo, Zip, Cleo) is the provider's own helper talking to a person. None of the flows let an AI agent hold an account, pass identity checks, appear as a payee or payer, or act under its own identity. The closest things are: - Monzo's "Manage apps", which lists third-party access grants but only as "Account Information... Ongoing session", with no stated scope, limit or expiry on the screen; - OKX's "Unlimited USDT" approval, a standing grant to a program with an all-funds warning; - Mercury budgets and Airwallex company cards, which give people (not agents) limited spending authority with limits, dates and policies. None give a non-human actor its own limits, preview, audit trail and undo. 2. **No rehearsal before commit.** Previews exist only as calculators or summaries: Wise's transfer calculator, Klarna's payment calculator, Zopa's loan explorer, Cleo's score simulator, and the live preview panels in Mercury and Airwallex. None let you (or an agent) run the whole action in a sandbox and see the predicted result, which checks would pass or fail, and what would happen next, without moving money. The final confirm screens state amounts but not the outcome of each check that will run. 3. **Few reasons when a check fails.** Most screens give a warning, not a reason. The exceptions: - Wise says why it needs an authority letter ("it seems like you are not a director listed"); - Monzo warns, before you submit, that a name change will likely be rejected and why; - CRED says how many points a score dropped and which factors moved it; - the only "name does not match" screen found is Australian (ANZ Plus), and it offers "What does this mean?" rather than the specific mismatch. Klarna explains declines only in generic FAQs ("Why was my purchase declined when I had enough purchase power available?"), not for the decision in front of you. No UK screen found shows a declined card payment or a blocked transfer with its specific reason and a fix. 4. **Undo is rare and narrow.** Only Monzo shows a timed undo after payment. Revolut and One state the opposite ("can't be unsent", "cannot be canceled"). 5. **No machine-readable record for the customer.** The only export seen is Revolut's "Generate chat PDF" and Starling's statements on closure. No flow offered a structured record of what was checked, decided and why. 6. **Scam defences assume a human under social pressure** (phone calls, loved ones, romance, "no one asked me to deposit"). None address an instruction arriving from an automated source, such as a compromised agent, injected instructions, or a payee that is itself a bot. ##### Method note OBSERVATION, 2026-10-31: searched with Mobbin's flow and screen search, iOS by default and web for business spend controls, using app-plus-task queries ("Monzo savings pots", "Starling freeze card", "Kraken send crypto to external wallet") and generic task queries for payee checks, scam warnings, fraud alerts, credit scores, closing and assistants. Five or fewer flows, or eight or fewer screens, per call. Image links returned by Mobbin expire and are not recorded; only `mobbin_url` links are kept. ### research/finance/PRESENCE-AND-SOLUTION-BANKS.md `research/finance/PRESENCE-AND-SOLUTION-BANKS.md` · sha256 `51a8d324d24dd923eb252975e690f4d15a6c33f13e73b58066263cec5cc33de3` #### EDGE Finance: Physical Presence and Solution Banks Research note for **JudeAI | EDGE ID** (apps: EDGE Finance, EDGE Clinic, EDGE Public Consortium, Judiverse). Prepared 31 October 2026. This is research, not legal, financial or regulatory advice. Nothing here says that EDGE is, or may become, a bank, an identity provider or a payment firm. Any such step needs authorisation and specialist advice. --- ##### 0. How to read this document ###### 0.1 The two intentions this note serves - **Intention A, physical presence and proximity:** EDGE acts in the physical world. EDGE points and devices come to the person, and identity is confirmed by proximity: tapping or scanning an identity document or a phone, using edge hardware for good. - **Intention B, solution banks ("from think tanks to solution banks"):** agents generate solutions; people are empowered and funded to carry them out; better lending structures and money movement create jobs and access to resources. ###### 0.2 Evidence labels | Label | Meaning | |---|---| | **OBSERVED** | Read directly in a primary source on 31 October 2026: a government, regulator, legislature, central bank or standards page, or an organisation's own published report about itself. | | **CLAIM** | Stated by a vendor, a trade body about its own members, a news outlet or a commentator; or a self-reported figure nobody has independently checked. | | **AGREEMENT** | A law, regulation, standard or formal specification. Status is given (in force, published, draft, consultation). | | **PREDICTION** | Our design idea or forecast. Assumptions are listed. It is not evidence. | | **UNKNOWN** | We looked and could not confirm. The line says what is missing. | ###### 0.3 Method and limits - Sources were found with web search and read with a page fetcher on 31 October 2026. Some PDFs were downloaded and converted to text so the figures could be read exactly (National Wealth Fund annual report, Responsible Finance appendix, British Business Bank finance markets report). - Several official pages refused automated access (HTTP 403) or were too long to read in full. Where we could not read the primary text, the line is marked CLAIM or UNKNOWN and says so. - During the session a shared search budget ran out for a time and later reset. A few topics (venture studio evidence, some EU wallet detail) therefore rest on fewer sources than we would like. These are flagged. - Figures are quoted as the source gives them. Different sources count different things (for example, One Login "users" versus "identities proven"); we say which is which. ###### 0.4 Acronyms (first use) | Acronym | Meaning | |---|---| | AML | Anti-Money Laundering | | ARF | Architecture and Reference Framework (EU Digital Identity Wallet technical rulebook) | | BBB | British Business Bank | | BLE | Bluetooth Low Energy | | BoE | Bank of England | | CDFI | Community Development Finance Institution (a lender to people and businesses that mainstream banks turn down) | | CEN | European Committee for Standardization | | cVRP | commercial Variable Recurring Payment | | DCMS | Department for Culture, Media and Sport | | DPIA | Data Protection Impact Assessment | | DSIT | Department for Science, Innovation and Technology | | DVS | Digital Verification Services (UK statutory term for identity and attribute checking services) | | EMT | E-Money Token (an EU stablecoin type under MiCA) | | eIDAS 2 | EU Regulation (EU) 2024/1183, the European Digital Identity Framework | | EUDI Wallet | European Digital Identity Wallet | | FCA | Financial Conduct Authority | | FSMA 2023 | Financial Services and Markets Act 2023 | | GDS | Government Digital Service | | GBTD | GB Tokenised Deposits (UK Finance industry pilot) | | ICAO | International Civil Aviation Organization | | ICO | Information Commissioner's Office | | ISO/IEC | International Organization for Standardization / International Electrotechnical Commission | | LCF | Life Chances Fund | | mDL | mobile Driving Licence | | MiCA | EU Markets in Crypto-Assets Regulation (EU) 2023/1114 | | NFC | Near Field Communication (short-range radio used by passport chips, contactless cards and phones) | | NWF | National Wealth Fund | | OfDIA | Office for Digital Identities and Attributes | | PACE | Password Authenticated Connection Establishment (passport chip access protocol) | | PSR | Payment Systems Regulator | | SME | Small and Medium-sized Enterprise | | UKAS | United Kingdom Accreditation Service | | UKPI | UK Payments Initiative (operator of the first commercial VRP scheme) | | VRP | Variable Recurring Payment (an open banking payment the customer authorises once, within limits, for repeated use) | --- ##### 1. Summary 1. **Proximity identity is already standardised and government-backed.** Passport chips (ICAO Doc 9303), mobile driving licences (ISO/IEC 18013-5) and the EU wallet all support in-person, device-to-device checking. The UK now has a statutory register of certified identity providers and a trust mark, "UK CertifID". (Sections 2.1 to 2.6.) 2. **The UK state already uses "identity at a counter".** The Post Office checks documents in person for GOV.UK One Login, and banking hubs run by Post Office staff now number over 250. Branch closures since 2015 are reported at 6,896. The access-to-cash rules under FSMA 2023 require large banks to assess and fill gaps. (Sections 3.1 to 3.5.) 3. **Edge hardware can keep identity data on the device.** Apple's Wallet ID design keeps keys in a secure chip and presents data without server calls; ISO 18013-5 "device retrieval" makes this the standard pattern. Deepfake injection attacks are rising, and a separate European test standard (CEN/TS 18099) now covers them. (Section 4.) 4. **There are working models for turning ideas into funded action, with jobs evidence.** CDFIs report 7,399 new jobs from 2025 lending; the British Business Bank forecasts 39,000 additional jobs from 2025/26 finance; the National Wealth Fund reports over 26,500 jobs created or supported in the year. Outcome-based contracts (Life Chances Fund) hit an average of 87% of targets. Most of these figures are self-reported or modelled. (Section 5.) 5. **Money movement has known pain points with numbers.** Late payments are said to cost £11bn a year; card scheme and processing fees rose at least 25% since 2017; the SME debt finance gap is estimated at £1.6bn to £3.5bn a year. New rails (commercial VRP live since June 2026; tokenised deposit pilots completed; sterling stablecoin regime from 2027) could help. (Section 6.) 6. **A solution bank design is offered as PREDICTION** in Section 7, with inputs, selection, instruments, people, outcome measures and governance. --- ##### 2. Intention A (1): Proximity and in-person identity verification ###### 2.1 Passport chips and ICAO Doc 9303 - **AGREEMENT (published).** ICAO Doc 9303 Part 11 sets the security mechanisms for machine-readable travel documents, including Passive Authentication (checking the issuing state's digital signature over chip data) and PACE (a protocol that protects chip access). [source](https://www.icao.int/Security/FAL/TRIP/PublishingImages/Pages/Publications/Doc%209303%207th%20Part11.pdf), accessed 2026-10-31. - **CLAIM.** A 2025 Common Criteria security target cites an 8th edition of Part 11, so the 7th edition (2015) we could read may not be the latest. [source](https://www.commoncriteriaportal.org/nfs/ccpfiles/files/epfiles/2023-40_ST-Lite.pdf), accessed 2026-10-31. - **UNKNOWN.** We could not open the current ICAO edition list to confirm which edition is in force. - **OBSERVED.** The UK government's own identity app asks the user to photograph the passport, scan the biometric chip with the phone, then scan their face. [source](https://www.gov.uk/guidance/using-the-govuk-id-check-app), accessed 2026-10-31. - **OBSERVED.** The Home Office design system publishes a reusable pattern, "Ask users to scan the passport chip". [source](https://design.homeoffice.gov.uk/design-system/patterns/ask-users-for/passport-details/scan-chip), accessed 2026-10-31. - **CLAIM.** A patent filing argues that a chip read alone does not prove the person physically holds the document unless the chip supports clone detection, and pairs NFC with optical checks. [source](https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/11681883), accessed 2026-10-31. **What it means for EDGE (PREDICTION).** An EDGE point that reads a passport chip (signature check) and matches a live face to the chip photo is using the same pattern government already uses. *Assumptions:* the device supports PACE; face matching meets the liveness standards in 2.6; the person consents. ###### 2.2 UK digital identity and attributes trust framework, certified providers, trust mark - **AGREEMENT (in force).** Part 2 of the Data (Use and Access) Act 2025 creates a statutory framework for digital verification services: a trust framework, supplementary codes, a register, an information gateway and a trust mark. [source](https://www.legislation.gov.uk/ukpga/2025/18/part/2/enacted), accessed 2026-10-31. - **OBSERVED.** Most Part 2 measures came into force on 1 December 2025, and the statutory DVS register replaced the earlier non-statutory one; 48 providers with 57 services had applied at launch. [source](https://enablingdigitalidentity.blog.gov.uk/2025/12/01/dvs-measures-in-the-data-act-have-come-into-force/), accessed 2026-10-31. - **OBSERVED.** OfDIA's first annual report (14 July 2026) says the register lists 46 providers and 64 certified services; five providers were removed on 1 April 2026 when beta certificates expired. [source](https://www.gov.uk/government/publications/office-for-digital-identities-and-attributes-2026-annual-report/ofdia-2026-annual-report-on-the-operation-of-part-2-of-data-use-and-access-act-2025), accessed 2026-10-31. - **OBSERVED.** The same report says trust framework version 1.0 was published on 9 June 2026, to take effect from 1 September 2026 subject to UKAS accreditation of certifying bodies. [source](https://www.gov.uk/government/publications/office-for-digital-identities-and-attributes-2026-annual-report/ofdia-2026-annual-report-on-the-operation-of-part-2-of-data-use-and-access-act-2025), accessed 2026-10-31. - **OBSERVED.** The trust mark is named "UK CertifID", designated in March 2026; only certified and registered services may use it. [source](https://www.gov.uk/government/publications/office-for-digital-identities-and-attributes-2026-annual-report/ofdia-2026-annual-report-on-the-operation-of-part-2-of-data-use-and-access-act-2025), accessed 2026-10-31. - **OBSERVED.** The information gateway (letting public bodies share data with registered providers at a person's request) is not yet in force; the code of practice is aimed for end of 2026. [source](https://www.gov.uk/government/publications/office-for-digital-identities-and-attributes-2026-annual-report/ofdia-2026-annual-report-on-the-operation-of-part-2-of-data-use-and-access-act-2025), accessed 2026-10-31. - **CLAIM.** The same report cites survey and market estimates: 77% of UK consumers have used digital identity services; 275 firms; about £2,027m annual revenue. These are estimates reported by government, not audited counts. [source](https://www.gov.uk/government/publications/office-for-digital-identities-and-attributes-2026-annual-report/ofdia-2026-annual-report-on-the-operation-of-part-2-of-data-use-and-access-act-2025), accessed 2026-10-31. - **CLAIM.** A vendor blog reports 65 services from 47 providers on the register at 30 September 2026. Not checked against the live register. [source](https://didit.me/blog/uk-digital-identity-trust-framework/), accessed 2026-10-31. - **OBSERVED.** The employer's guide to right to work checks (1 October 2026) tells employers to use a provider from the OfDIA register for digital checks. [source](https://assets.publishing.service.gov.uk/media/6ad9dfe29ea4857e15ab3294/Employer_s_guide_to_right_to_work_checks.pdf), accessed 2026-10-31. ###### 2.3 GOV.UK One Login, GOV.UK Wallet and the national digital ID plan - **OBSERVED.** GOV.UK One Login has over 23 million users, and passkeys were rolled out to them in September 2026; nearly 1 in 10 daily sign-ins use passkeys. [source](https://www.gov.uk/government/news/millions-of-people-to-benefit-from-simpler-more-secure-way-to-sign-in-to-government-services), accessed 2026-10-31. - **CLAIM.** Trade press reports, citing a parliamentary answer, over 15 million identities proven through One Login by March 2026. We did not read the parliamentary answer itself. [source](https://idtechwire.com/gov-uk-one-login-expands-passkeys-to-more-than-23-million-users/), accessed 2026-10-31. - **OBSERVED.** GOV.UK Wallet sits inside the One Login app. The first credential (digital HM Armed Forces Veteran Card) launched in October 2025 and more than 15,000 veterans had added it by January 2026. [source](https://gds.blog.gov.uk/2026/01/21/making-the-governments-first-digital-wallet-a-reality/), accessed 2026-10-31. - **OBSERVED.** Private testing of a digital driving licence began in December 2025, with wider rollout planned "later in 2026"; registered DVS partners were to test sharing licence data. [source](https://gds.blog.gov.uk/2026/01/21/making-the-governments-first-digital-wallet-a-reality/), accessed 2026-10-31. - **UNKNOWN.** Whether the digital driving licence was publicly live by 31 October 2026. One secondary guide (September 2026) said it was not. [source](https://passrates.uk/guide/digital-driving-licence-uk), accessed 2026-10-31. - **OBSERVED.** The government consulted (10 March to 5 May 2026) on a national digital ID; it proposes digital right to work checks become mandatory by the end of this Parliament and that the ID share limited attributes such as "over 18" where possible. [source](https://www.gov.uk/government/consultations/making-public-services-work-for-you-with-your-digital-identity/making-public-services-work-for-you-with-your-digital-identity), accessed 2026-10-31. - **OBSERVED.** The government explainer says the digital ID will be free, stored on the person's phone, and that face-to-face help and physical alternatives are being considered for people without phones. [source](https://www.gov.uk/government/publications/digital-id-scheme-explainer/digital-id-scheme-explainer), accessed 2026-10-31. - **CLAIM.** Law firms report the government said in January 2026 the digital ID would not be mandatory for citizens to hold. [source](https://www.lewissilkin.com/insights/2026/03/23/uk-government-opens-consultation-on-digital-id-scheme-102mnlk), accessed 2026-10-31. - **UNKNOWN.** Whether the consultation response has been published. **What it means for EDGE (PREDICTION).** The government itself says people without phones need face-to-face help. That is the gap EDGE points could fill, as a registered DVS route or as a partner to one. *Assumptions:* EDGE (or a partner) is certified against trust framework 1.0 and listed on the register; the government chooses to use third parties for assisted routes. ###### 2.4 Post Office identity services - **OBSERVED.** In August 2023 GDS and the Post Office announced in-person identity checks for GOV.UK One Login, saying most people would be within 10 miles of a branch offering it. [source](https://www.gov.uk/government/news/government-and-post-offices-team-up-so-people-can-access-public-services-in-person), accessed 2026-10-31. - **CLAIM.** The Post Office said about 1,000 branches would offer the check; the postmaster scans the photo ID and takes a photo, and the result is emailed. It also said 99.7% of the population lives within three miles of a Post Office. [source](https://www.mynewsdesk.com/uk/post-office/pressreleases/post-office-secures-new-government-contract-for-in-person-identity-checks-3269708?site=uk), accessed 2026-10-31. - **CLAIM.** Post Office EasyID, built with Yoti, is a reusable digital identity app; the Post Office describes it as certified under the UK trust framework for right to work, right to rent and criminal record (DBS) checks. [source](https://corporate.postoffice.co.uk/media/v2ploen3/ppr_verifying-your-identity_may2025.pdf), accessed 2026-10-31. - **UNKNOWN.** The current (2026) number of branches offering One Login in-person checks, and volumes. ###### 2.5 eIDAS 2 wallets and mobile driving licences - **AGREEMENT (in force).** Regulation (EU) 2024/1183 establishes the European Digital Identity Framework; the Commission says member states must provide EUDI Wallets by the end of 2026, and service providers legally required to identify customers will have to accept the wallet. [source](https://digital-strategy.ec.europa.eu/en/policies/eudi-regulation), accessed 2026-10-31. - **AGREEMENT (in force).** Five implementing regulations on wallets were published in the Official Journal on 4 December 2024, covering integrity and core functions, protocols and interfaces, person identification data and attestations, trust framework and certification. [source](https://ec.europa.eu/digital-building-blocks/sites/spaces/EUDIGITALIDENTITYWALLET/pages/844759146/EU+Digital+Identity+Wallet+milestone+reached+as+Commission+adopts+implementing+regulations), accessed 2026-10-31. - **AGREEMENT (in force).** Implementing Regulation (EU) 2024/2977 sets rules for person identification data and attestations for wallets. [source](https://eur-lex.europa.eu/eli/reg_impl/2024/2977/oj/eng), accessed 2026-10-31. - **CLAIM.** Commentators give the exact deadline as 24 December 2026, with regulated private-sector acceptance about a year later. We could not read the regulation text itself (page returned empty). [source](https://fintech.global/2026/07/13/eidas-2-0-deadlines-loom-two-tracks-one-trust-chain/), accessed 2026-10-31. - **UNKNOWN.** Which member states will meet the December 2026 deadline. - **AGREEMENT (published).** ISO/IEC 18013-5:2021 defines the mobile driving licence. A phone and a reader start a session by NFC tap or QR code, then exchange data over BLE, NFC or Wi-Fi Aware; "server retrieval" over the internet is a separate option. [source](https://www.vde-verlag.de/iec-normen/preview-pdf/info_isoiec18013-5{ed1.0}en.pdf), accessed 2026-10-31. - **CLAIM.** Researchers (one a member of the drafting committee) warn that in 18013-5 a server-retrieval token can be sent in the clear during engagement, because proximity was wrongly treated as trust. [source](https://pomcor.com/2023/10/27/overview-of-iso-iec-18013-5-innovations-and-vulnerabilities-in-the-mdl-standard/), accessed 2026-10-31. **What it means for EDGE (PREDICTION).** "Proximity is not trust" is the key lesson. An EDGE point should verify issuer signatures and device binding every time, not assume a person is genuine because they are standing there. *Assumption:* EDGE readers implement reader authentication and refuse server-retrieval tokens sent in the clear. ###### 2.6 Liveness and deepfake defences - **CLAIM.** iProov's 2025 threat report (vendor) says native virtual camera attacks rose 2,665% and face swap attacks rose 300% in 2024. [source](https://www.iproov.com/reports/threat-intelligence-report-2025-remote-identity-attack), accessed 2026-10-31. - **AGREEMENT (published, technical specification).** CEN/TS 18099:2024 is a test method for biometric *injection* attacks (fake video fed into the system), separate from presentation attacks (masks, photos) covered by ISO/IEC 30107-3. [source](https://www.evs.ee/en/cen-ts-18099-2024), accessed 2026-10-31. - **CLAIM.** A standards listing shows a 2026 revision of CEN/TS 18099 in development. [source](https://isme.me/en/project/show/isme:proj:108766), accessed 2026-10-31. - **OBSERVED.** The ICO says biometric data is special category data when used to uniquely identify someone; there is no specific Article 9 condition for identification, so explicit consent is often used; a DPIA is highly likely to be required. [source](https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/lawful-basis/biometric-data-guidance-biometric-recognition/how-do-we-process-biometric-data-lawfully/), accessed 2026-10-31. **What it means for EDGE (PREDICTION).** In-person checks remove most *injection* attacks (no virtual camera between person and sensor) but not presentation attacks. EDGE devices should be tested against ISO/IEC 30107-3, and any remote fallback against CEN/TS 18099. *Assumption:* the device camera pipeline is sealed (signed firmware, no external video input). --- ##### 3. Intention A (2): Physical banking in the UK ###### 3.1 Branch closures - **CLAIM.** Which? (consumer group) counts 6,896 branch closures since January 2015 (about 70% of branches open then), roughly 53 a month, including scheduled closures. [source](https://www.which.co.uk/money/banking/switching-your-bank/bank-branch-closures-is-your-local-bank-closing-a28n44c8z0h5), accessed 2026-10-31. - **CLAIM.** The same tracker counts 433 closures in 2025 and 314 scheduled for 2026 so far. Its brand-level figures for 2026 add to 308, so the total may be slightly inconsistent. [source](https://www.which.co.uk/money/banking/switching-your-bank/bank-branch-closures-is-your-local-bank-closing-a28n44c8z0h5), accessed 2026-10-31. - **OBSERVED.** NatWest (June 2026) committed £50m to its branch network, said it would pause further closures "until at least 2029" after nine more in 2026, and has 336 branches. [source](https://www.natwestgroup.com/news-and-insights/news-room/press-releases/our-updates/2026/jun/natwest-commits-50m-to-uk-branch-network-with-no-further-branch-.html), accessed 2026-10-31. ###### 3.2 Banking hubs (Cash Access UK and LINK) - **OBSERVED.** Cash Access UK (owned by banks) says over 250 banking hubs are open and over 140 deposit services, supporting nearly one million customers a month. [source](https://www.cashaccess.co.uk/), accessed 2026-10-31. - **OBSERVED.** In March 2024 LINK had recommended 117 hubs, of which 37 were open; hubs have a Post Office–run counter for customers of any bank; communities can request a review. [source](https://link.co.uk/news/link-recommends-eight-new-banking-hubs-across-the-uk), accessed 2026-10-31. - **CLAIM.** A parliamentary answer reported over 280 hubs announced and over 240 open (August 2026), and a government target of 350 by the end of this Parliament. We read this through a secondary archive, not Hansard. [source](https://westminsterbrief.co.uk/archive/pq/1110), accessed 2026-10-31. - **CLAIM.** Hub model: Post Office staff run the counter; community bankers from different banks attend on different days in a private room. [source](https://www.finextra.com/newsarticle/41694/uk-to-roll-out-more-shared-banking-hubs), accessed 2026-10-31. ###### 3.3 Mobile vans and community bankers - **OBSERVED.** NatWest says it runs the UK's only mobile branch service, reaching 577 communities a week and supporting 144,000 customers in 2025, and plans more stops. [source](https://www.natwestgroup.com/news-and-insights/news-room/press-releases/our-updates/2026/jun/natwest-commits-50m-to-uk-branch-network-with-no-further-branch-.html), accessed 2026-10-31. - **CLAIM.** News reports say Lloyds ended its mobile vans after average visits fell about 90% since 2018 to around 14 customers per stop, replacing them with community bankers. The year and exact figures were not confirmed from a Lloyds source. [source](https://www.cityam.com/?p=2108893), accessed 2026-10-31. ###### 3.4 Shared counters through the Post Office - **CLAIM.** The Post Office says over 30 banks, building societies and credit unions partner with it, so 99% of UK bank customers can use their accounts at a branch; it has over 11,500 branches (figures from 2023). [source](https://www.mynewsdesk.com/uk/post-office/pressreleases/post-office-secures-new-government-contract-for-in-person-identity-checks-3269708?site=uk), accessed 2026-10-31. ###### 3.5 Access-to-cash rules (FSMA 2023) - **AGREEMENT (in force since 18 September 2024).** FCA Policy Statement PS24/8, made under FSMA 2023, requires designated banks and building societies to assess local cash access, identify significant gaps and provide extra services without undue delay; cash access must be free for personal current account customers. [source](https://fca.org.uk/publications/policy-statements/ps24-8-access-cash), accessed 2026-10-31. - **CLAIM.** Law firm summaries say the rules apply to 14 designated firms, with LINK entities designated as the coordination body, and assessments must be done within 12 weeks. [source](https://www.hoganlovells.com/en/publications/uk-access-to-cash-fca-publishes-final-rules), accessed 2026-10-31. - **UNKNOWN.** The precise distance thresholds (we could not read the full PS24/8 PDF in this session). **What it means for EDGE (PREDICTION).** The physical network is shrinking at bank level but being rebuilt as *shared* infrastructure (hubs, Post Office counters, vans). An EDGE point is most realistic as a *tenant or partner* inside shared spaces (hubs, libraries, clinics, community centres) and as a mobile unit, not as a new branch network. *Assumptions:* Cash Access UK or LINK would accept third-party services in hubs; EDGE has the regulatory permissions for any money service offered. --- ##### 4. Intention A (3): Edge hardware and privacy-preserving on-device processing - **OBSERVED.** In Apple Wallet IDs, the authentication key lives in the iPhone's Secure Element (a tamper-resistant chip); presentation uses ISO 18013-5 device retrieval with no server calls, so Apple and the issuer do not see each presentation; the holder must consent to each one. [source](https://support.apple.com/guide/security/security-of-ids-in-apple-wallet-secb569bf393/web), accessed 2026-10-31. - **OBSERVED.** Apple's ID Verifier lets an iPhone read ISO 18013-5 or ISO 23220-1 mobile IDs with no extra hardware; it has a "Display Only" mode (shows name, age, photo, age-over-N, blocks screenshots) and a "Data Transfer" mode needing entitlement and a legal requirement; currently for US businesses. [source](https://developer.apple.com/wallet/id-verifier/), accessed 2026-10-31. - **OBSERVED.** Apple's guide says a first NFC tap sets up a secure BLE link and that session data is encrypted at the application layer, independent of the radio. [source](https://support.apple.com/guide/security/security-of-ids-in-apple-wallet-secb569bf393/web), accessed 2026-10-31. - **CLAIM.** A patent describes a secure element answering a yes/no age question without revealing the date of birth. [source](https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/10579997), accessed 2026-10-31. - **CLAIM.** A vendor passport SDK lists BAC, PACE, Extended Access Control and Active Authentication and matches faces against the chip portrait. [source](https://axonwireless.com/technology/epassport), accessed 2026-10-31. - **CLAIM.** One kiosk patent sends the face comparison to a central server, the opposite of on-device privacy. [source](https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/11308743), accessed 2026-10-31. - **UNKNOWN.** How many deployed identity kiosks do face matching entirely on-device, and how they retain biometric data. We found no independent survey. **"Edge hardware for good": proposed EDGE point design (PREDICTION).** | Component | Proposed choice | Why | Assumption | |---|---|---|---| | Reader | NFC plus camera, ISO 18013-5 reader and ICAO 9303 chip reader | Covers passports, ID cards and mobile IDs | Devices certified where certification exists | | Key storage | Secure element or hardware security module on the device | Keys never leave hardware | Hardware available at reasonable cost | | Face match | Done on the device; template deleted after the match | ICO special-category rules; data minimisation | Accuracy acceptable on device | | Liveness | Tested to ISO/IEC 30107-3; sealed camera pipeline | Defeats masks and video injection | Test lab available | | Output | Signed yes/no attribute ("is over 18", "matches document") rather than raw data | Same idea as the UK digital ID proposal | Relying parties accept attribute-only answers | | Audit | Tamper-evident local log; only hashes leave the device | Proves what happened without central biometric stores | Governance agreed (Section 7.6) | | Human | A trained person present for assisted cases | Government says people without phones need face-to-face help | Staffing funded | --- ##### 5. Intention B (4): Models that turn ideas into funded action ###### 5.1 Community Development Finance Institutions (CDFIs) - **CLAIM (self-reported by members to the trade body).** Responsible Finance reports all-sector CDFI lending of about £389.9m in 2025 (up 21%) across 148,397 loans. [source](https://responsiblefinance.org.uk/wp-content/uploads/2026/05/260513-DD-Responsible-Finance-Appendix.pdf), accessed 2026-10-31. - **CLAIM (self-reported).** SME and start-up CDFI lending was about £181.4m in 2025, with 5,741 new businesses created, 7,399 new jobs created and 6,627 jobs safeguarded; weighted average interest rate 7%. [source](https://responsiblefinance.org.uk/wp-content/uploads/2026/05/260513-DD-Responsible-Finance-Appendix.pdf), accessed 2026-10-31. - **CLAIM (self-reported).** For micro and small business lending (excluding start-up loans) the weighted average rate was 14.95% and average loan about £97,700. [source](https://responsiblefinance.org.uk/wp-content/uploads/2026/05/260513-DD-Responsible-Finance-Appendix.pdf), accessed 2026-10-31. - **CLAIM.** A Responsible Finance press release headline says "over 14,000 jobs created", which appears to combine new and safeguarded jobs. [source](https://www.politicshome.com/news/uk/economy/financial-sector/press-release/responsible-finance/100547/over-14000-jobs-created), accessed 2026-10-31. ###### 5.2 Community wealth building - **CLAIM (from the council and the consultancy that designed the model).** In Preston, anchor institution spend retained in the city rose from 5% to 18.2% of procurement (about £112.3m), and in Lancashire from 39% to 79.2%, between 2012/13 and about 2017/18. [source](https://cles.org.uk/?p=36784), accessed 2026-10-31. - **CLAIM.** Preston City Council's own guide gives £488.7m retained in Lancashire; a co-operative news site gives £448.7m. The figures conflict. [source](https://www.preston.gov.uk:443/article/1791/The-definitive-guide-to-the-Preston-model), accessed 2026-10-31. - **UNKNOWN.** An independent evaluation attributing job or wage outcomes to the Preston model. We did not find one. ###### 5.3 Social impact investment and outcome-based finance - **OBSERVED.** The Life Chances Fund (£70m, DCMS) ran 29 social outcomes partnerships; the 18 completed projects achieved an average of 87% of primary outcome targets (range 20% to over 100%); 48,904 people achieved at least one outcome; £41.6m of £48.6m committed central outcome funding was paid. [source](https://www.gov.uk/government/publications/achieving-outcomes-life-chances-fund-final-report/life-chances-fund-final-report-executive-summary), accessed 2026-10-31. - **UNKNOWN.** Whether the outcome-payment model performed better than conventional grants. The executive summary does not say. - **OBSERVED.** The Better Futures Fund (announced 14 July 2025) is £500m over ten years for children and families, aiming to raise a further £500m from councils, social investors and philanthropists, and to reach up to 200,000 children. [source](https://www.gov.uk/government/news/largest-fund-of-its-kind-to-support-vulnerable-kids-families), accessed 2026-10-31. - **CLAIM.** Secondary sources say Round 1 offers up to £37m and Round 2 opens in summer 2027. [source](https://www.sibgroup.org.uk/?p=2897), accessed 2026-10-31. ###### 5.4 Challenge prizes - **OBSERVED.** The Longitude Prize on antimicrobial resistance (£8m, May 2014 to September 2022) was won by Sysmex Astrego for a point-of-care test; Challenge Works says it cut urinary infection test waits from 2–3 days to minutes. [source](https://challengeworks.org/longitude-prize/), accessed 2026-10-31. - **OBSERVED.** The Longitude Prize on Dementia (£4.42m, September 2022 to March 2026) gave its £1m grand prize to CrossSense for AI smart glasses. [source](https://challengeworks.org/longitude-prize/), accessed 2026-10-31. - **UNKNOWN.** Jobs created by challenge prizes. Prize evidence is mostly about solutions found, not employment. ###### 5.5 Venture studios - **CLAIM.** A 2023 press release cites a 2022 Global Startup Studio Network report saying studio start-ups have a "30% higher success rate". Methodology not seen. [source](https://www.businesswire.com/news/home/20230711329736/en/Matt-Keezer-Momentum-Ventures-CEO-Recognizes-the-Success-and-Impact-of-Leading-Venture-Studios/), accessed 2026-10-31. - **CLAIM.** An investor network warns the studio data sets are small, self-reported and subject to survivorship bias. [source](https://angelinvestorsnetwork.com/alternative-investments/how-venture-studios-build-startups), accessed 2026-10-31. - **UNKNOWN.** Independent, peer-reviewed jobs or return figures for venture studios. Treat all studio performance claims as unproven. ###### 5.6 Public development banks - **OBSERVED.** The British Business Bank reports about £9.4bn of finance supported in 2025/26 (£1.3bn public, £4.3bn private capital attracted, £3.7bn guaranteed lending), about 38,000 businesses reached, and forecasts 39,000 additional jobs over the life of the finance (a forecast, not a count). [source](https://www.british-business-bank.co.uk/news-and-events/news/british-business-bank-supports-ps94bn-finance-smaller-business-202526-returning-statutory-profit), accessed 2026-10-31. - **OBSERVED.** The National Wealth Fund (£27.8bn capitalisation) reports for 2025/26: 27 transactions, £4.1bn deployed, £4.9bn private finance mobilised, over 26,500 jobs created or supported; since 2021: £9.5bn deployed, £19.1bn mobilised, over 76,000 jobs. [source](https://assets.publishing.service.gov.uk/media/6aa2ce144e17456e2cdf536c/National_Wealth_Fund_Limited_Annual_Report_and_Accounts_2025-2026.pdf), accessed 2026-10-31. - **OBSERVED.** The NWF states its jobs metric is "the number of jobs supported or created from those projects", assessed as expected impact; it targets over 200,000 jobs by 2050. [source](https://assets.publishing.service.gov.uk/media/6aa2ce144e17456e2cdf536c/National_Wealth_Fund_Limited_Annual_Report_and_Accounts_2025-2026.pdf), accessed 2026-10-31. **Evidence quality on jobs (OBSERVED from the sources above).** CDFI jobs are self-reported by lenders; BBB jobs are a lifetime forecast; NWF jobs are expected "created or supported" figures for projects. None of these is an independent count of net new jobs. A solution bank should not copy this weakness (see 7.5). --- ##### 6. Intention B (5): Money movement and lending problems ###### 6.1 Late payments - **OBSERVED.** The government says late payments cost the economy £11bn a year and 38 businesses close each day because they are not paid on time; it gives no direct source for either figure in the release. [source](https://www.gov.uk/government/news/time-to-pay-up-government-unveils-toughest-crackdown-on-late-payments-in-over-25-years), accessed 2026-10-31. - **OBSERVED.** Measures announced on 24 March 2026: a 60-day cap on payment terms when large firms pay smaller suppliers; mandatory statutory interest of 8% above Bank Rate; stronger Small Business Commissioner powers including fines; a ban on withholding construction retentions; board-level reporting for persistent late payers. [source](https://www.gov.uk/government/news/time-to-pay-up-government-unveils-toughest-crackdown-on-late-payments-in-over-25-years), accessed 2026-10-31. - **OBSERVED.** The government says the bill entered Parliament in May 2026. [source](https://www.gov.uk/government/news/largest-crackdown-on-late-payments-in-over-25-years-as-landmark-bill-enters-parliament), accessed 2026-10-31. - **UNKNOWN.** The bill's current stage and commencement date. A Commons debate on late payments took place on 15 October 2026. [source](https://hansard.parliament.uk/commons/2026-10-15/debates/4B9DA7ED-EFA4-49A4-B995-D0792C2B56F5/LatePayments), accessed 2026-10-31. ###### 6.2 SME payment cost and speed - **OBSERVED.** The PSR's final report (6 March 2025) found Mastercard and Visa raised core scheme and processing fees to acquirers by at least 25% since 2017, costing UK businesses at least £170m extra a year. [source](https://www.psr.org.uk/mr22-1-10-scheme-and-processing-fees-final-report/), accessed 2026-10-31. - **CLAIM.** A law firm note (30 July 2026) reports final PSR directions to card schemes on transparency and pricing governance. We did not read the directions. [source](https://finreg.aoshearman.com/uk-psr-market-review-final-directions-to-card-schemes-on-information-transparency-and-complexity-and-pricing-governance), accessed 2026-10-31. - **OBSERVED.** The government is consulting on merging the PSR into the FCA. [source](https://www.gov.uk/government/consultations/a-streamlined-approach-to-payment-systems-regulation-consultation), accessed 2026-10-31. ###### 6.3 SME lending gaps - **OBSERVED.** The British Business Bank reports gross SME bank lending (excluding overdrafts) of £68bn in 2025, up 9%, with challenger and specialist banks supplying 60%. [source](https://www.british-business-bank.co.uk/news-and-events/news/smaller-business-lending-markets-showing-signs-improvement-finds-latest-british-business-bank), accessed 2026-10-31. - **OBSERVED.** The same report says Ethnic Minority-led, female-led and disability-led businesses show higher levels of discouragement (expecting rejection, not knowing where to look, process too slow). [source](https://www.british-business-bank.co.uk/sites/g/files/sovrnj166/files/2026-03/report-small-business-finance-markets-2026.pdf), accessed 2026-10-31. - **CLAIM.** A January 2026 ministerial letter (read through a secondary archive) cites BBB analysis of an annual SME debt finance gap of £1.6bn to £3.5bn (2023 data, assuming 20,000 to 50,000 viable businesses and average loans of about £80,000), and a separate start-up debt gap of £8.2bn a year that cannot be added to it. [source](https://www.beyondthevote.co.uk/committee-publication/c5a00d58-f390-4dfa-8dff-8babb5704133), accessed 2026-10-31. - **CLAIM.** Other government announcements give the gap as £1.6bn to £4.1bn. The difference in method is not explained in what we read. [source](https://www.digit.fyi/500m-boost-to-smes-and-scaleups-from-uk-gov-and-british-business-bank/), accessed 2026-10-31. ###### 6.4 Open Banking payments and variable recurring payments - **OBSERVED.** Open Banking Limited reports UK open banking has passed 1 billion payments and 100 billion API calls (September 2026). [source](https://www.openbanking.org.uk/), accessed 2026-10-31. - **OBSERVED.** In September 2026 there were about 2.92 billion successful API calls, of which about 169 million were payment initiation calls. [source](https://www.openbanking.org.uk/api-performance/), accessed 2026-10-31. - **OBSERVED.** The CMA mandated the nine largest banks to offer VRP for "sweeping" (moving money between a customer's own accounts); other VRP uses are not mandated. [source](https://www.openbanking.org.uk/variable-recurring-payments-vrps/), accessed 2026-10-31. - **OBSERVED.** The FCA welcomed the launch of the UK Payments Initiative commercial VRP scheme on 2 June 2026 and plans to consult on a long-term framework by end of 2026, subject to new powers. [source](https://www.fca.org.uk/news/statements/open-banking-launch-uk-payments-initiative-scheme), accessed 2026-10-31. - **CLAIM.** UKPI says initial uses include payments to government, utilities, charities and financial services. [source](https://www.ukpaymentsinitiative.co.uk/uk_banks_and_fintechs_join_forces_to_launch_new_payments_scheme/), accessed 2026-10-31. - **OBSERVED.** The National Payments Vision (HM Treasury, November 2024, updated July 2026) sets the government's payments strategy. [source](https://www.gov.uk/government/publications/national-payments-vision), accessed 2026-10-31. ###### 6.5 Stablecoins and tokenised deposits **UK** - **OBSERVED.** The FCA published final rules on stablecoin issuance and cryptoasset custody on 30 June 2026; they apply to firms authorised under the new regime from 25 October 2027. [source](https://www.fca.org.uk/publications/consultation-papers/cp25-14-stablecoin-issuance-cryptoasset-custody), accessed 2026-10-31. - **OBSERVED.** The Bank of England's November 2025 consultation proposed holding limits for systemic sterling stablecoins of £20,000 per individual and £10m per business. [source](https://www.bankofengland.co.uk/paper/2025/cp/proposed-regulatory-regime-for-sterling-denominated-systemic-stablecoins), accessed 2026-10-31. - **OBSERVED.** The Bank's policy statement (22 June 2026) replaced holding limits with a temporary issuance guardrail per systemic stablecoin, and plans to finalise its Code of Practice by end-2026; systemic stablecoins will be regulated jointly with the FCA. [source](https://www.bankofengland.co.uk/paper/2026/ps/sterling-denominated-systemic-stablecoin), accessed 2026-10-31. - **CLAIM.** Trade press reports the guardrail is £40bn per coin and the backing split is 70% short-term gilts and 30% unremunerated Bank deposits. We could not read those numbers in the truncated primary page. [source](https://www.theblock.co/news/regulation/2026-06-22-bank-of-england-drops-proposed-individual-holding-caps-for-sterling-stablecoins-sets-40-billion-issuance-guardrail-405545), accessed 2026-10-31. - **OBSERVED.** UK Finance's GBTD pilot (Barclays, HSBC UK, Lloyds, Monzo, Nationwide, NatWest, Santander) completed retail live pilots, announced September 2026: two remortgages where funds were locked and released automatically at completion, and one marketplace purchase with funds released on exchange of goods. [source](https://www.ukfinance.org.uk/tokenised-sterling-deposits), accessed 2026-10-31. **EU** - **AGREEMENT (in force).** MiCA applied fully from December 2024; the transitional period for existing crypto service providers ends on 1 July 2026; ESMA keeps an interim register, including e-money token issuers. [source](https://www.esma.europa.eu/esmas-activities/digital-finance-and-innovation/markets-crypto-assets-regulation-mica), accessed 2026-10-31. - **CLAIM.** A non-official tracker counts 25 EMT issuers in ESMA's register at 14 October 2026; the EBA has not designated any EMT as "significant". [source](https://casptracker.eu/de/e-geld-token-liste-mica/), accessed 2026-10-31. **What it means for EDGE (PREDICTION).** The conditional-release pattern in the GBTD pilot (money locked, released when a condition is met) is exactly what outcome-based funding needs: release tranches when milestones are verified. *Assumptions:* tokenised deposits or conditional VRP become commercially available to non-bank partners; the condition is verified by a trusted, auditable source (for example an EDGE point check). --- ##### 7. Intention B (6): How a solution bank could work (all PREDICTION) Everything in this section is a design proposal. It is not evidence and not advice. Each part lists assumptions. ###### 7.1 Inputs: where solutions come from | Input | Example | Assumption | |---|---|---| | Agent-generated proposals | JudeAI agents draft solutions to stated local problems (late payment cash gaps, cash access gaps, clinic access) | Agents are governed: sources cited, uncertainty stated, human review | | Community requests | Like LINK's community request route for hubs (Section 3.2) | People can submit problems in person at an EDGE point, not only online | | Public challenges | Government or council "challenge" briefs, in the style of challenge prizes (5.4) | Public bodies publish problems with outcome prices | | Existing evidence | Evaluated programmes (Life Chances Fund, CDFI models) | Evidence is current and relevant locally | ###### 7.2 Selection: deciding what to fund - Score each proposal on: evidence that it works; size of the gap it fills (using sources such as the BBB finance gap or Which? closure data); cost per outcome; who carries it out locally; risk of harm. - Two-key rule: an agent may recommend; only a named human panel (including people from the affected community) may approve. - Publish every decision and its reasons. - *Assumptions:* enough local panel members can be recruited and paid; scoring criteria are agreed in advance and not changed after results. ###### 7.3 Funding instruments | Instrument | Use | Precedent in this note | Assumption | |---|---|---|---| | Small grants / prizes | Early testing of an idea | Longitude Prize (5.4) | Philanthropic or public money available | | CDFI-style loans | People starting or growing a business | CDFI 2025 figures (5.1) | Partner CDFI or bank; EDGE does not lend without authorisation | | Outcome payments | Paid only when a verified result happens | Life Chances Fund, Better Futures Fund (5.3) | A public or philanthropic outcome payer commits | | Guarantees | Making mainstream lenders lend to thin-file borrowers | BBB guarantee schemes (5.6) | Guarantee provider partners | | Conditional money release | Tranches released on verified milestones | GBTD conditional payments (6.5) | Tokenised deposits or VRP available to partners | | Fast invoice payment | Paying small suppliers in days, not 60 | Late payment reforms (6.1) | Large buyers join; cost covered | ###### 7.4 People empowered - **Enactors:** individuals and small organisations who carry out the solution, paid for their work. - **Verifiers:** trained local people who staff EDGE points, check identity and confirm milestones in person. - **Panels:** community members who approve funding. - *Assumption:* the in-person verifier role is itself a source of local jobs, in line with the government's own view that face-to-face help is needed (2.3). ###### 7.5 Outcome measurement: jobs, time, money, lives | Measure | How measured | Avoids the weakness in 5.6 by | Assumption | |---|---|---|---| | Jobs | Count of paid roles verified (payroll or HMRC data with consent), held for 6 and 12 months | Counting actual, lasting jobs, not forecasts | People consent to data sharing | | Time | Days to payment, days to funding decision, travel time to service | Measuring before and after | Baselines recorded at start | | Money | Cash released to enactors; cost per outcome; repayments; fees saved | Publishing cost per verified outcome | Accounts audited | | Lives | Agreed wellbeing or health outcomes (with EDGE Clinic where relevant) | Using validated measures, not self-chosen ones | Ethics approval where health data is involved | | Counterfactual | Compare with similar areas or people not funded | Showing what would have happened anyway | A comparison group exists | ###### 7.6 Governance - Every agent recommendation, human decision, money movement and in-person verification is logged and traceable to its evidence. - Biometric data never leaves the EDGE device (Section 4); only signed yes/no results and hashes are stored centrally. - Independent evaluation by a body not paid by EDGE. - Regulated activities (lending, payments, identity verification as a registered DVS) are carried out by authorised partners, or by EDGE only once it holds the right permissions. - *Assumptions:* partners accept shared logs; regulators accept the model; funding for independent evaluation is ring-fenced. ###### 7.7 Biggest risks (PREDICTION) 1. Agent-generated proposals that look good but are wrong. *Mitigation:* evidence scoring and human panels. 2. Exclusion of people who cannot use phones. *Mitigation:* EDGE points and mobile units. 3. Biometric misuse or breach. *Mitigation:* on-device processing; ICO DPIA. 4. Overclaiming impact (as with forecast job figures). *Mitigation:* verified job counts and counterfactuals. 5. Regulatory overreach (acting as bank, payment firm or identity provider without permission). *Mitigation:* partner-first approach and legal advice. --- ##### 8. Open questions (UNKNOWN) 1. Is the GOV.UK digital driving licence publicly live, and will GOV.UK Wallet credentials be readable by third-party in-person readers? 2. When will the DVS information gateway come into force? 3. Which EU member states will meet the December 2026 wallet deadline? 4. Will Cash Access UK or LINK allow third-party services such as identity checks inside banking hubs? 5. What is the bill name and stage for the late payment legislation, and when does it commence? 6. What independent evidence exists on jobs created by venture studios and challenge prizes? 7. Has the government published its response to the 2026 digital ID consultation? ### research/finance/REFERENCE-CARDS.md `research/finance/REFERENCE-CARDS.md` · sha256 `b0ef8bb6bd46822bc9a9de54b9c73fe7c2291c00ca414cb678095914f8873b2e` #### EDGE Bank Cards: the reference and the build, element for element Lane: Claude-Jegudiel 00:5531102026-1, 1 November 2026. Files: runtime/bank_app.py (Cards tab, card and payee tap patterns), runtime/screen_patterns.py (pick_rows, ledger_items), runtime/web_assets.py (.ux-head, .ux-cardstage, .ux-cardp, .ux-switch, .ux-pick, .ux-cardthumb, the quiet evidence control). **Statement class.** The "Reference" column is an OBSERVATION of Mobbin screenshots read on 1 November 2026 through the Mobbin search tools (`search_flows`, platform ios, queries "Starling Bank freeze card", "Starling Bank card controls", "Monzo card settings"). The "Build" column is what runtime/bank_app.py draws from the EDGE Bank ledger (runtime/domains/bank.py). Product names belong to their owners. Masked personal details in the screenshots are not reproduced. ##### Chosen reference | Surface | App | Mobbin flow | Screen id | Why this one | |---|---|---|---|---| | Cards tab | Starling, iOS, flow "Cards" | https://mobbin.com/flows/e2a47669-cab7-4f04-85e9-776db5a61b59 | `c71f082f-6203-4a6f-9a83-af030042e676` | The whole card screen on one page: card art, name, masked number, the lock switch, one list of controls | | Card controls rows | Starling, iOS, flow "Card controls" | https://mobbin.com/flows/862b50ac-eec9-4f9a-8583-9ed5e1138992 | `c0cef99f-0a21-46bd-a69f-d7fdf14a6698`, `153a10f0-66b8-4df5-835f-0af522472d51` | Each control is a row: icon, title, state, one sentence of what it does | | Gambling block row | Starling, iOS, flow "Gambling controls" | https://mobbin.com/flows/f3d1f086-6660-4cc2-925a-5a51ce98c44f | `b61b49f3-651e-4023-bda3-c78b6afe5be9` | Row copy "Turn on the gambling block · Stop card payments to gambling operators" | | Limits (not built, linked) | Starling, iOS, flow "Card and payment limits" | https://mobbin.com/flows/f19b5e70-c6ce-4a54-9f58-5d7bc2e5f84f | `5c563c61-…`, `5c707cc8-…` | Single contactless limit with a slider, £0.00 means contactless off; the estate's journey set-card-limits-and-controls covers it | | Also read | Monzo, iOS, flow "Card detail" | https://mobbin.com/flows/3c3788f7-d1f2-4d40-8e5c-8ce40636a509 | `7721f518-…`, `cd94cb0a-…` | Card art, Reveal PIN and Freeze card as two large buttons, card details with Reveal; Controls (Magstripe ATMs, Limits), Replacement (Order new card) | | Also read | Monzo, iOS, flow "Limits" | https://mobbin.com/flows/3b192e59-2820-4262-83bc-93712be6a62c | `badfaacb-…` | Limits grouped Using your card / ATM withdrawals / Outbound / Inbound, each with "left £…" | Starling was chosen over Monzo because its Cards screen is one scrolling page with a switch and a list, which is what the estate's single `/bank/cards` route can draw without a sheet; Monzo's freeze is a large button on a card sheet. ##### Cards tab `/bank/cards`, element for element (Starling screen c71f082f) | # | Reference element (Starling) | Build (EDGE Bank) | Data | |---|---|---|---| | 1 | Eyebrow "Personal", title "Cards", link "All cards" right | `.ux-head`: eyebrow Personal (Demo pill for the demo tenant), h1 Cards, link All cards → `/finance/cards` (the cards capability) | tenant | | 2 | Portrait card centred: brand name, chip, scheme mark | `.ux-cardp`: portrait 150×236, brand gradient, the approved EDGE Bank app mark top left, last four bottom. No chip, no scheme mark, no drawn art. Frozen: greyscale with a "Frozen" tag | `card`, `card_status` | | 3 | "Personal card" label | `.ux-cardp-name` "Personal card" ("Personal card · frozen" when frozen) | `card_status` | | 4 | Masked number "•••• •••• •••• ____" | `.ux-cardp-mask` •••• •••• •••• 4417 | `card` | | 5 | White rounded list, row "Card unlocked" with lock icon and a switch | `.ux-switchrow`: lock icon (snowflake when frozen), "Card unlocked" / "Card frozen", sub line, `.ux-switch` role=switch. The switch is a form POST to `/finance/freeze-unfreeze-card/toggle/do/continue` with `record.step_2`, which runs the journey's operation `finance/freeze-or-unfreeze-a-card/toggle-freeze` (receiver estate-ledger:bank). The person lands on the journey's Confirm step; no new API | `card_status` | | 6 | Row "Card details ›" | Row → `/finance/cards/card` (cards capability, screen card-detail). Sub line: holder and expiry | `card_holder`, `card_expiry` | | 7 | Row "PIN reminder ›" | Row → `/finance/view-pin-securely` (step-up, reveal for ten seconds, reveal logged) | | | 8 | Row "Card & currency controls ›" (opens the Card controls sheet) | Drawn as the second list "Card controls" below (element 12–15), because `/bank/cards` has one route and no sheet | | | 9 | Row "Card & payment limits ›" | Row → `/finance/set-card-limits-controls` (contactless, online, abroad, limit) | | | 10 | Row "Going abroad? ›" | Row → `/finance/set-card-limits-controls` (the abroad control lives in that operation) | | | 11 | Row "Virtual wallets ›" / "Mobile wallets ›" | Absent: no wallet journey exists in spec/units/finance (never a dead control) | | | — | Row "Cancel & replace card ›" | Row "Cancel & replace card", sub "Lost, stolen or damaged" → `/finance/report-lost-stolen-card` | | ##### Card controls list (Starling screen c0cef99f, 153a10f0), element for element | # | Reference element | Build | Data | |---|---|---|---| | 12 | Row "Online payments", switch, "Your card is unlocked and ready to use online." | Row "Online payments · Shops and websites" → `/finance/set-card-limits-controls`. No switch: the ledger holds no per-control state, so none is claimed | none yet | | 13 | Row "Card present" (contactless and chip and PIN), switch | Row "Contactless · Tap to pay, up to your single-payment limit" → `/finance/set-card-limits-controls` | none yet | | 14 | Row "ATM withdrawals", switch | Row "Cash withdrawals · Cash machines in the UK and abroad" → `/finance/cards/spending-controls` (field controls.cash_enabled) | none yet | | 15 | Row "Gambling payments" switch (c0cef99f) and the "Turn on the gambling block" row (b61b49f3) | Row "Gambling block · Stop card payments to gambling operators" → `/finance/handle-gambling-block-request` | none yet | | — | "Magstripe payments", "Unblock PIN/CVV" | Absent: no journey | | Switches for rows 12–15 are not drawn until the ledger holds the state. Needed in runtime/domains/bank.py (not this lane's file): `card_controls(engine, actor) -> {"online": bool, "contactless": bool, "cash": bool, "gambling_block": bool, "abroad": bool, "single_limit_pence": int}` read from a `bank_card_controls(tenant, account, control, enabled, limit_pence, at)` table, written by the receivers of `finance/set-card-limits-and-controls/set-contactless-online-abroad-limit`, `finance/cards/set-spending-controls` and `finance/handle-a-gambling-block-request/enable-block`. With that read, each row becomes a `.ux-switchrow` posting to its journey step the way the freeze switch does. ##### Journey steps that name a card or a payee | Step | Before | Now (Starling "Payments" recent payees, Monzo "Pay someone"; same row anatomy as the Cards list) | |---|---|---| | `/finance/freeze-unfreeze-card/open` "Which card?" | one free text box "Type the name or reference of the card" | "Your cards": a row per card from the ledger (card thumb, "Personal card", account · unlocked/frozen); a tap selects it (radio bound to `record.step_1`, value "Card ending 4417"); "Or enter a reference" with the text box below; Continue sends it | | `/finance/set-card-limits-controls/open` "Which card controls?" | free text | the same card rows | | any finance step whose single text field asks for a payee or recipient | free text | "Recent payees": initials, name, sort code · account number · name matched; a tap fills the field | | any finance `id` field bound to `card.*` or `payee.*` (for example cards/card-evidence "Card") | a list of recent records linked to `?pick=`, which nothing read | the ledger's cards or payees as rows; the radio carries the id | | other `id` fields in any application | the `?pick=` link | the same radio rows over the estate record store | No script runs (CSP `default-src 'self'`), so a row tap selects and Continue sends; the person is never asked to type what the bank already knows. ##### Evidence control on journey screens `open-…-evidence` controls are authored as Button.Destructive in several finance units (freeze, gambling block), which drew a red primary button under Continue. `.ux .controls .btn[data-control*="eviden"]` now draws it as a quiet underlined text link. The one-line change for runtime/web.py (not this lane's file), in `render_screen` where `BUTTON_CLASS` is looked up: `"btn btn-link" if "evidence" in str(c.get("id")) else BUTTON_CLASS.get(c.get("component"), "btn")`. ##### Walk `python3 tools/journeys_http.py --demo --only /tmp/cards-lane/only.txt --shard 1/1` on the live release 7a2a0ea, 1 November 2026 06:36Z: 9 journeys, 9 pass (freeze-or-unfreeze-a-card, set-card-limits-and-controls, report-a-lost-or-stolen-card, view-pin-securely, handle-a-gambling-block-request, add-a-payee-with-name-check, send-money, cards, dispute-a-card-transaction). Output ledger/evidence/journeys-demo-rerun-1.jsonl on the host. Found on the way: the demo tenant's ledger seed was silently skipped because an earlier row keyed by the same account id had been relabelled tenant `demo-walker-legacy` (`INSERT OR IGNORE` on the primary key), so `toggle-freeze` answered NO_ACCOUNT for the demo customer. The legacy rows for that one account were removed from `.cache/demo.sqlite` (demo data) and the seed now runs; `_seed_demo` in runtime/domains/bank.py should also check the account id, not only tenant and person. ### research/finance/REFERO-FLOWS-ALL.md `research/finance/REFERO-FLOWS-ALL.md` · sha256 `ef1e18395b33eefc90d60ac18ad594f4480f035a36e8233cbef2a86dafc8cb56` #### EDGE Finance research: real product flows from Refero **What this is.** Real screens and user flows from existing products, recorded by Refero (a library of real app and website screenshots, each journey described step by step). They show the floor (what users already expect) and good customer experience from any industry worth borrowing. **Statement class.** Every flow is an OBSERVATION by Refero of a real product, searched 2026-10-31. Descriptions are Refero's wording. Product names belong to their owners. **How it was gathered.** Semantic searches for this application's tasks on iOS and web. Results from other industries are kept on purpose: they are often the best customer experience to learn from. Refero does not hold many incumbents (and few government, enterprise or NHS systems), so this is a sample, not a census. **Products in this sample:** PayPal (35), District (32), Mercury (25), Family (19), Acctual (17), Wealthsimple (15), Cashcaded (15), Phantom (14), Copperx (12), Fuse (10), Airbnb (10), Lovable (10), Apple Stocks (10), GlossGenius (9), WhatsApp (9), Revolut (9), GoFundMe (9), Dropbox (8), Runey (8), Blackbird (7), Cake Equity (7), 222 (7), Stocktwits (7), Artsy (6), Wise (6), ElevenLabs (6), Fresha (6), Google Translate (6), Wabi (6), Cohere (6), TIDAL (6), UGLYCASH (5), Pinterest (5), Instagram (5), Column (5), OpenAI (5), Fernand (5), TikTok (5), Drinkit (4), Aboard (4), X (4), DailyArt (4), Apple TV (4), Adobe Photoshop (4), Reown (4), ElevenReader (3), Turf (3), Gizmo (3), Fourmula.ai (3), OpenSea (3), Exa (3), Givingli (3), Genie (3), Subo (3), Omio (3), On (3), Kickstarter (3), Vocabulary (3), Steel (3), Google Gemini (3), Granola (3), Understory (3), Spyglass (3), Grammarly (3), Rox (3), Teal (3), Shop (3), Apple Books (3), Tinder (3), Meta AI (3), Sora (3), Zendesk (3), n8n (3), ElevenMusic (2), Square (2), Claude (2), Voicenotes (2), Grok (2), Navan (2), Polar (2), TravelPerk (2), ROVR (2), Farfetch (2), Spoil me (2), Instacart (2), Orbit (2), Queue (2), RAD Weather (2), DoorDash (2), Doctronic AI (2), Trip (2), TIDE (2), Willow (2), AVNIER (2), MyMind (2), Bump (2), Train Fitness (2), Moonly (2), Chance AI (2), Plane Finder (2), Fingerprint (2), Tidal (2), Loowner (2), The Outsiders (2), Moises (2), Foodllama (2), Literal.club (2), Podimo (2), Twitch (2), WishWish (2), HYPE (2), Ground News (2), komoot (2), Gumroad (2), Brilliant (2), Binge (2), Pinkllama (2), Linktree (2), Poppy (2), Mozi (2), BeReal. (2), Сontra (2), Kit (2), Parallel (2), Product Hunt (2), Mapbox (2), Spotify (2), Doji (1), Tilt (1), Ada (1), Volv (1), Nooka (1), Peerlist (1), Intercom (1), Dopper (1), Exactly.ai (1), time2book (1), Introspect (1), HBO Max (1), Preply (1), Comet (1), Abode (1), pushr (1), Urban Outfitters (1), Partiful (1), The Leap (1), Asos (1), ten ten (1), Faire (1), Handshake (1), Melius (1), Apple Image Playground (1), Pool (1), YouTube Music (1), Curater (1), Waterllama (1), Picnic (1), Miles (1), Cushion (1), Butter (1), Modyfi (1), Nudea (1), Kann (1), Teachable (1), IKEA (1), UNHINGED (1), Riverside (1), Carbon (1), Stemz (1), Shopify (1), Typeform (1), Anam (1), Pi (1), Resend (1), ChatGPT (1), Wittl (1), Napper (1), Daze (1), Stardust (1), Programa (1), Manus (1), Verse (1), Denim (1), Linear (1), Kin (1), Haptic (1), Tripsy (1), Ballpark (1), Stripe (1), Appwrite (1), Loom (1), Homerun (1), Figma (1), LEGO (1), Chargetrip (1), Raycast (1), Bond (1), Care (1), Tripadvisor (1), Record Club (1), Sunsama (1), Calendly (1), Dribbble (1), Digg (1), Cosmos (1), ZARA (1) ##### Task: AI agent setup permissions **ElevenMusic: Artist Identity Setup** (ios, 3 screens, AI Tool, Entertainment, Music Streaming) [Refero](https://refero.design/flows/12441) > Happy path for setting up an artist identity during onboarding. The user starts on a “Become an artist” screen where they can add a stage name and profile photo, or skip this step for later. Tapping the photo control opens a media picker with Photos/Collections tabs and a close action. After selecting an image, the user returns to the setup screen with the chosen photo applied. Steps: Artist Intro → Photo Picker → Artist Intro With Photo **Comet: Agent card gallery** (ios, 3 screens, AI Tool, Utility) [Refero](https://refero.design/flows/10948) > Happy-path summary: The user starts on a mobile "card gallery" that shows two active AI agent cards. From this gallery the user inspects either card, brings one into focus (front), optionally drags to reorder the two cards, and returns to a side-by-side overview. Steps: Card gallery overview → Bring card to front / partial overlap → Updated overview / reordered cards **OpenAI: Developer onboarding and API key setup** (web, 26 screens, Business, Development Tools, AI Tool) [Refero](https://refero.design/flows/4178) > Happy-path onboarding for a new developer on the OpenAI platform. Entry: public developer overview (not signed in). The user chooses to sign up and proceeds through a multi-step, linear onboarding wizard: provide email, set password, verify email, enter personal details, optionally invite team members, create a project and API key, review code snippets, and optionally buy credits. Exit: user completed onboarding and is redirected back to the developer docs/overview. Assumptions: user is not logged in, using a desktop browser, and has permission to create an organization/project and generate API keys. Steps: Entry — developer overview → Create account — email entry → Email entered → Email validation error → Email corrected → Create account — password entry → Password meets requirement example → Password masked → Verify email prompt → Tell us about you → Invite your team — add emails → Multiple team invites added → Invite error — duplicate or pending invite → Make your first API call — name key → API key form filled → API key revealed — curl example → API key code snippets — Node tab → API key code snippets — Python tab → Add API credits (optional) → Exit — return to developer overview/docs **n8n: AI Agent automation editor** (web, 7 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9520) > This is a visual, node-based workflow editor where a user composes an "AI Agent" automation on a canvas and uses an AI assistant in a right-side panel to generate or modify the workflow and documentation. Steps: Canvas overview → Open assistant templates → Enter custom prompt → Assistant begins generation → Assistant expands checklist → Review generated documentation → Finalize configuration and run **Pinterest: Collaborator management** (ios, 3 screens, Reference, Inspiration Board, Social Networking, Lifestyle) [Refero](https://refero.design/flows/2493) > Happy path: A board owner opens the Collaborators screen to view current collaborators and invited people, then adjusts collaborator permissions and invitation settings. Entry: user opens Collaborators from a board. Exit: permission level and invitation toggles are saved and the UI returns to the Collaborators list. Steps: Collaborators list → Permissions options (default) → Permissions — updated **Tripsy: Onboarding and First Launch** (ios, 8 screens, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/2326) > First-launch onboarding for a travel organizer (iOS). The user is introduced to the product via a multi-slide carousel that highlights key features (flight updates, collaboration, documents). During or immediately after onboarding the app requests runtime permissions (notifications and location). An upsell (Tripsy Pro) promotional modal can be shown before the user lands on the main 'My Trips' screen. Steps: Notification Permission Prompt → Welcome / Intro Slide → Feature Slide — Flight Details → Feature Slide — Collaboration → Feature Slide — Documents → Location Permission Prompt → Promotional Upsell (Tripsy Pro) → My Trips — Empty State Home **Dropbox: Folder access management** (web, 4 screens, Collaboration Tools, Business, Productivity) [Refero](https://refero.design/flows/13031) > A logged-in user manages sharing and permissions for a team folder in a cloud file manager. The flow starts from the folder’s action menu, opens an access management dialog, changes a collaborator’s permission level, and confirms the update with a success message. **Reown: Manage team member roles** (web, 3 screens, Development Tools, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/5734) > Happy path: an administrator opens the organization's Members area, opens the per-member actions menu for a specific member, and promotes that member to an admin. Steps: Members list overview → Open per-member actions → Apply role change and reflect update ##### Task: AI chat assistant **PayPal: PayPal Assistant Conversation** (ios, 16 screens, Finance) [Refero](https://refero.design/flows/7266) > Happy-path walkthrough of a logged-in mobile user opening the Message Centre, launching the PayPal Assistant, asking for transaction help, receiving an automated answer with detailed steps and links, giving feedback, and ending the conversation. Steps: Message Centre (empty) → Assistant greeting and quick options → User quick reply sent → Assistant 'no transactions found' → User declines search change → User enters free-text question → Assistant processing response → Assistant provides step-by-step answer → Related links and feedback prompt → User taps thumbs up (positive feedback) → Feedback input (user typing) → Send feedback → Open End Conversation action → Confirm end conversation → Message Centre updated → End state and return to idle **Granola: Review Generated Email** (ios, 2 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14157) > The user is viewing an AI-generated follow-up email within a mobile assistant conversation. The flow starts with an existing prompt and generated email response, then continues when the user selects the positive feedback control. It ends with the response marked as helpful. The user is assumed to be in an active conversation on a mobile device; authentication and permission state are not visible. Steps: Generated Email Review → Positive Feedback Submitted **GlossGenius: Analytics with AI Growth Analyst** (web, 4 screens, Business) [Refero](https://refero.design/flows/9979) > Happy path: A signed-in user opens the Analytics section from the left navigation and sees key performance indicators (net revenue, projected revenue, service/retail revenue) for the selected timeframe and staff filter. A persistent right-hand panel titled 'AI Growth Analyst' offers suggested prompts and an input area. The user selects (or types) a suggested prompt to set a revenue goal; the assistant shows brief processing states and then returns a written recommendation with numeric targets and tactical suggestions. Steps: Analytics overview with AI suggestions → User submits suggested prompt → Assistant processing state → Assistant provides recommendation **Granola: Copy Generated Response** (ios, 2 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14156) > A user views a generated follow-up email in a mobile AI conversation and copies the response to the clipboard. The flow starts on an open conversation with the generated message visible and ends with confirmation that the message was copied. The user is assumed to be on a mobile device with access to the conversation; no additional authentication or permission state is visible. Steps: Generated Email Response → Copy Confirmation **Linktree: Insights and AI Tips** (web, 6 screens, Business, Social Networking) [Refero](https://refero.design/flows/12410) > Logged-in user is viewing the Insights section of a Linktree-style dashboard on desktop. The entry point is the main analytics dashboard, with a date-range filter set to the last 7 days. The user can review performance metrics, open an AI tips modal, and launch a right-side AI chat assistant to generate content ideas. The flow ends with the AI assistant actively prompting the user for topic input, while the underlying insights dashboard remains visible in the background. Steps: Insights Dashboard → Personalized Tips Prompt → AI Chat Opened → AI Analysis in Progress → AI Recommendation Response → Brainstorming Prompt **Cohere: Tool-assisted Charting with Python Interpreter** (web, 6 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3837) > Happy path: A logged-in user opens the Chat workspace in Tool Use mode, selects helper tools (Python Interpreter, Calculator, Web Search), types a single prompt requesting a plot of average 1-bedroom rental prices for the top 5 most expensive North American cities in Jan 2024, and sends it. The system invokes tools (web search then Python interpreter), generates code, executes it, and returns a matplotlib bar chart image embedded in the conversation along with code and references. Steps: Tool selection & prompt composer → Prompt submitted → Assistant begins reasoning & tool invocation → Python code produced and executed → Plot rendered inline → Tool result preview & continued interaction ##### Task: API keys developer settings **Willow: Willow onboarding intro** (ios, 3 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/13227) > A short onboarding carousel for Willow shown to a new or returning user before entering the main experience. The user can swipe through several promotional intro screens and tap the primary Get started button to continue. **Fuse: New wallet signup and secure key setup** (ios, 22 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8729) > Happy-path onboarding for a crypto wallet: user is introduced to product pillars (Spend / Save / Invest), chooses to create an account using Sign in with Apple, and runs through a 3-part key setup (Device Key — biometric, 2FA Key — iCloud, Recovery Key — email OTP). After key creation the wallet is provisioned, a short promo for a paid membership is shown, and the user arrives at the empty wallet home. Steps: Onboarding carousel — Spend → Onboarding carousel — Save → Onboarding carousel — Invest → Welcome / Sign-up choices → Sign in with Apple sheet → Signing-in progress → Secure your wallet — Device Key intro → Face ID permission prompt → Creating Device Key (progress) → Create 2FA Key (iCloud) prompt → Recovery Key — email entry → OTP entry — keyboard & autofill → OTP verification and provisioning → Locked state — Unlock CTA → Promo modal — free transactions → Fuse Plus membership detail → Wallet provisioning complete — home (empty) → Empty wallet CTA — Receive → Default bottom navigation & quick actions → End state — ready wallet **Poppy: Recommended Connection Setup** (ios, 4 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/12177) > Happy-path onboarding/settings sequence for a logged-in user named Connie. The user opens the profile/settings area, chooses a recommended connection type (Fitness), is shown a system Health permission request, and returns to the recommendation screen after granting access. Entry is from the profile/settings menu; exit is back to the recommendation step with the selected connection highlighted and a Next button available. Steps: Profile and Settings → Connection Recommendations → Health Permission Request → Connection Selected **komoot: New user onboarding and consent setup** (ios, 39 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8281) > Happy-path onboarding for a first-time mobile (iOS) user. Entry: app launch -> privacy/tracking consent banner. User can Accept All, Refuse, or Customize preferences. If user customizes they can toggle categories and open vendor details. After consenting, the flow presents sign-in options (social or email). Choosing email leads to email validation then account creation (password). Post-signup the user progresses through product onboarding cards that request push notifications and access to device integrations (Apple Health). The flow ends with device connections (optional) and a completion screen, then the main map/routes home screen. Steps: Privacy notice (top-level) → Customize preferences (categories visible) → Category list (scroll) → Analytics expanded (vendor row visible) → Vendor info modal (Firebase) → Sign-in options → Enter email (initial) → Email validation error → Email accepted (proceed) → Create account (password) → Onboarding hero / intro → Onboarding card: Navigation → Onboarding card: Notifications → iOS push permission dialog → Device integrations list → Health Access (iOS modal) → Integration connected state → Onboarding completion → Main map / routes home → Optional: Connect additional devices (post-onboarding) ##### Task: CRM contacts pipeline **Genie: Contact Search Add** (ios, 2 screens, Social Networking, AI Tool) [Refero](https://refero.design/flows/12420) > A user searches for people by name and adds them from a contact search list. Entry is a search screen with the keyboard open and a query already entered (“Dan”). The user can scroll through matching results and tap Add next to a person. After tapping, the action changes to Sent for that person, indicating the request/invite was sent. Steps: Search Results → Updated Results **Lovable: Missing Connector Request** (web, 4 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12342) > Happy-path for requesting a missing connector from the Connectors directory. The user starts in the Connectors section, filtered to the Messaging category, and uses the ‘Missing a connector?’ form in the lower-left panel to request an integration. The visible state assumes the user is already signed in and has access to the connectors catalog. After entering a connector name and submitting, the form confirms the request was received and disables further submission. Steps: Browse Messaging Connectors → Enter Connector Request → Submit Request → Request Confirmed **WhatsApp: Invite Contacts to WhatsApp** (ios, 4 screens, Social Networking, Collaboration Tools) [Refero](https://refero.design/flows/12603) > Happy path for inviting a contact to WhatsApp from a contacts list. The user enters the invite screen from within WhatsApp, browses or searches contacts, taps Invite for a contact who is not on WhatsApp, confirms sending an SMS invite in a bottom sheet, and receives an invite sent confirmation. Steps: Contact List → Search Contacts → Invite Confirmation → Invite Sent **Acctual: Select or Add Invoice Recipient** (web, 2 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8833) > Happy path: From the Invoices area the user opens a Create invoice modal to choose who the invoice is for. The modal offers a searchable field for contacts (and companies) and a live invoice preview at the right. The user searches for an existing contact or types a new identifier (phone/email/company) and is offered a one-click suggestion to add that entry as a customer. After selecting or creating the recipient the user continues with Next to fill invoice line items. Steps: Open Create invoice / Choose recipient → Type to add new contact suggestion **Sora: Cameo access recipient selector** (ios, 4 screens, AI Tool, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/7824) > Mobile modal for granting 'Cameo access' to other users. Entry: user opens the Cameo access modal (overlay) while signed in on a mobile device. The modal contains a searchable list of accounts (avatar, username, follower count) and a text field labeled 'Find people...'. The user can type to search (handles, names, numbers). Exit: user confirms selections by tapping Done, which closes the modal. Assumptions: user is authenticated and allowed to manage cameo access; device is a smartphone (iOS-style modal and keyboard behavior). Steps: Open Cameo access → Search — no results → Search — name results → Select recipients and confirm ##### Task: access review permissions audit log **TIDAL: Track Report Reason Selection** (ios, 3 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/13140) > A mobile modal for reporting a track, starting with a searchable list of report reasons and ending with a selected reason shown after search. The user opens the report dialog, searches or browses categories, and selects the matching reason. Steps: Browse Report Reasons → No Match Search State → Filtered Result **Apple Image Playground: Report Generated Image** (ios, 9 screens, AI Tool, Design & Media Editing) [Refero](https://refero.design/flows/12798) > A user reviews a generated image in Image Playground, opens the overflow menu, and chooses Report a Concern. They select concern categories, optionally add more information, review what will be included in the report, accept the data-sharing prompt, and submit. The report then submits successfully and returns them to the image editor, where the generated image remains visible with the usual controls. Steps: Image Review → Action Menu → Concern Category → Personal Characteristics Check → More Information → Submit Report → Data Consent → Submitting → Return to Editor **Granola: Share Note with Editor Access** (ios, 8 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14162) > The user starts on a note titled “Dashboard design and analytics” and opens its sharing controls. From the sharing panel, they choose to add people, search for a contact by name, select the matching attendee, and change the access role from Viewer to Editor. They return to the sharing summary, where the selected person is listed as having access. The flow is shown on a mobile device. The user is assumed to be signed in and authorized to share the note; no confirmation or notification result is visible after returning to the summary. Steps: Open Note Sharing → Review Sharing Options → Open Recipient Selector → Search for Contact → Select Contact → Open Permission Options → Assign Editor Access → Review Granted Access **Denim: Gallery permissions onboarding** (ios, 5 screens, Design & Media Editing) [Refero](https://refero.design/flows/5929) > Happy-path discovery and permission gating inside the Gallery tab on an iPhone. Entry: user opens the app and lands on the Gallery tab (bottom navigation). The screen shows seasonal hero cards, a “Get the Best Experience” section with action cards (Connect Account — Apple Music, Enable Alerts), and a Free Artworks carousel below. Two system permission dialogs can be triggered from the Gallery: (1) Apple Music / media library access and (2) push notifications. Exit: user remains in the Gallery; if permissions are granted the app can personalize covers from playlists and schedule alerts. Steps: Gallery Home → Apple Music / Media Permission Prompt → Gallery (after music permission) → Push Notifications Permission Prompt → Gallery with Free Artworks **TIDAL: Playlist Privacy Change** (ios, 3 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/13116) > A playlist details screen shows an empty playlist owned by the current user. The user opens the privacy control, confirms making the playlist public in a modal dialog, and returns to the playlist screen where the privacy state changes from Private to Public. Steps: Playlist Details → Public Confirmation → Public Playlist State **TikTok: Activity Status Visibility** (ios, 2 screens, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/2798) > Happy path: A logged‑in mobile user opens the Inbox to review notifications and then opens the activity status information modal to review or change who can see their active/last‑active presence. Steps: Inbox / Notifications → Activity Status Modal **Lovable: People Management** (web, 5 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12297) > This is the workspace People management area for the Refero Design workspace. The user enters from the left navigation on the People section and can view members, filter/search the list, and perform member actions. The happy path shown is selecting one or more people in the table, opening the bulk actions bar, changing a selected member’s role from a dropdown, and returning to the list. Assumptions: the user is logged in, has permission to manage workspace members, and is viewing this on desktop. Steps: Member List → Select Members → Bulk Actions → Role Selection → Error Feedback **Lovable: People Management** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12294) > This is the workspace People management area for a workspace called Refero Design. The user enters from the left navigation under Members & access > People. The screen shows a list of workspace people with tabs for All, Invitations, Collaborators, and Requests, plus search and role filtering. The user can also export the list, invite by link, or invite members. In the shown interaction, the user opens the role filter, selects Owner, and the table updates to show only matching people. Assumption: the user is logged in and has access to manage people in the workspace. Steps: People List → Role Filter Menu → Filtered Results **Product Hunt: Product Team Access Request** (web, 3 screens, Business, Social Networking) [Refero](https://refero.design/flows/11461) > A logged-in user views Notion’s Product Hunt profile on the Team tab and requests access to be added to the product team. Because the product page already has an owner, the system opens an access-request modal. After the user submits the request, the modal confirms that the request was sent and the team section changes to indicate a pending request. The flow begins on the product team profile and ends with the request awaiting owner approval. Steps: Notion Team Profile → Access Request Prompt → Request Confirmation ##### Task: account closure **Spotify: Account Closure** (ios, 8 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/2200) > Happy path: A logged-in user navigates from their Account screen to permanently close their account. Steps: Account overview → Retention / First prompt → Confirm account identity → Account details + Continue → Consequences / Acknowledgement → Consent confirmed → Email sent confirmation → Waiting / Final state **ROVR: Account Deactivation** (ios, 3 screens, Music Streaming) [Refero](https://refero.design/flows/13977) > An authenticated user opens their Profile settings, chooses Delete account, reviews a permanent-deletion warning, and selects Continue. The system deactivates the account and displays a confirmation message. The flow is shown on a mobile device and assumes the user is logged in with permission to manage their own account. The flow begins on the Profile screen and ends on the account-deactivated confirmation state. Steps: Profile Settings → Deletion Warning → Deactivation Confirmation **Voicenotes: Account Deletion Request** (ios, 6 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14276) > The flow starts with a signed-in user on the Settings screen. The user opens Account, selects Delete account, and confirms the intent in a native confirmation dialog. The system then opens an external Typeform deletion request in an in-app browser. The user reviews the permanent-deletion warning, completes the required account email, full name, and deletion reason fields, and submits the form. The flow ends on the external form’s completion screen. The screenshots do not show the account being immediately deleted or the user being signed out; they only confirm submission of the deletion request. **HYPE: Account Deletion** (ios, 8 screens, Lifestyle, News) [Refero](https://refero.design/flows/12254) > An authenticated user starts on the Account section, opens Profile, and scrolls to the account-management actions. They select Delete Account, review the permanent-deletion warning, choose a reason, and enter additional feedback when selecting Other. After continuing, a final confirmation modal asks whether to permanently delete the account. Confirming returns the user to the Hype welcome screen, where login, sign-up, or guest access options are available. The screenshots imply that the account deletion succeeds, although no explicit success message is shown. Steps: Account Overview → Profile Details → Profile Account Actions → Deletion Warning → Deletion Reason → Custom Feedback → Final Deletion Confirmation → Signed-Out Welcome **Bond: Permanent Account Deletion** (ios, 4 screens, AI Tool, Social Networking) [Refero](https://refero.design/flows/14404) > The flow begins with a signed-in user viewing their Settings screen. The user selects Delete Account, reviews an initial confirmation dialog, then proceeds to a second warning that explains the action is irreversible and data will be permanently deleted. After confirming permanent deletion, the interface transitions to the Bond welcome screen, where the user can continue into the unauthenticated entry experience. The screenshots do not show an explicit success message, so the account deletion result is inferred from the transition to the welcome screen. Steps: Settings → Deletion Confirmation → Permanent Deletion Warning → Welcome Screen **On: Account deletion** (ios, 4 screens, Marketplace) [Refero](https://refero.design/flows/9612) > Happy path: a logged-in user navigates from Help & support to the Delete account flow, reads the consequences, confirms intent via a checkbox, taps the Delete account CTA, and is returned to the logged-out welcome screen. Steps: Help & support → Delete account (explanation) → Delete account (confirm) → Logged-out welcome **PayPal: Account deletion and data removal** (ios, 13 screens, Finance) [Refero](https://refero.design/flows/7271) > A logged-in user navigates from Privacy/Manage your data to request deletion of their PayPal data and close their account. The flow: user selects "Delete your data / close account", reads eligibility and pre-steps, gives explicit consent, reviews a notice about linked services, chooses an email for updates, provides exit reasons (optional), completes a quick security check (SMS code), and receives a final confirmation that the request was received. Entry: Manage your data settings. Exit: confirmation screen stating request received and that an email will follow. Assumptions: user is authenticated, has a verified mobile number for SMS, has an email on file, and can complete any outstanding account requirements (balances, limitations) outside this flow. Steps: Manage your data list → Deletion info & prerequisites → Consent acknowledgement (unchecked) → Consent acknowledgement (checked) → Linked services notice → Choose email for updates → Exit feedback (reason selection) → Quick security check → Enter security code (empty) → Submit security code → Processing / success animation → Request received confirmation → Final status / completion **Phantom: Account management — remove wallet** (ios, 4 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3696) > Happy path: A logged-in user with multiple wallet accounts opens the Your Accounts list, selects a specific account to manage, navigates to the Edit Account screen, initiates account removal, confirms the removal, and is returned to the app home screen with the remaining account set as active. Entry: Your Accounts list. Exit: Home screen with remaining account active. Assumptions: user is authenticated, there are at least two accounts present, and the UI shows account-level actions (edit, remove). Steps: Accounts list → Edit Account → Confirm removal → Home (post-removal) — active account updated **Phantom: Manage Wallet Accounts** (ios, 4 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3699) > Happy path: A signed‑in user opens the Your Accounts list on a mobile device, chooses an existing account to edit, navigates into the Edit Account screen, selects Remove Account, confirms the destructive action in a confirmation screen, and is returned to the accounts list with the removed account deleted and another account selected. Steps: Your Accounts (list) → Edit Account → Remove Account (confirmation) → Your Accounts (updated list) **District: Account Deletion Review** (ios, 4 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12883) > A logged-in mobile user opens Profile, then navigates to Account Settings and initiates account deletion. The entry point is the Profile screen; the exit point is either canceling the deletion and returning to Account Settings, or confirming deletion from the warning modal. The flow assumes a signed-in user on a mobile device with access to the profile menu. Steps: Profile Menu → Account Settings → Delete Account Confirmation **222: Account management - delete** (ios, 3 screens, Lifestyle, Dating & Relationships) [Refero](https://refero.design/flows/9330) > Happy path: A signed-in mobile user opens the Account screen to manage settings. From the Manage section they choose "delete account", see a destructive confirmation modal, and (if they confirm) the app transitions to a full-screen illustrative/empty state. Steps: Account overview → Delete confirmation modal → Post-deletion / empty state **DailyArt: Account Deletion** (ios, 6 screens, Education, Magazines & Newspapers) [Refero](https://refero.design/flows/3344) > Happy path: A signed-in user opens Settings, navigates to Account Details, chooses Delete Account, selects a reason, confirms intent on a warning screen (with subscription notice) and completes deletion which signs them out. Steps: Settings - Signed In → Account Details → Delete Account — Reason (initial) → Delete Account — Reason Selected → Final Confirmation & Subscription Notice → Settings - Signed Out **BeReal.: Account Deletion** (ios, 6 screens, Social Networking) [Refero](https://refero.design/flows/4945) > Happy-path: A logged-in user opens Settings, navigates to the "Other" section, taps Delete Account, reviews the deletion confirmation (which explains immediate logout and a 15-day scheduled deletion), optionally selects or types a reason for leaving, confirms deletion, and is then prompted to re-verify their phone number. Steps: Settings (entry) → Other (delete entry) → Deletion confirmation message → Choose 'Other' reason and input → Enter custom reason and confirm → Phone number verification (re-auth) **ElevenReader: Account deletion and confirmation** (ios, 5 screens, AI Tool, Books) [Refero](https://refero.design/flows/8988) > Happy path: A logged-in user navigates from their account overview into the settings list, selects the 'Delete account' option, confirms the deletion by typing their email address exactly, and taps the destructive CTA. The UI requires an exact email match before enabling the red Delete action. After tapping Delete, the user receives an inline success confirmation banner and is effectively signed out. Steps: Account overview → Settings list (account actions) → Delete confirmation (initial) → Delete confirmation (email entry) → Deletion success feedback **X: Deactivate account flow** (ios, 6 screens, Social Networking, Entertainment, News, Video Streaming) [Refero](https://refero.design/flows/2930) > Happy path: A signed-in user navigates from Settings → Your account to permanently start account deactivation. The user reads the deactivation details, taps Deactivate, confirms their account password, accepts a confirmation modal, and sees a final success screen. Steps: Settings — Account entry → Deactivate your account — information → Confirm your password — input (disabled submit) → Confirm your password — submit → Confirmation modal → Deactivated — success acknowledgement **Podimo: Account Deletion** (ios, 4 screens, Entertainment) [Refero](https://refero.design/flows/12286) > Happy path starts in the Settings screen of the Podimo mobile experience, where a signed-in user can access personal account options. The user selects Delete account, sees an informational warning, then a stronger confirmation dialog. If the user confirms deletion, the account is removed and the session ends, returning the user to the welcome/sign-in screen. Assumption: the user is logged in on a mobile device and has permission to delete their own account. Steps: Settings → Deletion Warning → Final Confirmation → Welcome Screen **Drinkit: Account deletion and confirmation** (ios, 3 screens, Food & Drink) [Refero](https://refero.design/flows/11016) > Happy path: A signed-in user opens the Settings screen to manage profile, notifications, and account-level actions. From the bottom of the Settings list the user selects "Delete account." The app presents a prominent confirmation modal that explains what will be removed and offers a destructive "Remove" CTA. After confirmation the app displays an in‑context banner/toast saying the account will be deleted in a few days and updates the Settings screen to show an "Account is being deleted" status. Entry: Settings. Exit: account deletion scheduled (user sees confirmation/feedback). Steps: Settings overview → Delete confirmation modal → Deletion scheduled feedback **Plane Finder: Account Deletion** (ios, 4 screens, Travel & Booking) [Refero](https://refero.design/flows/10417) > Happy-path: A signed-in user opens the app Settings, navigates to Contact Support, and submits a request to delete their account. The UI presents a confirmation modal describing consequences (subscriptions unaffected, irreversible) and offers Cancel or Continue. If the user continues, a success acknowledgement modal is shown. Entry: Settings screen. Exit: Acknowledgement modal is displayed and dismissed, returning the user to Settings. Assumptions: mobile (iOS-style) settings sheet, user has access to Contact Support (some screens show logged-in state and others show a Sign In button), network/email links available for contacting support. Steps: Settings - Landing → Contact Support - Options → Deletion Confirmation Modal → Acknowledgement Modal **Pinterest: Account Deletion** (ios, 4 screens, Reference, Inspiration Board, Social Networking, Lifestyle) [Refero](https://refero.design/flows/2470) > Happy path: A signed-in mobile user navigates from Account management to permanently delete their account. From Account management they choose "Delete your data and account", read a final confirmation screen, tap Continue, and a bottom-sheet modal asks for a reason. After selecting a reason the primary action (Send email) becomes enabled; tapping it sends an email to the account address with the final step to complete deletion. Steps: Account management → Delete confirmation screen → Reason selection (initial) → Reason selected and confirm **Tinder: Account Pause and Deletion Options** (ios, 3 screens, Dating & Relationships) [Refero](https://refero.design/flows/3476) > Happy path: A logged-in user navigates to Settings -> Delete Account to either pause or permanently delete their account. The Settings list contains a 'Delete Account' entry. Tapping it opens a screen that offers 'Pause Account' (prominent gradient CTA) and a secondary 'Delete Account' option. When the user taps 'Pause Account', the app hides the user's profile, returns them to the main discovery view, and shows a clear paused-state screen with a 'Jump back in' CTA to unpause. Steps: Settings list → Pause vs Delete decision → Paused state confirmation **Wealthsimple: RESP Account Closure** (web, 5 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8518) > Happy path for closing an individual account (Joint RESP) from the Accounts section. Entry: user starts at Accounts list (logged in, desktop). The user selects an account, opens its details, taps the Close account control, confirms in a modal, and receives a success confirmation. Exit: user is returned to the Accounts list where the closed account appears in a "Closed accounts" section with a closed date. Assumptions: user is authenticated, using a desktop layout, and the UI enforces/communicates prerequisites for closure (e.g., $0 balance and no pending transfers). Steps: Accounts list → Account details → Closure confirmation modal → Success confirmation modal → Accounts list (updated) **Spotify: Account Closure and Data Deletion Process** (web, 5 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/58) > This flow guides a logged-in Spotify user through the steps required to close their account and delete all associated data. Starting from the user account overview page, the user navigates to the account closure section, confirms account details, acknowledges the consequences of deletion, and finally receives instructions to verify the closure via email. The flow assumes the user has permissions to manage their account and is using a desktop device. Steps: User Account Overview → Initiate Account Closure → Confirm Account Details for Closure → Acknowledge Closure Consequences → Email Verification Instruction **Care: Close Membership Account** (web, 3 screens, Job Search, Lifestyle) [Refero](https://refero.design/flows/12240) > Happy path for closing a Care.com membership. The user starts on a membership management screen while signed in, chooses to close their account, reviews cancellation options and reason selection, then completes the closure and reaches a confirmation screen. Steps: Membership Overview → Cancellation Review → Closure Confirmation **Tripadvisor: Account closure and confirmation** (web, 4 screens, Travel & Booking) [Refero](https://refero.design/flows/7090) > Happy-path: A logged-in user navigates from their account settings to the account closure process, selects a reason for leaving (or an alternative action such as changing email preferences or disconnecting a social login), submits a closure request, and sees a confirmation that their account has been closed and they are being logged out. Steps: Account Info (Settings) → Account Closure — choose reason → Account Closure — provide details → Closure confirmation and logout **Airbnb: Account deactivation** (web, 6 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6013) > Happy-path summary: A logged-in user navigates from Account settings > Login & security to permanently deactivate their account. The flow is presented as a three-step wizard (1. Select reason, 2. Confirm, 3. Done). The user picks a reason, confirms consequences, submits the request, and is shown a success screen. After closing the success state the interface returns to the public discovery/search area (user is effectively logged out or unable to access account-specific features). Steps: Account settings – Login & security → Select reason – initial → Select reason – choice made → Confirm deactivation → Account deactivated – success → Public discovery – post-deactivation **Record Club: Account closure & confirmation** (web, 2 screens, Music Streaming, Social Networking) [Refero](https://refero.design/flows/5783) > Happy path: A logged-in user opens Settings → Account to manage credentials or permanently close their account. The Account tab exposes three primary actions: Reset password, Change email address, and Delete account (red CTA). When the user triggers account deletion, they are redirected to the public marketing/landing site and shown a temporary success toast reading "Successfully deleted account." Entry point: Settings (user must be signed in). Exit point: Public landing/marketing (user returned to site with login/join CTAs visible). Assumptions: desktop viewport; user is authenticated when on Settings. The UI does not show an intermediate confirmation modal in these screenshots; only the pre-deletion controls and the post-deletion landing with success feedback are visible. Steps: Settings — Account tab → Public landing — post-deletion confirmation **Kit: Account deletion via in-product support** (web, 4 screens, Business, Communication Tools) [Refero](https://refero.design/flows/2302) > Happy path: A signed-in user on desktop is in their Settings > Login Details screen, opens the on-screen support chat, and requests permanent account deletion. The support agent replies asking for confirmation and additional details. After no user response, the agent cancels the account and closes the ticket. When the user later attempts to access the product they land on an account-closed screen that offers a reinstatement CTA and contact link for Customer Success. Steps: Open support chat and request deletion → Support asks for confirmation and details → Agent follows up and closes ticket → Closed-account landing with restore option **Shop: Shop Account Deletion** (web, 4 screens, Marketplace) [Refero](https://refero.design/flows/13287) > The flow begins on a Shop account deletion screen for an authenticated user. The user reviews the warning that deletion is permanent and selects the Delete Shop account button. The system then requests a six-digit verification code sent by email. The user enters the code, after which the deletion attempt is processed. In the shown outcome, the attempt fails and the user is instructed to try again or contact support. The screenshots do not show a successful deletion state. Steps: Deletion Warning → Enter Verification Code → Complete Code Entry → Deletion Failed **MyMind: Account Deletion** (web, 4 screens, Productivity, Reference, Inspiration Board) [Refero](https://refero.design/flows/13327) > A logged-in user opens the account settings and chooses to close and delete their account. The system presents a prominent warning modal requiring the user to type DELETE before the destructive action becomes valid. After confirmation, the account is scheduled for deletion and the user is shown a reactivation option before the displayed deletion date. The flow begins on the account settings screen and ends on the deletion-scheduled confirmation screen. Steps: Account Settings → Deletion Warning → Confirm Deletion → Deletion Scheduled **Sunsama: Account Deletion** (web, 6 screens, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/11603) > A signed-in user navigates to Account Management and chooses to delete their account. The system presents a feedback form and explains that deletion is permanent. After the user enters a reason, the destructive action becomes available. The user submits the deletion request, sees a processing state, and is then taken to a 404 screen before arriving at the login screen. The flow assumes the user has an active Sunsama account and sufficient permissions to manage or delete it. Steps: Account Management → Deletion Feedback → Deletion Confirmation → Deletion Processing → Unavailable Destination → Login Screen **Doctronic AI: Account Deletion Request** (web, 4 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13707) > The flow begins in the signed-in user's account settings, where the user selects the option to delete their account. The system opens an account deletion request form with the user's email address and phone number prefilled. The user enters a reason for deletion and can then submit the request. The final screenshot shows the completed form before submission; no confirmation or processing state is visible. **Grammarly: Account Deletion Confirmation** (web, 2 screens, Paperwork) [Refero](https://refero.design/flows/1883) > Happy-path: A logged-in desktop user navigates to Account Settings (left navigation). From the Profile/Account Settings area they select the Delete Account action. A centered confirmation modal overlays the screen, shows the account email for context, and offers two choices: cancel and return to saved documents, or continue with account deletion. Steps: Account Settings (Profile) → Delete Account Confirmation Modal **Calendly: Account Deletion** (web, 5 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/14057) > A logged-in user starts in Profile settings and chooses to delete their account. The system opens a confirmation dialog requiring a verification code sent to the account email, a deletion reason, and optional feedback. After the user enters the code, selects a reason, and confirms deletion, the account is removed and a completion screen is shown with a link back to the Calendly homepage. The flow is shown on a desktop-sized interface, and no permissions beyond access to the user’s own profile settings are indicated. **Shop: Shop Account Deletion** (web, 4 screens, Marketplace) [Refero](https://refero.design/flows/13288) > A signed-in Shop user starts the account deletion process from the account settings area. The interface first explains that deletion is permanent and identifies the types of data that may be removed. After the user selects “Delete Shop account,” the system requests a verification code sent by email. The user enters the six-digit code, and the system confirms that the deletion request has been received. The account data is scheduled to be deleted within 30 days. Steps: Deletion Warning → Enter Verification Code → Complete Code Entry → Deletion Request Received **Dribbble: Account deletion and redirect** (web, 3 screens, Communication Tools, Job Search) [Refero](https://refero.design/flows/783) > Happy path: A signed-in user goes to Settings → General, clicks the 'Delete Account' entry in the left navigation, is shown a centered confirmation card with a strong warning and two actions (Nevermind / Delete my account), and after confirming deletion they are redirected to the public homepage where a banner or top message indicates the account is being deleted and the user is logged out. Steps: Settings — General → Delete confirmation card → Homepage — post-deletion **Digg: Account Deletion** (web, 4 screens, News) [Refero](https://refero.design/flows/14353) > The user starts logged in on a desktop Digg experience with the Account settings modal open on the Security section. They select the account deletion option, review the warning, type the required confirmation phrase, and submit the deletion request. The flow then returns to the public home feed, indicating the account action completed or the user was redirected after submission. The screenshots do not show an explicit success message, error state, or sign-out confirmation. Steps: Security Settings → Deletion Warning → Confirm Deletion → Home Feed Redirect **Cosmos: Account Deletion** (web, 4 screens, Reference, Social Networking, Inspiration Board) [Refero](https://refero.design/flows/14207) > The user starts in a signed-in desktop session and opens account settings over a content feed. From Account details, they choose to close the account, review the permanent-deletion warning, and continue. The system indicates that an email has been sent, then the user reaches the signed-out welcome screen with an “Account deleted” confirmation. The flow assumes the user has permission to manage their own account and that any required email confirmation occurs outside the shown screens. Steps: Account Details → Deletion Confirmation → Email Sent Feedback → Deleted Account Sign-In **Product Hunt: Permanent Account Deletion** (web, 5 screens, Business, Social Networking) [Refero](https://refero.design/flows/11470) > A signed-in user opens their account Settings and chooses to permanently delete their account and associated data. The system presents a confirmation modal requiring the user to enter the displayed username, then processes the deletion. After completion, the user is returned to the public Product Hunt home experience in a signed-out state. The screenshots do not show an error state or an explicit completion message. Steps: Account Settings → Deletion Confirmation → Username Verification → Deletion Processing → Public Home **ZARA: Account deletion confirmation** (web, 4 screens, Marketplace) [Refero](https://refero.design/flows/1622) > Logged-in user (desktop) navigates from their Profile area to permanently delete their account. Steps: Profile — Account options → Delete your account — Warning → Confirm account deletion — Credential prompt → Confirm account deletion — Credentials filled **Fresha: Account deletion confirmation** (web, 8 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/2384) > Happy-path: A logged-in user navigates from Settings to permanently delete their Fresha account. The flow collects an optional reason, requires explicit acknowledgement of consequences (two checkboxes), asks the user to re-enter credentials (email + password) to confirm identity and intent, then shows a completion screen confirming deletion. Steps: Settings (entry) → Why are you leaving? → Reason selected → Acknowledge consequences (unchecked) → Acknowledge consequences (checked) → Re-authenticate to confirm → Credentials entered → Deletion success confirmation ##### Task: apply for loan credit **District: Add Card** (ios, 4 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12836) > Happy path for adding a payment card. The user enters from an Add Card screen, optionally chooses to scan the card to autofill details, grants camera access when prompted, confirms or reviews the scanned card number and expiry date, selects a card nickname category, and submits the card. Steps: Add Card Entry → Camera Permission → Card Scan → Review and Save **Mercury: Create company card** (web, 4 screens, Business, Finance) [Refero](https://refero.design/flows/11068) > Happy path: A logged-in admin or employee opens the Cards area, starts the Create card drawer, completes the form (cardholder, nickname, credit/debit, virtual/physical, source account, controls and limits), and submits. The system shows a live preview while editing, then displays a success/activation screen with options to view details or create another. Steps: Cards overview / select Create → Create a card — basics → Create a card — account & review → Confirmation — card created **PayPal: Request Payment** (ios, 5 screens, Finance) [Refero](https://refero.design/flows/7264) > Happy path: A logged-in user opens a contact's History, chooses "Request Money", enters an amount, reviews the request, and confirms. Steps: Contact History / Actions → Amount Entry (empty) → Amount Entry (filled) → Review Request → Request Sent / Success **District: Wallet Balance and History** (ios, 3 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12877) > A wallet/balance screen showing the current District Money balance, an Add money action, and a transaction history list. The user can filter history by All, Debit, Credit, or Refunds. Entry is from the back navigation into the balance/history view; exit is via the back button or by selecting Add money. Steps: Balance Overview → Debit Filter Applied → Credit Filter Applied **Introspect: Friend decision response** (ios, 3 screens, Finance) [Refero](https://refero.design/flows/11952) > Happy path for a recipient responding to a friend’s product decision request on mobile. Steps: Decision Request → Sending State → Completed State **GoFundMe: Fundraiser Support Chat** (web, 6 screens, Fundraising) [Refero](https://refero.design/flows/11747) > This is the fundraiser management area for a logged-in organizer. The user starts on the main Fundraiser dashboard for a specific fundraiser, then opens a right-side support chat with the GoFundMe helper bot. In the chat, the user asks how to get money from the fundraiser, and the bot explains that transfers must be set up on desktop and gives step-by-step guidance. The flow ends with an end-chat confirmation dialog. Steps: Fundraiser Dashboard → Open Help Chat → Ask a Question → View Transfer Instructions → Continue Support Conversation → End Chat Confirmation **Airbnb: Artist experience request** (ios, 11 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6413) > Happy path for requesting entry to a limited/lottery artist event. Steps: Event detail / Request CTA → Guests selection → Application question prompt → Text entry (keyboard active) → Answer review → Request details summary → Before you submit (rules & process) → Submission (start animation) → Submission animation sequence → Confirmation screen → Exit / Return **Mercury: Request money - customer selection** (web, 4 screens, Business, Finance) [Refero](https://refero.design/flows/11048) > Happy path: A logged-in desktop user begins the "request money" wizard, chooses or creates a customer, then selects the invoice/request type. Steps: Customer selector → Add customer (blank) → Add customer (filled) → Request method selection **OpenAI: Billing overview** (web, 2 screens, Business, Development Tools, AI Tool) [Refero](https://refero.design/flows/4183) > Overview of the Billing section (desktop). Entry: user navigates to Organization > Billing > Overview while signed in. The screen shows current trial credit balance, primary CTAs (Add payment details, View usage) and a set of tiles linking to Payment methods, Billing history, Preferences, Usage limits, and Pricing. Happy path: user inspects the credit balance, optionally clicks Add payment details to enter a payment method or clicks View usage to see consumption. Exit: the user either proceeds to add a payment method, opens the usage breakdown, or navigates to one of the billing sub-sections. Assumptions: user is authenticated with billing access, using a wide/desktop viewport. Steps: Billing overview (default) → Credit info tooltip **DailyArt: Checkout Payment Entry** (ios, 2 screens, Education, Magazines & Newspapers) [Refero](https://refero.design/flows/3332) > This is the final payment step of a mobile checkout flow. The screen lets the user choose a payment method (credit/debit card or PayPal), enter card details (card number, expiration date, security code), read a short privacy notice, check a required terms-and-conditions box, and submit the order via the PLACE ORDER button. Steps: Payment method & card input (initial) → Inline validation error (invalid card) **Lovable: Plans and credits management** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12301) > Happy path for managing a workspace subscription from the Plans & credits area. Entry is the workspace settings sidebar on the Plans & credits screen. The user can review current plan status, compare Pro/Business/Enterprise options, and initiate an upgrade or credit purchase. If upgrading, the flow continues to a Stripe-style checkout where the user enters payment details and confirms the subscription. Exit is the updated Plans & credits screen showing the new active plan and refreshed credit balance. Steps: Current Plan Overview → Checkout and Payment → Updated Subscription Status **OpenAI: Add payment method and configure credits** (web, 5 screens, Business, Development Tools, AI Tool) [Refero](https://refero.design/flows/4194) > Happy path: A logged-in user in an organization navigates to the Billing section, opens the Add payment details modal, fills card and address information (including selecting a country), submits the details, then configures an initial credit purchase and optional automatic recharge. Steps: Billing overview → Add payment details (empty) → Country selection → Add payment details (filled) → Configure payment / initial credit **Omio: Discount card selection** (ios, 2 screens, Travel & Booking) [Refero](https://refero.design/flows/5166) > This is a short selection and search flow for adding discount cards to the user's profile/booking. Steps: Discount card list → Search - numeric entry **Revolut: Add Money - Card Top-up** (web, 3 screens, Business, Finance) [Refero](https://refero.design/flows/1343) > User is on an Add Money flow (amount already chosen in the background) and opens a 'How to add money' modal that lists two methods: Debit or credit card, and Regular bank transfer. Choosing the card option navigates to a Card details screen where the user must enter card number, expiry date and CVV to top up the visible amount (Add £100.00). The happy path: user provides a supported debit card and the CTA becomes active to complete the top-up. The screenshots show the initial method selection, an empty card-entry form with the CTA disabled, and a populated card form that triggers a validation error indicating credit cards are not supported. Steps: Choose add-money method → Card details (empty) → Card validation error **On: Apply gift card at checkout** (ios, 5 screens, Marketplace) [Refero](https://refero.design/flows/9603) > A single-screen modal-style mobile checkout where the user selects a payment method, optionally applies a gift card, and completes payment. Steps: Payment selection and gift card collapsed → Enable gift card entry → Enter gift card details → Apply submission (loading) → Validation error shown **District: Offers Search and Selection** (ios, 3 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12822) > A logged-in mobile user opens an Offers overlay from a prior screen, searches within unlocked offers, and can either browse all available offers, inspect details, or apply an offer. The flow starts with the offers list and search bar, includes an empty-state result when the search term has no matches, and returns to filtered offer results when a matching keyword is entered. The flow ends when the user either closes the overlay or selects/apply an offer from the list. Steps: Offers List → No Results State → Filtered Offer Results **District: Movie Offer Application** (ios, 4 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12821) > Happy path on a movie detail screen where the user views information about a film, opens the offers panel, searches for a voucher, applies an eligible offer, and returns to the movie detail screen with the offer marked as applied. Steps: Movie Details → Offers List → Offer Search → Offer Applied **Drinkit: Apply Welcome Drink Offer** (ios, 3 screens, Food & Drink) [Refero](https://refero.design/flows/11009) > Happy path: a logged-in user (recognized by name) scans a personalized feed of promotions, opens a promotion card to read details, and applies the offer. Steps: Promotions feed (list view) → Promotion detail → Offer applied confirmation & updated feed **Turf: Send Friend Request** (ios, 2 screens, Sports) [Refero](https://refero.design/flows/9425) > Happy-path for sending a friend request from another user's profile. Entry: user opens someone else's profile (mobile, logged-in). The profile shows header (avatar, name, username), two primary actions (Add friend / Taunt), a Stats card (games, wins, high score) and an Activity list of past games and picks. Exit: the Add friend control transitions to a Requested state indicating a pending friend request. Steps: Profile (not friends) → Profile (request sent) **Asos: Customer Care Search & No-Results** (ios, 2 screens, Marketplace) [Refero](https://refero.design/flows/2940) > Mobile customer care entry that surfaces a prominent search box and categorized FAQ topics. Entry: user opens the Customer Care section (mobile browser). The primary path shown is: user enters a query in the help search (example: "visor") and is taken to a search results screen. Exit points: refine search and resubmit, select a FAQ category or question, tap header navigation (menu, account, basket), or close the help overlay. Assumptions: user is on a mobile browser (portrait), may be signed in or out (header shows account icon but no explicit sign-in state), and has basic browsing permissions. Steps: Customer Care landing → Search results (empty state) **Artsy: Select billing country** (ios, 5 screens, Marketplace) [Refero](https://refero.design/flows/7610) > Happy path: A user adding billing details on an Add Credit Card screen opens the Country selector, finds their country using scrolling or the search field (handles mistyped input and empty results), selects the country, and returns to the Add Credit Card screen with the country populated so they can save the card. Steps: Add Credit Card — billing section → Country list — browse alphabetically → Country search — no-match state → Country search — filtered results → Add Credit Card — country populated **ten ten: Add friend by PIN** (ios, 6 screens, Communication Tools, Social Networking) [Refero](https://refero.design/flows/2268) > Happy path for adding a friend to the product using a short, shareable PIN. Steps: Empty friends / Add prompt → Add options bottom sheet → PIN entry (prefilled example) → PIN entry (empty / clipboard) → Request sent confirmation → Main screen showing pending friend **222: Friend compatibility invite** (ios, 7 screens, Lifestyle, Dating & Relationships) [Refero](https://refero.design/flows/9322) > Happy-path for discovering compatibility with a friend. Steps: People list — Compatibility entry → Discover compatibility — Friend picker → Filter friends — Type query → Invite friend — Send invite → Picker shows invited state → People list — Compatibility slot created → Nudge friend — Send reminder **Genie: Friends Management** (ios, 2 screens, AI Tool, Social Networking) [Refero](https://refero.design/flows/12430) > Happy path for managing friends on a mobile social networking screen. Steps: Review Pending Request → Browse Friends List **District: Apply Offer Code** (web, 3 screens, Entertainment, Food & Drink, Travel & Booking) [Refero](https://refero.design/flows/13007) > The user is on an Offers screen during checkout or booking. They can either manually enter a coupon/promo code in the input field at the top or choose from a list of suggested offers below. The entry point is the offers panel opened from a broader booking/payment context, and the exit point is either a successful offer application or an inline validation error if the entered code is invalid. **Faire: Request Unlisted Product** (web, 5 screens, Business, Marketplace) [Refero](https://refero.design/flows/13083) > Happy path on a brand storefront where a buyer requests products that are not currently listed. Entry is the brand shop/product listing page. The user opens the brand actions menu, chooses Request unlisted product, fills in one product name, URL, and quantity in the modal, then submits. The modal closes, a confirmation toast appears, and the storefront remains visible. Steps: Brand Storefront → Open Request Form → Enter Product Details → Submit Request → Request Confirmed **Handshake: Job application via modal** (web, 5 screens, Social Networking, Job Search) [Refero](https://refero.design/flows/8788) > Happy path for applying to a job listing from the job detail view on desktop. Entry: signed-in user lands on a specific job listing (Design Researcher). The user reviews job details, clicks the Apply control, attaches required documents (resume and optional cover letter), answers a work-authorization question, and submits. Exit: the listing updates to show a successful application state (green banner with applied date) and offers a Withdraw application CTA. Assumptions: user is signed in and has at least one document (resume/cover letter) available to attach. Steps: Job listing — primary view → Application modal — attach resume → Application modal — add cover letter and confirm authorization → Submission acknowledgement (toast) → Job listing — applied state **Lovable: Start Project Prompt** (web, 6 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12303) > A logged-in user lands on the Lovable home screen with a left sidebar, a central prompt composer, and a template gallery below. The user can start by typing or using voice input in the composer, optionally connect tools, then submit the prompt to generate a project. The flow ends when the prompt is fully entered and ready to build/submit, with the system showing the composed request and available build controls. Steps: Home Prompt Start → Template Inspiration → Prompt Refine → Voice Capture → Transcribing State → Prompt Ready to Build **Melius: Request a Template** (web, 4 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/14095) > The user is in the Templates area and has searched for “pppp,” which returns no matches. From the empty-results prompt, they open a “Request a template” modal, enter a description of the desired template outcome, and submit the request. The modal closes and a “Request sent” confirmation appears. The flow assumes an authenticated user with access to the Templates area, using a desktop browser. The user remains on the Templates screen after submission. Steps: Search with No Results → Open Request Form → Describe Template Need → Request Confirmation **Copperx: Savings Interest Signup** (web, 2 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8902) > A simple, two-step happy path where a logged-in user navigates (via the left navigation) to the Savings Space and expresses interest in a new savings product. Steps: Savings Space — promotional card → Savings Space — confirmation state **Steel: Wallet Credit Purchase** (web, 15 screens, AI Tool, Development Tools) [Refero](https://refero.design/flows/14283) > Happy-path flow for adding paid credits to a logged-in account wallet. The user starts from a session-related deposit requirement, proceeds to the Wallet settings area, selects a credit amount, adds a payment method, reviews the purchase, and submits payment. The system processes the transaction, confirms that credits were added, and returns the user to the Wallet view with the updated balance. The screenshots show a desktop interface and an authenticated account with permission to manage billing and wallet credits. **Revolut: Find and open API endpoint** (web, 5 screens, Business, Finance) [Refero](https://refero.design/flows/1319) > Happy path: A developer lands on the Revolut Docs home, opens the global search, types a query, reviews suggestion or result cards, and navigates into a specific API endpoint documentation (Retrieve all transactions). Steps: Docs home / hero → Search overlay with suggestions → Search empty result (no matches) → Search results grouped by API → Endpoint detail — Retrieve all transactions **Lovable: Manage plan and redeem credit** (web, 4 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12302) > Logged-in workspace admin is viewing the Plans & credits area and opens a centered modal to manage billing. From the plan summary, the user can choose to downgrade, redeem a gift card, view coupon code options, edit billing details, or go to invoices and payments. The happy path shown continues into gift card redemption, where the user enters a code and submits it. If the code is valid, credit would be added and applied to subscription payments; in the final screen shown, the entered code is rejected with an inline validation error. The flow starts from the Plans & credits settings screen and ends either by closing the modal, navigating to billing/invoices, or seeing a gift-card validation result. Steps: Manage Plan → Gift Card Entry → Code Review → Invalid Code Error **Lovable: Help and Resources Center** (web, 5 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12206) > This is a help and support entry experience for Lovable. The user starts on a public support/resources landing page with navigation, a central help search prompt, resource cards, FAQs, and a build prompt at the bottom. The user can also open a centered search modal, enter a query or numeric text, and review an AI-generated response with suggested follow-up chips. Assumption: the user is on desktop and not required to be logged in to access the visible help content. Steps: Help Landing Page → Search Modal Opened → Query Entered → Search Suggestions → Assistant Response **ElevenLabs: Request License for Iconic Voice** (web, 6 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/8622) > Happy-path: A visitor on desktop browses the ElevenLabs "License Legendary Voices" marketplace, opens a specific voice detail (John Wayne™), opens the in-page modal to request a license, completes the request form (contact + project details + estimated start date), agrees to terms and submits. The flow ends with an in-page confirmation toast. Steps: Marketplace listing / Voice discovery → Open Request License modal → Choose estimated start date → Complete contact and project fields → Acknowledge terms and submit → Show submission confirmation ##### Task: approval workflow **District: Closed Support Rating** (ios, 2 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12841) > A mobile support chat ends after the user’s issue is resolved and the conversation is marked closed. The entry point is an active support conversation with recent agent messages. The exit point is a closed chat state where the user can rate the assistance or start a new chat if they still need help. ##### Task: book GP appointment **District: Salon Service Booking** (ios, 8 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12874) > Happy path for booking a salon service and reaching payment method selection. The user starts on a salon profile, chooses to book a service, selects one service and an available date/time slot, reviews the appointment details, and opens the payment method picker. **District: Event Booking Confirmation** (ios, 5 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12872) > Happy-path ticket booking for an event, starting on a booking review screen and ending on a confirmed order with an accessible QR code ticket. **Doctronic AI: Appointment Payment Confirmation** (web, 2 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13700) > The user is on the final payment step of a doctor appointment booking. The appointment details are shown alongside payment options. The user can use saved payment information, enter card details manually, or choose another payment method. In the shown path, the user selects saved payment information, reviews the populated card and billing fields, and can confirm the booking for $39.00. The flow starts at the payment step and ends when the user submits the booking confirmation. The screenshots do not show the post-submission result. **District: Movie Booking Support Request** (ios, 7 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12840) > A user opens a movie-support conversation that already contains a booking summary. They confirm the booking, are prompted to describe the issue, open the attachment menu, and grant photo-library access. A booking confirmation containing a QR code and ticket details is shown as supporting context. The user types a request to replace cheese popcorn with salt, then sends the message with the booking image attached. The flow starts in an existing support chat and ends with the issue submitted to the support assistant. ##### Task: budget spending insights **District: Restaurant Savings Calculator** (ios, 9 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12859) > User starts on a restaurant listing’s Offers tab and reviews available dining discounts for a specific date and meal period. They can open the sample bill panel, adjust guests, date, and meal type, and optionally enter a custom bill amount to recalculate savings. The experience is modal-driven on mobile, with entry from the offers list and exit via close controls back to the restaurant offers screen. **Wabi: Submit Personal Analysis** (ios, 8 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12052) > This is a mobile analysis experience where the user enters a personal feeling, dream, or life situation, then submits it for interpretation by a panel of analysts. Steps: Prompt Entry → Text Composed → Submit Analysis → Processing State → Results Begin → Analyst Responses → Detailed Reading → Expanded Analysis View **Pool: Place Details to Directions** (ios, 2 screens, Utility) [Refero](https://refero.design/flows/11980) > Happy path starts from a saved or viewed place detail screen for Citizens Bank Park, showing venue information, actions like call/website/tickets, and a directions button. The user can open navigation from the place detail view, which transitions into the maps interface focused on Citizens Bank Park and nearby points of interest. Steps: Place Details → Map Results **Plane Finder: Breakdown category selector** (ios, 3 screens, Travel & Booking) [Refero](https://refero.design/flows/10403) > User views an SFO Performance screen with a top chart and a “Breakdown” card below. The Breakdown card shows a ranked list of items (default: Airlines) with icon/flag on the left, short code below the name, and a numeric count on the right. A compact category selector (label + caret) in the card header opens a small popover to choose other breakdown dimensions (Airlines, Airports, Countries). Selecting a different dimension closes the popover and replaces the list contents and header summary accordingly. Steps: View Breakdown (Airlines) → Open Category Picker → View Breakdown (Countries) **Spyglass: Export and Rate Analysis** (web, 3 screens, Business, Inspiration Board) [Refero](https://refero.design/flows/13461) > A user reviews a generated advertising analysis in the Spyglass workspace and interacts with the response action toolbar. The sequence begins with the completed response visible, then shows a PDF download confirmation, a hover tooltip for the positive-feedback control, and finally the positive-feedback control selected. The user appears to be logged in and viewing an existing conversation or research workspace. The flow ends with the response retained in the conversation and feedback recorded. Steps: Review Generated Analysis → Identify Helpful Response Control → Submit Positive Feedback **Rox: Account news insight analysis** (web, 12 screens, Business, Productivity, AI Tool) [Refero](https://refero.design/flows/10763) > Happy path: A logged-in user on desktop opens the command composer, selects a company account (switching from default Ramp to Apple), picks a saved command (“Analyze insight”), and runs it. The assistant prompts for a news source; the user selects a suggested action to retrieve recent news. The assistant runs a three-step process (search → analyze → synthesize) with visible progress, then returns a bullet-style summary with source attributions and follow-up suggested actions. The user can then see the completed session persisted in Recent sessions. Steps: Main composer — pick account → Account selector — open list → Account selector — filter/search → Launch saved command — Analyze insight → Assistant prompts for source or permission → Select suggested action — summarize → Background stepper — searching → Background stepper — analyze & synthesize → Result rendered — initial bullets → Full summary and suggested next actions → Assistant finalization and persistent composer → Home with Recent sessions updated **Exa: Usage Spend Review** (web, 3 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/12024) > This is a logged-in account usage and billing screen, opened from the left navigation on the Usage section. The happy path shows the user reviewing spend data in the Spend tab, optionally changing the Billing Cycle filter, and then viewing the updated pricing breakdown and invoice summary. The interface assumes the user has access to usage analytics for the current workspace/account and is on a desktop layout. The entry point is the Usage section; the exit point is either staying on the same analytics screen after reviewing spend or selecting a different billing cycle to inspect another period. Steps: Review Spend Overview → Change Billing Cycle → View Empty Period **GoFundMe: Search Results Category Filter** (web, 3 screens, Fundraising) [Refero](https://refero.design/flows/11734) > A logged-in user on GoFundMe searches for “paris” from the top search field and browses result types using the category filter. The initial results show nonprofit organizations with a secondary location filter. The user opens the category dropdown and changes the result type to Profiles, which updates the results grid to people profiles matching the same query. Entry is the global search/results view; exit is either selecting a result or refining the search/filter. Steps: Nonprofit Search Results → Category Dropdown → Profile Search Results **Cashcaded: Transaction Period Review** (ios, 2 screens, Finance) [Refero](https://refero.design/flows/14145) > The user reviews spending activity in a Transactions view and compares summaries across different time ranges. The first screen shows a daily breakdown for Sunday, September 27, including total spending, over-budget status, chart activity, category totals, and individual transactions. The user switches to the Month view, after which the interface updates to show September 2026 totals, monthly budget information, a year-positioned chart, expanded category totals, and monthly transactions. The flow begins in the Transactions analytics view and ends with the selected Month summary. The screenshots imply a mobile device and an existing data set; login state, permissions, and whether the user manually initiated the view are not visible. Steps: Daily Transaction Summary → Monthly Transaction Summary **Cashcaded: Filter Daily Spending by Category** (ios, 2 screens, Finance) [Refero](https://refero.design/flows/14144) > The user starts in a daily Transactions view for Sunday, September 27, where total spending, a chart, category totals, and transaction records are visible. They select the Food category chip to focus the summary and transaction list on food-related spending. The flow ends with the Food category highlighted, the total updated from $229 to $76, and the matching grocery transaction displayed. The screenshots suggest a mobile device and a user who is already viewing transaction data; login state and permissions are not visible. Steps: Daily Transaction Summary → Food Category Results **Foodllama: Nutrition Insight Completion** (ios, 7 screens, Health & Wellness) [Refero](https://refero.design/flows/11779) > This is a guided content experience in the Foodllama mobile interface. The user starts on a home/dashboard screen showing a featured insight card and a progress summary for today’s insights. Tapping or swiping into the card opens a full-screen lesson sequence with multiple steps, each presenting one nutrition tip with illustration, title, short explanation, and a progress indicator plus a Leave action. The sequence ends on a completion screen with celebratory confetti, a trophy graphic, and options to tell friends or finish. After completion, the user returns to the dashboard, where the insight count updates. Assumptions: user is already inside the app, on a mobile device, and viewing personalized insight content. Steps: Insight Card Dashboard → Intro Tip Screen → Carb Tip Screen → Strength Tip Screen → Completion Celebration → Return Dashboard → Updated Insight State **Binge: Viewing Statistics Breakdown** (ios, 3 screens, Entertainment) [Refero](https://refero.design/flows/12972) > A signed-in user views a personal viewing breakdown screen, then opens a display filter to switch between All, Movies, and Shows. The screen updates the chart and stats based on the selected category. Entry is from a stats or profile area; exit is via back navigation. Steps: Overview Stats → Display Filter Menu → Filtered Stats View **Moonly: Tarot chat with recommendations** (ios, 9 screens, Lifestyle) [Refero](https://refero.design/flows/9906) > Happy-path: From the Practice tab the user opens a "Go deeper" card to pick an area of life (Situation selected) and taps Get Insight. The system opens a one-to-many chat with an automated astrologer named "Luna". The bot returns a short tarot-spread summary, then a multi-part reading (several message blocks), asks a follow-up question, the user types and sends a reply, and the bot responds with empathetic text plus two recommended audio exercises (card list). Steps: Select life area (Go deeper) → Chat loading / quick summary → Longform reading displayed → Follow-up question prompt → Open keyboard and compose → Send user reply → Bot acknowledges and responds → Resource recommendations surfaced → Ongoing conversation state **Brilliant: Chart-based question selection** (ios, 2 screens, Education) [Refero](https://refero.design/flows/4320) > This is a two-step question-and-explanation interaction for a chart-based multiple-choice question. Steps: Question with chart and answer reveal → Explanation bottom sheet **Omio: Recommended train selection** (ios, 2 screens, Travel & Booking) [Refero](https://refero.design/flows/5171) > User is viewing search results for a one-way trip from Barcelona to Madrid (Wed, Apr 9, 1 adult) on a mobile device. The screen shows a transport-mode summary (train / bus / plane) with prices and durations, filter/sort chips (Sort: Cheapest, Stops, Departure), and a list of results. One result is marked RECOMMENDED and shows operator, train number, departure/arrival times, stations, duration, transfers, and price. The user can open an informational modal explaining why the train is recommended, then dismiss it with a Got it button to return to the results. Steps: Results list with recommended card → Recommendation information modal **Artsy: Set Artwork Budget Preference** (ios, 4 screens, Marketplace) [Refero](https://refero.design/flows/7622) > Happy-path: A signed-in user opens Account > Preferences and updates their Artwork Budget preference. Steps: Account - Preferences list → Artwork Budget - Editor (initial) → Artwork Budget - Option selected → Account - Confirmation toast **District: Restaurant Q&A Assistant** (ios, 7 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12846) > Happy path for an in-context restaurant question assistant. The user starts from a restaurant detail context and asks about Hotel Rishi. The interface shows the current topic, displays the user’s question as a highlighted prompt, and returns an AI-generated response with related action cards (Food Menu, Bar Menu, Beverage Menu). The user can rate the answer, ask a follow-up by text or voice, or submit a new query such as finding nearby breakfast options. The flow ends with a refreshed response and the follow-up input ready for another question. **Cashcaded: Switch to Monthly Summary** (ios, 4 screens, Finance) [Refero](https://refero.design/flows/14143) > The user starts on a daily spending view for September 27, 2026. They open the period selector, switch from Day to Month, and save the selected month. The interface then returns to the main summary view showing aggregated September 2026 spending and records. The flow assumes a mobile user viewing existing financial records; no login, permissions, or account setup state is visible. Steps: Daily Summary → Day Picker → Month Selection → Monthly Summary **Chance AI: Image Insight Recommendations** (ios, 2 screens, AI Tool) [Refero](https://refero.design/flows/7354) > Mobile result screen after an image has been analyzed. The UI presents quick external links, a short explanation and two specialist agent cards (OOTD Fitchecker and Product Value Recognizer) that the user can switch to for deeper analysis. Below the agent suggestions are a generator attribution notice, inline feedback controls (headphones, thumbs up/down, copy, flag), Share and Create Webpage actions, a collapsible Sources area, and an input bar with a camera shortcut. Entry: user has already submitted or selected an image for analysis. Exit: user either taps a specialist agent to switch into a specialized analysis, shares/creates a webpage, provides feedback, opens Sources, or types a follow-up query in the input bar. Assumptions: user is on a mobile device (touch input), image content has already been processed and AI-generated results are displayed; the user interface allows tapping agent cards and the various controls to continue the interaction. Steps: Result Summary with Specialist Suggestions → Feedback, Share and Input Controls **Pinkllama: Consume and complete a recommended insight** (ios, 5 screens, Health & Wellness) [Refero](https://refero.design/flows/5659) > Happy-path: a logged-in mobile user browsing a "For You" feed of personalized cycle insights. The user taps a recommended card to open a short multi-step content piece (carousel style), swipes through the slides, and completes the content. On completion the app shows a celebratory confirmation with a share CTA; when the user dismisses or finishes, the feed updates the card as completed and increments the daily progress badge. Steps: Feed — For You recommendations → Insight detail — multi-step content → Completion — celebratory confirmation → Feed — card marked completed → Feed — next recommended insight **Pinkllama: Cycle Insights** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/5656) > Compact home/dashboard shows a friendly hero illustration, quick shortcuts (categories) and a cycle summary card that indicates the current phase (Luteal phase) and a short progress bar plus a CTA “See all Cycle Insights.” Happy path: user taps the cycle card or the “See all Cycle Insights” CTA → a full-screen overlay opens with a timeline (Day 1–28) and explanatory content broken into sections (Period, Follicular phase, Fertile window, Luteal phase) → user scrolls through sections for details → user closes overlay to return to dashboard. Steps: Dashboard — Cycle summary card → Cycle overview (overlay header) → Details — Period and Follicular phase → Details — Fertile window and Luteal phase **Cashcaded: Cashcaded+ Onboarding and Trial** (ios, 31 screens, Finance) [Refero](https://refero.design/flows/14116) > A mobile onboarding and subscription flow for Cashcaded, a spending tracker that turns purchases into visual memories. The user answers three preference questions, receives a product explanation, previews historical spending experiences, selects a currency and visual board, reviews the Cashcaded+ feature offering, chooses a subscription plan, confirms an Apple in-app purchase, and reaches the initial home dashboard. The flow begins with a new or unconfigured user on an iPhone and ends with an active seven-day trial and an empty yearly spending dashboard. No login, permissions, or account creation screens are shown; the user is assumed to have an Apple account configured for in-app purchases. Steps: Retail Therapy Question → Food Preference Question → Memory Preference Question → Memory-Based Introduction → Product Purpose → Expense Tracking Introduction → Spending History Preview → Second History Preview → Currency Selection → Board Selection → Premium Feature Overview → Subscription Plan Selection → Premium Benefits Details → Subscription Value Review → Plan Comparison → Monthly Plan Selection → Apple Purchase Confirmation → Purchase Success Alert → Trial Activated → Empty Spending Dashboard **Mercury: Money In Insights review** (web, 3 screens, Business, Finance) [Refero](https://refero.design/flows/11058) > Logged-in desktop user lands on the Insights section (left nav > Insights). The 'Money in' tab is active. The main canvas shows a large Money in total ($1.9M), a Monthly average ($477,366), and a bar chart plotted across months. A date-range picker (top-right) is open and highlights a multi-month selection. Three small insight cards appear under the chart (Runway and cash position; Money out trends; Money in trends). A floating feedback widget (thumbs up / thumbs down) is available in the lower-right. Steps: Insights — Money In Overview → Provide feedback — mark helpful → Feedback tooltip dismissed — compact widget **Steel: Set Spending Budget** (web, 4 screens, AI Tool, Development Tools) [Refero](https://refero.design/flows/14300) > The user starts on the authenticated “Limits & alerts” settings screen for a selected workspace or project. They open the budget editor, enter a spending cap of $10, save it, and return to the settings screen. The current usage card updates to show the new $10 total budget, and the “New alert” action is available for configuring notifications. The flow assumes the user has permission to manage spending settings and is using a desktop browser. **Meta AI: Weekend Activity Suggestions** (web, 9 screens, AI Tool) [Refero](https://refero.design/flows/12751) > User starts on the assistant home screen, enters a query about free things to do nearby for the weekend, and waits while the system searches and composes an answer. The response appears in the main conversation area with a right-side panel that can switch between Steps and Sources. From there, the user can review how the answer was assembled, inspect cited sources, and hover or click source items to preview the referenced content. Steps: Home Prompt → Searching State → Steps Panel Opened → Draft Answer Displayed → Citation Hover → Sources Tab Opened → Source Preview Card → Image Preview → Source Link Hover **Cohere: Billing Usage Insights** (web, 4 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3851) > A desktop (wide-screen) Billing & Usage interface for an authenticated user navigating from the left-hand Settings menu to review recent usage and costs. Steps: Usage tab overview → Review detailed usage rows → Open endpoint filter → Apply endpoint filter **Linktree: Insights Dashboard and Content Assistant** (web, 2 screens, Business, Social Networking) [Refero](https://refero.design/flows/12411) > Logged-in user is viewing the Linktree Insights area from the left sidebar. The main dashboard shows performance cards such as most clicked links, visitor sources, lifetime totals, and social posting/connect prompts. On the right, an AI assistant drawer is open for generating personalized Instagram content. The user can review analytics, follow prompts to connect accounts or explore insights, and interact with the AI assistant. The flow starts on the Insights dashboard and ends with the user reviewing or continuing within the assistant panel or dashboard cards. Steps: Insights Overview → AI Prompt Refinement **Claude: Interactive Budget Dashboard Creation** (web, 3 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/1821) > Happy-path: A logged-in user (Robert) uses the chat assistant to convert a screenshot of a personal monthly budget into an interactive React dashboard component. Steps: Assistant Home / New Chat → Request Dashboard from Image → Review Generated Component and Preview **Runey: Add Expense** (web, 5 screens, Business, Productivity, Project Management) [Refero](https://refero.design/flows/11708) > A logged-in user starts on the Expenses dashboard, opens the New Expense modal, fills in basic and optional expense details, submits the form, and returns to the updated Expenses dashboard. The flow assumes access to the business finance area on desktop, with permissions to create expenses. It ends with the new expense saved, analytics updated, and the expense appearing in the list. Steps: Expenses Dashboard → Basic Expense Entry → Expanded Expense Details → Completed Expense Form → Updated Expenses Dashboard **Runey: Expense Category Management** (web, 4 screens, Business, Productivity, Project Management) [Refero](https://refero.design/flows/11709) > A logged-in user is on the Expenses dashboard, viewing spending summary cards, a spending chart, and a list of recent expenses. From this screen, the user can review spending by date range, inspect transactions, and open the expense categories dialog. In the dialog, the user can add a new category by typing a name and confirming with the Add button. The flow ends with the new category appearing in the category list in the modal, while the main dashboard remains dimmed in the background. Steps: Expenses Dashboard → Categories Dialog → Enter Category Name → Add Category **GlossGenius: Staff Performance Analytics** (web, 3 screens, Business) [Refero](https://refero.design/flows/9980) > Happy path: A signed-in user navigates to the Insights -> Analytics area to review salon revenue and performance metrics for the current month. The dashboard loads in an "All staff" default state showing aggregate metrics (net revenue, projected revenue, service vs retail split, team sales, booking utilization, average ticket, total bookings). The user can open the staff filter in the top-right, select an individual team member, and the dashboard updates to show metrics scoped to that person. Steps: Analytics overview (All staff) → Open staff selector → View selected staff metrics **Spyglass: Audience Scaling Recommendations** (web, 10 screens, Business, Inspiration Board) [Refero](https://refero.design/flows/13462) > A logged-in user reviews an existing AI-generated advertising analysis, asks a follow-up question about target audience segments, and receives a detailed audience targeting and scaling recommendation. The flow begins in an existing conversation and ends with the assistant presenting campaign structure recommendations and asking whether to create performance-focused creative variants. Steps: Review Existing Creative Analysis → Inspect Voice Input → Activate Voice Recording → Enter Audience Question → Submit Follow-Up Request → Monitor Brand Research → Review Analysis Progress → Inspect Persona Results → Review Audience Strategy → Review Campaign Scaling Plan **Rox: News Insight with Suggested Actions** (web, 2 screens, Business, Productivity, AI Tool) [Refero](https://refero.design/flows/10766) > Happy path: A user opens an AI-generated news analysis for a selected account (Apple). The interface shows the analysis text, a short list of suggested actions rendered as tappable cards, and a bottom composer/tool bar that includes an account tag and a text input with a submit control. Steps: Analysis Reading → Choose Action or Compose Query **Cohere: Market overview chat** (web, 3 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3856) > Happy path: A user composes a natural-language prompt in the Cohere Playground chat area, optionally configures JSON mode/schema and advanced parameters from the right-hand panel, sends the prompt, and reads the generated market overview. Entry: user lands on the Playground with Chat selected and an empty chat editor. Exit: the user has a multi-section, generated response (market size, drivers, innovations, challenges) visible in the chat and the editor ready for follow-up. Assumptions: user is on a desktop viewport, logged in, and has access to model/parameter controls on the right-hand panel. Steps: Prompt composition & schema setup → Submit prompt and view error / initial generation → Review expanded assistant response ##### Task: business account onboarding company verification **Fuse: Virtual Bank Account Signup** (ios, 24 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8730) > Happy path: From an empty Fuse Wallet the user opens a promotional card to create a virtual bank account (USD or EUR), chooses a currency, supplies an email, and is routed to a third-party identity verification (Bridge → Persona). The user completes the Basic Information step (country, name, birthdate, SSN, address with autocomplete) and submits. Steps: Wallet Home (empty) → Currency Selection (USD shown) → Currency Selection (EUR shown) → Bridge - Email capture → Bridge - Email entered → Connecting / Processing → Full-screen Loading → Persona Consent (Getting started) → Required submissions list → Country selection → Basic Information (form start) → Typing name fields → Date picker (scrolling) → Date selected → SSN entry placeholder → Typing SSN (masked) → SSN visible (toggle) → Address autocomplete → Address details (expanded) → Verification failure **UGLYCASH: New user onboarding and verification** (ios, 19 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/11202) > End-to-end new-user onboarding for a trading app: marketing carousel -> sign-in options -> phone verification (WhatsApp/SMS) -> Apple sign-in fallback -> username + referral -> legal consent -> biometric enablement -> first-time home. Steps: Welcome carousel - follow traders → Welcome carousel - trade everything → Welcome carousel - trading clubs → Welcome carousel - custody ownership → Sign-in selection → Phone entry (channel chooser) → Phone entry (send via SMS) → Code entry (empty) → Code entry (auto-filled) → Linked-account conflict modal → Apple Sign In prompt → Apple signing in → Choose username → Username claimed (input filled) → Referral code entry → Accept terms and conditions → Enable Face ID prompt → Face ID OS permission → First-time Home (authenticated) **Doji: Account creation & likeness submission** (ios, 47 screens, AI Tool, Marketplace) [Refero](https://refero.design/flows/6210) > Onboarding sequence for a mobile product that guides a new user from an initial hero carousel through phone verification, image review, permission prompts, brand-preference selection, and into the live app while a background 'likeness' creation job runs. Steps: Welcome hero — slide 1 → Welcome hero — slide 2 → Welcome hero — slide 3 → Phone entry — country & number (initial) → Country picker modal → Phone entry — number typed → Submit phone number → Code entry — waiting → Code typed — verification attempt → Verifying code → Review uploaded images → Notifications prompt (in-flow) → Creating likeness — progress screen → Select brands — entry → Select brands — selection & save → Creating likeness — animated progress → Creation continues — background state → Onboarding complete — return to home → Brand selection — long list navigation → Saved preferences and final progress **Blackbird: New user onboarding and wallet setup** (ios, 33 screens, Food & Drink) [Refero](https://refero.design/flows/9017) > Happy-path new user onboarding for Blackbird's payment wallet. Entry starts at the marketing splash where user taps 'Go', then they provide a phone number (country selector possible), receive an SMS verification code, optionally enter a referral, provide an email, add a card, and then walk through 3 short onboarding screens that explain how to tap the Blackbird puck and check out. The flow asks for location and notification permissions before landing on the authenticated wallet home. Steps: Welcome splash → Phone number entry (empty) → Phone number entry (filled) → Verification code - entry screen → Verification code - partial entry → Verification success -> Referral prompt → Email collection → Add card - input → Card added confirmation → Onboarding - Tap in → Onboarding - Puck image → Onboarding - Tell server → Onboarding - Checkout & leave → Wallet home (initial) + request location → System location permission → System notifications permission → Wallet home - populated → Check-in / Tap interaction → Payment/receipt preview → Final wallet state **Drinkit: Phone verification and profile setup** (ios, 8 screens, Food & Drink) [Refero](https://refero.design/flows/11000) > Happy-path: A user on the venue/product landing screen taps the profile/avatar control to create or update their account. The app opens a modal to collect a phone number, sends a one-time code via WhatsApp or SMS, and shows a code-entry screen. After successful verification the user is prompted to enter a display name. Once the name is saved the modal closes and the Profile screen displays the saved phone number and name. Steps: Landing / Product view → Phone number input (initial) → Phone number input (filled) → Code entry screen (empty) → Code entry (in progress) → Name input (initial) → Name input (confirmed) → Profile screen (final) **Artsy: Profile Completion Checklist** (ios, 12 screens, Marketplace) [Refero](https://refero.design/flows/7617) > Happy-path: a logged-in user from their Account/Profile card taps Complete Profile and is guided through a 3-step onboarding to add a profession, upload a profile image, and send an ID verification email. Each step shows progress (1/3 → 3/3) and offers Save & Exit. After completing all steps the user sees a confirmation checklist and returns to their Account with an updated avatar and profession. Assumptions: user is authenticated, has camera/photo permissions, and an email address tied to the account. Steps: Account — Prompt to complete profile → Add your profession — input focused → Add your profession — typing → Profession saved — proceed → Add a profile image — placeholder → Select source — photo library → Pick photo — preview → Profile image set — proceed → Verify your ID — send email → ID verification — confirmation → Completion — checklist → Account — updated profile **Ada: Onboarding and Account Creation** (ios, 19 screens, Health & Wellness) [Refero](https://refero.design/flows/4094) > New-user onboarding that introduces the product (5 illustrated slides) and guides the user through creating an account. Steps: Onboarding - Intro (1/5) → Onboarding - Causes (2/5) → Onboarding - Next steps (3/5) → Onboarding - Get started (4/5) → Onboarding - Final slide (5/5) → Consent - Confirm you're ready (initial) → Consent - Checkboxes selected → Create account - Empty form → Create account - Inline validation errors → Create account - Corrected email → Create account - Password entry and show/hide → Manage your privacy - Default toggles → Manage your privacy - Agree or save choices → Notifications & tracking opt-in → Check your inbox - Verification sent → Tips to find your email → Verification email - External message → Email verification success (web) → Account verified → Welcome Home **Airbnb: Phone sign-in, verification, and onboarding** (ios, 32 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6385) > Happy-path sign-in and first-time onboarding. Entry: user taps Log in / Sign up from the app's bottom nav. The flow supports phone-based authentication (primary), with alternatives (email, Apple, Google, Facebook). The user selects country, enters a phone number, receives an SMS, chooses verification delivery if needed, enters the code, completes a short security check, and then is guided through permission prompts (notifications) and a short onboarding/feature tour before arriving at the main Explore/home screen. Assumptions: user is on a mobile device, not previously authenticated, has a reachable phone number, and allows keyboard/numeric input. Steps: Sign in / Choose method → Privacy / Cookie prompt → Country/Region selector → Phone number entry → Phone keyboard active → Code entry (prompt) → More verification options → Entering verification code (partial) → Entering verification code (filled) → Alternate sign-in (email) → Security check intro → Landing skeleton (post-auth) → Turn on notifications prompt → OS notification permission → Onboarding screen — intro → Onboarding — Experiences preview → Onboarding — Services preview → Pricing change snackbar → Explore / Home content → Final content with badge **Copperx: New user sign-up and KYC onboarding** (web, 18 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8892) > End-to-end new user registration and initial onboarding for a financial product (marketing -> signup -> verification -> KYC -> dashboard). Entry: user arrives from marketing/marketing hero or direct signup link. Happy path: user chooses sign-up method (Google or email), verifies email with a 6-digit code, creates a user-password account (with inline password-strength feedback), selects account type and services, fills personal/KYC-related fields, accepts a third-party terms modal when linking bank accounts, and lands on the authenticated dashboard. Exit: user reaches the dashboard (account created but KYC incomplete). Assumptions: user is not signed in at entry; user has access to the email used for signup; device is a desktop browser; user can upload or provide KYC documents later. Steps: Marketing landing / signup entry → Signup method selection → Email entered (confirmation) → OTP entry (empty) → OTP entered (partial) → Account creation form (basic) → Password feedback (weak) → Password feedback (improving) → Password feedback (strong) → Services & account type selection → Service selection (multiple chosen) → Personal information / KYC form → Third-party terms modal (Bridge) → Submission error feedback → Dashboard landing (KYC prompt) → Dashboard left navigation → Empty transactions state → Ongoing verification reminder **Acctual: Account onboarding and payments setup** (web, 26 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8817) > Happy-path onboarding for a payments/invoicing product. Entry from the marketing/home hero (promotional landing) into account creation. The user either authenticates with Google or supplies email + password, then selects who will be paid (Individual or Business). For a Business, the user supplies company name and contact name, triggers an email verification, clicks the verification link in their inbox, then returns to complete branding and business details (logo, postal address, tax ID). After submitting those details the user reaches a confirmation screen and is taken to the Payments area where they can continue to configure receiving methods and identity verification. Steps: Marketing / Landing Hero → Create account (empty) → Email entered → Password reveal toggle → Choose who gets paid → Business selected & fields displayed → Verification pending state → Verification email (inbox) → Make it yours (branding) → Address modal (open) → Country selector list → Country search no-results → Country selection corrected → Add address → Branding preview & inputs → Tax ID entered → You're all set → Payments overview → Payments onboarding CTA → Landing in Payments workspace **Acctual: Company document verification** (web, 4 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8828) > Summary: This is a modal-based, multi-step document submission sequence for verifying a company's identity. Steps: Select document type and upload (Legal presence) → Upload ownership structure document → Upload company details / proof of address → Review uploaded documents and continue **PayPal: Business account setup** (web, 20 screens, Finance) [Refero](https://refero.design/flows/7417) > Happy-path user journey for converting or creating a PayPal business account and completing required verification. Entry: signed-in personal account (user visible on account settings). The user is prompted to set up a business account, selects account type (Individual/Business/Nonprofit), chooses whether to create a new business account or convert an existing personal account, fills registration fields (name, email, password, phone), completes phone verification (SMS code), provides personal identity and address information, confirms a normalized address, completes business profile (industry, sales, currency, website, registered address), optionally uploads supporting documents, and finally lands on the business dashboard. Exit: user reaches the PayPal business dashboard with prompts to complete profile/verification. Steps: Account settings (start) → Business account intro → Select account owner type → Individual option expanded → Create vs Convert choice → Convert selection state → Create account - registration form (blank) → Create account - form filled → PSD2 phone verification prompt → Enter confirmation code (empty) → Confirmation code entered → Phone verified - success → Share personal details - form → Date picker validation → Personal details - address filled → Address confirmation modal → Tell us about your business (blank) → Business details populated → Add supporting info → Merchant dashboard (landing) **Square: Merchant onboarding and account setup** (web, 38 screens, Business, Finance) [Refero](https://refero.design/flows/1952) > Happy-path merchant onboarding from marketing into the Square dashboard. Entry: visitor clicks through marketing/CTA to sign up. The flow collects account credentials, business entity details, business category, identity verification (SSN), payout preferences, and bank account details. It then shows a summary of chosen products, offers app download, and lands the user in the Square dashboard. Steps: Marketing / Signup entry → Create account (empty) → Create account (filled) → Submission loading → Business entity question → Entity type options → Business details form → Business details (filled) → Business category picker → Category search results → Full SSN requested → Verification failure / Retry → Choose payout frequency → Payout option selected → Link bank account (empty) → Link bank account (filled) → Selections overview → Welcome / Next steps → Download Point of Sale → Dashboard (first sign-in) **Mercury: Business account onboarding & setup** (web, 67 screens, Business, Finance) [Refero](https://refero.design/flows/11018) > End-to-end new business account onboarding for Mercury. Entry begins on the public marketing site and proceeds into an in-product sign-up widget where the user provides name, email and password, then sets basic company metadata (company name, callsign), enables two‑factor authentication (2FA), optionally opts into SMS updates, and is guided to the in-app dashboard. Steps: Marketing landing → Get started — name entry → Get started — name filled → Create account — blank credentials → Create account — credentials filled → Password visibility tooltip → Password visibility toggled → Company name input → Company name — filled → Callsign setup → 2FA — Mercury app (QR) option → 2FA — other authenticator option → 2FA validation — incomplete code error → 2FA validation — code entry → Onboarding checklist — progress → Get mobile updates — phone on file → Checklist — all required completed → Dashboard — welcome view → Invite recipient modal → Dashboard — default state **GlossGenius: New account signup & onboarding** (web, 27 screens, Business) [Refero](https://refero.design/flows/9925) > Happy-path new user registration and first-run onboarding on a desktop layout. Entry: marketing / product landing or a direct "Start free trial" CTA. The user fills an account creation form (name, email, phone, password), completes phone/email verification, optionally adds a payment method for trial continuation, answers short setup questions (specialty, marketing source, existing software, work setting), is offered free data transfer, and lands on the product dashboard with a welcome modal and product walkthrough video. Exit: user arrives at the main dashboard with setup checklist visible. Assumptions: user is on desktop, not previously logged in, and has access to the email and/or phone used for verification. Steps: Marketing landing / product overview → Start free trial — account form → Account form (confirmation state) → Verify your phone number (prompt) → Phone verification — alternative actions → Verify your email (prompt) → Verify email (code entered) → Verify email (confirm + alternate options) → Add payment method for trial continuity → Onboarding — choose your specialty → Onboarding — marketing source & questions → Onboarding — marketing source selected → Onboarding — existing software options → Onboarding — confirm 'Not using any' software → Onboarding — work setting accordion → Onboarding — work setting selected → Offer: free data transfer service → Dashboard — initial welcome modal (compact) → Welcome video modal (expanded) → Authenticated dashboard — next steps **Pinterest: Business account conversion** (ios, 18 screens, Reference, Inspiration Board, Social Networking, Lifestyle) [Refero](https://refero.design/flows/2547) > A mobile, in‑app sequence that converts a logged‑in personal account into a business account and walks the user through an 8‑step business onboarding (profile name, website, business type, ads interest) followed by a short product tour and an option to create the first Pin. Steps: Account management → Add account — options → Convert to a business account — confirm → Onboarding — Add profile name (input) → Onboarding — Profile name (filled) → Onboarding — Add site (input) → Onboarding — Describe your business (select) → Onboarding — Ads interest (select) → Onboarding — Tour slide: Share ideas → Onboarding — Tour slide: Grow your audience → Onboarding — Tour slide: Build a profile → Onboarding — How would you like to start? → Photo picker — grid → Photo picker — preview → Pin created — success modal → Profile view — initial state → Profile view — engaged state → Post-creation tips & confirmations **Nooka: Account creation, onboarding, and subscription** (ios, 50 screens, AI Tool, Education, Books) [Refero](https://refero.design/flows/8112) > End-to-end new user acquisition: from welcome -> sign-in / sign-up -> email verification -> short onboarding questionnaire -> subscription trial purchase -> return to main content. Steps: Welcome / Entry → Login (email) / Social alternatives → Sign Up — empty form → Sign Up — email input → Sign Up — password entry → Email Verification screen → Verification polling → Onboarding — Intro slide 1 → Onboarding — Question slide (choose friction) → Onboarding — Confirmed selection → Free Trial overview → Subscription modal — summary pricing → Subscription modal — Pro (default) yearly selected → Subscription modal — Plus tier selected → Subscription modal — Ultra tier selected → Native App Store confirmation → Purchase success modal → Subscription confirmation overlay → Tap to Join overlay (open mic prompt) → Home — authenticated content **WhatsApp: Add Verified Profile Link** (ios, 5 screens, Social Networking, Collaboration Tools) [Refero](https://refero.design/flows/12614) > Happy path for adding a verified social link to a WhatsApp profile. The user starts on the WhatsApp Profile screen, opens Links, confirms adding WhatsApp to Accounts Center, and returns to Links with an Instagram verified link added. Steps: Profile Overview → Links Options → Accounts Center Confirmation → Link Added → Success Feedback **Voicenotes: Email Verification** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14272) > The user starts in Settings with a pending email-verification prompt. They open the verification step, enter the six-digit code sent to their displayed email address, and submit it. The system validates the code, shows a success confirmation, and returns the user to Settings with the pending verification prompt removed. The user is assumed to be signed in on an iOS mobile device and able to access the specified email inbox. **Grok: Account creation and verification** (ios, 19 screens, AI Tool) [Refero](https://refero.design/flows/6076) > Happy-path new user onboarding that starts on the app's welcome screen and ends at the main Chat/Voice landing. Entry: app welcome / sign-in screen. Exit: authenticated user landing screen. Assumptions: user is on an iOS device, not signed in, and has network connectivity. Flow uses social sign-in (X/Google/Apple) or email sign-up; the screenshots follow the email sign-up path with verification, password entry, Cloudflare CAPTCHA, and an optional SuperGrok subscription upsell before landing in the app. Steps: Welcome / Sign-in prompt → Sign-in options including Email → accounts.x.ai — Log into your account → accounts.x.ai — Create account entry → Email input screen → Email typed (in progress) → Verify your email (code input) → Code validation error state → Code input with keyboard → Code accepted / proceed → Complete sign up — initial form → Filling name and password → Password validated (strength OK) → Cloudflare CAPTCHA challenge → Form ready — submit → SuperGrok upsell — monthly selected → SuperGrok upsell — yearly selected → Post-signup landing (empty state) → Main Chat/Voice interface **Instagram: Professional account setup** (ios, 25 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/2899) > Happy-path: A logged-in user is guided through an in-app onboarding flow to convert a personal account into a free professional account. The flow begins with a short carousel that explains benefits, then collects a category (and whether to display it), asks whether the account is Business or Creator, offers to follow contacts from other accounts, and finishes with a 5-step setup checklist that links to professional tools (Insights, Ads) and shows progress. Steps: Intro: Switch to professional account → Carousel: Learn about followers → Carousel: Reach more people → Carousel: Get new contact options → Choose category (entry) → Category list (scroll) → Category list & Done CTA → Display on profile toggle (off) → Category selected (Editor example) → Toggle display on profile (on) → Saving / Loading → Select account type (Business or Creator) → Follow accounts you know (import list) → Following state confirmation → Setup checklist (progress: 3 of 5) → Introduce yourself (create post) → Explore tools and insights → Finalizing setup (progress updates) → Post-setup profile view → Professional dashboard and tools **X: Account creation and onboarding** (ios, 41 screens, Social Networking, Entertainment, News, Video Streaming) [Refero](https://refero.design/flows/2903) > Happy-path sign-up and onboarding sequence for a new user. Entry from app launch (splash) → welcome screen offering Google/Apple/create options → create account form (name + phone or email + DOB) → verification (email/phone code) → suggested follows → interest selection (categories + subtopics) → follow additional recommended accounts → finished on the home timeline. Steps: Launch / Splash → Welcome / Sign-up options → Create account (empty form) → Name entry / Phone field activated → Phone entry with dial pad → Country code selection → Phone + DOB available (calendar trigger) → Date picker / Select DOB → Email entry (alternate path) → Email validation error → Verification code entry (sent) → Enter code and proceed → Suggested follows — single recommendation → Suggested follows acknowledged → Interest categories grid → Multiple interests selected → Interest subcategories (chips) → More subtopic selections → Follow recommended accounts list → Finalize onboarding — timeline **ElevenReader: New account onboarding** (ios, 21 screens, AI Tool, Books) [Refero](https://refero.design/flows/8941) > Happy-path onboarding and first-run experience for ElevenReader. Steps: Welcome / Auth choices → Create account form (initial) → Password requirements → Consent checkboxes → Submit sign up → Verification code entry → Autofill suggestion → Verification failure → Successful verification → Date of birth picker → Voice selection intro → Iconic vs in-house voices → Confirm voice and continue → Loading / transition → First-run Home → Notification permission prompt → Subscription upsell overlay → Plan selection → Purchase processing → Onboarding complete — Active Home **Turf: Phone Sign-In & Onboarding** (ios, 38 screens, Sports) [Refero](https://refero.design/flows/9419) > New-user sign-in and onboarding using phone verification. Steps: Welcome / Splash → Phone Number Input (empty) → Phone Number Input (filled) → Submitting Number (loading) → Verification Code Entry (empty) → Verification Code Entry (partial) → Verification Code Entry (processing) → Name Input (empty) → Name Entry (filled) → Contacts Permission Prompt → Contacts Shared Confirmation → Initial App Skeleton / Loading Home → Onboarding Card: Welcome → Onboarding Card: Get Ticket → Onboarding Card: Pick a Side → Onboarding Card: Win Real Cash → Onboarding Card: Coming Soon (Ask Coach) → Onboarding Card: You're Ready → Home: Ticket List Visible → Home: Interact With Ticket **PayPal: New account signup (email + phone verification)** (ios, 53 screens, Finance) [Refero](https://refero.design/flows/7224) > End-to-end new-account onboarding for a mobile payments product (PayPal). Steps: Welcome / Splash → Enter email (empty) → Email typed → Submitting email → Phone number entry (initial) → Phone input focused → Phone number entry (partial) → Phone number complete → Enter SMS code (modal) → SMS code entry in progress → Create password → Email verification entry → Email code typed → Email code validation (processing) → Notifications opt-in screen → OS notification permission → Contacts sync intro → OS contacts permission → Choose contacts to share → Completion and Home **Gizmo: Phone verification, permissions & contacts onboarding** (ios, 29 screens, Entertainment, Social Networking, AI Tool) [Refero](https://refero.design/flows/10006) > Happy-path new user onboarding that uses phone-based account creation followed by optional contact sync and a short product tutorial. Entry: app launch / splash. Exit: signed-in home feed with brief walkthrough overlays. Assumptions: user is on a mobile device, not signed in, has network connectivity and a phone number that can receive SMS, and the app requests standard iOS permissions (notifications, contacts). Steps: Welcome / Splash → Phone Number Entry (initial) → Numeric Keyboard Active → Country Selector List → Country Search → Country Selected → Phone Number Typed → Request SMS Code → Code Entry (empty) → Code Entry Error → Code Entry (partial/full) → Birthday Collection → Date Picker → Notifications Permission Prompt → Contacts Sync Explanation → Contacts Permission (OS Dialog) → Contacts Sharing Options → Find Friends List → Compose Invite Message → Signed-in Home Feed + Tutorial **ElevenLabs: Email sign-up & first-run onboarding** (ios, 12 screens, AI Tool) [Refero](https://refero.design/flows/10728) > Happy-path onboarding and first-run. Entry: user arrives at the app Home while logged out. The user chooses to create a new item which triggers the sign-in / sign-up sheet. They pick email sign-up, complete the form, receive a 6-digit verification code by email (possible iOS autofill), verify it, confirm age (18+), optionally answer a "Where did you hear about us?" question, then land back on the Home screen as a signed-in user. During first-run the OS permission prompts for notifications and app tracking appear. Exit: Home (signed-in). Assumptions: iOS device (keyboard autofill shown), user initially logged out, user has access to the provided email address, Terms/Privacy check is required to enable sign-up. Steps: Home — Logged out → Sign in / Sign up options → Sign up with email — empty form → Sign up with email — form filled → Enter verification code — empty → Verification email (external) → Enter verification code — filled → Age confirmation → Referral question → Home — First-run: Notifications prompt → App Tracking Transparency prompt → Home — Signed in / Ready **Navan: New user signup and company onboarding** (web, 18 screens, Travel & Booking) [Refero](https://refero.design/flows/9346) > Happy-path flow for a new user who visits the marketing site, signs up with a business email, verifies that email, creates a personal account and then begins company onboarding. Steps: Marketing homepage (entry) → Sign up — Enter work email → Email input accepted (illustration change) → Validation: missing marketing checkbox → Checkbox selected — continue enabled → Email rejected: personal email blocked → Alternate business email entered → Check your inbox (verification sent) → Onboarding intent selection → Create your account — form → Create account submitted → Authenticated landing — 'Where to?' → Pricing / Create company account → Welcome onboarding modal → Onboarding side panel — task checklist → Resume account setup tooltip → Admin dashboard access tooltip → Final landing with Account setup indicator **Peerlist: Create Company Profile** (web, 6 screens, Job Search, Business, Social Networking) [Refero](https://refero.design/flows/1102) > Happy path: A logged-in user (with permission to create a company) initiates creating a company profile via a modal. They choose an organization profile type, complete the profile form (display name, website, tagline, location, handle, size, etc.), then trigger an email verification step. After entering the verification code and accepting terms, the profile is created and the user is taken to the newly created company profile overview. Steps: Choose profile type → Company profile form (empty) → Company profile form (filled) → Open verify work email modal → Enter verification code → Company profile overview **Intercom: New account signup and verification** (web, 17 screens, Business, Communication Tools) [Refero](https://refero.design/flows/674) > Happy-path new user onboarding for Intercom's 14-day free trial. Entry from a marketing/landing hero into a multi-step signup wizard: enter work email (or continue with Google), provide full name, create a password (minimum 10 chars), add company name, and select company size. After submission the system attempts automated account setup and sends a verification email with quick links to open common mail clients. Exit points: successful email-sent state (primary exit) or an error state requiring manual verification via chat. Assumptions: user is not signed in, on a desktop browser, and has access to the email account they entered. Steps: Marketing hero / CTA → Enter work email → Email entry confirmation → Collect full name → Name validation error → Valid full name entered → Create password (instructions) → Password validation error → Valid password entered → Collect company name → Company name confirmed → Select company size → Company size selected → Account setup in progress → Verification email sent → Waiting / transition state → Manual verification fallback **Polar: New account & org setup** (web, 23 screens, Business, Finance) [Refero](https://refero.design/flows/11379) > Happy-path onboarding that takes a visitor from the public marketing hero into a fully provisioned test account and blank dashboard. Entry: user clicks Get Started on the marketing hero. Exit: user lands on the account dashboard in test mode. Assumptions: user is on desktop, not previously authenticated (new user), has an email address for verification, and is comfortable providing basic personal, organization, and product details. Steps: Marketing Hero → Get Started Modal → Sign Up Modal (options) → Sign Up Email Entered → Verification Code Entry → Invalid Code Error → Verification Email → Enter Code and Sign In → Personal Details → Country Selection → Business Details — Start → Currency Dropdown → Currency Confirmation → Business Mode Selected → Organization Creation → Product Details Form → Product Details — Filled → Product Validation Note → Review & Submission → Account Dashboard (Test Mode) **Dopper: Business account registration & onboarding** (web, 7 screens, Marketplace) [Refero](https://refero.design/flows/7533) > Happy-path: A visitor starts on the public storefront, navigates to the "For Businesses" / Business orders area, chooses to register, completes a multi-field business registration form (personal + company + address + login), receives an email confirmation requirement, then signs in to reach the business account dashboard (overview with contact info and address book). Steps: Public storefront (entry) → Business orders landing (login vs register) → Registration form (empty state) → Registration form (validation feedback) → Registration submit and confirmation required → Sign in and land on dashboard → Account overview and address book **Aboard: Employee sign-up and verification** (web, 6 screens, Collaboration Tools, Paperwork) [Refero](https://refero.design/flows/406) > Happy path: An unauthenticated visitor arrives at the product marketing entry screen, starts account creation, completes a short sign-up form, receives a 6-digit verification code by email, submits the code, and lands in the authenticated workspace home (dashboard). Steps: Marketing entrance → Sign-up form (empty) → Sign-up form (filled & submit) → Verification code entry → Invalid code error → Workspace dashboard (first sign-in) **TravelPerk: New account signup** (web, 9 screens, Productivity, Travel & Booking) [Refero](https://refero.design/flows/5196) > Happy-path: an unauthenticated desktop user lands on the marketing homepage, opens the primary navigation, selects the Get started action and enters a three-step signup. Step 1 collects work email and personal name + acceptance of terms, step 2 collects company details (job title, phone, company name, company size, currency), and step 3 shows an email-verification prompt with quick links to open common mail clients. The flow ends at a sign-in screen (or waiting for the user to verify their email). Steps: Homepage — Hero → Primary navigation / action modal → Sign up — Step 1 (Basic info) → Sign up — Step 1 (Filled) → Sign up — Submitting (loading) → Sign up — Step 2 (Company details) → Sign up — Step 2 (Filled) → Email verification prompt → Sign in (post-signup / default auth screen) **Exactly.ai: Account signup to prompt composer** (web, 13 screens, Business, AI Tool) [Refero](https://refero.design/flows/5679) > Entry: Public marketing/home hero with a primary CTA to start. Happy path: user clicks the CTA, opens a centered Create account dialog, chooses to continue with email (or Google), enters name and email, receives a 6-digit verification code by email, submits the code, then completes a short 3-step profile onboarding (profession, how they heard about the product, primary use case). After completing onboarding the user lands in the image-creation workspace where template cards and a prompt input are available. Assumptions: user is on a desktop viewport, not signed in at start, has access to the provided email address and can receive verification codes. Steps: Marketing / Hero entry → Create account (initial) → Create account (form filled) → Enter verification code (blank) → OTP error feedback → Enter verification code (retry) → Profile setup — profession (1 of 3) → Profile setup — profession options continued → Profile setup — 'Other' text input → Profile setup — acquisition source (2 of 3) → Profile setup — primary use (3 of 3) → Onboarding complete — workspace revealed → Prompt entry and image creation **Fourmula.ai: Account Sign Up and Dashboard Entry** (web, 7 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/11833) > This happy path starts on a marketing landing page for an AI visual generation product, then moves into account creation and lands on the authenticated workspace dashboard. Steps: Marketing Landing → Create Account → Account Filled → Submitting Sign Up → Dashboard Empty State → Email Verified Notice → Ready to Create **Cake Equity: New company onboarding** (web, 22 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/6105) > End-to-end new-company onboarding for an equity management product. Entry begins from marketing or an onboarding CTA, then guides a new user (assumed logged out, acting as a founder/administrator) through: selecting account type, filling company details, inviting teammates (co-pilots), creating an account (signup, email verification, password), adding shareholders and choosing whether to enter holdings as raw share counts or percentages, providing the total number of company shares, and landing in the product dashboard with an onboarding checklist. Exit is the authenticated dashboard with the company's cap table seeded. Assumptions: user is a company representative (founder/finance/legal), has access to email for verification, and can supply shareholder information. Steps: Marketing landing → Account type selection → Account type selected (focused) → Company details form (blank) → Company details filled → Invite co-pilots (empty) → Role selection dropdown → Permissions selection dropdown → Invite row completed → Account signup form → Email verification code input → Password setup (blank) → Password fields populated → Password mismatch error → Add your shareholders (initial draft table) → Shareholders validation (required fields) → Switch to percentage view → Bulk / totals hint and Next → Total number of company shares (input) → Onboarding dashboard (seeded) **Mercury: Formation document upload and validation** (web, 4 screens, Business, Finance) [Refero](https://refero.design/flows/11020) > Happy path: a logged-in user uploads a company's formation document from a desktop browser. The screen lists acceptable formation document types and provides a drag-and-drop upload area. After the user drops a file the system uploads it, runs automated checks (entity name, document type, filing date) and shows progress and status. If the document passes checks the user can continue with onboarding; if issues are found the system displays a warning with itemized problems and guidance. Steps: Document instructions and upload area → File added / upload started → Automated checks in progress → Validation results and issues **time2book: Studio onboarding and account creation** (web, 8 screens, Business, Fitness, Productivity) [Refero](https://refero.design/flows/6305) > Happy-path onboarding for a fitness studio platform. Steps: Landing / Marketing Hero → Sign in / Choose authentication → Email entry confirmation → Check your inbox / Email verification prompt → Account setup — role selection (Client) → Account setup — role selection (Provider) and fields → Provider form filled & submitted → Onboarding / Setup guide ##### Task: buy crypto **Family: Create account and wallet** (web, 31 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/6184) > Happy-path onboarding for creating a new Family account and a new wallet from the marketing landing entry. Entry: user begins on the public landing and chooses a platform (mobile or desktop). The flow proceeds to an email / phone sign-in that creates an account if one doesn't exist, followed by email verification, password creation, optional account customization (avatar + name), optional wallet customization (icon + wallet name), and ends at the Wallets dashboard showing the newly created wallet. Assumptions: user is on desktop (web) and not previously authenticated; they have access to the email address provided to receive the verification code. The user may skip customization steps; the system supports creating default names/icons if skipped. Steps: Landing / Intro → Choose Platform → Welcome — Identifier Input → Submit Identifier → Verification Code Prompt → Enter Verification Code → Verification Error → Set Password → Submit Password → Customize Account — Avatar Picker → Choose Emoji/Icon → Avatar Action Menu → Upload or Select Image → Set Account Name → Customize Wallet — Default Icon → Wallet Avatar Menu → Pick Wallet Color/Icon and Name → Finalize Wallet Creation → Wallets Dashboard — List → Post-onboarding options **Acctual: Add flexible payment option** (web, 10 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8824) > Happy path for adding a flexible payment option (crypto wallet or bank account) from the Payments area. Entry: user is signed in and navigates to Payments > Receive. The user opens the Actions menu, selects Add flexible payment option, chooses either Crypto wallet or Bank account, fills required fields (name, wallet address OR currency, account details, address), corrects validation errors, and saves the payment option for later or activates it. Exit: modal/slide-over is closed and the new payment option is added to the Payments list or an inline error toast is shown. Steps: Payments landing → Open Actions menu → Add option - default crypto → Switch to Bank account tab → Choose currency → Show ACH/Wire and account fields → Account number validation error → Enter account holder address → Form validation success → Return to crypto tab with toast error **Fuse: Receive crypto and update balance** (ios, 5 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8732) > Happy path: A user opens an empty wallet (logged in) and wants to receive assets. From the wallet home they invoke the primary action (+), choose Receive, pick between Fiat or Crypto, open the crypto receive modal that shows a QR code and address, then copy the address (or scan) to send funds. After funds arrive the home balance updates. Entry: wallet home. Exit: wallet home shows updated balance. Assumptions: user is authenticated and allowed to receive crypto; device has camera or clipboard available; network and asset type restrictions are handled elsewhere. Steps: Wallet Home — Empty Balance → Quick Actions Menu → Receive Options → Crypto Receive — QR & Address → Wallet Home — Balance Updated **UGLYCASH: Buy Crypto with Apple Pay** (ios, 7 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/11223) > Happy path: A logged-in user on the Portfolio tab taps the Transfer button to add funds, chooses Apple Pay from the Add funds sheet, selects a token from the token list, enters a USD amount using the large keypad, and completes the purchase via Apple Pay Confirm UI. Steps: Portfolio overview → Add funds options → Token selection / search → Buy screen — amount entry (initial) → Amount entered — show payment action → Apple Pay authentication → Processing payment **Family: Swap ETH to token** (ios, 4 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/2703) > Happy-path: a logged-in wallet user swaps a portion of their Ethereum balance for a different token (Tether / USDT). Steps: Enter source amount → Open token selector → Search for token → Select token and update amounts **Fuse: Token selector** (ios, 2 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8735) > A compact, full-screen modal for selecting a cryptocurrency token. Entry: user invokes a token selector (e.g., from a send/swap/receive flow). The modal shows a searchable list of tokens with icon, token name, on-chain balance, and USD equivalent. The user can either tap a token row to choose it or type into the search field to filter. Exit: modal closes and the chosen token is returned to the parent context, or the modal remains open showing a no-results state if the search yields nothing. Assumptions: user is on a mobile device, signed into a wallet with at least one token (Solana is shown), and the client can show token balances and USD values. Steps: Token list (default) → Search input — no results **PayPal: Add Currency to Wallet** (web, 4 screens, Finance) [Refero](https://refero.design/flows/7386) > Happy-path: A logged-in desktop user in Wallet navigates from the Wallet overview to add an additional currency (US Dollar). Steps: Wallet overview → Currency selection list → Addition confirmation → Updated wallet balances **OpenSea: Add wallet funds** (web, 2 screens, Marketplace, Web3 & Crypto) [Refero](https://refero.design/flows/1906) > Happy path: A logged-in user (wallet connected) is viewing an NFT detail and needs funds to buy or make an offer. From the NFT details the user opens the wallet drawer, taps Add Funds, and is presented with an Add funds modal defaulting to the 'Deposit crypto' tab that shows the user's wallet address and a Copy button. Steps: Wallet drawer (My wallet) → Add funds modal — Deposit crypto **Blackbird: Buy $FLY Purchase Flow** (ios, 11 screens, Food & Drink) [Refero](https://refero.design/flows/9031) > User starts from the wallet/home screen (shows USD balance, profile prompts and recent activity). From Recent Activity the user opens My Activity, then taps Buy $FLY. The Buy $FLY landing explains use cases and offers three payment channels (USDC, debit, credit). Tapping a buy option opens a bottom sheet modal with preset denomination tiles, an amount due summary (including a bonus line), a selected payment method, and a swipe-to-buy control. The user can either pick a preset denomination or open a numeric keypad to enter a custom USD amount (minimum $10). Entering $10 maps to 1,000 $FLY and shows bonus (+100) and total $FLY received (+1,100). After swiping to confirm the purchase the system attempts the transaction and returns an error: purchases must be unlocked by using Blackbird Pay at a restaurant. Steps: Wallet Home → My Activity → Buy $FLY Landing → Purchase Modal (presets) → Numeric Keypad (enter USD) → Amount Confirm (USD -> $FLY) → Ready to Buy (swipe) → Error Modal → Purchase Modal (post-error) → Enter Confirmation (alternate view) → Final State (blocked purchase) **UGLYCASH: Share Token Summary** (ios, 3 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/11207) > Happy path: A logged-in mobile user viewing another user's profile opens one of that user's open positions (the RUBIA token), inspects the token/trade detail, taps the Share control, and then copies the visual share card to the clipboard. Steps: Profile — Open position selection → Token detail — Open share preview → Share modal — Copy and return **Wealthsimple: Nvidia stock detail and order preparation** (web, 3 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8510) > Happy path: A desktop (wide), logged-out user lands on the NVDA stock detail screen to review price history, market fundamentals and news, and to prepare a limit buy. Steps: Stock overview — annual chart and fundamentals → Stock overview — monthly timeframe and extended news → Order preparation — compact chart with buy panel **Stocktwits: Add symbols to watchlist** (web, 3 screens, Finance) [Refero](https://refero.design/flows/7210) > Logged-in desktop user is on their profile page and uses the left-hand symbol search to find and add market symbols (stocks/crypto) to a personal watchlist. Steps: Profile - initial view → Search input focused / partial entry → Autocomplete results with add controls **Stocktwits: Add symbol to watchlist** (web, 5 screens, Finance) [Refero](https://refero.design/flows/7209) > Happy-path: A logged-in user on desktop (profile visible) opens the left watchlist/search panel, searches for a symbol (example: "bitcoin"), selects a matching symbol from the suggestions, and adds it to their watchlist. Entry: user is on their profile page with the left navigation visible. Exit: the app shows a confirmation toast and the watchlist UI reflects the addition. Assumptions: user is authenticated (profile shown), using a wide-screen desktop layout, and has a visible left-side watchlist/search panel. Steps: Profile landing with empty watchlist → Open search / view trending suggestions → Filter suggestions by typing → Add symbol to watchlist → Return to profile / watchlist updated **Phantom: Buy Bitcoin with third‑party provider** (ios, 11 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3710) > Happy-path: User opens a Bitcoin asset detail, taps Buy and enters an amount (or selects a preset). The app validates the amount (enforces a $5 minimum), shows the USD ↔ BTC conversion and enables Buy once valid. The user then chooses a payment method (Credit & Debit Card), views partner providers and live quotes, picks a provider (Stripe), and is routed to an external Stripe/Link checkout where they must sign in/confirm payment. Entry: asset detail screen (logged-in user). Exit: redirected to external provider checkout (Stripe/Link) to complete payment. Assumption: user is logged in to the wallet, has at least one payment method available via partners, and network quotes update on a short timer. Steps: Asset overview → Buy screen — empty amount → Validation — below minimum → Preset selected — valid amount → Toggle input unit → Payment method — entry → Payment method options → Provider quotes list → Provider checkout (Stripe Link) — summary → Stripe Link — sign up or sign in → Return to app or finalize quote **Phantom: Bookmark page from swap screen** (ios, 3 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3725) > Happy path: A user is on a mobile browser viewing a decentralized exchange swap screen (jup.ag) with a three-dot browser overflow control visible. The swap UI shows 'Selling' (USDC) and 'Buying' (SOL) fields and a prominent Connect Wallet CTA; a transient 'Wallet Disconnected' notification is present. The user opens the browser overflow menu and toggles the page bookmark. Steps: Swap screen (wallet disconnected) → Overflow menu opened (bookmark available) → Bookmark confirmed (menu state toggled) **Phantom: Receive Address & Home Navigation** (ios, 5 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3703) > Happy-path: A logged-in mobile wallet user opens the Receive list, selects a network (Ethereum), views the full receive screen with QR code and full address, and then returns to the account home where quick actions and token balances are shown. Steps: Network receive list → Network address detail (Ethereum) → Return to home (onboarding state) → Home - populated balance view → Home - zero balance / default state **Fuse: Token Selector** (ios, 2 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8736) > Happy path: A mobile user opens a coin-selection modal to choose a cryptocurrency from their wallet. The modal shows a search field at the top and a vertical list of available tokens (icon, name, token balance and fiat value aligned right). The user taps the search field, types to filter the list, and the keyboard appears. When the desired token is visible, the user can tap the token row to choose it (modal expected to close and return the selection). Steps: Token list → Search active **Family: Token Overview — Empty Wallet** (ios, 2 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/2697) > Single-screen token overview for a mobile crypto wallet showing the Tokens tab with Ethereum selected and a zero balance. Steps: Tokens tab — empty state (selected token) → Tokens tab — selected token with market data visible **Fuse: Token Receive Selection** (ios, 2 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8740) > User opens the "You receive" selector to choose which cryptocurrency to receive into their wallet. Steps: Token list (initial) → Search and filter tokens **UGLYCASH: Deposit USDC to Portfolio** (ios, 4 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/11221) > Happy path: A logged-in mobile user on the Portfolio tab taps Transfer to add funds. An Add funds modal appears offering multiple funding methods (Apple Pay, Uglycash, USDC). The user selects USDC -> the app shows a Deposit crypto sheet with a Solana USDC wallet QR code and full address plus a prominent Copy wallet address button. After completing (or simulating) a deposit, the modal closes and the Investment Portfolio balance is updated to reflect the incoming funds. Steps: Portfolio overview → Add funds — choose method → Deposit crypto — QR & address → Portfolio updated **Phantom: Share Bitcoin to Messages** (ios, 3 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3711) > Happy path: A logged-in user on an iOS device views a Bitcoin asset detail screen, taps the Share CTA, opens the iOS share sheet, selects Messages (or a contact) and ends up in the Messages compose view with a link preview of the asset inserted. Steps: Asset Detail — Bitcoin → iOS Share Sheet → Messages Compose with Preview **Kann: In‑game Coin & Life Purchase** (ios, 6 screens, Education) [Refero](https://refero.design/flows/9159) > User opens the in‑game store to buy coins or lives. The screen lists a "Buy Lives" option (1 Life for 2 coins) and several coin packs with prices. The user taps a coin pack "Buy Coins" button -> the in‑app purchase begins, the UI shows a processing state, and the device displays the system App Store confirmation sheet (Double Click / side button). After the user confirms with the side button, the App Store reports success and the app shows a success alert. Steps: Store landing (coin + lives list) → Store — alternate copy state → Local processing state → Platform purchase confirmation (App Store) → Purchase success acknowledgment → Return to store (post‑purchase) **Phantom: Receive Ethereum Address** (ios, 2 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3704) > Happy path: a logged-in mobile wallet user navigates to their Ethereum receive address to receive tokens or collectibles. The screen displays a large QR code, explanatory text, the truncated hex address, a copy affordance, and a prominent Share button. From this screen the user can either copy the address (which shows an in-line "Copied" confirmation) or tap Share to invoke a share action (share sheet not shown). Steps: Address Display → Copy Confirmation **Family: Wallet Ready — Backup Prompt** (ios, 2 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/2668) > Happy-path onboarding screen shown after a wallet is created or imported on a mobile device. Entry: user lands on a "Your wallet is ready." confirmation screen that displays an account card (name, partial address, balance) and a prominent "Back Up Now" button. The screen also contains an explanatory callout about backing up the wallet’s Secret Recovery Phrase. Exit: user either taps the Back Up Now CTA to begin the backup flow (not shown) or taps the copy control on the account card to copy the address. Assumption: user is on a mobile device and already has or just created a wallet (logged-in state). Steps: Wallet ready — backup prompt → Copy feedback — toast **Phantom: Token Selector** (ios, 2 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3708) > This is a mobile token-selection modal used to pick a cryptocurrency token from a list or by searching. Entry: user opens the token picker from a parent screen (e.g., send, swap, or portfolio action). Happy path: user opens modal, optionally types into the search field, then taps a token row to choose it. Exit: the modal closes and the selected token is returned to the caller. Assumptions: the user is on an iOS-style mobile device, has access to their wallet/token list, and the selector displays available tokens and their balances. The modal supports keyboard input and shows an empty-state when no matches are found. Steps: Token list with search → Search input and empty state **Phantom: Token search and results** (ios, 3 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3709) > Happy path: a signed-in user on iOS starts at their wallet's Account Home, opens the token search, types a query ("btc"), and views filtered token results. Steps: Account Home → Search — Trending → Search Results **Reown: Wallet Receive & Activity** (web, 4 screens, Development Tools, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/5711) > Happy path: a user opens an embedded wallet modal (AppKit demo) to view an account, request a receive address (QR + copy), switch to a different account/address, and inspect recent transaction activity. Steps: Wallet overview (account selected) → Receive view (QR + copy) → Switch account / Selected account view → Activity tab (transaction history) **OpenSea: Add WETH to make offer** (web, 4 screens, Marketplace, Web3 & Crypto) [Refero](https://refero.design/flows/1913) > User attempts to place an offer on an NFT but lacks wrapped ETH (WETH). Steps: Make an offer → Add funds — Swap for WETH → Swap settings → Token selector **Acctual: Add Crypto Wallet Payment Method** (web, 12 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8822) > Happy path for adding a crypto wallet as a payment method from the Payments area (desktop, logged-in user). Entry: user is on Payments > Receive. User opens the "Add flexible payment option" modal, chooses Crypto wallet, uses a third-party wallet connector to log in (MetaMask selected), approves connection and signs to prove ownership, then returns to the modal where the wallet address is populated. The user selects a blockchain network and an asset (USDC on Optimism in this example) and saves the configuration. Exit: the Payments list updates to show the new payment method in a pending state. Steps: Payments dashboard (Receive) → Add flexible payment option modal (initial) → Wallet connector chooser → Filter wallet list → Connector: Connect instruction → Connector: Sign-in / ownership prompt → Wallet connected — address populated → Network dropdown (expanded) → Network selected (Optimism) → Asset dropdown (search) → Asset selected and ready to save → Payments dashboard — new method pending **Acctual: Connect crypto payout wallet** (web, 6 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8823) > Happy path: From the Payments settings screen the user opens the 'Add flexible payment option' modal, chooses 'Crypto wallet', clicks Connect, selects a wallet provider from a third‑party connector, completes the wallet connection and signature to prove ownership, and returns to the modal with the wallet address populated and connection status shown. Steps: Open add payment modal → Wallet provider list → Search providers → Initiate wallet connect → Sign to confirm ownership → Connected wallet configured **Family: Create and Add Wallet** (web, 4 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/6193) > Logged-in user starts on the 'Your Wallets' list inside a family/account area. From this overview they can create a new wallet via a prominent '+ New Wallet' control (or choose the Import link). Clicking '+ New Wallet' opens a centered modal where the user can pick an avatar/color and enter a name. Submitting the modal creates a new wallet, closes the modal, and adds the new wallet card to the top of the wallets list showing a generated address and starting balance. Steps: Wallets Overview → New Wallet Modal — Customize → New Wallet Modal — Name Entered → Wallet Added to List **Twitch: Buy Twitch Bits** (web, 6 screens, Social Networking, Entertainment, Video Streaming) [Refero](https://refero.design/flows/13587) > A logged-in Twitch viewer purchases a 300 Bits package from a live channel. The user opens the Bits purchase selector, chooses the 300 Bits option, reviews the purchase in the checkout dialog, and attempts to pay by card. The first card submission fails validation, after which the user provides acceptable card details and proceeds through Revolut authentication. The payment completes successfully, and the Bits selector remains visible with the purchased 300 Bits balance reflected. The flow assumes the user is signed in, has access to a supported payment method, and can authorize a 3-D Secure payment. Steps: Select Bits Package → Review and Enter Payment → Correct Card Validation Error → Authorize Payment → Confirm Successful Purchase → View Updated Bits Balance **Reown: Wallet Selector** (web, 2 screens, Development Tools, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/5701) > Preview of a wallet-selection component inside a design/demo environment. Steps: Wallet browser (grid) → Search results — empty state **Steel: Add Card and Buy Credits** (web, 5 screens, AI Tool, Development Tools) [Refero](https://refero.design/flows/14284) > The user starts from a credits or billing area and opens an “Add payment method” modal to purchase $10.00 in credits. They choose Card, enter valid card and billing details, save the payment method, and review a purchase confirmation before buying the credits. The flow ends at the purchase confirmation step, with the final transaction pending the user selecting “Buy $10.00.” The screenshots imply an authenticated user with permission to manage payment methods and purchase credits, using a desktop-sized interface. No payment success state is shown. **Family: Connect Wallet Modal** (web, 3 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/6178) > This is a mobile (iPhone) preview of a Connect Wallet modal shown inside a Theme Builder preview. Steps: Connect Wallet modal displayed → Share prompt appears (transient) → Copy-link confirmation **Acctual: Internal transfers - add wallet & networks** (web, 2 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8851) > Happy path: From the Payments section the user opens Internal transfers and either adds a network from the network selection modal or edits/creates an internal wallet record. Steps: Select network → Wallet details and network assignment **Family: Import Wallet via Recovery Phrase** (web, 5 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/6194) > Happy-path: A logged-in user opens the Wallets area, chooses to import an existing wallet, submits a recovery phrase (or switches to private key), the system validates the input and either shows an error for an invalid mnemonic or adds the imported wallet to the user's wallet list. Steps: Wallets list → Import screen (blank) → Enter recovery phrase → Validation feedback (error) → Wallets list (import success) **OpenSea: Place Offer on NFT** (web, 3 screens, Marketplace, Web3 & Crypto) [Refero](https://refero.design/flows/1909) > Happy path (intended): A logged-in user with a connected wallet views an NFT's detail page, opens the "Make an offer" modal, enters a WETH amount, chooses offer duration, and confirms to submit an offer. Steps: NFT detail view → Make an offer modal — initial state → Make an offer modal — validation state ##### Task: calendar scheduling meeting **Genie: Chat Reminder Setup** (ios, 3 screens, Social Networking, AI Tool) [Refero](https://refero.design/flows/12426) > A chat-style assistant screen where the user types a reminder request, sends it, and receives confirmation that a reminder has been created for tomorrow at 9:00 AM. The flow begins in an active conversation with the keyboard open and ends with the assistant confirming the reminder and offering to save reference photos to a vault. Assumptions: user is already in the conversation, signed in, and has permission to create reminders. Steps: Compose Reminder → Message Sent → Reminder Confirmed ##### Task: care plan **Poppy: Poppy Account Setup and Subscription** (ios, 42 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/12176) > A first-time user sets up Poppy on an iPhone. The user enters a phone number, verifies it with an SMS code, provides a preferred name, selects the areas where Poppy can help, and configures notification behavior. The user is then shown Sprout and Bloom subscription options, selects the Bloom annual plan, and completes the purchase through Apple Pay. After confirmation and a short preparation state, the user reaches the personalized Poppy dashboard. The screenshots show a successful paid onboarding path; the visible “Not now” option on the subscription screen is an alternative exit that is not followed. Steps: Welcome and Phone Entry → Phone Number Input → Submit Phone Number → Phone Verification Request → Enter Verification Code → Complete Verification → Name Entry → Confirm Name → Choose Support Areas → Confirm Support Preferences → Customize Daily Outreach → Sprout Subscription Option → Bloom Subscription Option → Compare Subscription Features → Confirm Bloom Annual Plan → Apple Pay Subscription Review → Authorize Subscription → Dismiss Purchase Confirmation → Prepare Personalized Experience → Personalized Dashboard ##### Task: case management **Figma: Manage team member permissions and resend invite** (web, 3 screens, Design & Media Editing, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/4990) > Happy path: A signed-in user opens the team "Members" modal to review and manage team collaborators. From the modal they inspect the list of members, open the action/role dropdown for a specific member, and trigger an action (in this case "Resend" an invite). The flow ends with an on-screen confirmation toast saying the invite email was sent again. Steps: Open Members modal → Open member role/action menu → Receive confirmation ##### Task: code repository pull request **Wabi: Invite Collaborator** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12045) > Happy path for inviting another user to join a shared space from a collaboration screen. The user starts on a member management view showing current participants and invitation options. They can search for a person by username, tap an Invite action next to a matching result, and return to the main view where the member count updates. Steps: Member Overview → Search Active → Search Results → Invite Confirmed → Updated Member Count **Dropbox: Share File Request Link** (web, 3 screens, Collaboration Tools, Business, Productivity) [Refero](https://refero.design/flows/13056) > This is a Dropbox-style file requests area for a signed-in user with access to manage requests. The user starts on the File requests list, opens a sharing dialog for an existing request, copies the request link, and receives confirmation that the link is ready to share. The flow assumes the user is already authenticated and has permission to share the request. **Dropbox: Share File Request** (web, 6 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/13057) > A logged-in Dropbox user opens the File requests area, starts sharing an existing file request, selects a recipient by email, and sends the request. The flow begins on the File requests list and ends with the modal closing and a confirmation toast indicating that the Personal files request was shared. Steps: File Requests List → Share Request Dialog → Search Recipient → Select Recipient → Confirm Request Details → Share Confirmation ##### Task: compliance dashboard audit **Copperx: Account Dashboard - Transactions** (web, 3 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8899) > Happy-path: A logged-in user on desktop views their account dashboard to check balance, perform quick actions (Deposit, To bank, Send, Get Paid), and review recent transactions. The user can open the transactions overflow menu to reveal options (Show/Hide Dates, Download Statement) and toggle date visibility. Steps: Dashboard Landing → Open Transactions Overflow → Toggle Dates On **YouTube Music: Report a comment** (ios, 5 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11336) > User is viewing the comments sheet for a video (comments are sorted by Top or Newest). From this sheet the user opens the per-comment actions menu, chooses to report a specific comment, picks a reason from a modal list, and receives a submission confirmation. Entry: comments sheet is open for a video. Exit: confirmation modal is dismissed and the user returns to the comment list. Steps: Open Comments and locate comment → Open per-comment actions → Choose report reason → Submit report → Acknowledge confirmation **Dropbox: Member Activity Report Generation** (web, 4 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/13063) > Happy path in the Admin console Activity section: the user reviews activity records, opens the report generation dialog, optionally excludes file operations, and submits the report request. Steps: Activity Review → Report Options → Submit Report → Generation Confirmation **ElevenLabs: Developer Usage Analytics** (web, 2 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/8713) > Happy path: A signed-in developer navigates to Developers → Usage to inspect API consumption and credit usage for a selected time window. The interface presents an Analytics panel with controls for metric selection (e.g., Credits), date range, granularity, log scale, cumulative toggle, breakdown/filter options, and an Export CSV action. The user can switch visualization modes (line vs cumulative area) and review related sections below (API Requests, High priority concurrent requests) which display aggregated values. Entry: user opens the Developers area (desktop layout) and selects the Usage tab. Exit: user either exports CSV, adjusts filters, or navigates back to another Developers tab (Overview, API Keys, Webhooks). Assumptions: user is authenticated with developer-level access; using a desktop or large-screen browser; has permission to view usage analytics and export data. Steps: Usage - Default (Per-interval Credits) → Usage - Cumulative Credits View **Runey: Invoice PDF Download** (web, 2 screens, Business, Productivity, Project Management) [Refero](https://refero.design/flows/11704) > A logged-in user is viewing a completed invoice inside a project workspace and downloads it as a PDF. The entry point is an invoice detail/preview screen showing invoice metadata, line items, totals, and attachments. The exit point is a successful download confirmation toast after the PDF export action. Steps: Invoice Preview → Download Success **Tinder: Safety tips carousel** (ios, 3 screens, Dating & Relationships) [Refero](https://refero.design/flows/3438) > A three-card safety tips carousel presented as a centered modal overlay on the main Tinder screen. Each card contains an illustration, a bold headline, and two short guidance items. Navigation is linear via a prominent gradient CTA: 'Next' on the first two cards and 'Got it' on the final card. A three-dot progress indicator shows which card is active. The modal sits above the app's bottom navigation and dimmed background. Entry is when the modal is presented over the main screen; exit is when the user taps the final CTA ('Got it') which closes the modal and reveals the underlying main screen. Steps: Respect & Boundaries → Scams & Red Flags → Verification & Reporting ##### Task: connect bank account open banking **PayPal: Bank selection and search** (ios, 4 screens, Finance) [Refero](https://refero.design/flows/7240) > Mobile 'Link your bank account' selection flow (happy path: user finds and selects their bank). Entry: user reaches the 'Link your bank account' screen (mobile, iOS keyboard patterns visible). The screen shows a searchable list of popular banks and a prominent search field. The user can either pick a bank from the visible list or tap the search field to type the bank name (or partial name). Search shows an empty state message when no input/results and updates the list with filtered matches as the user types. Exit point (not shown): user taps a bank result to proceed to the bank linking/authentication step (next screen not included). Assumptions: user is on a mobile device, logged-in state and permissions to link a bank are already granted by prior flow steps. Steps: Popular banks list → Search field focused — empty state → Search with numeric input — no results → Search with partial text — filtered results **PayPal: International bank transfer setup** (ios, 24 screens, Finance) [Refero](https://refero.design/flows/7244) > Happy path for sending money to a new recipient via bank deposit. Entry starts from the "Send" screen (user tapped "Send to someone new"). The user chooses a destination country, enters the sending amount and currency, selects a receiving option (Bank Deposit), reviews fees, provides recipient bank details and contact address, provides sender address, adds a payment method (debit/credit card) with billing address, and attempts to continue. Exit point is either advancing to the transfer confirmation step (not shown) or stopping the flow when a validation error requires correction or the user quits. Steps: Send entry / New recipient → Country selection → Country search → Select United States → Enter amount and currency → Open receiving options → Show fees for chosen option → Confirm receiving option and continue → Recipient bank details → Contact's details (recipient) → Contact's details — phone & email → Sender's details (address) → Payment method selection → Add new card — input fields → Billing address entry → Card validation fails → User attempts to exit flow → Card type selector (picker) → Card details populated (example) → Billing address completed and re-validate **Phantom: Create New Wallet Account** (ios, 7 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3695) > Happy path: From the Your Accounts list the user taps Add / Connect Wallet, selects Create New Account, types a custom account name, and confirms creation. The app then opens the new account's home (balances and actions). Returning to the accounts list shows the new account added and highlighted. Assumptions: user is authenticated and on a mobile device; permissions to create accounts are available; this is the standard "create new" path (not importing or connecting hardware). Steps: Your Accounts (start) → Add / Connect Wallet Options → Create Account (name input shown) → Enter Account Name → Create Account (confirm) → New Account Home → Your Accounts (new account added) **Wealthsimple: Add & verify bank account** (web, 18 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8504) > Happy-path: A logged-in user on the Move funds area chooses to Add money and links a Canadian bank account. The flow alternates between in-product modals: choose bank (or search), optionally authenticate using the bank's online credentials, or fall back to a manual connect where the user provides transit + account numbers and uploads a void cheque. The flow shows progress screens while connecting and ends with a submission confirmation (account submitted for review). Steps: Move funds hub → Add money entry modal → Bank selection list → Search input (by routing number) → Bank result selected → Bank sign-in info → Bank credential sign-in (empty) → Bank credential sign-in (filled) → Connecting securely → Informational tip (Did you know) → Progress countdown → Manual connect intro → Enter transit + account numbers (empty) → Enter transit + account numbers (filled) → Verify ownership instructions → File attached → Submission processing → Submission confirmation **Wealthsimple: Open Chequing Account** (web, 15 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8496) > Happy-path onboarding for opening a new deposit account starting from a logged-in user's dashboard. Steps: Dashboard — Add an account → Account selection → Product details — Chequing → Intro — Tell us about yourself → Form — Legal name & date of birth (empty) → Form — Legal name & DOB (completed) → Form — Phone number (empty) → Form — Phone number (completed) → Form — Residency, tax residency & citizenship (initial) → Validation — Citizenship requirement error → Form — Residency & citizenship (resolved) → Form — Social Insurance Number (empty) → Validation — SIN format error → Form — SIN masked and accepted → Validation — SIN temporary / expiry notice **Wise: Personal account signup and card onboarding** (web, 18 screens, Finance) [Refero](https://refero.design/flows/371) > Happy-path sign-up and onboarding to order a card. Entry: user arrives from the marketing/landing hero and clicks the primary CTA to start registration. The user is not logged in and proceeds through a modal-style multi-step registration: enter email (or social SSO), choose account type (Personal/Business), confirm country of residence, verify phone with SMS 2FA, create a password, see a registration success screen, choose next actions, complete personal details (name, DOB, address) and lands on the Cards dashboard where they can order a card. Assumptions: desktop browser, user has a valid email and phone number, user intends to open a personal account and order a card. Exit: user arrives on Cards area (dashboard) and can continue ordering a physical card or other actions. Steps: Landing / Hero → Create account — Email entry → Email submitted → Account type selection → Country of residence → Phone entry (2FA) — empty → Phone entry (2FA) — filled → Enter 6‑digit code — prompt → Code validation error → Code accepted → Create password — empty → Create password — filled → Registration success → Post-signup choices → Tell us about yourself — form → Form — filled and month picker → Form confirmed — profile saved → Cards dashboard **Mercury: Initial deposit bank connection** (web, 5 screens, Business, Finance) [Refero](https://refero.design/flows/11026) > This flow guides a logged-in user through connecting a US bank account (via Plaid) so they can schedule their first deposit. Steps: Onboarding checklist — Connect account → Plaid consent / intro modal → Select your institution → Bank login redirect (Wise example) → No compatible accounts error **Aboard: Employee Bank Details Setup** (web, 6 screens, Collaboration Tools, Paperwork) [Refero](https://refero.design/flows/757) > Happy-path: A signed-in HR/manager user opens an employee profile, selects Edit for Bank details, chooses an account type from a dropdown, fills the required fields (routing/account/bank name as applicable), and saves. Entry: employee profile (Profile tab) where Bank details shows "No bank details set." Exit: employee Profile updated with new Bank details populated. Assumptions: user is authenticated with permission to edit employee payroll/bank info; using a desktop browser (modal centered, full-page profile layout). Steps: Employee profile (Bank details empty) → Bank details modal — blank form → Account type dropdown — options visible → Account type selection — US bank chosen → Bank details modal — type-specific fields → Profile updated — Bank details populated **Fingerprint: Find and open account-related documentation** (web, 4 screens, Business, Development Tools) [Refero](https://refero.design/flows/11143) > Happy path: a user lands on the Fingerprint Docs home screen, opens the site-wide search/Ask AI modal, searches for "account" (or a related term), and selects an "Account Sharing" / "Account Takeover" article. The flow ends with the full documentation article loaded, including code examples and step-by-step instructions. Steps: Docs Home / Discover → Search Modal (empty) → Search Results → Documentation Article **Spoil me: Create wishlist and add first items** (ios, 37 screens, Fundraising) [Refero](https://refero.design/flows/5343) > Happy-path new-user onboarding that creates a Spoil.me wishlist (vanity link), verifies identity via an email magic link, imports product photos from the device, adds item details (price, crowdfunding toggle), and lands the user in their wishlist dashboard. Entry: first-time visitor (not signed in) at the hero screen. Exit: activated wishlist/profile screen showing created wishes. Steps: Hero - Choose vanity link → Typing vanity handle → Vanity handle entered → Email input - prompt → Email validation (invalid) → Email input - valid → Confirmation - check inbox → Resend active state → Open verification email (inbox) → Waiting for verification → Photo picker - select images → Photo selection confirm → Item details - edit → Alternate suggestions - browse images → Confirm item images → Add another wish - title input → In-app notifications prompt → OS-level notification permission → Optional payout connection (Stripe) → Wishlist dashboard / onboarding state **Verse: Connect Streaming Account and Load Playlists** (ios, 7 screens, Social Networking, Design & Media Editing) [Refero](https://refero.design/flows/4000) > Happy path for connecting a music streaming account (Spotify shown) from an in-app "Add a song" modal so the app can export songs to a user playlist. Entry: user opens the Add a song modal inside the mobile app (iOS). If no streaming account is connected the modal shows a Connect Streaming Service CTA. Exit: streaming account is authorized and the user's playlists become available in the Add a song modal so the user can pick a playlist to export songs to. Assumptions: user is inside the host app and wants to link an external streaming account; the streaming service (Spotify) is installed or available on the device; the user is able to authenticate with the streaming service when prompted. Steps: Add a song (no streaming connected) → Select streaming provider → System open-app confirmation → Spotify authorization (intro) → Spotify authorization (consent) → Connection success confirmation → Add a song (playlists loaded) **Wabi: Connect Gmail in Settings** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12042) > Happy path for connecting a Gmail account from the Settings screen to enable a power-up. The user starts inside Settings, triggers the Gmail connection, passes through Google sign-in and consent, and returns to Settings with Gmail enabled. Assumption: the user is already inside the product on a mobile device and has at least one Google account available to choose from. Steps: Settings and Power-ups → Permission Prompt → Account Selection → Google Access Consent → Settings Enabled State **Phantom: Manage Wallets** (ios, 4 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3730) > Compact wallet/account management screen. Steps: Accounts list — initial state → Account selection / grouped other accounts → Secondary accounts reordered / visibility indicated → Finalized account order and persistent CTAs **222: Account → How 222 Works** (ios, 4 screens, Lifestyle, Dating & Relationships) [Refero](https://refero.design/flows/9327) > Logged-in account area (vertical list of grouped items) → user selects “how 222 works” under About → app presents a full‑screen, branded informational viewer with multiple consecutive text pages. The viewer uses a minimal header, a back affordance and a persistent “tap to continue” instruction; each tap (or gesture) advances to the next text page. Steps: Account (settings list) → How 222 Works — Page 1 → How 222 Works — Page 2 → How 222 Works — Final Pages **Mozi: Scan-to-Connect Contact** (ios, 4 screens, Social Networking, Travel & Booking) [Refero](https://refero.design/flows/4150) > Happy-path: A user scans another person's Mozi QR code with the iPhone camera, the system recognizes the code and offers to open the Mozi connection. The app loads the connection, shows a success confirmation modal, and navigates to the newly connected person's profile in 'My People'. Steps: Camera QR Detection → Open My People (Loading) → Connection Success Confirmation → Person Profile Opened **Google Gemini: Sign in and launch Gemini** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/3182) > Happy path: user opens the Gemini welcome screen, selects which Google account to use (or uses the default shown), confirms by tapping the primary CTA, sees a brief loading/transition, and lands in the Gemini assistant home where they can type or use voice/camera input. Entry: user reaches the Gemini welcome screen (not yet confirmed to a specific account). Exit: authenticated assistant home ready to accept input. Assumptions: mobile (iOS) context, device has internet connectivity, user has at least one Google account on device, and the app can access account chooser/authentication flows. Steps: Welcome + Compact account prompt → Account chooser (expanded) → Selected account confirmation → Initialization / Loading → Assistant home — Ready to interact **Binge: Apple Sign In in Settings** (ios, 5 screens, Entertainment) [Refero](https://refero.design/flows/12971) > Happy path in Settings on iPhone. The user starts on a settings screen with account sign-in and general preferences visible. They tap Sign in with Apple, review the Apple sign-in sheet, choose email-sharing options, and continue. After confirmation, the sheet closes and the settings screen updates to show the signed-in profile and account stats. Steps: Settings Overview → Apple Sign In Sheet → Signing In → Sign In Complete → Signed-In Settings **PayPal: Bank Selection for Linking** (web, 3 screens, Finance) [Refero](https://refero.design/flows/7379) > A signed-in user is presented with a modal to choose a bank to link to their PayPal account. Steps: Bank Selection List → Search — No Matches → Search — Single Match **Wealthsimple: Accounts overview and view toggle** (web, 3 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8498) > Logged-in account overview (desktop). The screen is an accounts dashboard showing a zero balance placeholder and primary actions (Add an account, Move an account). The main region lists individual accounts and grouped account sections (e.g., Non-registered) with 'Action required' indicators. A small view control (Group view / List view) lets the user change how accounts are presented. Right rail contains promotional content and a Watchlist call-to-action. Steps: Accounts overview (grouped view) → Open view menu → Switch to list view **Rox: Select or add account context** (web, 3 screens, Business, Productivity, AI Tool) [Refero](https://refero.design/flows/10761) > Happy path: A signed-in desktop user sets the current account/context for an assistant by opening the account selector, typing to filter the list, and picking an existing account (or adding a new one when there are no matches). Steps: Open account selector → Search yields no matches → Search and select existing account **Revolut: Revolut Login and Welcome Onboarding** (web, 9 screens, Finance, Business) [Refero](https://refero.design/flows/76) > This flow starts from the Revolut marketing homepage for potential or returning users, moves to the login page where the user can enter their phone number or use alternative login methods, select their country code if needed, enter a passcode for security, and finally arrives at the Revolut web app onboarding welcome modal and main account home screen. Steps: Landing and Marketing → Login Screen with Phone Number Entry → Country Code Selection → Filtered Country Code Search → Phone Number Entered Ready to Continue → Passcode Entry Prompt → Passcode Entry Completion → Welcome Onboarding Modal → Main Home Dashboard **Wealthsimple: Manage linked account nickname** (web, 3 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8519) > Happy-path: An authenticated desktop user opens the Accounts area, selects the Linked accounts tab, edits the display name (nickname) for an already linked bank account via an inline edit control, and sees the updated nickname reflected in the list. Steps: Linked accounts list → Set nickname modal → Updated linked accounts list **Fourmula.ai: Account and billing management** (web, 3 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/11846) > Happy path across account management and onboarding-related account actions. The user starts on a logged-in account/billing screen, can review current credits and subscription options, optionally choose a paid plan or top up assets, and can also access destructive account deletion via a confirmation dialog. A separate entry point shows the sign-up screen for creating a new account. Assumptions: desktop web experience, dark theme, user is either logged in for the billing/delete screens or logged out for the create-account screen, and actions are available based on current auth state. Steps: Subscription Overview → Delete Confirmation → Create Account ##### Task: contract management **Acctual: Create bill from uploaded invoice** (web, 2 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8845) > Happy path: A user creates a new bill by uploading a supplier invoice PDF. Steps: Start new bill — upload area → Preview and review extracted details **Queue: Subscription Cards List** (ios, 3 screens, Entertainment, News) [Refero](https://refero.design/flows/6239) > A mobile (iPhone) list of subscription cards where each card represents a recurring item (podcast, news subscription, website). Steps: List — default state → Reveal destructive action → List — updated / scrolled view **Subo: Find Subscription** (ios, 3 screens, Finance) [Refero](https://refero.design/flows/8588) > Happy path: a signed-in mobile user opens the Subscriptions screen (list of active subscriptions). From this list they tap the persistent search bar at the bottom, which focuses and raises the keyboard so the user can type a query. If the entered query does not match any subscription, the screen shows an empty-search state that prompts the user to change or clear the search. Steps: Subscriptions List → Search Field Focused → No Results / Empty State **Apple TV: Manage AMC+ subscription** (ios, 3 screens, Entertainment, Video Streaming) [Refero](https://refero.design/flows/11102) > Happy path: a signed-in user viewing an active Apple TV Channel subscription (AMC+) that was started via a free offer. Entry: user opens the subscription details for AMC+. The screen shows the subscription card, renewal date and price after the free offer, and primary actions (See All Subscriptions, Cancel Free Trial). If the user taps to cancel, a confirmation sheet appears. After confirming cancellation, the details screen updates to show an expiring state with an option to re-subscribe. Exit: user either keeps the subscription (closes modal) or completes cancellation and sees the expiring-subscription state. Assumptions: user is logged into their account, viewing subscriptions on a mobile device, and has permission to manage their own subscriptions. Steps: Subscription Details (Active free trial) → Confirm Cancellation Modal → Subscription Expiring (Post-cancellation) **Google Gemini: Report legal content from Gemini chat** (ios, 18 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/3187) > Happy path: A signed-in user starts in a Gemini conversation, opens the message actions menu on an assistant reply, and chooses to report the content. The app links to Google Legal Help. The user selects a legal reason, taps Create request, completes the online ‘Report a legal removal issue’ form (country, personal details, contact email), attaches screenshots/files from the device photo picker, provides a written explanation, confirms the perjury declaration, types a signature (digital), and submits. The flow ends with a submission confirmation screen telling the user they will receive a confirmation email. Assumptions: user is signed in (avatar visible), using a mobile browser/webview on iOS, has the required evidence files in their photo library, and can receive email confirmations. Steps: Gemini conversation - message actions visible → Message options menu → Legal Help landing → Select reason to report → Create request CTA → Form – country selector (tap) → Form – country list (choose) → Form – fill identity fields → Allegedly infringing content section → Photo picker – browse library → Photo picker – confirm selection → Files attached in form → Explain why content is unlawful → Perjury declaration checkbox → Signature and date → Submit request → Submission success → End state / wait for confirmation **Cake Equity: Create seed raise** (web, 18 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/6148) > Guided wizard for creating a new capital raise (example: seed stage). Steps: Raise landing → Disclaimer / Start → Start raise — metadata → Start raise — values filled → Confirm pre-money valuation → Add raise details — initial → Add raise details — filled → Company signers (nomination) → Select first signer → Add company signer — form → Add company signer — filled → Board member roles selection → Role added (alternate) → Role validation error → Primary signer added confirmation → Bank details — entry → Bank details — filled → Raise summary (Seed stage) **Kickstarter: Manage Pledge & Cancel** (ios, 5 screens, Fundraising) [Refero](https://refero.design/flows/4644) > Happy-path for a logged-in backer on a mobile device to manage and cancel an existing pledge. Entry: user opens a project where they are already a backer and taps Manage. Exit: user confirms cancellation and returns to the project overview with an in-context confirmation banner. Steps: Project overview (backer state) → Manage your pledge (details) → Action sheet (management options) → Cancel confirmation → Project overview (cancellation confirmed) **Instagram: In-chat order creation & management** (ios, 13 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/2893) > Seller creates and manages an order directly from a one-to-one chat. Entry: user opens chat with a contact and uses the message composer menu to start an order. The flow opens a bottom-sheet modal where the seller enters a monetary amount using a numeric keypad, optionally adds a description, then taps Create order. The order is posted back into the chat as an “Order created” message with an order number and View order button. Tapping View order opens Order Details where the seller can progress the order through statuses (Mark as paid → Mark as shipped). Each status change updates the order timeline in Order Details and posts status bubbles back into the chat. Steps: Chat — open composer menu → Create order — numeric keypad → Amount entry — incremental update ($1.26) → Amount entry — incremental update ($12.60) → Amount entry — target amount ($126.00) → Add description → Create order — publish to chat → Order Details — open from chat → Mark as paid — in progress → Chat — order paid message → Order Details — mark shipped → Chat — order shipped message → Order Details — final timeline **Сontra: Create and Send Project Proposal** (web, 14 screens, Job Search) [Refero](https://refero.design/flows/4880) > Happy-path: A logged-in freelancer (desktop) with no active projects starts a new paid project proposal from the Projects & invoices area. The user opens a multi-step wizard (Proposal → Contract → Review → Sign), fills client basic information and project name, configures payment method and invoice schedule, adds a project scope, selects the contract template, reviews contract and project summary, then either saves the proposal as a draft or proceeds to sign/send. Steps: Projects - Empty state → Projects - CTA tooltip → Proposal — Basic information (start) → Proposal — Basic information (filled) → Payments details — Billing method → Payments details — Invoice configuration → Payments details — Form completed → Project scope — Add description → Project scope — Completed → Contract — Select template → Contract preview and signer info → Review — Project details summary → Draft Projects list — Draft created → Draft Projects — Context and next actions ##### Task: course enrolment university **WhatsApp: Add Community Member** (ios, 5 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12678) > Happy path in a community management area. The user starts on the Community info screen as the community owner, opens Add members, selects one contact, confirms the add, and returns to Community info where the member count updates and a success message appears. **Wabi: Invite Members** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12046) > This is a mobile collaboration/invite screen reached from a shared workspace or group context. The user can either share the space, invite others directly, search for users by username, or send an invite link. The happy path starts with the user already inside the invite interface and ends when matching users are found and the user can tap Invite next to a person. Assumption: the user is logged in and has permission to invite others. Steps: Invite Overview → Search Active → Searching → No Match Result → Matched Users ##### Task: credit score report **Lovable: Referral Invite Modal** (web, 2 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12350) > A centered referral modal opens over a dimmed dashboard/sidebar view. The user is shown a shareable invite link and QR code with instructions for earning credits through referrals. The intended entry point is a referral/share action from within the product, and the exit point is either closing the modal or copying the invite link. Steps: Referral Details → Link Copied State **Willow: Referral Card Sharing** (ios, 2 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/13241) > A logged-in user views a referral card screen for sharing an invite link or QR code. The entry point is likely from a rewards/referral section, and the exit points are back navigation, copying the referral link, downloading the card, or opening the system share sheet. The screen assumes the user already has an active account and referral code available. Steps: Referral Card View → Copy Confirmation **TIDAL: Artist Credits Browser** (ios, 4 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/13137) > A mobile music credits screen for an artist, showing a list of tracks associated with Steve Lacy and allowing the user to filter credits by role using pill tabs such as All, Songwriter, Performer, Producer, and Production team. The user enters from the artist credits view, changes filters to narrow the visible track list, and can exit via the back button. Steps: All Credits View → Songwriter Filter → Performer Filter → Production Team Filter **TIDAL: Artist Biography** (ios, 2 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/13136) > A music biography screen shows a long, scrollable artist article in a dark theme. The user enters from an artist or music profile context and can read the biography, follow linked artist/song names, and scroll through the full text. A back button returns to the previous screen. A persistent mini player stays docked at the bottom, allowing playback control while reading. Steps: Biography Reading → Biography Continued **Bump: Report profile/name** (ios, 6 screens, Map & Navigation, Social Networking) [Refero](https://refero.design/flows/3164) > Happy path: a logged-in user views another person's profile (Danny). From the profile the user opens the more-actions menu, selects Report, steps through the reporting flow by choosing the problematic field (Name) then choosing a reason category and submitting. After submission the system shows a confirmation toast and returns the user to the profile. Steps: Profile - primary view → Actions sheet → Report - select field → Report - choose reason → Confirmation toast → Final profile state **Train Fitness: User Profile Actions** (ios, 3 screens, Health & Wellness, Social Networking) [Refero](https://refero.design/flows/7143) > Happy path: A signed-in user views another person's profile feed (list of workout/activity cards). From the profile header the user opens the overflow (ellipsis) menu, chooses the Report User action, and receives immediate in-app feedback confirming the report. Steps: Profile Feed (Activity List) → Overflow Action Sheet → Confirmation Toast **ElevenLabs: Generation Details** (ios, 2 screens, AI Tool) [Refero](https://refero.design/flows/10742) > User inspects the details of a completed image-generation job. The screen lists read-only metadata (model, number of generations, aspect ratio, resolution), shows the textual prompt in a rounded container with a copy control, and displays reference image thumbnails. The happy path: user reviews settings and reference images, copies the prompt to the clipboard, and returns to the previous screen. Steps: Details Overview → Copy Confirmation **Moonly: Compatibility Check** (ios, 8 screens, Lifestyle) [Refero](https://refero.design/flows/9909) > Happy-path flow for checking compatibility between the user and another person using a birth-chart based system. Steps: Birth Chart (Entry) → Discover Compatibility Intro → Partner Data — Selection Mode → Partner Grid — Browse → Confirm & Calculate → Compatibility Results — Overview → Category Breakdown → Detailed Insight Card **Chance AI: Palmistry Camera-to-Report** (ios, 6 screens, AI Tool) [Refero](https://refero.design/flows/7363) > Happy path: A mobile user launches the Discover / Home area, chooses the Palmistry AI Snap and taps the camera CTA to open the camera in Palmistry mode. The UI provides an on-screen palm outline guide and then detects key points on the hand after capture. The image is processed by the visual reasoning model and the app returns a formatted "Palmistry Reading Report" the user can read, scroll, and share. Steps: Discover Home / Camera CTA → Camera - Palmistry Mode (Alignment Guide) → Hand Detection Feedback → Processing / Visual Reasoning → Palmistry Reading Report (Initial) → Full Report & Actions **RAD Weather: Weather Intelligence Details** (ios, 2 screens, Weather) [Refero](https://refero.design/flows/13805) > A user reviews a detailed weather and environmental analysis screen. The flow begins at the top of a dark-themed details view, showing threat status, humidity analysis, sensor readings, and a temperature chart. The user scrolls down to inspect moon-phase information, including illumination, viewing conditions, and upcoming phase dates. The user can finish by selecting Done. Steps: Environmental Analysis Overview → Moon Phase Details **Understory: Generate Report** (web, 8 screens, Business, Scheduling Tools) [Refero](https://refero.design/flows/12725) > Logged-in user in the Insights > Reports area creates an Excel report by opening the empty-state report action, choosing a report type, selecting a date range, and submitting the form. The modal closes while the report is generated in the background, then the new report appears in the reports list with a status indicator and a success toast/message. Steps: Reports Empty State → Generate Report Modal → Report Type Selection → Date Range Setup → Date Picker Adjustment → Ready to Generate → Generation In Progress → Report Created **Dropbox: Create Activity Report** (web, 7 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/13062) > From the Admin console’s Members section, the user opens a row actions menu for a deleted member, starts creating an activity report, selects a date range in a calendar picker, and submits the report request. Steps: Members List → Row Actions Menu → Report Date Range → Calendar Selection → Date Range Updated → Submit Report → Request Confirmation **GoFundMe: Report Content Submission** (web, 4 screens, Fundraising) [Refero](https://refero.design/flows/11736) > Happy path for reporting content from a public profile. The user starts on a profile/feed-style page, opens the report modal, selects a reason from a dropdown, and submits the report. The flow assumes the user is already logged in and has access to the report action. It ends with a confirmation modal indicating the report was received and will be reviewed by support. Steps: Profile Overview → Select Report Reason → Submit Report → Report Confirmation **Grammarly: Writing Performance Report** (web, 2 screens, Paperwork) [Refero](https://refero.design/flows/1878) > Happy path: a user working in the editor opens a centered Performance modal to inspect writing quality metrics (text score, word count, readability) and then exports a PDF report of those metrics. Entry: user is editing a document and the assistant/score panel is visible in the right-hand rail. Exit: the user either closes the modal after inspection or initiates (and waits for) a PDF report export. Assumptions: user is on a desktop-class viewport (editor with right sidebar visible) and has access to the performance/details control in the assistant; the document contains text and the modal can be opened from the sidebar. Steps: Open Performance modal → Export PDF report **Tidal: Report incorrect track credits** (web, 4 screens, Music Streaming) [Refero](https://refero.design/flows/3237) > Happy path: A signed-in desktop user browsing an artist's credits opens a track's context menu, selects "Report error," fills out a "Report incorrect credits" form (one or more track entries plus a description), completes the reCAPTCHA, and submits. The flow ends on a simple confirmation/thank-you screen. Steps: Open track context menu → Report incorrect credits — empty form → Report incorrect credits — populated form → Submission confirmation **ElevenMusic: Artist Issue Report** (web, 5 screens, Music Streaming, Entertainment, AI Tool) [Refero](https://refero.design/flows/12580) > This happy path starts on an artist profile in the music service, with the user already signed in and viewing the Danger Twins artist page. From the profile, the user opens a report modal, selects a reason from a dropdown, adds optional details, and submits the report. The flow ends back on the artist page with a confirmation message indicating the report was received and will be reviewed. Steps: Artist Profile → Report Dialog → Reason Selected → Submit Report → Submission Confirmed **Curater: Report AI-Generated Content** (web, 4 screens, Inspiration Board) [Refero](https://refero.design/flows/13885) > The user starts on a pin detail view with an author contribution and selects the visible Report action. A modal asks them to choose an issue type, and the only available option is AI-Generated. After selecting it, the interface presents a confirmation step explaining that the vote will count toward flagging the content and that multiple reports will be reviewed by moderators. The user submits the report and returns to the pin detail view, where a success toast confirms completion. The flow assumes the user can view the pin and access its reporting control; no additional authentication or permission state is visible. Steps: Pin Detail View → Report Type Selection → Report Confirmation → Submission Confirmation **Teal: Final Resume Check** (web, 4 screens, Productivity, Job Search) [Refero](https://refero.design/flows/2721) > Happy path: a signed-in user working in the Resume editor opens a "Final Resume Check" side panel to review completeness, analysis, and job-matching scores before exporting. Steps: Resume Editor — Guidance → Open Final Resume Check → View Analysis Breakdown → Inspect Job Matching **Shop: Product Reporting** (web, 6 screens, Marketplace) [Refero](https://refero.design/flows/13269) > A user views a product detail screen and reports the listing. The user opens the product actions menu, selects the report option, chooses a report reason, optionally adds supporting details, and submits the report. The flow begins on a product detail view and ends with a confirmation toast indicating that the report was sent. Steps: Product Detail → Choose Report Reason → Select Other Reason → Add Report Details → Review Report Details → Report Confirmation **Tidal: Artist Credits and Track Liking** (web, 2 screens, Music Streaming) [Refero](https://refero.design/flows/3234) > Desktop, logged-in user (profile visible in left nav) is viewing a credits list for a single artist. The screen shows a header "Credits for [Artist]", a row-based table of tracks with columns (#, Title, Role, Artist, Album, Year, Time), and filter chips across the top (All, Songwriter, Producer, Engineer, Production team). A persistent mini player sits at the bottom showing the currently playing track. Steps: Browse artist credits → Like/favorite a track from list ##### Task: crypto wallet send receive **PayPal: Send money to contact (currency & payment type)** (ios, 25 screens, Finance) [Refero](https://refero.design/flows/7235) > Happy path: A logged-in user starts on the 'Send and Request' hub, searches for or selects a recipient, picks the destination currency, enters an amount, adds an optional message/sticker, chooses a payment type (Goods & Services or Friends & Family), and attempts to send money. Steps: Send & Request hub → More options sheet → Choose a currency (list) → Search currency input → Filtered currency result → Amount screen – initial → Enter amount using keypad → Amount updated with conversion → Message composer (text + stickers) → Compose message text → Sticker gallery → Search recipients → Recipient profile preview → Amount screen for selected recipient → Confirm recipient currency modal → Choose payment type → Payment type selected (friends & family) → Sending attempt / processing → Recipient cannot accept payments error → End / exit point **PayPal: Send money to contact (currency conversion)** (ios, 12 screens, Finance) [Refero](https://refero.design/flows/7263) > Happy path: Logged-in user opens the Send and Request flow, selects a suggested contact (Bennett Marcus), chooses a currency for the recipient, enters an amount on the numeric keypad, reviews payment details (funding source, goods/services, shipping address, conversion method), confirms the send, and sees a success confirmation. Steps: Send and Request - landing → Amount screen (initial) → Amount screen (recipient currency active) → Enter amount (numeric keypad) → Open Review before sending → Choose funding source (modal) → Shipping address selection → Confirm shipping preference → Currency conversion method → Final review and send → Transaction confirmation → Post-send options **Fuse: Send SOL to external wallet** (ios, 14 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8734) > Happy-path: A logged-in user on the Wallet home screen uses the floating + button to start a Send to wallet flow. The user selects Send → To wallet, chooses Solana as the coin, selects or scans a recipient address, enters an amount using the numeric keypad, reviews the transfer (including on-chain fee sponsorship), authenticates, and sends. The flow exits back to the Wallet home with a transient Sending indicator and then a Sent confirmation; the displayed balances update. Assumptions: user is authenticated, has network connectivity, has a positive balance in the selected coin, and the device supports camera access and biometric authentication. Steps: Wallet Home → Quick Actions → Send Destination Choice → Select Coin → Choose Recipient (manual/QR) → Camera Permission Prompt → QR Scanner → Recipient Auto-filled → Enter Amount (keypad) → Amount Confirmation → Review Transfer → Authenticate & Send → Broadcasting / Sending State → Sent Confirmation & Updated Balance **Phantom: Send Ethereum to Another Account** (ios, 13 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3718) > Happy-path mobile wallet send flow: user starts on the account overview (logged-in mobile wallet), chooses a token to send, selects or pastes a recipient, enters an amount (with conversion between ETH and fiat), reviews the transaction summary including network fees, and finally sends the transaction. The app shows a sending state and success confirmation, then exposes the transaction in Recent Activity and allows viewing the TX on Etherscan. Steps: Account Overview → Select Token → Recipient Input → Enter Amount — Keyboard → Validation — Insufficient Balance → Enter Amount — Fiat Mode → Summary → Network Fee Info Overlay → Sending (Progress) → Sent Confirmation → Recent Activity → Transaction Detail (In-app) → Etherscan Transaction Details **Family: Send ETH to Watched Wallet** (ios, 18 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/2707) > Happy-path: From the wallet home the user opens the action sheet and chooses Send. They pick a recipient (from watched wallets or by pasting/scanning), select a token (Ethereum), enter an amount using the on-screen numeric keypad (optionally Use Max), review a confirmation screen that shows the ETH amount and network fee with selectable speed presets (Normal/Urgent), then confirm the transaction (biometric). The system submits the transaction, shows a 'Starting your transaction' loader, returns to the wallet with the updated balance and adds a new entry to the activity list. Assumptions: user is unlocked and on the Ethereum asset screen, user has a small ETH balance, watched wallets are already configured. Steps: Wallet Home - Action Sheet → Send — Enter Recipient → Send — Recipient Selected → Token List → Collectibles Tab (Empty) → Amount Entry — Keypad (Initial) → Use Max — Toggle → Amount Entry — Max Applied → Max Adjustment Info Modal → Confirm — Review Transaction → Adjust Fee Speed → Confirm — Final Approval → Transaction Starting Loader → Return Home — Balance Updated → Activity List — New Transaction → Transaction Detail Modal → Detail — Confirmation & History → Final State — Updated Balances & Feed **Copperx: Email Payment to New Contact** (web, 9 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8896) > Happy-path: a logged-in user on the Copperx dashboard initiates a 'Send' action, chooses to send by email, adds a new contact (payee), enters the amount and purpose, confirms the payment, and receives a success confirmation. Steps: Dashboard (start) → Choose Send Method → Send Money Form (empty) → Recipient Dropdown → Add Payee (empty) → Add Payee (filled) → Confirm Payment → Success Confirmation → Post-send Dashboard State **PayPal: Send money with currency conversion** (web, 11 screens, Finance) [Refero](https://refero.design/flows/7408) > Happy path for sending a one-off payment to a saved contact using desktop. Steps: Select recipient → Enter amount (initial) → Enter amount (converted) → Add optional note → Quick security check (method) → Enter security code (empty) → Enter security code (filled) → Review payment summary → Choose funding source (modal) → Currency conversion choice (modal) → Confirmation **Family: Receive address and QR** (ios, 3 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/2673) > Happy path: a logged-in mobile wallet user with an empty token balance wants to receive Ethereum (or ERC‑20 tokens). Entry: wallet home (Tokens view). The user opens the action menu (+), selects Receive, and is taken to a Receive screen that displays their address (truncated), a QR code, a copy affordance, and a Share Address CTA. Exit: user copies the address or shares it and returns to the Tokens view. Assumptions: user is authenticated, a wallet/address already exists, and device is a mobile phone. Steps: Tokens (home) - Empty state → Action sheet - Send / Swap / Receive → Receive - QR and address **Fuse: Receive address (QR & copy)** (ios, 2 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8733) > Happy path: a logged-in user on an iPhone with an empty wallet opens the Receive action to display their Solana receive address as a QR code and address string, copies the address, and returns to the Wallet home where a short "Copied" confirmation is shown. Steps: Receive modal — QR & address → Wallet home — copy confirmation **Phantom: Ethereum Receive Address** (ios, 2 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3705) > Happy path: A user views their Ethereum receiving address (QR + truncated hex) and shares it using the device's native share sheet. Steps: View Ethereum Address → Native Share Sheet **Family: Save Sender To Address Book** (ios, 3 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/2698) > User is on the wallet's transaction history (history list showing recent received Ethereum transactions). From there they open a transaction's action menu (context sheet) and choose to add the sender address to their address book. Entry: user is on the History tab, logged into the wallet on a mobile device. Exit: the address is added and a transient confirmation toast appears. Assumption: user has access to transaction details and the app persists entries to an address book. Steps: History List → Transaction Actions Sheet → Confirmation Toast **Family: Token send initiation** (ios, 4 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/2672) > Happy path: A logged-in user on a mobile (iOS) Ethereum wallet lands on the Tokens view, opens the quick actions menu, selects Send, and is taken to an address entry screen where they can paste or scan a recipient address. The flow entry point is the Tokens list for a single wallet; the exit point is either successful navigation to the next send step (amount/confirm) or an inline validation error if the entered address is invalid (in the screenshots the app blocks sending to the user's own address). Assumptions: user is authenticated, Ethereum network selected, device is a mobile phone, and the wallet has no token balances (empty state shown). Steps: Tokens list / empty state → Quick actions bottom sheet → Send — recipient entry → Send — validation error for self-address **WishWish: Copy Wallet Address** (ios, 2 screens, Entertainment, Kids) [Refero](https://refero.design/flows/14244) > A logged-in user opens the Settings screen and copies their wallet address using the copy icon. The interface displays a temporary confirmation toast, “Copied to clipboard,” while remaining on the same screen. The flow starts on the authenticated Settings screen and ends with the wallet address available in the device clipboard. No navigation or permission change is shown. **Fuse: Add External Recovery Key** (ios, 8 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8748) > Happy-path: A logged-in user opens Keys & Recovery, taps + to add a Recovery Key, selects "Crypto wallet", pastes or scans a public key, reviews the new key, triggers a proof-of-ownership signature via their external wallet, and returns to the Keys & Recovery screen where the new recovery key appears and the recovery key counter increments. Steps: Keys & Recovery (list) → Add Recovery Key (choose type) → Crypto Wallet (empty input) → Crypto Wallet (input validated) → Review Recovery Key → External Wallet Signature → Adding Recovery Key (progress) → Keys & Recovery (updated) **Family: Send ETH from Wallet** (web, 14 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/6199) > Happy path: A signed-in user opens the Family > Your Wallets list, opens the wallet context menu and starts a Send operation to another address/contact. They paste or search for a recipient, select the recipient, choose the asset and balance to send, enter an amount (or use max), review transaction details (including fee estimate and source wallet), acknowledge a first-time-recipient warning, confirm to send, see a sending state, and finally see a success confirmation. Entry: Your Wallets list. Exit: Updated wallets list with refreshed balances. Assumptions: user is authenticated, has multiple wallets and at least one token balance; the interface verifies address lookups and shows a confirmation step before broadcasting a transaction. Steps: Wallets List → Context Menu - Send → Send - Recipient Entry → Recipient Resolved → Select Token to Send → Choose Specific Balance → Amount Entry → Asset Confirmation → Review Transaction (Initial) → Review Transaction (Acknowledge) → Transaction Sending → Transaction Sent Confirmation → Return to Wallets - Balances Updated → End State **Reown: Send Token from Wallet** (web, 9 screens, Development Tools, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/5713) > Happy-path summary: A signed-in, connected wallet user views their wallet summary, opens the send modal, selects a token, enters an amount and recipient, previews the transaction, and is prompted to approve the transaction from their wallet provider. Steps: Wallet overview → Open Send modal → Send modal initial state → Select token → Enter amount → Add recipient and preview → Review send details → Wallet approval prompt → Show error / transaction failed **Copperx: Receive USDC via QR/address** (web, 5 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8898) > Happy-path flow for receiving USDC into a Copperx wallet. Steps: Dashboard — Get Paid entry → Get Paid modal — method selection → Receive payment — network selection → Deposit USDC — QR and address display → Dashboard — updated balance and transactions **Family: Wallet Actions Menu** (web, 3 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/6192) > Happy-path: A signed-in user visits the "Your Wallets" area, selects the actions (ellipsis) menu for a specific wallet, chooses "Copy Address," and receives a confirmation toast. Steps: Wallets Overview → Actions Menu Opened → Copy Confirmation **Family: Manage Wallet Visibility** (web, 3 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/6195) > User is on a logged-in desktop/large-screen 'Wallets' area of an account called "Family." The main screen shows a list of existing wallets (name, short address, balance), a prominent "New Wallet" button, and an "Import" link. From the list the user can open a context menu (three-dot) for each wallet to perform actions (edit, send, view activity, copy address, hide, export, etc.). The flow shows the user opening the context menu for one wallet and choosing "Hide Wallet," resulting in the wallet disappearing from the primary list, a "View 1 Hidden Wallet" affordance appearing, and a transient confirmation toast "Wallet Hidden." Entry: user is signed in and viewing Wallets. Exit: wallet is placed in a hidden state and the UI returns to the Wallets list with a confirmation message. Steps: Wallet list overview → Open wallet context menu → Hide wallet and show confirmation **Copperx: Dashboard Sent Transaction — Cancel** (web, 3 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8912) > Logged-in desktop dashboard -> open a specific sent transaction -> cancel that transaction. Entry: authenticated user lands on the account Dashboard (top navigation shows user avatar and a KYC banner). The user inspects their recent transactions and selects a 'Send' transaction. A right-side transaction details panel (modal/drawer) opens showing status and metadata, with a Cancel Transaction action available when the status is initiated. Exit: user cancels the initiated transaction, the status updates to Canceled in the details panel and a success toast appears; the transaction list reflects the canceled state. Assumptions: desktop web, user is authenticated and has permission to view and cancel their own transactions, KYC is not completed (banner visible) but does not block these actions. Steps: Dashboard overview → Transaction details — initiated → Transaction details — canceled **Copperx: Add Contact with Wallet** (web, 7 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8904) > Add contact flow starting from the Recipients list. Steps: Recipients list (Entry) → Add New Contact - initial form → Enable wallet address input → Network selection → Enter/paste wallet address → Confirm Details modal → Contact saved success ##### Task: direct debit **District: District Money Top Up** (ios, 5 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12876) > This is a wallet top-up and balance management journey for District Money. It starts from a District Money intro screen, then moves to an amount entry screen, a payment method selection screen, and finally the District Money balance and transaction history screen. Assumption: the user is already inside the District experience and can access the wallet; the flow is shown on mobile and appears to support only certain payment methods for adding funds. Steps: District Money Intro → Amount Entry → Payment Method Selection → Payment Method Restrictions → Balance and History **Gizmo: Direct message - compose, react, share liked post** (ios, 12 screens, Entertainment, Social Networking, AI Tool) [Refero](https://refero.design/flows/10028) > Private one-to-one conversation where the user composes text, sends messages, receives short replies, reacts, and shares a liked post into the chat. Entry: user opens a conversation with a contact (Danny Go). Exit: shared content card and final text message delivered and marked as seen. Steps: Conversation open (empty) → Start composing message → Message sent → Incoming reply → Compose follow-up question → Confirm total → Recipient short confirmation + typing → Open emoji / reaction picker → Add reaction → Open content menu → Select liked item → Send post with message **Exa: Add Billing Funds** (web, 8 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/12025) > This is a billing top-up journey in a logged-in account on desktop. The user starts on the Billing section with an existing remaining balance and can add funds, choose or enter a payment method, submit the purchase, and receive confirmation. Steps: Billing Overview → Add Funds Details → Card Entry → Payment Submission → Processing State → Payment Method Confirmation → Success Confirmation → Updated Billing Overview **Givingli: Schedule & Pay for Gift** (ios, 8 screens, Lifestyle, Scheduling Tools) [Refero](https://refero.design/flows/8194) > User schedules a pre-configured gift to a selected recipient and completes payment. Steps: Gift review / recipient → Checkout with payment placeholder → Add payment modal opened → Start card entry → Numeric entry for card fields → Card & billing populated → Confirm card and return to checkout → Attempt schedule & payment failure **Orbit: Change subscription payment method** (ios, 4 screens, Finance) [Refero](https://refero.design/flows/9996) > Happy path: A signed-in mobile user edits an existing subscription to change its payment method (and optionally adjusts the payment date), then saves the changes. Steps: Open Edit Subscription → Payment Methods list → Confirm Payment Method → Save Edited Subscription **DailyArt: Checkout Payment Details** (ios, 2 screens, Education, Magazines & Newspapers) [Refero](https://refero.design/flows/3333) > Final payment step of a mobile checkout where the user provides payment details, chooses between credit/debit card or PayPal, accepts terms, and submits the order. Entry: user arrives at the payment section during checkout (mobile view). Happy path: user selects credit/debit card, fills card number, expiration date and CVC, checks the "I agree" checkbox, then taps Place Order to submit. Exit: payment submission is attempted (leading to order confirmation or error handling). Assumptions: mobile device (iOS status bar visible), user has reached the checkout summary and is ready to provide payment; network connectivity is available. The UI shows supported card logos, a PayPal option, privacy notice, and the required terms-and-conditions checkbox. Steps: Payment method & empty card fields → Filled card fields and consent checked **RAD Weather: Support Credit Purchase** (ios, 8 screens, Weather) [Refero](https://refero.design/flows/13815) > The flow starts on the RAD Weather Dispatch screen, where the user can view a generated weather card and open the sharing or support experience. The user selects Share, chooses an optional support tier, and selects the Minimal Input contribution for €0.99. The system presents the native App Store payment sheet, where the user confirms the one-time purchase using the device side button. After payment processing, the system confirms that the purchase was successful and presents an in-product credit-received acknowledgement. The user continues through the acknowledgement until the Dispatch screen is restored, now showing a green status indicator. The flow assumes the user is on a supported iOS device with an App Store account and an available payment method. Steps: Dispatch → Insert Credits → App Store Payment → Purchase Processing → Purchase Success → Credit Received → Support Acknowledgement → Updated Dispatch **District: Booking payment selection** (web, 4 screens, Entertainment, Food & Drink, Travel & Booking) [Refero](https://refero.design/flows/13009) > A booking checkout payment modal opens over a dimmed booking review page. The user can choose among payment methods in a left-side menu, including UPI QR, card entry, and netbanking. Depending on the selection, the right panel shows either a QR code to scan, a form to add a new card, or a saved cards list. The flow assumes the user is already logged in and is completing payment on desktop web. It starts from booking review and ends when the user either scans the QR code, saves/uses a card, or selects a stored payment method. **District: Payment Method Selection** (web, 4 screens, Entertainment, Food & Drink, Travel & Booking) [Refero](https://refero.design/flows/13017) > A centered payment-method modal appears over a blurred checkout or booking screen. The user can expand one of several accordion sections to choose a payment method: saved cards, add a new card, netbanking, UPI, or pay later. In the card section, a saved card can be selected and a CVV is required before the Pay button becomes active. **Givingli: Add card and confirm payment** (ios, 6 screens, Lifestyle, Scheduling Tools) [Refero](https://refero.design/flows/8196) > Happy-path: user is on the Delivery / Checkout summary for a scheduled gift and needs to add a new credit card to complete the purchase. Steps: Delivery / Checkout summary → Payment modal — choose Link or card → Card entry — numeric keyboard → Card details filled → Confirm card entry → Delivery summary with saved card **District: Payment Method Settings** (ios, 3 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12881) > Happy-path view of the Payment settings screen where a signed-in user manages saved payment methods on a mobile device. Entry is from a higher-level settings area; exit is back to the previous screen after adding, renaming, or deleting a payment method. The screen groups methods into Cards, UPI, Wallets, and Pay Later. Some items are already saved and expandable, while others show ADD actions. One state shows a card expanded to reveal Rename card and Delete card actions, and a later state confirms removal with a success message and the list updated. Steps: Payment Settings Overview → Card Actions Expanded → Deletion Success State **Blackbird: Pass purchase & gift checkout** (ios, 15 screens, Food & Drink) [Refero](https://refero.design/flows/9028) > User views a single-pass product (Daily Scoop Club LA) then proceeds to an in-context checkout (BLACKBIRD PAY) to either buy for themselves or gift the pass. From the product card the user can open the checkout modal, toggle use of in‑product credit ($FLY), enter a promo code, or choose 'Gift this pass' which opens a recipient form and date picker. The user attempts to complete payment via a swipe-to-buy control; the attempt results in a card-declined error. Steps: Product detail / pass card → Checkout modal (initial) → Toggle $FLY on → Open promo code input → Enter promo code → Promo code applied state → Open Gift flow → Fill recipient & select duration → Open date picker → Select send date → Gift Checkout ready → Swipe-to-buy control → Processing state → Card declined error → Error dismissal / recovery **Fresha: Add payment and confirm booking** (ios, 8 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4955) > Happy-path: A signed-in mobile user on the booking "Review and confirm" screen adds a payment method to pay a required deposit, then returns to the review screen with the new card selected and confirms the booking. Steps: Review and confirm (no payment) → Select payment method (sheet) → Add card — start entering name → Add card — complete details → Card save — immediate error shown → Card save — processing retry → Select payment method — saved card available → Review and confirm (card selected) **Blackbird: Blackbird tap-to-pay tutorial** (ios, 3 screens, Food & Drink) [Refero](https://refero.design/flows/9021) > Three‑screen onboarding that teaches a diner how to pay using a Blackbird puck at a restaurant. Entry: user opens the "How it works" onboarding within the app (device with NFC). Happy path: user reads each screen then advances with the circular arrowors CTA until tapping the final Continue button to exit onboarding. Assumptions: the user has the Blackbird app installed and is viewing an in‑app instructional carousel on a phone with NFC capability. Exit: onboarding marked complete and control returns to the main app (ready to use the payment feature). Steps: Tap in on arrival → Inform the server → Checkout confirmation & finish **Column: ACH Transfer Creation** (web, 9 screens, Business, Finance, Productivity) [Refero](https://refero.design/flows/1932) > Happy-path: A logged-in user (Sandbox) starts on the Transfers list, opens the Create Transfer control, chooses an ACH Transfer, creates a new counterparty via a modal (including routing number lookup that resolves bank name), fills required fields (bank account, amount, effective date, company/receiver metadata, advanced options) and submits the transfer. The user is returned to the Transfers list where the newly created transfer appears with an initial status. Steps: Transfers List (empty) → Create Transfer Menu → Create Transfer Form (Type Selection) → New Counterparty Modal (empty) → New Counterparty Modal (user input) → Routing Lookup and Validation → Counterparty Created and Selected → Complete Transfer Details → Submit and Confirm Transfer Listing **PayPal: Add Payment Card** (web, 8 screens, Finance) [Refero](https://refero.design/flows/7384) > Happy path: A logged-in user from the Wallet area chooses to add a payment method, selects to link a card, fills card number, expiry and CVV, optionally adds or edits the billing address, and submits the form. The UI shows card network detection, inline help for CVV, and a pre-filled billing address with an option to add a new one. In this particular run the submission resulted in an account-level error shown as a red banner. Steps: Wallet overview → Add method choice → Card form initial → Card network detection → CVV help tooltip → Add new billing address (expand) → Billing address entered → Account-level error shown **Farfetch: Address to Payment Checkout** (web, 12 screens, Marketplace) [Refero](https://refero.design/flows/3550) > Happy-path checkout starting from the shopping bag. Steps: Shopping Bag Review → Delivery Address (empty) → Delivery Address (filled) → Delivery Method Review → Payment Options Shown → Cryptocurrency Selected (example) → Card Payment Fields → PayPal Selected → Redirecting to External Provider → Redirect Interruption → Interruption Modal Shown → Fallback / Session Recovery **District: Movie Ticket Booking** (web, 31 screens, Food & Drink, Travel & Booking, Entertainment) [Refero](https://refero.design/flows/13002) > A user in Mumbai browses the details of Spider-Man: Brand New Day, checks available cinemas and showtimes, applies filters, selects a screening, chooses two seats, and completes payment by card. The journey begins on the movie detail screen and ends with a confirmed booking containing a QR code and ticket information. The user is assumed to be able to browse without signing in; payment is completed using a saved or newly entered card. Steps: Movie Details → Browse Showtimes → View Seat Pricing → Filter by Format → Filter by Show Time → Filter by Price → Review Filtered Results → Open Seat Map → Inspect Seating Layout → Choose Seats → Select Payment Method → Add Card → Review Card Details → Choose Saved Card → Enter CVV → Submit Card Payment → Confirm Gateway Payment → Wait for Booking → Booking Confirmation → Access Digital Tickets **Column: Check Deposit via Remote Capture** (web, 5 screens, Business, Finance, Productivity) [Refero](https://refero.design/flows/1934) > Happy-path: A signed-in user (Sandbox environment, desktop) navigates to the Transfers area, opens the Create Transfer control, selects Remote Deposit Capture, uploads front/back check images, fills required bank and check fields, and returns to the Transfers list where the new deposit appears. Steps: Transfers list (empty) → Create Transfer menu → Remote Deposit Capture form (initial) → Populate deposit details and uploads → Transfers list (with new entries) **Trip: Multi-City Flight Booking** (web, 22 screens, Travel & Booking) [Refero](https://refero.design/flows/12090) > The flow begins on Trip.com’s Flights landing screen and ends on a booking-cancelled screen after payment times out. The user configures a multi-city itinerary from Shanghai to Guangzhou and then Guangzhou to Tokyo, selects travel dates and passenger details, searches available flights, filters and chooses flights for both legs, enters traveler and contact information, reviews baggage and booking terms, and proceeds to payment. The screenshots show a single-adult, economy-class booking. No successful payment or completed booking is shown. Steps: Flight Search Landing → Choose Multi-City Search → Select First Origin → Select First Destination → Select Travel Dates → Set Passengers and Cabin → Load Flight Results → Review First-Leg Results → Filter First-Leg Results → Choose First Flight → Choose Second Flight → Enter Booking Information → Acknowledge Fare Availability → Open Passenger Details → Review Name Guidance → Complete Passenger Details → Confirm Passenger Information → Review Completed Checkout → Select Payment Method → Payment Timeout and Cancellation **Instacart: Payment Methods Management** (web, 3 screens, Food & Drink, Marketplace, Food Delivery) [Refero](https://refero.design/flows/10102) > This is the Edit / Manage Payment Methods area of a logged-in user's account settings (left navigation has Payment methods selected). Happy path shown: user views saved payment methods, chooses to delete a saved Visa card, confirms deletion in a modal, and returns to the add-payment-method list with the saved card removed. Steps: Saved methods and add options → Delete confirmation modal → Post-deletion state **Mercury: Delete payment from inbox** (web, 4 screens, Business, Finance) [Refero](https://refero.design/flows/11059) > Happy path: a user in the Payments > Inbox list opens an individual payment's detail drawer, chooses to delete that payment, confirms in a modal, and the item is removed from the list with a success toast. Steps: Payments inbox list → Payment detail drawer → Delete confirmation modal → Post-deletion list state **Airbnb: Remove saved payment method** (web, 5 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6012) > Happy path for removing a saved payment method from Account Settings > Payments on a desktop layout. Steps: Payments overview → Open per-card actions → Remove confirmation → Confirm removal and feedback → Updated payments state **Navan: Manage travel payment methods** (web, 3 screens, Travel & Booking) [Refero](https://refero.design/flows/9459) > Admin is in Settings → Payments to manage travel payment methods on a desktop. The happy path shown: view an existing card, trigger deletion, confirm deletion in a modal, and return to an empty-state list with a primary CTA to add a payment method. Steps: Payment methods list → Delete confirmation modal → Empty-state with add CTA **Cohere: Remove payment method** (web, 3 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3897) > Happy path: A logged-in user navigates to Billing & Usage → Payment, reviews their stored card, chooses to delete it, confirms deletion in a modal, and returns to the Payment tab where the card is removed and the interface shows a free-account/empty-payment state with CTAs to add a payment method or get a production key. Steps: Payment overview with stored card → Delete confirmation modal → Post-deletion / empty payment state **OpenAI: Manage payment methods** (web, 2 screens, Business, Development Tools, AI Tool) [Refero](https://refero.design/flows/4195) > Happy path: a signed-in user navigates to Organization > Billing > Payment methods to view and manage stored credit cards. The screen shows card tiles with last four digits, expiry, a 'Default' badge where applicable, and action links: 'Set as default' and 'Delete'. An 'Add payment method' button is available to add new cards (button visible but not activated in the screenshots). Steps: Payment methods list (multiple cards) → Payment methods list (updated) **Preply: Manage payment methods** (web, 3 screens, Education) [Refero](https://refero.design/flows/9231) > Happy-path: A logged-in user visits Settings > Payment methods to remove an existing saved credit/debit card. From the Payment methods screen they click Remove, confirm deletion in a modal, and the UI updates to show no saved card and an Add card action. Steps: Payment methods list → Remove confirmation modal → Updated payment methods / empty state ##### Task: dispute transaction chargeback **PayPal: Cancel sent money request** (ios, 6 screens, Finance) [Refero](https://refero.design/flows/7265) > Happy path: A logged-in user sees a pending incoming request on their Home activity list, opens the request details, chooses to cancel the request, confirms the cancellation in a modal, and receives a success acknowledgement. The flow entry point is the Home activity list showing a pending request; the exit point is the Home activity list updated to show the request as canceled. Assumptions: user is authenticated, has network connectivity, and has permission to cancel their own sent requests. Steps: Home — Activity list → Request detail — Pending → Confirm cancel — Modal → Cancel acknowledgement — Success modal → Request detail — Canceled → Home — Activity list updated **PayPal: Send or Request Money to Contact** (web, 7 screens, Finance) [Refero](https://refero.design/flows/7388) > Logged-in desktop flow for sending or requesting money to a saved contact. Steps: Select recipient → Payment card — enter amounts → Currency selector modal → Payment card — amounts confirmed → Choose payment type — initial → Choose payment type — friends & family selected → Blocking error — account problem **Column: Initiate transfer return** (web, 3 screens, Business, Finance, Productivity) [Refero](https://refero.design/flows/1931) > Happy path: A logged-in user (desktop) lands on a specific Transfer detail view for an incoming transfer. From this screen the user can open a Create Return dialog, choose a return code, optionally edit description/addenda, and submit. After submission the UI shows a success banner and a new Return section on the same Transfer view that lists the created timestamp, status and the chosen return code and description. Steps: Transfer details → Create return modal → Return confirmation and details **District: Gift Card Redemption** (ios, 5 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12868) > User starts from a restaurant bill screen where a payment offer is available and opens the gift card redemption option from the wallet source sheet. They can enter a 16-digit gift card code and 6-digit PIN, then attempt to claim it into District Money. The flow assumes the user is already in the payment context and can use the gift card option on a mobile device. It exits after either successfully claiming the card or receiving an invalid code/PIN error. **Square: Requesting payment from conversation** (web, 5 screens, Business, Finance) [Refero](https://refero.design/flows/1943) > Happy path: A logged-in merchant (Refero) in the Messages inbox opens a customer's conversation, launches the Request Payment modal, enters an amount, and attempts to send a payment request. The system prevents sending because the merchant has not completed identity verification, then inserts a payment-request card into the conversation marked as failed. Steps: Conversation view / Composer → Request payment modal (empty) → Request payment modal (amount entered) → Identity verification required → Conversation updated — Payment request inserted (failed) **Fresha: Membership purchase and checkout** (ios, 7 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4956) > Happy-path purchase of a salon membership on a mobile device. Entry: user is on the salon's profile under the Buy tab. The user taps to buy a membership, selects a specific membership plan, reviews membership details and the list of services it can redeem, then proceeds to checkout and attempts to pay with a saved card. Exit: either successful purchase (not shown) or payment failure (shown). Assumptions: user is signed in, using a mobile device, and has at least one saved payment method (Visa •••• 4292). Steps: Buy tab — Membership entry → Select a membership → Review details (membership) → Redeemable services list → Review and checkout — summary → Payment attempt → Payment failure / error **District: Event Booking Checkout** (web, 8 screens, Entertainment, Food & Drink, Travel & Booking) [Refero](https://refero.design/flows/13014) > User starts on a booking review screen for an event ticket purchase. They review the order summary, expand pricing details, complete billing information, accept terms, and proceed to payment. The payment step opens a payment-method modal with options such as UPI QR, card, and netbanking; selecting a saved card can trigger a CVV prompt and processing state. If payment authorization fails, the user is returned to a booking failed screen with a retry option. **AVNIER: Checkout and Payment** (web, 6 screens, Marketplace, Lifestyle) [Refero](https://refero.design/flows/12003) > This is a purchase checkout journey for a fashion/e-commerce site. It begins with the cart drawer opened from the store and continues to the full checkout page. The user reviews items, optionally adds an accessory, enters shipping/contact details, selects a shipping method and payment option, applies a discount or gift-card code, and then attempts payment. Assumptions: the user is already browsing the store, has items in the cart, and can access checkout on desktop. The final state shown includes both a successful filled-in checkout setup and an error state when card payment is refused. Steps: Cart Drawer → Checkout Details → Discount Code Entry → Shipping Calculated → Payment Form Ready → Payment Refused **PayPal: Open a transaction dispute** (web, 15 screens, Finance) [Refero](https://refero.design/flows/7409) > End-to-end happy-path for reporting a problem against a transaction from the activity/transaction detail screen. Steps: Transaction detail → Choose issue category → List items — empty → List items — filled → Pick issue — primary options → Pick issue — subtype selection → Provide other details — empty → Date picker → Provide other details — filled → Seller contact — asked → Seller contact — selection → Add documents — empty → Add documents — uploaded + notes → Review entered info → Confirmation / Case filed **PayPal: Activity transaction search** (web, 3 screens, Finance) [Refero](https://refero.design/flows/7382) > Happy-path user journey for locating transactions from the Activity section. Steps: Activity landing / default list → Search with numeric ID — no results → Search by keyword — results shown **PayPal: Transaction Conversation Thread** (web, 6 screens, Finance) [Refero](https://refero.design/flows/7380) > Happy-path: A logged-in user navigates from an Activity transaction (a refund) into the transaction-specific conversation thread, composes a message, sends it, and sees the updated chat with timestamps and the contact header. Steps: Transaction detail (Refund) → Conversation intro (amount shown) → Conversation history (scroll) → Compose message (draft) → Send message (message posted) → Updated conversation with contact header **Gumroad: Process sale refund** (web, 4 screens, Business, Marketplace) [Refero](https://refero.design/flows/9247) > Seller views the Sales list, opens an individual sale in a right-hand details panel, attempts to refund the purchase using a Refund control which asks for confirmation, and the system returns an error banner reporting insufficient balance. Steps: Sales list → Sale details panel → Refund confirmation modal → Insufficient balance error **Teachable: Transaction inspection and refund** (web, 5 screens, Business) [Refero](https://refero.design/flows/9629) > Happy-path: A logged-in school owner or admin views the Transactions overview (list + summary totals), opens an individual transaction to inspect buyer, pricing and fee breakdowns, chooses to refund the transaction, confirms the refund in a modal, and returns to the Transactions list where totals and the individual transaction’s earnings update to reflect the refund. Steps: Transactions overview → Transaction detail view → Refund confirmation modal → Refund success and updated transaction → Transactions list with updated totals **WhatsApp: Report Business Chat** (web, 5 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/13168) > Happy path on WhatsApp Desktop/Web for reporting a business chat. The user is already logged in and viewing an existing conversation with District By Zomato. The flow starts from the chat header overflow menu, opens the report dialog, optionally includes blocking the business, submits the report, and ends with a confirmation toast while returning to the chat view. **Brilliant: Streak Charges Confirmation** (ios, 3 screens, Education) [Refero](https://refero.design/flows/4324) > Happy-path: a mobile user is shown a confirmation screen after acquiring Streak Charges, with an option to continue or read more. The user can tap the inline link (“What are Streak Charges?”) which opens an in-app help article (FAQ) presented as a sheet. The FAQ is scrollable and includes sections explaining what Streak Charges are, how to get and use them, visibility location, and impact on progress. The article ends with a “Did this answer your question?” feedback control (three emoji responses) and a Close control in the sheet header. Steps: Confirmation / Reward Acknowledgement → Streak Charge FAQ — Header → FAQ — Visibility, Impact, and Feedback **WhatsApp: WhatsApp Plus Subscription Cancellation** (ios, 7 screens, Social Networking, Collaboration Tools) [Refero](https://refero.design/flows/12616) > This is a mobile subscription management and cancellation path for WhatsApp Plus. It starts on the subscription overview screen, where the user can review active status, benefits, billing details, payment method, and a cancel subscription action. The user then opens the cancellation form, selects a cancellation reason, submits the cancellation, and receives a confirmation message that benefits remain active until the end of the current billing period. The flow assumes the user is already subscribed and viewing their account’s subscription settings on a phone. Steps: Subscription Overview → Subscription Details → Cancellation Reason → Reason Selected → Processing Cancellation → Cancellation Confirmation → Canceled Subscription Overview **DailyArt: Tip Jar Donation** (ios, 4 screens, Education, Magazines & Newspapers) [Refero](https://refero.design/flows/3334) > Happy path: A user on the DailyArt welcome/home screen taps a prominent "Tip Us" CTA which opens a bottom sheet (Tip Jar) listing donation amounts. The user selects an amount which triggers iOS purchase authorization (App Store / Side Button confirmation). The flow ends with the purchase attempt result shown and the Tip Jar still visible. Steps: Welcome / Home → Tip Jar — Amount Selection → Platform Purchase Authorization → Result — Error Feedback **Trip: Canceled Booking Refund Support** (web, 8 screens, Travel & Booking) [Refero](https://refero.design/flows/12086) > Happy path starts on a canceled hotel booking details screen for a logged-in Trip.com user. The user reviews cancellation and refund information, then opens Hotels & Homes Service Chat from the booking context to ask about refund timing and what to do if the money has not returned. The chat shows an automated response with refund details, the user follows up with an additional question, and the system returns more guidance about contacting the hotel if needed. The flow ends in the chat with support information visible; no booking changes are made. Steps: Canceled Booking Details → Service Chat Opens → Refund Question Sent → Refund Answer Returned → Follow-up Question → Follow-up Processing → Refund Escalation Guidance → Additional Deposit Advice **Gumroad: Membership charge refund** (web, 3 screens, Business, Marketplace) [Refero](https://refero.design/flows/9249) > Happy path: Seller opens the Sales section, selects a specific sale (a membership) from the list, views the sale details in a right-hand detail panel, initiates a charge refund from the Charges section, confirms the refund in a modal, and observes the Charges entry update to show a Refunded state and related receipt options. Steps: Sales list and membership detail → Refund confirmation modal → Refund completed and updated charges **Mercury: Transaction comment management** (web, 4 screens, Business, Finance) [Refero](https://refero.design/flows/11036) > Happy-path: A signed-in user opens a transaction from the Transactions list to read or manage comments in the right-hand detail panel. From the transaction detail the user opens the per-comment actions menu, chooses Delete, confirms in a modal, and the comment is removed and replaced by a “This comment was deleted.” placeholder. Steps: Transaction detail with comments → Open comment actions → Show delete confirmation → Comment removed and thread updated **IKEA: Contact Support Chatbot Interaction** (web, 10 screens, Marketplace, Home & Decor) [Refero](https://refero.design/flows/8273) > A desktop customer support scenario starting on the Contact us landing area and continuing through the site’s chatbot. Steps: Contact us landing → Open chat — welcome message → Chat loading → Chatbot initial prompts → Select cancel order → Provide order number → Email confirmation → Confirm same email → Clarification and next steps → Chat end and survey ##### Task: doctor video consultation **District: Restaurant Question Assistant** (ios, 5 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12847) > Happy-path mobile assistant experience for asking a restaurant-related question and continuing with a follow-up. Entry is a restaurant detail or assistant overlay with an active query context (“Asking about Hotel Rishi”). The user asks a question by voice, the system transcribes and processes it, shows an answer with suggested next actions, and keeps a persistent follow-up input available. The flow ends when the user has a ready-to-send follow-up request, such as asking for nearby breakfast options. ##### Task: document editor collaboration **Google Gemini: Export Chat Response to Google Docs** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/3188) > Happy path: A user in the Gemini Introduction chat opens the contextual actions for a specific assistant reply, chooses Export to Docs, waits for the system to create a document, and then opens the newly created Google Doc to review the exported text. Steps: Chat message with action bar → Contextual actions menu → Document creation progress → Creation success confirmation → Google Docs viewer with exported text ##### Task: email inbox **Givingli: Send a Card to a Contact** (ios, 12 screens, Lifestyle, Scheduling Tools) [Refero](https://refero.design/flows/8216) > Happy-path: A logged-in user starts at an empty Inbox and chooses to send a digital card. They select one or more recipients from their contacts, pick a card from categories or popular cards, personalize the card (choose template and message), optionally add a gift, review delivery options (send now or schedule), then send. The flow ends at a send-confirmation screen with sharing options and a return to the Inbox where the sent item appears. Steps: Empty Inbox → Select Recipients (list) → Recipient confirmed → Card Gallery → Card Detail → Choose a Template → Card Editor (Message tab) → Add a Gift → Review & Delivery → Send Confirmation → Inbox Updated → Conversation Detail (recipient) ##### Task: expense management receipts **District: Restaurant Bill Split** (ios, 6 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12867) > Happy path for reviewing a restaurant bill, optionally splitting it equally with other guests, and proceeding toward payment. The user enters from a restaurant bill screen, can keep the bill as-is or turn on bill splitting, choose how many guests to split with, confirm the split, and then continue to payment. **Cashcaded: Receipt Sharing Preparation** (ios, 2 screens, Finance) [Refero](https://refero.design/flows/14136) > The user opens a Share sheet for a purchase record and can copy the displayed social-media tag before sharing the receipt image. The flow begins on a mobile device with a share modal already open over a purchase detail screen and ends with confirmation that the tag was copied. The user appears to be viewing an existing receipt or transaction; no login, permissions, or account state are visible. The final Share button remains available but is not shown as selected. Steps: Share Preview → Copy Confirmation **Runey: Invoice Sharing** (web, 3 screens, Business, Productivity, Project Management) [Refero](https://refero.design/flows/11705) > A logged-in user views an invoice, opens the Share action, sends the invoice to an email recipient from a share modal, and then returns to the invoice view with a success toast confirming delivery. Steps: Invoice Detail → Share Invoice → Sent Confirmation **Orbit: Subscription import and review** (ios, 5 screens, Finance) [Refero](https://refero.design/flows/9991) > Happy path: A user chooses to import a bank statement to find subscriptions, follows in-app instructions to export and save a statement to Files, selects the file from the iOS Files picker, waits while the app parses the file, then reviews a list of detected subscriptions which are staged for adding. Steps: Get started → Bank statement instructions → Select file (Files picker) → Parsing / Finding subscriptions → Add subscriptions (review) **Runey: Invoice Payment** (web, 5 screens, Business, Productivity, Project Management) [Refero](https://refero.design/flows/11706) > This flow starts on an invoice detail screen for project INV-0001 inside a workspace labeled Website Redesign. The user opens the Add Payment side panel, enters a payment amount and selects a payment method/date, then submits the payment. The invoice status updates from Pending to Paid, the remaining balance reaches zero, and a success confirmation appears. Assumptions: the user is already signed in, has permission to view and edit the invoice, and is using a desktop layout. Steps: Invoice Review → Add Payment Panel → Payment Amount Update → Paid Confirmation → Settled Invoice View **Cashcaded: Replace Expense Photos** (ios, 7 screens, Finance) [Refero](https://refero.design/flows/14123) > The flow starts on an expense entry with an existing food photo attached. The user opens the image actions, chooses to replace the attachment, enters the camera interface, and is prompted to grant photo-library access. After allowing access, the user selects up to four photos from the library and confirms the selection. The expense entry returns with the replacement images displayed as a carousel. The flow assumes an iOS mobile device, an existing expense draft, and system-level permission prompts that may appear when photo access has not yet been granted. The user is not shown signing in or saving the expense in the provided screens. Steps: Expense Entry With Image → Image Actions → Camera Capture → Photo Access Permission → Photo Picker → Updated Expense Image → Image Carousel Review **Cashcaded: Create Expense Record** (ios, 37 screens, Finance) [Refero](https://refero.design/flows/14121) > Happy path for creating and saving a dated expense record. The user starts on an empty 2026 daily dashboard, creates a new entry, captures and attaches food photos, adds a map location, link, amount, rating, note, and title, then saves the record. The flow assumes a mobile iOS device with camera, photo-library, and location access available; the user is not shown signing in or selecting permissions beyond the photo-library prompt. The flow exits to the dashboard, where the saved entry contributes to the daily total, rating, and category summaries. Steps: Empty Daily Dashboard → New Entry Editor → Camera Capture → Photo Attached → Photo Actions → Additional Photo Capture → Photo Access Prompt → Photo Picker → Multiple Photos Attached → Photo Carousel Review → Location Confirmation → Location Search → Location Applied → Link Entry → Amount Entry → Rating Selection → Note Entry → Required Title Validation → Title Completion → Saved Dashboard Summary **Cashcaded: Delete Expense Record** (ios, 3 screens, Finance) [Refero](https://refero.design/flows/14137) > The flow starts on the detail view for expense record #0004, showing its image, metadata, amount, rating, and available actions. The user initiates deletion through the trash icon, reviews an irreversible-action confirmation sheet, and confirms. The system then closes the detail view and returns to the expense overview, where the deleted record is no longer visible in the displayed collection. The screenshots show a mobile device layout. Authentication and permissions are not visible, so the user is assumed to have access to manage the displayed record. Steps: Expense Details → Deletion Confirmation → Expense Overview **Cashcaded: Restaurant Expense Entry** (ios, 2 screens, Finance) [Refero](https://refero.design/flows/14130) > The user is editing an expense record for a restaurant visit. The flow starts on a transaction detail form with an existing link field focused and a text keyboard visible. The user then enters or edits the transaction amount, which changes from $0.00 to $25.00 and switches the keyboard to numeric input. The visible flow ends with the completed amount ready to be saved. The user appears to be on a mobile device, with the record already open and editable. No login or permission state is visible. Steps: Transaction Details → Amount Entry **Cashcaded: Share Purchase Receipt** (ios, 6 screens, Finance) [Refero](https://refero.design/flows/14135) > The user starts on a purchase detail view for an Apple AirPods Max expense and opens the sharing controls. A share preview appears with a branded receipt image and an optional “Tag us @cashcaded” copy control. The user reviews the preview, taps Share, selects Messages from the native iOS share sheet, and sends the receipt image to a suggested contact. After returning, the sharing surface displays a Done confirmation. The flow assumes an iOS device with the purchase record available and Messages configured; login state and permissions are not visible. Steps: Purchase Detail → Share Preview → Full Receipt Preview → Native Share Sheet → Message Composer → Share Complete **Cashcaded: Edit Expense Link** (ios, 2 screens, Finance) [Refero](https://refero.design/flows/14129) > A logged-in mobile user reviews an expense or receipt record and edits its link field. The flow starts on a record-detail screen showing a food and restaurant entry with an image, metadata, total, rating, note option, and Save button. The user activates the Link field, enters a URL using the mobile keyboard, and remains on the detail screen with the updated value ready to save. The visible sequence ends before the user taps Save, so persistence is assumed but not shown. Steps: Expense Record Details → Enter Reference Link **Cashcaded: Add Expense Note** (ios, 2 screens, Finance) [Refero](https://refero.design/flows/14131) > The user is viewing an expense or record detail screen for a restaurant transaction. They focus the note field, enter a personal note, and leave the record ready to save. The flow starts on an editable detail view and ends with the note entered but not yet saved. The user appears to be on a mobile device; authentication and permissions are not visible. The record already contains a title, date and time, location, link, currency, amount, and a four-star rating. Steps: Focus Note Field → Enter Note **District: Booking Review Voucher Offer** (ios, 5 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12831) > User is on a movie ticket review screen and is shown a promotional dining voucher experience before continuing to payment. The flow starts on the booking review screen, opens a voucher modal, unlocks the voucher with a celebration state, and returns to the booking summary where the voucher is marked as added. It ends with the payment section visible and a clear action to choose a payment method. Steps: Booking Review → Voucher Reveal → Voucher Unlocked → Voucher Details → Checkout Summary **Loowner: Add new item with photo and receipt** (ios, 13 screens, Marketplace) [Refero](https://refero.design/flows/6597) > Happy-path for adding a new owned item from an empty-states entry point. User starts on the "All your stuff" empty screen and taps the primary CTA to add an item. They pick the purchase date, enter the cost using a large numeric keypad, add optional photo(s)/receipt (invoking camera/gallery and permission prompt if needed), enter a name, then save. After save the app shows a success confirmation and returns to home where the new item appears in Quick access and in the item detail view. Steps: Empty state / Entry → New Item — Choose purchase date → New Item — Enter cost (empty) → New Item — Enter cost (value entered) → New Item — Name & images screen (keyboard open) → New Item — Confirmed name preview → Add photo — camera permission → Add photo — choose source → Image attached preview → Save item → Home updated — Quick access → Open item detail → Item detail — edit / add more **Runey: Receipt Creation** (web, 7 screens, Business, Productivity, Project Management) [Refero](https://refero.design/flows/11707) > Happy path begins on the Invoices dashboard, where the user opens the New menu and chooses to create a receipt. The user enters the receipt editor, selects a client, picks or confirms dates, adds a line item, reviews payment details, and creates the receipt. The flow ends on a completed receipt view with a success confirmation. Steps: Invoices Dashboard → Choose Receipt Option → Receipt Editor → Client Selection → Date Selection → Line Item Entry → Created Receipt **GlossGenius: Send sale receipt to client** (web, 3 screens, Business) [Refero](https://refero.design/flows/9964) > Happy path: A logged-in user (desktop) opens the Clients section, selects a client with a recorded sale, views the sale details in a right-hand panel, and sends a receipt to the client via a contact method (text or email). Steps: Clients list & Sale details → Choose contact method → Receipt sent confirmation **PayPal: Invoice search and results** (web, 4 screens, Finance) [Refero](https://refero.design/flows/7401) > User is on the Invoicing area of a PayPal account looking at the Invoices tab. The interface shows a search bar with a persistent filter (Creation date: Past 180 days), an invoices table (columns: Customer, Invoice no., Invoice date, Due, Status, Viewed, Amount), and a Create New button. The happy path: user searches for an invoice by number or customer name, the system presents autocomplete suggestions while typing, and then returns matching rows in the table (or a no-results empty state). Steps: Invoices list (default) → Search with no matches → Autocomplete suggestions → Search returns matches **Mercury: Manage Expense Categories** (web, 4 screens, Business, Finance) [Refero](https://refero.design/flows/11077) > Happy path: An authorized user opens Settings → Company → Categories to view and manage the list of transaction categories. From the list they click Create category, an inline row appears at the top for entering a category name and selecting applicability across Reimbursements / Card spend / Other. After providing values they save; the UI shows a confirmation toast and the list returns to its normal state with the new item present. Steps: Categories list (overview) → Open inline create row → Provide name and applicability → Save and confirm creation **Runey: Payment Template Setup** (web, 4 screens, Business, Productivity, Project Management) [Refero](https://refero.design/flows/11683) > Logged-in user opens Settings > Payments to manage reusable payment templates for invoices, quotes, and receipts. The main entry point is the Payments section in the settings sidebar. The user can create a new template from an empty state, fill in template name, choose a payment method, add payment detail fields, and save it. After saving, the template appears in the list and a confirmation toast is shown. The user can also toggle whether payment details appear on documents. Exit point is the updated Payments settings screen with the new template visible. Steps: Empty Payment Templates → Template Form → Filled Template Form → Saved Template List **Acctual: Create bill from invoice** (web, 7 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8844) > Happy-path: A signed-in user navigates to the Bills list, clicks Add bill, uploads an invoice (drag or browse), and the system extracts invoice data into a right-side New bill panel. The user reviews and edits the prefilled fields (contact, invoice #, dates, amount, currency, memo, asset), then marks the bill 'Ready for payment'. The drawer closes and the bill appears in the Bills list under the 'Ready for payment' tab (count updates). Entry: Bills overview. Exit: Bill recorded and visible in Ready for payment. Assumptions: user is authenticated, has access to the Bills section, and has contacts/assets available to select. The screenshots are a desktop/larger viewport layout with a right-side drawer for bill creation. Steps: Bills overview (empty) → New bill — upload area and form → Invoice preview with auto-extracted fields → Select contact/vendor → Choose due date → Finalize bill details and ready action → Bills overview (Ready for payment) **Understory: Manage categories** (web, 4 screens, Business, Scheduling Tools) [Refero](https://refero.design/flows/12710) > Logged-in desktop admin interface on the Experiences section. The user starts on an empty Manage categories screen, opens a modal to create a category, enters a category name, and saves it. After saving, the modal closes and the new category appears in the list with an experience count and overflow actions. Steps: Empty Categories State → New Category Modal → Category Name Entered → Categories List **Mercury: Remove transaction attachment** (web, 4 screens, Business, Finance) [Refero](https://refero.design/flows/11034) > Happy path: A user opens a bank transaction in the Transactions list, reviews the transaction detail side panel, removes an attached invoice file, confirms the removal in a modal, and returns to the transactions list with the attachment removed. Steps: Open transaction detail → Open attachment actions → Confirm removal → Attachment removed and panel updated ##### Task: eye test vision **Google Translate: Lens Text Translation** (ios, 2 screens, Reference, Utility) [Refero](https://refero.design/flows/12142) > A user is viewing a document through Google Lens with text recognized and translated from English to Spanish. The screen shows the live camera/document view with translated text overlays, a language selector at the top, and a bottom sheet offering actions for the translated text. The user is assumed to be on a mobile device, already inside Google Lens, with camera access granted. The flow starts on the translated document view and ends after the user dismisses or responds to the feedback prompt. Steps: Translated Document View → Feedback Confirmation ##### Task: file sharing storage **WhatsApp: Export Chat and Share** (ios, 5 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12669) > Happy path for exporting a WhatsApp chat and sharing the exported ZIP file through iOS Messages. ##### Task: fraud alert suspicious transaction **UGLYCASH: P2P USD Send** (ios, 8 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/11227) > Happy path: A logged-in user with limited balances starts from the Portfolio tab, taps Transfer, reviews the 'Move Money Globally' introduction and taps Continue, enters an amount on a full‑screen numeric keypad, chooses a source balance (or is prompted to verify identity), then selects a recipient by searching for a member tag/name. Steps: Portfolio - Transfer CTA → Transfer Intro → Amount Entry - Keypad (initial) → Amount Entry - Insufficient Feedback → Select Balance Modal → Recipient Search - Empty Recents → Search No Results → Search Retry / Keyboard Input **Column: Simulate a Transfer in Sandbox** (web, 4 screens, Business, Finance, Productivity) [Refero](https://refero.design/flows/1935) > Happy path: a signed-in operator opens the Transfers list, clicks the Simulate button to open a Sandbox modal, fills the required fields (account, transfer type/direction, amount) and sends a simulated ACH/Wire. The modal shows a note that simulated ACHs may take 5–10 minutes. After sending, the modal closes and the Transfers table refreshes to show the new simulated transfer(s) with their Type, Status, Direction and Created date. Steps: Transfers list → Open Sandbox modal → Configure simulation details → Transfers list updated **Artsy: Create artist alert with detailed filters** (ios, 18 screens, Marketplace) [Refero](https://refero.design/flows/7597) > Happy-path: A signed-in user on an artwork detail taps Create Alert, configures filter criteria (medium, rarity, price, size, ways to buy, color, etc.), reviews the selected filters, sets notification preferences, and saves the alert. Steps: Artwork Detail → Create Alert (start) → Filters — Medium (initial) → Filters — Medium (multi-select) → Filters — Rarity + Price (open) → Price Range — set numeric → Filters — Size & Ways to Buy → Filters — Custom Size (open) → Filters — Custom Size inputs → Filters — Enter numeric values → Filters — Color & Ways to Buy selection → Set Filters (apply) → Create Alert — review filters → Create Alert — submit → Saved Confirmation → Alerts list (Favorites tab) → Edit / Remove (implicit) → End state — alert visible in list **HYPE: Upcoming Product Details** (ios, 2 screens, Lifestyle, News) [Refero](https://refero.design/flows/12252) > This is a product detail screen for an upcoming sneaker release, shown from the Drops tab. The user enters from a product listing or drop navigation, reviews the item name, release date, image, and product details, then can tap Notify Me to subscribe for an alert. After tapping, the button changes to Notification On, confirming the alert is enabled. Steps: Review Drop Details → Notification Enabled **Tinder: Help Center — Account Appeals** (ios, 6 screens, Dating & Relationships) [Refero](https://refero.design/flows/3473) > Happy-path: From the in-app Settings screen the user taps Help & Support, which opens Tinder’s Help Center in a web view. From the Help Center landing the user either uses the search box or selects a category (Safety & Reporting). Inside that category they open the “Account Violations & Appeals” section and read an article (“My account was warned”). The article contains links to the Appeal Center and may surface an embedded submission form (text input + Submit). Steps: Settings — Help & Support → Help Center Landing → Safety & Reporting — Category List → Account Violations — Article List → Article — Read Guidance → Appeal / Feedback Submission **Ground News: Delete single alert** (ios, 4 screens, News) [Refero](https://refero.design/flows/7576) > Happy-path: a signed-in mobile user opens the Alerts tab, locates an alert they want to remove, uses a left-swipe gesture on the alert row to reveal action tiles (Share, Delete), selects Delete, confirms in a modal, and the alert is removed from the list. Entry: user is on the Alerts tab in the app's bottom navigation. Exit: alert row is removed and the Alerts list updates. Assumptions: iOS-style mobile view, bottom navigation visible, swipe gestures supported, user has permission to manage their alerts. Steps: Alerts list (default) → Swipe reveals actions → Initiate delete (destructive) → Confirm delete and update list **Apple Books: Gift Card Redemption** (ios, 4 screens, Books) [Refero](https://refero.design/flows/11945) > A modal redemption screen is opened from a purchase or account area. The user can either scan a gift card/code with the camera or enter a code manually. In this sequence, the user chooses manual entry, types a code, taps Redeem, and receives an error dialog stating the code is invalid. Assumptions: the user is on a mobile device and has access to the redemption screen; no login state is shown in the screenshots. Steps: Redemption Options → Manual Code Entry → Submit Code → Invalid Code Alert **komoot: Report tip comment** (ios, 4 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8373) > Happy path: From a tip/comment list the user taps the inline "Report" control for a specific tip. A bottom sheet titled "Submit report" slides up where the user picks one reason from a list (radio choices). The Submit button is disabled until a reason is chosen. After choosing a reason the user taps Submit, and a confirmation screen thanks the user and offers a "Got it" button to dismiss. Steps: Tip comment — trigger Report → Report reasons — no selection → Report reasons — select reason → Submission confirmation **Ground News: Alerts Inbox Cleanup** (ios, 3 screens, News) [Refero](https://refero.design/flows/7577) > Happy path: A signed-in mobile user opens the Alerts tab, sees one or more alert items, uses the trash (bulk delete) action in the top-right to remove all alerts, confirms the destructive action in a modal, and is returned to an empty-alerts state with a message indicating no new alerts. Entry: Alerts tab is active. Exit: Alerts list is cleared and empty-state message shown. Steps: Alerts list → Delete confirmation → Empty alerts state **PayPal: Manage Contact Block** (ios, 3 screens, Finance) [Refero](https://refero.design/flows/7243) > Happy-path: A logged-in user views a contact's profile (John Doe) from the Send/Request flow, chooses to block that contact, confirms the block in a modal, and the app shows a brief confirmation toast then returns to the Send/Request contact picker with suggested contacts. Entry: user tapped a contact to view details (from Send/Request). Exit: contact picker / suggested contacts view with a confirmation banner. Steps: Contact profile → Block confirmation modal → Block confirmation and return to contact picker **TikTok: Report account from profile** (ios, 6 screens, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/2760) > Happy-path: A logged-in iOS user viewing a public profile (hello_dongwon) opens the Share / More actions sheet, selects the Report action, chooses 'Report account' → picks a specific reason (Other) → reads the reason detail and taps Submit. Entry: user is on a user's profile. Exit: the app shows follow-up help/confirmation content (a 'Report violations' help article). Assumptions: user is logged in, has network access, and is on the profile-level (not inside an individual video). Steps: Profile overview → Share / actions sheet → High-level report choice → Select specific reason → Reason detail and submit → Help / confirmation content **District: Store Offer Notification Subscription** (ios, 3 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12873) > A user browses the Stores section, opens a flash deal offer card, reviews the offer details and terms, and taps Notify me to subscribe for alerts. **Meta AI: Report Restricted Content** (web, 9 screens, AI Tool) [Refero](https://refero.design/flows/12757) > A logged-in user browses the Vibes media feed and reports a post they believe promotes a restricted item. The user opens the post's overflow menu, selects Report, chooses Selling or promoting restricted items, specifies Weapons, reviews the report details, and submits it. The system confirms receipt of the report and returns the user to the feed with the reported vibe hidden. Steps: Browse Vibes Feed → Open Post Actions → Choose Report Reason → Select Restricted Items → Identify Promoted Item → Select Weapons → Review Report → Confirm Submission → Return to Feed ##### Task: government service identity verification **Airbnb: ID capture and name confirmation** (ios, 6 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6419) > Happy path for collecting a driver's license (front and back) using the device camera, reviewing each captured image, and extracting the legal name. Entry: user arrives at the 'Front of ID' camera screen (user likely in the middle of identity verification). Exit: user confirms the extracted legal name. Assumptions: user is on a mobile device, has a physical ID, camera permission granted, and is following on-screen prompts. The flow enforces distance/framing, allows retakes, guides the user to flip the card, and runs OCR to prefill name fields before final confirmation. Steps: Capture front of ID (camera) → Review front of ID → Flip to back instruction → Capture back of ID (camera) → Review back of ID → Confirm extracted legal name **Airbnb: Selfie verification capture** (ios, 4 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6420) > Happy path: A user who has already submitted an ID is prompted to take a live selfie so the system can compare it with the provided ID. Entry: instruction screen that explains the selfie step. The user taps Continue, granting camera access if required, and the in-app camera opens. The UI guides the user to center their face inside a circular mask, shows a progress ring while analyzing, and automatically captures the selfie. Exit: a completion confirmation appears with a check and a CTA to acknowledge (Got it!). Assumptions: mobile device with a front-facing camera; user has already provided ID (text on the intro screen); camera permissions may be requested before capture; automatic capture with visual guidance rather than manual shutter. Steps: Intro: selfie required → Camera open with guidance → Capture in progress → Capture success confirmation **Airbnb: Event booking and verification** (ios, 17 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6417) > User books a ticketed event (stand-up comedy) from the event details screen, chooses a date/time and number of guests, goes through a compact checkout (payment method selected), then is asked to complete bank authentication and identity verification. Steps: Event details → Select a time (list) → Calendar month view → Date selected confirmation → Times list for chosen month → Scroll for more slots → Select time slot → Confirm and pay (summary) → Payment method + price details → Processing reservation → Bank action required notice → Bank authentication (Revolut) → Identity verification start → Verification deadline & pending booking → Trips — pending reservation notice → Pending event card in Trips → End state: verification required **Volv: Claim username via phone verification** (ios, 19 screens, AI Tool, News) [Refero](https://refero.design/flows/2157) > User opens the side menu and selects "Claim your username" (or Login). A login modal appears requesting a phone number with a country selector. The user chooses a country (or searches for it), enters their phone number, and the system validates the number. If valid the app runs a reCAPTCHA verification and sends an SMS verification code. The user enters the code in a dedicated modal; on success they continue to a username claim form (username + full name). The system checks availability and either shows an inline error (username taken) or accepts the choice and returns the user to the menu showing the claimed username. Assumptions: user is not authenticated at start, using a mobile device, SMS-based auth is enabled, country codes list is preloaded. Steps: Open menu / Claim username CTA → Login modal (phone entry) → Country selector list → Country search → Phone entry focused → Invalid phone error → Valid phone entered → reCAPTCHA / bot check → Verification code sent notice → Enter code modal → Partial code input → Complete code entry attempt → Authentication failed (rate/suspicion) → Alternate code attempt → Successful code validation → Enter username and full name → Username already taken error → Username submit / saving → Success and menu update **Mercury: Upload and verify government ID** (web, 3 screens, Business, Finance) [Refero](https://refero.design/flows/11021) > Happy-path: A logged-in user (top-right account shown) manually verifies identity by selecting an ID type, uploading a color photo of their ID, and submitting a selfie if required. Entry: user lands on an Identification screen with instructions and an ID-type selector. Primary actions: choose identification type, upload or drag-and-drop an uncropped photo of the ID, optionally take a live photo of self. Exit: the system either accepts the uploads or shows validation warnings/errors and guidance; in the screenshots the system returns issues that may require a resubmission or follow-up. Assumptions: user is authenticated; they have access to photos or a camera; permissions to upload files and to take a selfie are available. Steps: Identification instructions and upload → Upload preview and processing → Validation results and remediation **Mercury: Identity verification with webcam selfie** (web, 4 screens, Business, Finance) [Refero](https://refero.design/flows/11022) > Happy path: A user completes manual identity verification by selecting an ID type, uploading a photo of the ID, and capturing a live selfie via webcam. The flow opens on an Identification form where the user picks an ID type and either uploads the ID image or drags and drops it. The user clicks Take photo to open a modal camera tool, captures a selfie, confirms the photo, and returns to the form where the uploaded selfie is shown. The user can optionally add a professional profile link and save to proceed into the final terms section. Steps: Identification form → Camera modal — live preview → Photo review → Form with uploaded selfie **Airbnb: Booking payment + verification** (web, 18 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6002) > Happy path (attempt): User starts on a listing and proceeds to Request to Book. The booking flow is a multi-step, two-column layout on desktop: left column collects booking steps (choose payment timing, add a payment method, message host, review request) and the right column shows a reservation summary/price breakdown. The user selects a split payment option (pay part now / part later), adds a debit card, writes a message to the host, and submits a booking request. The system attempts payment confirmation via the user's bank (embedded bank / PSP flow). If the bank auth times out/fails, the UI surfaces inline errors and suggests alternatives. Before finalizing, the system requires identity verification — the user is guided to upload government ID, shown a deadline, and given a final holding screen telling them they have 12 hours to complete verification. Assumptions: user is signed in, on desktop, has a valid card and access to bank/PSP authentication (and email). Steps: Listing details / start booking → Choose when to pay → Add a payment method (card entry) → Payment method confirmed, move to message → Write a message to the host → Review request → Confirm card required (bank flow trigger) → Bank connection loading → Bank auth (Revolut) prompt → Bank auth timed out → Inline retry toast → Bank couldn't confirm card modal → Processing / identity verification prompt → Choose ID submission method → Verification deadline notice → Pending verification landing → Final hold state — 12 hour notice → End state / next steps **Acctual: Country selector modal** (web, 2 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8826) > A modal-based country selector used during an identity verification step. Entry: user opens the 'Select country' control from the Verify your identity flow (user appears to be signed in; left navigation shows Payments). Modal covers the main view and shows an alphabetic list of countries with flags and a search field at the top. Happy path: user types or scrolls to find a country, taps a country row, the modal closes and the selected country is applied back to the verification form. Exit points: selecting a country (not shown) or closing the modal with the Close control. Steps: Open country selector → Search and empty result **District: State Selection** (ios, 3 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12814) > A state-selection modal opens from Edit Profile. The user can search within a list of states, tap a matching state to choose it, or dismiss the modal with the close icon. Steps: Open state picker → Search input state → Matched state result **Fingerprint: Identification events export** (web, 3 screens, Business, Development Tools) [Refero](https://refero.design/flows/11155) > Happy path: a signed-in user viewing the Identification events list requests an export of events for a specified date range. Steps: Events list → Export events modal → Export confirmation **District: Edit Profile and Select State** (ios, 4 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12813) > Logged-in user edits profile details on a mobile screen, including personal, optional, and invoice-related fields. The flow starts on the Edit Profile screen, continues into a state selector modal where the user can browse or search states, and ends back on the profile screen with the chosen state filled in and the option to save changes via Update profile. Steps: Edit Profile → State Picker → State Search → Profile Updated State **Meta AI: Meta AI Email Sign-In** (ios, 7 screens, AI Tool) [Refero](https://refero.design/flows/13401) > A signed-out user opens Meta AI and chooses to get started. The system requests a mobile number or email address, then sends a six-digit verification code to the submitted email. The user enters the code, authentication completes, and the user is taken through a brief loading screen to the Meta AI home screen. The screenshots show the email-code happy path; no device setup is selected. ##### Task: grant application **Fuse: Token swap from Wallet** (ios, 16 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8739) > Happy-path: From the Wallet home the user opens the global action (+) menu, chooses Swap, picks the asset to pay (SOL) and the asset to receive (USDT), enters a small SOL amount using the numeric keypad, reviews the quote and swap details (slippage, route, price impact, fees), adjusts slippage settings if desired, confirms the swap with biometric authentication, sees a success/processing state, and returns to Wallet with balances updated. Steps: Wallet Home → Quick Actions → Select Pay Asset → Swap Entry - Currency Picker → Receive Token List → Search Token → Swap Entry - Keypad → Review Button → Review Details → Slippage - Dynamic → Slippage - Manual (low) → Slippage - Manual (updated) → Updated Review → Execution & Confirmation → Wallet Updating → Final State **GoFundMe: Fundraiser Story Review** (web, 3 screens, Fundraising) [Refero](https://refero.design/flows/11742) > A multi-step fundraiser setup screen where the user writes or edits their personal story for donors, reviews AI-generated improvements, and accepts the suggested copy before continuing. The entry point is the story-writing step (step 4 of 6). The user is assumed to be in a logged-in fundraising setup context with permission to edit the story. The exit point is proceeding to the next setup step after reviewing or applying the enhanced story text. Steps: Story Draft Editor → Suggested Story Review → Updated Story Editor ##### Task: health onboarding profile **GoFundMe: Profile Sharing** (web, 3 screens, Fundraising) [Refero](https://refero.design/flows/11725) > Logged-in user profile page on GoFundMe with a share-profile modal opened from the profile screen. Entry is the profile page; exit is either closing the modal or copying the profile link. Steps: Profile Overview → Share Profile Modal → Copied Confirmation ##### Task: health record medical history **Anam: Session History Export** (web, 2 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12949) > This is an authenticated analytics/admin interface on the Sessions section, with the History tab selected. The user can review a paginated table of recorded sessions, filter or search the list, and export the sessions as CSV. The flow starts on the Sessions history screen and ends after the user downloads the session list, confirmed by a toast notification. Steps: Review Sessions → Export Sessions **Pi: Export personal chat history** (web, 2 screens, AI Tool) [Refero](https://refero.design/flows/4708) > Happy path: a logged-in user navigates to Profile > Manage history on a desktop browser, sees an explanation about portability and a primary CTA to download their personal conversation (PI) history, taps the CTA to start export, and the UI updates to a "Retrieval initiated" state showing that the export is being prepared and an estimated ready date. Steps: Manage history — Download prompt → Retrieval initiated — In-progress status ##### Task: insurance claim health **Wabi: Service Access Settings** (ios, 3 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12043) > This mobile settings experience lets the user enable or disable data connections for several services. Steps: Service Toggles → Health Permission → Settings Updated **Haptic: Enable Health Access** (ios, 3 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/2128) > User is prompted during onboarding to allow the app (Haptic) to access their iPhone Health data so the app can import activities (sleep, steps, weight). Steps: Onboarding: Health Access Prompt → System Health Access Modal — Initial → System Health Access Modal — Enabled ##### Task: international transfer currency exchange **PayPal: Peer Payment & Request** (ios, 7 screens, Finance) [Refero](https://refero.design/flows/7236) > Happy path: A logged-in user opens a contact profile, taps Send to start a payment/request. The user sees the Amount screen (sender currency = GBP) and is shown a notice that the recipient will receive USD. The user enters an amount using the on-screen keypad, checks the converted amount and rate, optionally opens the currency picker to change recipient currency, then either Request or Send. Choosing Request opens a confirmation bottom sheet; confirming creates a request and shows a success screen. Steps: Contact profile (start) → Amount screen — recipient currency notice → Amount screen — initial state → Amount entry with keypad → Amount entered — real-time conversion → Currency picker → Request confirmation and success **WhatsApp: Message Translation** (ios, 4 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12640) > Happy path for translating a chat message. **Wise: International Transfer Quote** (web, 3 screens, Finance) [Refero](https://refero.design/flows/354) > Happy path: An unsigned (logged-out) desktop visitor lands on the hero section to get a quick quote for sending money internationally. The user enters the amount to send, opens a currency selector for source/destination, uses the inline search to find a currency, selects it, reviews the recipient amount, estimated arrival time and savings, then can choose 'Compare price' or 'Get started' to proceed. Steps: Hero: Transfer Quote Widget → Currency Selector — No Match / Request Prompt → Currency Selector — Search & Select **PayPal: Request money from contact** (web, 6 screens, Finance) [Refero](https://refero.design/flows/7389) > Happy path: A logged-in user navigates to Send and Request, picks a contact, composes a money request (entering a send amount and optional note), selects which currency the recipient will be paid in, reviews the request, and confirms it. Steps: Select contact → Compose request - empty → Enter amount and note → Choose payout currency → Review request → Confirmation **Mercury: Internal funds transfer** (web, 7 screens, Business, Finance) [Refero](https://refero.design/flows/11044) > Happy path: A logged-in user starts from the dashboard, opens the Transfer funds modal, enters an amount, selects a source and a destination account (different accounts), reviews the details, confirms the transfer, and sees a success confirmation with an estimated arrival window. Steps: Dashboard (start) → Transfer form - initial → Select source account → Same-account validation → Select destination account (valid) → Review transfer → Transfer confirmation **Column: Book Transfer Creation** (web, 4 screens, Business, Finance, Productivity) [Refero](https://refero.design/flows/1933) > Happy path: A logged-in user in the sandbox environment navigates to Transfers, opens the Create Transfer control, selects a transfer type (Book Transfer), completes required fields (sender account, receiver account, amount) and submits the form. Steps: Transfers List → Create Transfer Menu → Create Transfer Form — Book Transfer → Validation Error — Insufficient Funds **Wealthsimple: Contact & Transfer Request** (web, 4 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8493) > Happy path: A desktop (likely not-logged-in) user visits the Contact Us hub, selects the option to transfer an account, completes the 'Get in touch' transfer form with personal and financial details, submits the form, and receives a confirmation/thank-you screen with suggested help-centre articles and contact options. Steps: Contact hub (support landing) → Get in touch — Transfer form (empty state) → Get in touch — Form filled → Submission confirmation (Thank you) **Wealthsimple: Account transfer onboarding** (web, 23 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8502) > Happy-path flow for transferring an external investment account into the user's Wealthsimple account. Steps: Registration success / CTA → Transfer intro / Select account type → Account type list → Account type chosen / Next → Select originating institution → Institution selected → Select institution division → Connect vs enter manually → Enter account number → Account number entered → Transfer details form (initial) → Choose destination account → Select transfer method → Provide estimated value & USD flag → USD account subscription prompt → Optional statement upload → Statement uploaded confirmation → Review summary and signature (empty) → Signature provided / Submit → Transfer submitted confirmation **Cashcaded: Currency Search** (ios, 3 screens, Finance) [Refero](https://refero.design/flows/14117) > The user enters a currency-selection interface with USD already selected. They can browse the available currency list or use the search field at the bottom to filter currencies. In the shown path, the user enters a numeric query that produces no visible results, clears or replaces it with “Dollar,” and sees matching currencies such as USD, AUD, CAD, HKD, and SGD. The flow begins in a currency picker and ends with filtered search results; no currency selection after filtering is shown. The user appears to be on a mobile device, with the on-screen keyboard open while searching. Login status and permissions are not visible or required by the interface shown. Steps: Currency List → Numeric Search → Filtered Results **PayPal: Currency Selection** (ios, 3 screens, Finance) [Refero](https://refero.design/flows/7234) > User opens a modal titled "Choose a Currency" from the Amount screen on an iOS mobile device. The modal contains a prominent search input, a scrollable list of currencies (each with flag and name), and shows the current selection with a check icon. When the search field is focused the on-screen keyboard appears. As the user types the list is filtered to matching currency names. Steps: Open Currency Selector & View List → Enter Search Query → View Filtered Result & Select **Blackbird: Rewards Wallet Overview** (ios, 2 screens, Food & Drink) [Refero](https://refero.design/flows/9018) > Happy path: a logged-in user opens the Wallet area to view their rewards balance, check recent activity, and see profile completion tasks. The main card shows the user's balance and a two-position currency toggle ($FLY / USD). The user can switch the toggle to view the balance in either unit (first screenshot shows the $FLY view, second shows USD). From this screen the user can also access profile completion steps (add payment method, upload profile photo), view recent activity details, and use bottom navigation (Wallet, Map, Check In). Steps: Wallet — Balance in $FLY → Wallet — Balance in USD **WhatsApp: Chat Message Translation** (ios, 8 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12639) > A user in an existing chat wants to translate an incoming English message. The user long-presses the message, chooses Translate, selects French as the target language, and waits while the required English and French language resources download. The translation result is then displayed alongside the original message. The flow assumes the user is already inside a chat and has access to the message translation feature. **Google Translate: Source Language Picker** (ios, 4 screens, Reference, Utility) [Refero](https://refero.design/flows/12132) > A modal language picker is opened to choose the source language for translation. The user can dismiss the picker, search for a language, or browse an alphabetized list. Some languages appear as already available, while others show a download icon indicating they may need to be downloaded before use. The flow starts when the language selector opens and ends when the user selects a source language or cancels the picker. Steps: Language List → Search Activated → Search Input → Search Result **Google Translate: Text Translation Input and Result** (ios, 4 screens, Reference, Utility) [Refero](https://refero.design/flows/12134) > A mobile translation interface where the user can enter source text by typing, pasting, or handwriting, then view the translated result between English and Thai. Entry starts on an empty input screen with the source/target language pair visible and the on-screen keyboard open or available. The user can switch input methods, enter text, and proceed to a translated result view. Steps: Text Entry → Keyboard Mode Switch → Handwriting Input → Translation Result **Google Translate: Word Translation Details** (ios, 3 screens, Reference, Utility) [Refero](https://refero.design/flows/12126) > A user views a translation result for an English word translated into Spanish, with pronunciation, copy, save, feedback, and example sentence options. The flow starts on a word detail screen and can end after the user rates the translation, with confirmation shown. Steps: Translation Detail View → Rate Translation Prompt → Feedback Submitted **Google Translate: Dictionary Translation Entry** (ios, 2 screens, Reference, Utility) [Refero](https://refero.design/flows/12125) > A bilingual dictionary/translation detail screen for the English word “butterfly,” showing its pronunciation, Spanish translation “mariposa,” example sentences, and actions like listen, copy, bookmark, rate, and add a new translation. Steps: Translation Detail → Saved Translation State **Acctual: Invoice Currency Selection** (web, 4 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8838) > Happy path: A logged-in user opens the Create Invoice modal on desktop, enters line items and selects the invoice currency from a dropdown. The modal shows a two-column layout: editable invoice form on the left and a live preview on the right. The user opens the currency selector (default USD), searches for EUR, selects it, and the currency badge updates in the form (and preview reflects the new currency symbol). Steps: Invoice Details (default USD) → Open Currency Selector → Search Currency (type filter) → Select Currency and Update **ROVR: Currency Selection** (web, 4 screens, Music Streaming) [Refero](https://refero.design/flows/14110) > A desktop storefront starts on the main catalog landing screen with the currency control visible in the header. The user opens the currency selector, searches for “United,” reviews the filtered country options, and selects a currency. The selector closes and the header currency indicator updates to the dollar symbol. The flow assumes a desktop user, with no visible sign-in requirement or special permissions. It begins on the storefront landing screen and ends with the selected currency applied to the storefront. Steps: Storefront Landing → Open Currency Selector → Search Currency Options → Apply Currency **Polar: Product Creation - Pricing & Currency** (web, 3 screens, Business, Finance) [Refero](https://refero.design/flows/11387) > Happy path: an authenticated user creates a new product from the admin sidebar, enters a product name, chooses a pricing model (One-time purchase or Recurring subscription), and adds currencies via a searchable dropdown. Entry: user navigates to Products > Catalogue and clicks Create Product (form shown). Exit: the UI shows the chosen pricing model and the selected currencies in the Currencies area (final save/submission not shown). Assumptions: user is signed in with permission to create products, using a desktop/larger viewport where the left navigation is visible. Steps: Open Create Product and open currency selector → Search currencies — no matches → Search and highlight currency by name **OpenAI: Translation example — choose code sample language** (web, 3 screens, Business, Development Tools, AI Tool) [Refero](https://refero.design/flows/4182) > Happy path: A user visits the 'Translation' documentation example, reads the prompt and sample response, then changes the displayed API request code sample language (e.g., from Python to JSON) to inspect the request in their preferred format. Steps: View Translation example → Open language selector → Select JSON format **Farfetch: Change shipping country and currency** (web, 4 screens, Marketplace) [Refero](https://refero.design/flows/3538) > Happy-path: A signed-in desktop user browsing the Menswear sale opens the global Language/Region selector (globe icon) from the site header, searches for and selects a target country (United States), and the site updates to reflect the new region/currency. Steps: Menswear listing (entry) → Language / Region modal - country list → Search inside modal → Apply selection and confirm **Airbnb: Checkout: Confirm and pay** (web, 4 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/5988) > User is on the booking checkout (Confirm and pay) screen for an experience. Steps: Enter payment details / view summary → Open currency selector → Select currency and confirm → Review updated totals and confirm payment ##### Task: inventory warehouse **District: Save to list** (ios, 4 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12818) > User opens a save sheet for an item and can either create a new list or save it into an existing Hotlist. The sheet shows the item being saved, available lists, a note field for each selected list, and a Done button to confirm. Steps: Open save sheet → Start note entry → Enter comment → Confirm save **Adobe Photoshop: Add Adobe Stock — Search and Empty Result** (ios, 4 screens, Design & Media Editing) [Refero](https://refero.design/flows/4780) > User opens an in-app modal to add Adobe Stock assets, can either browse category tiles or open the search input to type keywords or an asset ID. The flow shown is the happy path for performing a manual search: open search, type a query/ID, submit, wait for results, then receive a 'no results' empty state. Steps: Browse Stock Categories → Enter Search Term or ID → Searching / Loading Results → No Results Found **Nudea: Stockist locator** (web, 3 screens, Marketplace) [Refero](https://refero.design/flows/7657) > Public 'Stockists' section showing an interactive store‑locator (desktop layout). Steps: Stockists landing → Search no-results → Search success and select **Artsy: Contact Gallery Inquiry** (ios, 10 screens, Marketplace) [Refero](https://refero.design/flows/7605) > Happy path: From an artwork detail screen the user taps Contact Gallery, chooses what information they want (Shipping and/or Condition & Provenance), optionally adds a location for shipping, edits or adds a custom message, and sends the inquiry. After sending the app confirms and prompts the user to add artists to their collection. Steps: Artwork Detail → Contact Composer (Initial) → View Artwork Metadata → Select Shipping (reveals location) → Add Location (sheet) → Type Location (autocomplete) → Select Suggested Location → Location Applied in Composer → Select Additional Info → Send Inquiry & Confirmation ##### Task: investing buy stock **Apple Stocks: Share Stock to Message** (ios, 3 screens, Finance, News) [Refero](https://refero.design/flows/10703) > Happy path: A signed-in iOS user views details for the Dow Jones index in the Stocks app, opens the system share sheet, chooses Messages (or a suggested contact), and is taken into the iMessage composer with a stock-preview card attached and the keyboard ready for additional text. Steps: View Instrument Details → Open Share Sheet → Compose iMessage with Stock Card **Apple Stocks: Stocks Onboarding to Watchlist** (ios, 2 screens, Finance, News) [Refero](https://refero.design/flows/10702) > Happy path: A user opens the Stocks experience, reads a short welcome/onboarding summary of features, and taps the prominent Continue CTA. The app then presents the main watchlist / market overview showing multiple symbols with sparklines, current prices and change badges. A Business News panel is accessible from the watchlist screen and is shown as a rounded bottom sheet presenting a headline and article source. Entry: initial onboarding/welcome screen. Exit: user viewing a news article in the Business News bottom sheet. Assumptions: iPhone-style device in dark mode; user has at least one watchlist or default market symbols available; network access for market data and news. Steps: Welcome / Onboarding → Watchlist / Market Overview with News **Apple Stocks: Add Stock to Watchlist** (ios, 2 screens, Finance, News) [Refero](https://refero.design/flows/10710) > Happy path: A user opens the watchlist named “Tech”, invokes the stock search overlay, types a company name or symbol, reviews matching tickers in a results list, and uses the plus control to add a symbol to the watchlist. Steps: Open Search (focused input) → Search Results and Add **Apple Stocks: Add symbol from watchlist search** (ios, 4 screens, Finance, News) [Refero](https://refero.design/flows/10714) > Happy path: A signed-in user views their Stocks "My Symbols" watchlist and opens the in-screen search to locate a company (Sony). The search overlay (with text field, microphone and the system keyboard) appears over the watchlist. The user types or uses voice to enter a company name. Matching ticker results (multiple exchange listings) are returned below the search field with a + control to add a listing. Steps: Watchlist overview → Search overlay - keyboard open → Search overlay - voice / typed input → Search results - symbol list **Apple Stocks: Create and populate new watchlist** (ios, 9 screens, Finance, News) [Refero](https://refero.design/flows/10709) > Happy path: From the Stocks overview the user opens the watchlist menu, creates a new watchlist, names it, searches for stock symbols, adds one or more symbols, and finishes on the newly-created watchlist showing the added symbols. Steps: Watchlist Overview → Open New Watchlist Option → New Watchlist Modal (empty) → Enter Watchlist Name → Search Stocks Entry → Search Results Listing → Add Symbol to Watchlist → Search Cleared / Focus Returned → Final Watchlist View **Apple Stocks: Watchlist item quick actions** (ios, 3 screens, Finance, News) [Refero](https://refero.design/flows/10704) > Happy path: User opens the Stocks list (dark theme) and scans tickers and indices. A persistent 'Business News' bottom card overlays the lower portion of the list. The user swipes left on a specific stock row (BA) to reveal contextual actions (Share, Remove); tapping outside or completing an action closes the contextual actions and returns the list to its default state. Steps: Watchlist Overview → Reveal Row Actions → Dismiss Actions / Return **Apple Stocks: Add earnings event to Calendar** (ios, 5 screens, Finance, News) [Refero](https://refero.design/flows/10716) > Happy path for adding a company earnings date from the Stocks detail screen into the device Calendar. Entry: user is viewing a single-stock detail (NKE) on an iPhone with an earnings-date CTA visible. Exit: a Calendar event is created for the earnings date and the Stocks UI updates to show the event was added. Assumptions: user has a configured Calendar account on the device; the Stocks app may request calendar permission if not previously granted; device is unlocked and online. Steps: Stock Detail — Earnings CTA → Calendar Permission Prompt → New Event — Composer Opened → Event Details — Description & Save → Stock Detail — Added Confirmation **Apple Stocks: Watchlist Search & Add** (ios, 4 screens, Finance, News) [Refero](https://refero.design/flows/10705) > Happy path: A user opens their Stocks list, taps the search icon, enters a query, and either sees no results (for an invalid numeric query) or sees matching symbols and can add one to the watchlist. Steps: Watchlist Overview → Open Search Field → No Results State → Search Results & Add **Apple Stocks: Submit Stocks Feedback** (ios, 7 screens, Finance, News) [Refero](https://refero.design/flows/10713) > Happy-path: From the Stocks list the user opens the action menu, selects Provide Stocks Feedback, completes the Apple Stocks feedback form in an in-app browser (name, email, subject, feedback type, comments, selects feature, chooses device/OS, acknowledges policy) and submits. Steps: Open Stocks menu → Feedback landing / instructions → Fill basic fields → Validation triggered (missing feature) → Open feature picker → Complete device/OS selection → Submission and confirmation **Apple Stocks: Add Symbol to Watchlist** (ios, 5 screens, Finance, News) [Refero](https://refero.design/flows/10715) > Happy-path: From the Stocks list (My Symbols) the user opens contextual actions on a single symbol (SONY), chooses the Lists action, selects an additional watchlist (Tech) in a bottom sheet, and then views the Tech watchlist which now includes SONY. Steps: Stocks list (My Symbols) → Invoke contextual actions → Open Add to Watchlist modal → Select additional watchlist → View Tech watchlist **TikTok: Product Purchase Help Chat** (ios, 6 screens, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/2779) > Happy path: A mobile shopper reviews a product detail screen (images, price, color options, size selector) and then opens the seller's customer service chat to ask how to buy the item. The virtual assistant greets the user, offers recommended questions, and the user selects "How can I buy it?". The assistant returns step-by-step purchase instructions and presents the product card with a "Send" action; the user interacts further (feedback/thumbs-up). Steps: Product detail (hero) → Product options and details → Open seller customer service → Recommended questions and product card → Assistant replies with purchase steps → Follow-up and feedback **Adobe Photoshop: Add Adobe Stock** (ios, 4 screens, Design & Media Editing) [Refero](https://refero.design/flows/4781) > Happy path: From a host editor the user opens the "Add Adobe Stock" modal, explores category thumbnails or taps the search field, enters a query (example: "flame"), reviews suggested searches while the keyboard is visible, submits the search, sees a loading state, then views a scrollable grid of image results. Steps: Browse Categories → Search Input with Suggestions → Results Loading → Results Grid **Adobe Photoshop: Add stock asset to new canvas** (ios, 7 screens, Design & Media Editing) [Refero](https://refero.design/flows/4825) > Happy path: A signed-in user starts on the Home screen, chooses to create a new blank canvas, picks a size preset, opens Adobe Stock, browses categories, previews an asset, confirms adding it, and then places/edits that asset on the canvas. Steps: Home - Start new canvas → Choose canvas size → Open Adobe Stock (loading) → Browse stock categories → Preview asset and confirm → Asset placed on canvas → Return to Home with saved file **Podimo: Premium Subscription Purchase** (ios, 3 screens, Entertainment) [Refero](https://refero.design/flows/12256) > User lands on a subscription upsell screen for exclusive content, then initiates an Apple App Store subscription purchase. The flow starts from a marketing paywall with a free trial offer and ends with a successful purchase confirmation. Steps: Subscription Offer → App Store Confirmation → Purchase Complete **Adobe Photoshop: Insert Stock Asset** (ios, 7 screens, Design & Media Editing) [Refero](https://refero.design/flows/4779) > Happy path: From an open image editor canvas the user opens the Add image sheet, chooses Adobe Stock, searches for a flame asset, previews a result, confirms insertion, waits for the asset to download, and the asset is added as a new layer on the canvas. Steps: Open Add image sheet → Browse Adobe Stock → Enter search term → View search results → Preview and confirm asset → Download asset → Asset inserted as new layer **Wealthsimple: Enable options trading** (web, 13 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8512) > User is on a stock's options chain but does not yet have options trading enabled. From the options chain the user is invited to start options trading. The flow guides the user through a short onboarding sequence: an initial CTA, an investment experience questionnaire, an options risk disclosure acknowledgement, a review of which accounts to enable, two-factor authentication verification, and an optional educational module. Entry: user clicks a CTA from the options chain when options are not enabled. Exit: the user returns to the options chain with options trading enabled (UI now displays strategy builder and active buy/sell controls). Assumptions: the user is authenticated (signed in) and has at least one tradable account; the user has a working authenticator app for 2FA. Steps: Options chain (disabled state) → Prompt to explore options → Start options trading CTA → Investment experience (initial) → Investment experience (completed) → Options risk disclosure (read) → Options risk disclosure (acknowledged) → Review account setup → Re-check risk acknowledgement (if revisited) → Two-factor verification modal (empty) → Two-factor verification modal (code entered) → Optional education prompt → Options chain (enabled state) **Wealthsimple: Stock timeframe selector** (web, 2 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8514) > Happy path: A signed-in desktop user views a single-stock detail screen and changes the chart timeframe from intraday (1D) to a multi-year view (5Y). Steps: Stock detail — intraday chart → Stock detail — multi-year candlestick **Stocktwits: Search, Add to Watchlist, and View Sentiment** (web, 4 screens, Finance) [Refero](https://refero.design/flows/7219) > Happy path: a logged-in desktop user opens the Market Overview/dashboard, uses the global search to find a ticker (AAL), adds the ticker to their watchlist from the search results, then opens the ticker detail page to view sentiment metrics. Steps: Market Overview (Dashboard) → Search Modal — Suggestions → Search Results — Add to Watchlist → Ticker Detail — Sentiment Tab **Wealthsimple: Add money to investment account** (web, 8 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8505) > Happy-path: A logged-in user navigates to Move funds and chooses Add money to deposit funds from a linked bank account into an investment account. They pick a source account, pick a destination investment account, enter amount, choose a start date and optional recurrence, review a confirmation screen, then submit. The flow ends with a processing/confirmation modal and the system will later update the user's activity feed. Assumptions: user is signed in on a desktop device, has at least one linked bank account and at least one investment account, and has permission to initiate deposits in CAD. Steps: Move funds landing → Add money modal — initial → Select source account → Select destination account → Enter amount and frequency → Ready to continue → Confirm deposit details → Processing / acknowledgement **Stocktwits: Poll Reply with Price Prediction and Media** (web, 2 screens, Finance) [Refero](https://refero.design/flows/7214) > Happy path: A logged-in desktop user opens a poll/conversation modal and composes a reply. From the composer they optionally create a price prediction (symbol, target price, date, sentiment) and attach a media file. Steps: Compose reply with price prediction → Attach media and confirm upload **Stocktwits: Poll Reply with Price Prediction** (web, 11 screens, Finance) [Refero](https://refero.design/flows/7213) > Happy path: A user views a poll post in a thread, opens the poll in a modal, composes a reply, adds a structured price prediction (symbol, target price, target date), optionally attaches an image, and posts the reply. Entry: user is on a poll thread (logged in). Exit: reply appears inline in the thread with attached media and a price prediction card. Assumptions: user is logged in, has permission to reply, and the site supports in-reply prediction creation and media uploads. Steps: Thread view with poll → Poll modal opened → Compose reply text → Open create prediction widget → Open date picker → Select target date → Prediction panel populated → Upload media preview → Attached prediction + media visible → Submit reply → Reply displayed in thread **Cake Equity: Issuing shares to a shareholder** (web, 8 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/6132) > Happy path: A logged-in company user with permissions navigates from the Cap Table overview to the Shareholders area, opens the Issue shares form, selects an existing shareholder, fills transaction details (share type, issue date, quantity, amounts), configures optional vesting, and submits. The flow entry point is the Cap Table (Overview) screen; exit point is the Shareholders list updated to show the newly issued shares. Assumptions: desktop viewport, user is authenticated and has rights to issue shares, the shareholder already exists in the system. Steps: Cap Table overview → Open Shareholders or start issuance → Issue shares — select shareholder → Fill transaction details → Pick issue date (date picker) → Confirm date selection → Configure vesting and review → Submit and verify Shareholders list **Wealthsimple: Inspect stock chart with indicators** (web, 3 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8513) > User opens the NVIDIA (NVDA) instrument detail and inspects its price chart. The sequence shows an intraday (1D) chart, the user opening the Indicators control and selecting indicators, then viewing the chart re-rendered for a multi-year (5Y) range with active indicators (moving averages, RSI, volume). Entry: NVDA stock detail/chart. Exit: chart updated and displaying selected indicators for the chosen time range. Steps: Intraday price (1D) view → Open Indicators and select range → 5Y chart with active indicators **Cake Equity: Add transaction comment** (web, 4 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/6135) > Happy path: A signed-in user viewing a shareholder (John Doe) on the Cap Table opens the transaction comments modal for a specific transaction, types a new comment, and submits it. Steps: Open shareholder transaction list → View empty Transaction Comments modal → Compose new comment → Submit and confirm comment **Stocktwits: Sign In and Access Investor Feed** (web, 4 screens, Finance) [Refero](https://refero.design/flows/7222) > Happy path: A signed-out visitor lands on the Stocktwits marketing homepage, chooses to sign in, completes credential-based authentication in a split-screen login modal, and is redirected to their authenticated feed/dashboard. Steps: Marketing Homepage (Hero) → Login Modal (Empty) → Login Modal (Credentials Entered) → Authenticated Feed / Dashboard **Butter: Insert stock clip into editor** (web, 4 screens, Design & Media Editing) [Refero](https://refero.design/flows/3043) > Happy-path summary: The user is in a new project editor with a persistent left-side media library and a multi-panel canvas + timeline on the right. They browse categories or run a search in the library, review thumbnails, then add a selected clip to the timeline or canvas. Steps: Open media library (default view) → Select category results → Search yields empty placeholder → Search returns matches **Modyfi: Add Stock Image to Canvas** (web, 5 screens, Design & Media Editing) [Refero](https://refero.design/flows/480) > Happy path: From a new project canvas the user opens the global Add menu, chooses the "Add stock image" action, opens the Asset Browser, selects a stock thumbnail, and places that image into the project canvas. Steps: Open Add Menu → Select 'Add stock image' → Asset Browser — Empty State → Asset Browser — Thumbnails Visible → Canvas Selection & Inspector **Stocktwits: Earnings Call Chat Interaction** (web, 5 screens, Finance) [Refero](https://refero.design/flows/7218) > Happy path: A logged-in user on the StockTwits market overview opens the left-side earnings-call card, interacts with the Pre-Earnings Chat by typing a message, opens the emoji picker, browses emoji categories, (optionally) inserts an emoji, and sends the message. Steps: Open event chat → Compose message → Open emoji picker → Browse emoji categories → Send message and return to chat **AVNIER: Product Detail and Purchase Decision** (web, 2 screens, Marketplace, Lifestyle) [Refero](https://refero.design/flows/12001) > This is a product detail page for an outerwear item. The user enters from browsing the catalog or a direct product link, reviews imagery and product information, selects a size and color variant, and either sees a restock alert state when unavailable or adds the item to cart when available. The user is assumed to be on desktop, not logged in is acceptable, and no special permissions are required. The flow ends after the product is added to cart or after the user chooses to be notified when the item is available. Steps: Unavailable product state → Available product state ##### Task: invoice create send **Mercury: Create and Send Invoice** (web, 21 screens, Business, Finance) [Refero](https://refero.design/flows/11047) > Happy-path: A logged-in user starts from the dashboard, opens the 'request money' wizard, selects or creates a customer, builds an invoice (add items from catalog or create new items), reviews payer page and email previews, configures invoice details (dates, payment methods, destination account), optionally schedules the send, and then sends the invoice. Steps: Dashboard / Start Request → Request type & customer selector (initial) → Open customer dropdown → Customer selected → Add customer (blank) → Add customer (filled) → Return to request type with new customer → Invoice setup (line items and destination) → Edit customer (from invoice step) → Customer updated confirmation → Add catalog item modal (blank) → Add catalog item (filled) → Single item added to invoice → Multiple items and totals → Payer page preview → Email preview → Invoice details and payment methods → Review and send (immediate) → Schedule send → Confirmation / Invoice sent **Mercury: Pay selected invoice(s)** (web, 15 screens, Business, Finance) [Refero](https://refero.design/flows/11061) > Happy-path: A user starts in the Payments inbox, selects one or more invoices, and initiates a multi-step payment wizard that runs in a split view (invoice preview on the right). The system first attempts to match the invoice to an existing recipient. The user confirms the matched recipient (or searches/creates a new one), selects payment method/account, enters amount and send date, assigns GL codes to line items, supplies wire purpose and supporting details, adds memo/email options, and reviews the summary. The flow ends when the user schedules the wire and receives a confirmation screen. Assumptions: user is signed in with permission to send money, at least one sending account is available, recipient bank details exist or can be created, and invoice PDF is available to preview. Steps: Payments inbox → Select invoices → Recipient match (split preview) → Provide feedback → Feedback submitted / continue → Recipient details confirmation → Amount & send account → Categorization (line items) → Categorization (full view) → Wire purpose → Wire details & attachments → Payment details and memo → Review → Wire scheduled (confirmation overlay) → Final confirmation / next invoice **Mercury: Set up recurring ACH payment** (web, 33 screens, Business, Finance) [Refero](https://refero.design/flows/11038) > Happy path: A user on the dashboard initiates a bill/payment, creates a new recipient (or uploads a bill to prefill), selects ACH as the payment method, enters routing/account details, categorizes the invoice into GL line items, fills optional bill details and attachments, reviews the payment schedule, and confirms a recurring ACH payment. Steps: Dashboard / Start a payment → Start: Upload or select recipient → Create recipient — blank form → Create recipient — populated basic fields → Create recipient — address entry → Recipient created — choose payment method → Validation — required payment method → Payment method dropdown options → Select ACH → Enter recipient bank details → Categorization — single item (mismatch warning) → Categorization — multiple line items → Details — payment name, memo, email receipt → Details — add multiple receipt emails → Bill details — attachment upload → Bill details — fill invoice metadata → Review — payment schedule summary → Review — payment & recipient details → Review — accounting and attachments → Success confirmation **Grok: Chat to Voice Activation** (ios, 6 screens, AI Tool) [Refero](https://refero.design/flows/6077) > User opens a conversational assistant in Chat mode, composes a message, and attempts to send it. The system intercepts the send to request age confirmation (year picker modal). After confirming age, the message is submitted, the assistant generates a reply, and an in-context tooltip invites the user to try Voice mode via a swipe gesture. The user activates Voice mode and the interface transitions to a voice session state with audio controls and a "Connecting..." indicator. Steps: Home Chat / Compose area → Compose message → Age confirmation modal → Message submitted / Generating → Assistant reply + Voice onboarding → Voice session connecting / Controls **Instagram: In-chat Order Creation** (ios, 7 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/2891) > Happy path: A logged-in user in a one-to-one chat opens the composer menu, chooses Create order, enters a monetary amount using the numeric keypad, optionally adds a description, and taps Create order. The system generates an order record and inserts an "Order created" message card in the chat with an order number, description and total price. Entry: chat conversation (user visible). Exit: chat updated with order card and "View order" CTA. Steps: Chat — open composer menu → Create order modal — initial state → Enter amount — small value → Adjust amount — intermediate value → Finalize amount — target value → Add description and submit → Order created — chat confirmation **Gizmo: Invite Contacts** (ios, 2 screens, Entertainment, Social Networking, AI Tool) [Refero](https://refero.design/flows/10009) > User is on a mobile (iOS-style) Find Friends screen that lists personal contacts with an 'Invite' action for each contact. The user can invite an individual contact by tapping its Invite button or proceed with the screen-level Next button. Tapping Invite opens the device messaging composer as a modal overlay pre-filled with the selected recipient and a share card (preview + content). Steps: Find Friends (contact list) → New Message composer (modal) **Sora: Invite a friend** (ios, 4 screens, AI Tool, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/7848) > Happy-path summary: A logged-in iOS user navigates from their profile to an Invite screen that shows a ticket-style invite with a short join code and two primary CTAs: “Copy code” and “Share invite.” Tapping Share opens the system share sheet; choosing Messages opens a prefilled message composer containing a visual preview and the invite. The user can enter a recipient and send the message (or cancel at any step). Steps: Profile — invites entry → Invite ticket — view code → System share sheet → Message composer — send invite **BeReal.: Invite Contacts by Message** (ios, 3 screens, Social Networking) [Refero](https://refero.design/flows/4917) > Happy path: A logged-in mobile user opens the BeReal 'Add or search friends' area to invite people. The screen shows a search bar, an invite-link card, a section for phone contacts and a list of suggested public accounts/brands. The user taps the INVITE button next to a contact which opens the device messaging composer prefilled with a short invite and link. After sending, the app returns to the invite list and shows a transient "Invitation Sent" confirmation. Assumptions: user is signed in, on iOS, and the app has access to contacts; the device can open the native messaging composer. Steps: Invite & Suggestions List → Message Composer (native) → Return & Confirmation **District: Invite Guests After Booking** (ios, 7 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12863) > Happy path starts on a confirmed restaurant booking screen after a reservation is successfully created. The user chooses to invite guests, either by sharing a link or selecting contacts from the device. If contact access is requested, the user grants permission, reviews synced contacts, selects one or more people, and sends invitations. The flow ends back on the booking confirmation screen showing guest invitation status and options to share again, modify, cancel, or pay the bill. **222: Invite Contacts to Other Tables** (ios, 2 screens, Lifestyle, Dating & Relationships) [Refero](https://refero.design/flows/9311) > Happy-path: From the party screen the user opens an overlay titled “invite to other tables or restaurants,” selects one or more contacts (via chips or list rows), optionally types a short message, then taps SEND INVITES. Steps: Invite modal — selection overview → Search and confirm selection **Napper: Invite caregiver to baby profile** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/8784) > Happy-path: A logged-in parent on an iPhone invites a caregiver to a baby profile. Entry: the app's "Share your baby's profile" screen. The screen presents a QR code and a prominent "Send invite" CTA. Selecting Send invite opens the iOS share sheet. Choosing a channel (e.g., Messages) opens a prefilled composer with a link preview. Exit: after the user sends the message (or cancels), the share sheet/composer closes and control returns to the profile screen. Assumptions: user is signed in, has a baby profile (e.g., Frank), device is iPhone with system share capabilities, and app can generate a shareable invite link/QR. Steps: Share invite (QR + CTA) → System share sheet → Message composer with link preview **Daze: New conversation creation** (ios, 5 screens, Social Networking) [Refero](https://refero.design/flows/9150) > Happy-path for creating a new one-to-one conversation from the Messages list on a mobile device. Entry: user is on the Messages list (logged-in). Exit: user returns to Messages list where the newly created conversation appears with a preview/timestamp. Steps: Messages list → New chat — contact list → New chat — recipient selected → Conversation composer → Messages list — updated **Bump: Invite Friends — Other Services** (ios, 2 screens, Map & Navigation, Social Networking) [Refero](https://refero.design/flows/3181) > Simple invite screen that presents three ways to share an invite link (Share sheet, Copy link, iMessage). Steps: Invite options → Copy confirmation **Stardust: Partner Invite** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/5866) > Happy-path: A logged-in iPhone user opens the app's Partner Mode card and taps the CTA to invite a partner. The app prepares an invite payload (a short code plus a deep link) and opens the iOS share sheet. The user chooses Messages, which opens the Messages composer with a prefilled invite text containing the code and link; the user can then select a recipient and send. Steps: Partner Mode Card → System Share Sheet → Messages Composer with Prefilled Invite **Drinkit: Invite a Friend Reward** (ios, 4 screens, Food & Drink) [Refero](https://refero.design/flows/11010) > Happy-path: a logged-in user opens their Profile, taps the "Invite a friend" card (Friendzone), reads the invite details in a full-screen modal, then uses the OS share sheet to send a prefilled referral message via Messages. Entry: Profile screen (user authenticated). Exit: user arrives at the Messages composer with a prefilled invite and app preview image; from there the user can send the message or cancel. Assumption: user is signed in, device is iOS, the app can open the native share sheet and supply a prefilled message + preview image. Steps: Profile (Friendzone card) → Invite details modal → OS share sheet → Messages composer (prefilled) **222: Invite Contacts to Other Tables** (ios, 5 screens, Lifestyle, Dating & Relationships) [Refero](https://refero.design/flows/9312) > Happy path: From the party overview the user taps Invite Friends, which opens a bottom sheet / drawer titled “invite to other tables or restaurants.” The user selects one or more contacts (tapped contacts become chips and entries show checkmarks), optionally types a short message in the text field, then taps Send Invites. After sending the UI returns to the party overview where invited people are shown with a ‘YOUR +1’ badge under Other Tables or Restaurants (and the dining table list is unchanged otherwise). Assumptions: user is signed in, has contacts available, is on a mobile device and can invoke the system keyboard; the invite drawer slides up over the party screen. Steps: Open invite drawer → Select recipients → Compose optional message → Send invites → Return to party overview **Programa: Create and Send Invoice** (web, 14 screens, Project Management, Business) [Refero](https://refero.design/flows/13436) > An authenticated company user starts from an empty Invoices dashboard, creates an invoice for an existing project, adds a new client when no client is available, enters the client’s contact details, configures the invoice, adds a line item, previews the generated PDF, and sends the invoice by email. The flow ends on the issued invoice detail view and an updated invoice list showing the new invoice. **Сontra: Create and send invoice with inline project** (web, 12 screens, Job Search) [Refero](https://refero.design/flows/4881) > Happy path: A logged-in user starts from the Projects list, opens the New Invoice composer, creates a new project inline, fills recipient and line item details, previews the invoice, fixes a validation issue (minimum total), then sends the invoice. The platform performs a short review (noted as up to 24 hours) and shows the invoice and project in an "In review" state. Steps: Projects landing → New invoice (draft) form → New project modal (empty) → New project modal (filled) → Enter recipient email and name → Invoice preview → Validation error on send → Discard confirmation → Return to invoice (after cancelling discard) → Invoice sent — under review modal → Project details — invoice in review → Projects list updated **GlossGenius: Invoice creation and draft save** (web, 7 screens, Business) [Refero](https://refero.design/flows/9977) > Happy-path summary: From the Invoices list the user creates a new invoice, selects a client, adds a packaged item via the Add-to-invoice modal, sets a due date, and attempts to send the invoice. The system blocks sending because payment processing is not configured, saves the invoice as a draft, and returns the user to the invoices list where the new invoice appears with status "Draft." Entry point: Invoices listing. Exit point: Invoices listing showing the saved draft. Assumptions: user is signed in with permission to create invoices, a client record exists (John Doe), at least one package/product exists (Beauty Combo), and the user is on a desktop or large-screen layout. Steps: Invoices list (landing) → New Invoice (composer - blank) → Add to invoice modal (packages tab) → Confirm selection in modal → Edit Invoice (with line items) → Send invoice? (payment setup modal) → Invoices list (draft visible) **Acctual: Invoice creation and preview** (web, 3 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8839) > Desktop invoice creation flow (happy path). Steps: Start invoice — blank item → Add first line item → Add multiple items, note and discount **PayPal: Create and send invoice** (web, 34 screens, Finance) [Refero](https://refero.design/flows/7395) > Happy path: A logged-in merchant starts from an empty Invoicing list, creates a new invoice, selects currency, adds or creates a customer (including full contact and address via a right-side drawer), adds line items, applies a discount and tax rules, reviews the rendered invoice in the Preview panel, adds business information and logo, then saves/sends the invoice and returns to the invoices list showing the new invoice. Steps: Invoicing - empty state → New invoice canvas → Choose currency → Customer dropdown → Inline new customer form → Customer added - reveal details → Customer contact details → Customer address tab → Customer selected on invoice → Open new discount modal → Discount applied → Open tax options → Preview tab - review invoice → Open business information drawer → Upload logo and save business info → Business info persisted → Final preview review → Send / save invoice (implicit) → Invoicing list - new invoice row → Final confirmation / list state **Acctual: Invoice Onboarding & Login** (web, 4 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8856) > Summary (happy path): A visitor arrives on the marketing homepage and chooses to sign in. The user authenticates (Google or email/password) on the Welcome Back screen and is directed to the Invoices dashboard. On the dashboard the user sees onboarding cards prompting them to add company details, choose where to get paid, and send an invoice — plus a table/list showing an invoice draft. Steps: Marketing landing — hero and CTAs → Welcome back — sign in options → Welcome back — credentials entered → Invoices dashboard — drafts and onboarding **PayPal: Create invoice — add due date & discount** (web, 9 screens, Finance) [Refero](https://refero.design/flows/7398) > Happy-path: A logged-in merchant creates a new invoice for an existing customer, configures line items, sets payment terms (due date), adds a reusable discount, and reviews the updated invoice summary and totals. Steps: Invoice draft — items & summary → Open invoice metadata → Select due date option → Confirm due date applied → Begin adding discount → Open new-discount panel → Fill discount details → New discount saved → Final review — discount applied **Acctual: Create and Send Invoice** (web, 35 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8830) > Happy-path: a logged-in user creates and sends a new invoice using a stepper modal on the Invoices screen. Entry: user clicks Create invoice from the Invoices list. The modal guides them through: terms (invoice number, issue/due dates), company details, selecting/adding a customer, payment details (add/select methods), and a final review/email step. A live invoice preview on the right updates as fields change. Exit: invoice is saved and appears in the Invoices list under Unpaid. Steps: Invoices landing → Terms — initial → Issue date picker open → Select specific issue date → Set due date → Company — basic → Company — additional details → Customer selection — start → Customer search options → Customer search results → Payment details — methods → Add payment method type → Bank currency selection → Add ACH bank details → Payment methods updated → Final review — ready to send → Edit recipients and message → Email template preview selected → Email preview with CTA → Back to Invoices list — result **Runey: Project Invoice Creation** (web, 14 screens, Business, Productivity, Project Management) [Refero](https://refero.design/flows/11700) > A logged-in project user starts from the “Website Redesign” project dashboard, creates a new invoice, customizes its visual presentation, reviews invoice settings, and saves the finished document. The flow begins from the project action menu and ends on the created invoice view, identified as INV-0001. The screenshots show an existing project, populated client and billing information, line items, payment details, and a total of $7,300. Steps: Project Dashboard → Invoice Editor → Style Presets → Sand Preset Applied → Customization Summary → Layout Options → Bold Header Applied → Updated Customization → Font Options → Typography Applied → Invoice Settings → Saved Invoice Preview → Created Invoice → Final Invoice View **PayPal: Create and Add Invoice Line Item** (web, 9 screens, Finance) [Refero](https://refero.design/flows/7397) > Happy-path: A logged-in seller creates a new invoice for an existing customer, adds a new product line (quantity, price, description), assigns a date and tax rate to that line, saves the line as a reusable item, and sees the invoice summary update (subtotal, tax, total). Steps: New invoice — initial screen → Items field focused → Choose item input mode → Enter item details → Select item date → Open new tax drawer → Define tax details → Tax applied and summary recalculated → Item added — collapsed view **PayPal: Add customer to invoice** (web, 8 screens, Finance) [Refero](https://refero.design/flows/7396) > Happy path for creating a new invoice and adding a customer. Steps: Open New invoice → Open Customer dropdown → Inline new customer inputs → Fill basic contact details → Open customer details slide-over → Switch to Address tab → Fill address and save → Customer attached to invoice **Mercury: Invoice customer edit** (web, 3 screens, Business, Finance) [Refero](https://refero.design/flows/11049) > A 3-step happy-path for creating an invoice where the user inspects the invoice preview, edits the selected customer's details, and returns to the invoice setup with the updated customer applied. Steps: Invoice setup (customer selected) → Edit customer form → Invoice setup with updated customer **Mercury: Invoice setup with catalog items** (web, 6 screens, Business, Finance) [Refero](https://refero.design/flows/11050) > Happy-path: A user sets up an invoice by selecting a customer, adding one or more line items (from an existing catalog or by creating a new catalog item in a modal), specifying quantities and unit prices, optionally adding a memo or attachment, choosing a destination account and toggling a secure virtual account. As items and values are entered the invoice preview on the right updates in real time to reflect line items and totals. Steps: Invoice setup (initial) → Open item selector / catalog list → Add New Catalog Item (blank modal) → Add New Catalog Item (filled) → Single item invoice preview → Multiple line items and final total **Acctual: Add and configure invoice payment method** (web, 7 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8840) > User is editing an invoice draft and adding a second payment method. Steps: Payment details — existing primary method → New method slot created → Select create new method → Creation form — choose wallet or bank → Bank account — select currency → Bank account — enter account details → Payment details — new method saved ##### Task: open bank account identity verification **Tilt: New user onboarding** (ios, 38 screens, Social Networking) [Refero](https://refero.design/flows/3063) > This is a new-user onboarding sequence that captures basic profile data, verifies contact method, invites friends, requests media permissions, and lands the user in the app camera/home screen. Entry: fresh install / first run (user is not signed in). Exit: the user arrives at the app's camera/home screen with the option to add friends. Steps: Welcome / Splash → Name — prompt → Name — typing → Name — completed → Age — prompt → Age — numeric entry → Phone — prompt with country code → Phone — suggestions & keypad → Phone — validated & Continue → Country code list → Country code search → Verification / processing → Invite friends — landing → Invite friends — search → Invite friends — add interactions → Invite via other apps → Final invite list / confirm → System permission — Camera → System permission — Microphone → Camera/home — main screen **Dropbox: Vault setup and biometric enablement** (ios, 9 screens, Business, Collaboration Tools, Productivity) [Refero](https://refero.design/flows/2467) > Happy path for setting up a protected “Vault” inside an existing account. Entry: user taps the Vault feature (onboarding card or account feature). Flow creates a vault, prompts the user to set a numeric PIN, then offers to enable Face ID for quick unlocking. Exit: user lands in the empty Vault with action tiles (Scan document, Upload photos, Create folder, Take photos). Steps: Vault introduction → Creating vault (progress) → Set PIN (empty) → Set PIN (partial) → Set PIN (multiple digits entered) → In-app Face ID offer → System Face ID permission dialog → Face ID authentication (animated) → Vault landing (empty state) **PayPal: Sign-in and biometric enablement** (ios, 7 screens, Finance) [Refero](https://refero.design/flows/7268) > Happy-path mobile sign-in for an existing PayPal account. Entry: user opens the app (logged out). Exit: user lands on the account home/dashboard (logged in). Assumptions: user has an existing account and network connectivity; device can show the on-screen keyboard and supports Face ID/biometric prompts. The flow: enter email (or switch to phone), tap Next → enter password → tap Login → brief setup/loading → optional Face ID prompt → arrive at account home. Steps: Enter email / identifier → Email typed with keyboard visible → Password entry screen (email shown) → Submit credentials / loading → Biometric prompt (Face ID) → Landing on account home → Password focused state (alternate layout) **Claude: Email + SMS verification onboarding** (web, 16 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/1813) > Happy-path sign-up and first-time onboarding for Claude. Steps: Landing — sign-in options → Email entered — CTA enabled → Magic link sent — wait for email → Enter verification code (fallback) → Code input — submitting → Email code error state → Phone number & age collection → Phone number entered — CTA enabled → Phone code send error → Enter SMS verification code → SMS code submitted → Assistant greeting — ask for name → Name entered — assistant response → Usage policy acknowledgement → Final onboarding acceptance → Main assistant screen (first session) ##### Task: pay invoice bill **PayPal: Bill Split** (ios, 13 screens, Finance) [Refero](https://refero.design/flows/7241) > Happy-path: From the app's Send & Request entry the user chooses "Split the bill" from More options, reads a short intro, then enters a total amount (currency selectable). They select participants (their own account is pre-selected), the app auto-calculates an even split, and the user can edit any person's share using an inline editor and numeric keypad. After adjustments the user reviews a final breakdown and taps Request now. The flow ends either with a success (not shown) or an error state where the request fails and the user may retry. Assumptions: user is signed in, has access to contacts/suggested users, and currency selection/keyboard are available. Device: mobile (iOS-style chrome). Steps: Send & Request entry → More options sheet → Bill split intro → Enter total amount (empty) → Amount entered → Choose participants - start → Add participants → Auto-calculated even split → Edit share (pencil) → Input new share amount → Adjusted split with reset option → Review split breakdown → Error / retry **District: Restaurant Offer and Bill Split** (ios, 10 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12866) > Happy-path restaurant offer and bill-splitting journey on a mobile device. The user enters from a restaurant detail screen, reviews offers, enters a bill amount, sees the discounted total, optionally enables bill splitting, chooses equal split and number of guests, then proceeds to the next payment/cart step after verification prompts. **District: Dining Save and Reserve** (ios, 3 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12857) > Happy path for a dining savings experience. The user starts from a dining discovery/home screen, then opens an in-context help or promo overlay that explains two options: reserve a table or pay a bill. The user can switch between the two tabs, view the instructional content, and proceed to either book a table or pay the restaurant bill. **222: Event invite & checkout** (ios, 13 screens, Lifestyle, Dating & Relationships) [Refero](https://refero.design/flows/9314) > Happy path: user opens an invite card from their invites list, reviews event details and the "Your Party" section, edits the +1 (replaces an existing +1), invites a new contact via the invite modal (with share options), sees the party list update, then proceeds to Pay & Confirm — reads payment guidelines, accepts them, and reaches the payment entry modal where they can pay via a link or enter card details. Entry: invites list. Exit: payment entry modal ready to continue. Assumptions: user is logged in, has at least one contact available, and has permission to invite/modify their +1 and complete a payment. Steps: Invites list (card overview) → Event detail — header + itinerary → Your Party overview → Open 'change +1' controls → Confirm +1 replacement → Invite modal — search & contacts → Select contact to invite → Send invites → Party list updated → Payment summary preview → Payment guidelines — full terms → Payment entry modal (link or card) → Ready to continue to payment **Instagram: Order lifecycle from chat** (ios, 5 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/2892) > Happy path: a seller sees an order card inside a 1:1 conversation, opens the Order Details screen, marks the order as paid, and the conversation is updated with an "Order paid" card. The seller can then open the updated Order Details view which now shows the Paid state and exposes the next action (Mark as shipped). Steps: Conversation – Order created card → Order Details – Waiting for payment → Order Details – Marking as paid (loading) → Conversation – Order paid card → Order Details – Paid (Next: Mark as shipped) **Acctual: Record invoice payment** (web, 7 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8842) > Happy path: A signed-in user opens the Invoices list (Unpaid tab), selects an invoice to view details in a right-hand preview panel, changes the invoice status from Unpaid to Paid, enters payment details manually (payment type, date, transaction link and memo), saves the update, and sees the invoice moved to the Paid tab with a success toast. Steps: Invoice list (Unpaid) → Open invoice detail → Open status selector → Select Paid and choose method → Enter payment details (manual) → Complete form and update → Updated invoice listed as Paid **Subo: Add Payment Method** (ios, 3 screens, Finance) [Refero](https://refero.design/flows/8592) > Happy path: A signed-in user opens the Payment Methods screen, types a name for a new payment method in a single-line input, and taps Add. The UI confirms the action by adding the new method as a list item above the input and clearing the input for additional entries. Steps: Payment Methods - empty input → Typing method name → Method added and list updated **ElevenLabs: Upgrade to Creator subscription** (ios, 5 screens, AI Tool) [Refero](https://refero.design/flows/10729) > Happy path: A signed-in mobile user on the Home screen taps the Try Creator control to view an Upgrade modal that compares Free vs Creator features. The user reviews the feature table, selects a billing option (monthly or annual via a toggle), and taps the primary CTA “Unlock Everything.” The app opens the App Store purchase sheet where the user authenticates using the device side button. After the transaction completes, the app displays a success confirmation. Steps: Home — Try Creator entry → Review features & choose billing → App Store purchase sheet → Purchase success acknowledgement → Return to subscription screen **District: Add Card for Booking Payment** (ios, 9 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12835) > The user is reviewing a movie booking and needs to select a payment method before completing payment. Starting on the booking review screen, the user opens the payment method selector, chooses to add a credit or debit card, and enters card details using the card-scanning option. The system requests camera access, scans the card, auto-fills the card number and expiry date, and enables the Add Card action. After the card is added, the user returns to the booking review screen, where the Visa card is selected and the Pay now action is available. The flow assumes the user has an active booking, is within the displayed payment timer, and has permission to add a payment card. The screenshots end before payment authorization or booking confirmation. Steps: Booking Review → Payment Method Selection → Card Entry → Camera Permission → Scan Card → Review Card Details → Add Card Processing → Updated Booking Review → Card Ready for Payment **ElevenLabs: Subscription Billing & Cancellation** (web, 6 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/8714) > Happy-path: a logged-in user on the Subscription section opens the Billing modal to manage their active Starter subscription, initiates cancellation, moves through a confirmation and feedback/retention sequence, and finishes with the subscription flagged as cancelled (scheduled to stop at end of the current billing period). Steps: Subscription overview → Billing modal → Cancellation confirmation → Cancellation reason & feedback → Retention offer → Billing modal (cancelled state) **Manus: Account creation, verification, and upgrade** (web, 23 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/10420) > Happy-path from landing to a paid Manus Pro account and into the product workspace. Entry: user starts at the Manus landing screen and chooses to sign in / sign up. The flow proceeds through sign-up via email (Cloudflare human check), set password with client-side validation, email code verification, optional phone verification (country selector and SMS code), plan selection, payment entry (7‑day free trial or paid tier), payment processing, confirmation modal, then final return to the Manus workspace with Pro access. Steps: Landing / prompt → Sign in / Sign up entry → Email entry success → Create account - password input → Password validation error → Password corrected attempt → Password accepted → Email verification input → Verification email received → Enter email verification code → Phone verification - country selector → Phone input ready → Phone code entry → Upgrade modal presented → Checkout - payment entry → Payment processing → Welcome / subscription confirmation → Workspace - initial Pro message → Workspace with cookie prompt → Ready to create **Acctual: Pay a bill from list** (web, 5 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8846) > Happy path: A logged-in user on desktop opens the Bills area, selects a vendor invoice from the 'Ready for payment' list, opens the invoice detail drawer to review the invoice PDF and editable payment fields, marks the bill as paid (or updates details) and returns to the Bills list where the bill now appears under 'Paid'. Steps: Bills list — Ready for payment → Select invoice → Open invoice detail drawer → Edit payment details / confirm → Bills list — Paid confirmation ##### Task: payroll **Mercury: Schedule Payment Amount & Recurrence** (web, 4 screens, Business, Finance) [Refero](https://refero.design/flows/11041) > This is the 'Amount' step of a multi-step payment scheduling flow where the user confirms the amount, chooses the sending account, sets the send/start date, and optionally configures a recurring payment. Entry: user arrives here after selecting a recipient. Exit: user clicks Next to continue to categorization/details/review steps. Assumptions: user is authenticated, has at least one funding account and a recipient already selected, and is on a desktop or large-screen device. The UI shows estimated arrival timing and allows toggling and configuring recurrence before moving forward. Steps: Confirm Amount & Select Source → Pick Send / Start Date → Enable Recurring Payment → Configure Recurrence End & Preview **Blackbird: Default payment preferences** (ios, 4 screens, Food & Drink) [Refero](https://refero.design/flows/9038) > Happy path: A signed-in user opens the Payment settings and launches the 'Set up Blackbird Pay' modal to configure default payment preferences. From the modal they toggle use of $FLY, open the Tip card to set a default tip percentage using a +/- control or a slider, confirm the change, and then save the updated defaults back from the Blackbird Pay modal. Steps: Open Blackbird Pay modal → Open Set default tip → Adjust tip percentage → Save updated defaults **Queue: Add Podcast Paid Support** (ios, 8 screens, Entertainment, News) [Refero](https://refero.design/flows/6232) > Happy-path: A signed-in user on iOS opens a podcast detail screen, opens the podcast options, selects "Track Paid Support", supplies an amount, selects frequency, pastes a membership link (granting iOS paste permission if needed), and saves. Steps: Podcast Overview → Open Podcast Options → Open Paid Support Modal → Set Amount (increment) → Change Frequency → Paste Link Permission → Confirm Paste and Save Link → Podcast Updated **Cashcaded: Apple Pay Tracking Setup** (ios, 4 screens, Finance) [Refero](https://refero.design/flows/14119) > A four-step instructional onboarding sequence explains how to automate expense tracking after Apple Pay purchases. The user enters from the existing expense-tracking interface and progresses through educational screens using “Show me how,” “Next,” and “Done.” The sequence assumes an iPhone user with access to Apple Pay, the Shortcuts feature, and the Cashcaded action referenced in the instructions. It ends after the user has been shown how to configure a shortcut to add transaction details, but the screenshots do not show the shortcut being created or confirmed in the operating system. Steps: Automation Introduction → Create Automation → Configure Shortcut → Transaction Details **Mozi: Add Work Entry** (ios, 5 screens, Social Networking, Travel & Booking) [Refero](https://refero.design/flows/4159) > Happy-path flow for adding a Work entry to a user's profile. Steps: Edit Profile (list) → Work editor (empty) → Enter details → Save enabled → Edit Profile (updated) **Aboard: Add employee salary** (web, 8 screens, Collaboration Tools, Paperwork) [Refero](https://refero.design/flows/756) > Happy path: From an employee's profile (Profile tab) the user adds a new salary record. Steps: Employee profile - Payroll empty → Open Add salary modal → Expand currency dropdown → Scroll currency list → Select currency and open pay period → Choose pay period → Fill salary form → Salary saved on profile **GlossGenius: Subscription cancellation and verification** (web, 19 screens, Business) [Refero](https://refero.design/flows/9985) > Happy path: a logged-in user navigates to Settings > Subscription, initiates cancellation, confirms via an in-product modal, then is redirected to a multi-step Typeform-style cancellation request. The form collects phone, email, business name, cancellation reason, optional feedback, payroll status, and a final confirmation that the user has exported needed data. After submission the user sees a thank-you screen noting the request will be processed within 48 hours. Steps: Subscription settings → Cancellation confirmation modal → Cancellation form landing → Phone number input (empty) → Phone number input (filled) → Email input (empty) → Email input (filled) → Business name input (empty) → Business name input (filled) → Pause alternative offer → Primary cancellation reason (list) → Primary cancellation reason (selected) → Open feedback input (empty) → Open feedback input (filled) → Payroll product question → Payroll question (answered) → Export data confirmation (empty) → Export data confirmation (checked) → Thank-you and next steps **GlossGenius: Product commission setup** (web, 6 screens, Business) [Refero](https://refero.design/flows/9959) > Happy-path: A logged-in manager opens Payroll > Commissions (Products tab) and creates a new product-level commission structure. From the empty-state Commissions screen they open a modal, choose the 'By Product' structure type, select products, enter a percentage rate, save the structure, confirm success, and return to the Commissions list where the new structure is visible and applied to listed team members. Steps: Commissions — Products empty state → Select commission type → Set products & rates — product list → Attempt to save — upgrade prompt → Success confirmation modal → Commissions list — created structure shown **GlossGenius: Service commission setup** (web, 6 screens, Business) [Refero](https://refero.design/flows/9960) > Happy path: A user opens the Commissions area from the left navigation and begins creating a new service commission structure. The user is presented with a modal wizard: first choose a commission type (By Member selected), then pick team members and enter percentage rates per member. After saving, a success confirmation is shown and the new commission structure appears in the Commissions list with a toast confirmation. Steps: Commissions landing → Select commission type → Select members & initial state → Enter member rates → Success confirmation → Commissions list updated **GlossGenius: Add team member and configure access** (web, 23 screens, Business) [Refero](https://refero.design/flows/9956) > Happy-path walkthrough for adding a new staff member from the Staff list. Steps: Staff list (entry) → New member form (blank) → New member form (filled) → Role & Permissions tab (default) → Exit without assigning roles confirmation → Open Assign Role panel → Select role and assign → Edit role permissions → Save permissions → Services tab (none assigned) → Schedule tab (default) → Open Manage Schedule panel → Select start time → Update end time and copy → Copy confirmation toast → Outside business hours warning → Save schedule → Completion screen → Staff list (invite pending) → Staff list (final state) **Wittl: Create job listing** (web, 35 screens, Job Search) [Refero](https://refero.design/flows/5364) > Happy-path: A logged-in recruiter starts from their inbox/dashboard and opens a modal wizard to create a new job listing. The wizard guides them through: initial job metadata (title, location, type, workplace), composing the job description with reusable sections (text/images/title blocks), building the candidate application form (fields and sections), defining hiring stages, and finally choosing whether to publish or keep the job as a draft. Entry: recruiter dashboard -> Create new job. Exit: job is created and the user is taken to the job details screen where the job status and editable fields are shown. Assumptions: user is authenticated with permission to create/edit jobs; desktop/laptop viewport (modal centered, wide editor area); a default job template and default hiring stages exist. Steps: Dashboard / Inbox → Get started — Metadata → Select job type → Select workplace → Confirm metadata and continue → Job posting — Preview / Editor → Job posting — Scroll / Edit → Insert section panel → Choose section type → Add image placeholder → Application form — default sections → Insert field panel → Select personal/contact field → Application form — field added → Insert section templates for form → Questionnaire section → Continue to Hiring stages → Hiring stages → Publish — choose visibility → Job details / Edit job **Mercury: Create and select recipient** (web, 5 screens, Business, Finance) [Refero](https://refero.design/flows/11039) > Happy path: A logged-in user (Mercury Demo account) starts a payment by either uploading a bill or selecting/creating a recipient. If the user creates a recipient, they complete a profile (Person or Business), optionally request tax documents, and enter a postal address. After creation the UI returns to the payment flow with the new recipient pre-selected and the user chooses a payment method to proceed. Steps: Start payment / choose recipient → Create recipient — Profile section → Create recipient — Profile populated → Create recipient — Address section → Recipient returned to payment — choose payment method ##### Task: policy management approval **Mercury: Payment approval review** (web, 3 screens, Business, Finance) [Refero](https://refero.design/flows/11062) > Happy path: A signed-in user with approval permissions opens the Payments area, filters to the Needs Approval tab, selects a payment row to view details in a right-side details panel, then declines the request. Steps: Needs Approval list → Open payment details drawer → Decline payment and confirm ##### Task: prescription refill repeat **Subo: Add Subscription** (ios, 4 screens, Finance) [Refero](https://refero.design/flows/8581) > Happy path for adding a recurring subscription (Apple Music) to the user's list. Steps: Blank Add form (initial) → Open Card Editor → Select Color and Icon → Complete Form & Confirm ##### Task: request money split bill **Revolut: Help centre article discovery** (web, 3 screens, Business, Finance) [Refero](https://refero.design/flows/1317) > Help centre entry → search → article view (happy path). Entry: user lands on the public Help centre (search input and topic cards visible). User can either browse topic cards (Personal / Business toggle) or type a query in the top search field. Selecting a suggested search or a card opens a detailed article view with breadcrumbs and the article body. Assumption: this is a public help/FAQ area; user may be signed in or out and is only reading help content. Steps: Help centre landing / browse topics → Search with autocomplete suggestions → Article detail view **Spoil me: Invite and Earn** (ios, 3 screens, Fundraising) [Refero](https://refero.design/flows/5430) > User views an “Invite and Earn” screen that shows a personal invite link/code, a simple earnings estimator, and controls to copy or share the invite. Entry: signed-in user opens the Invite and Earn section (iOS device). Exit: user either copies the invite link, opens the system share sheet to send the link, or cancels the modal. Assumptions: user is authenticated (invite code is shown), on an iOS device (native share sheet present), and has basic clipboard/share permissions. Steps: Invite overview and estimator → Invite code modal with actions → System share sheet **PayPal: Split a Bill** (web, 6 screens, Finance) [Refero](https://refero.design/flows/7407) > Happy path: A logged-in user navigates from the Send & Request hub to the "Split a bill" composer, enters a total bill amount, optionally includes their own share, adds one or more people to split with, and the system automatically calculates each person's share (with editable per-person inputs). Steps: Send & Request hub → Split composer — new request → Enter total and enable self-share → Edit a share amount → One participant added — auto calculation → Multiple participants — equal split shown **Mercury: Categorize Line Items** (web, 4 screens, Business, Finance) [Refero](https://refero.design/flows/11042) > User is on the Categorization step of a multi-step payment/expense entry flow (previous steps: Recipient, Amount; next steps: Details, Review). The screen lets the user choose whether to apply a single GL code to the whole amount or split the amount across multiple GL codes using separate line items. The user can add, edit, and remove line items (description, unit price) and assign a GL code for each. The UI shows the grand total from the Amount step and updates a running sum of line items; if the line items do not sum to the total, the interface highlights the discrepancy. Steps: Categorization - initial → Add first line item → Partial allocation warning → Complete allocation and proceed **GoFundMe: Donate and Share Confirmation** (web, 8 screens, Fundraising) [Refero](https://refero.design/flows/11729) > Happy-path fundraising donation journey on GoFundMe. Entry begins on a fundraiser detail view with the donor logged in, then moves into the donation checkout, donation confirmation, optional sharing, and receipt viewing. Exit returns the user to the fundraiser detail view. Steps: Fundraiser Overview → Donation Amount Setup → Currency Selection → Donation Payment Form → Donation Confirmation → Post-Donation Actions → Receipt Dialog → Fundraiser Return **Fernand: Create Split Inbox** (web, 5 screens, Business, Communication Tools) [Refero](https://refero.design/flows/11892) > Happy path for creating a new split inbox view from the Split inbox settings modal. The user starts in an existing shared inbox configuration, opens the create-inbox form, defines a name and one or more filters, chooses sorting, and submits the new inbox. The user is assumed to already be logged in with permission to manage inbox settings. The flow ends with the new split inbox added to the inbox list and visible in Split inbox settings. Steps: Split Inbox Settings → New Inbox Setup → Add Filter Condition → Filter Applied → Inbox Created **Instacart: Help Center Chat with Virtual Assistant** (web, 5 screens, Food & Drink, Marketplace, Food Delivery) [Refero](https://refero.design/flows/10063) > A desktop help-center conversation with a virtual assistant presented in a centered chat modal. Steps: Chat opens with suggested topics → User selects a suggested topic → User enters free-text request → Assistant indicates it's composing → Assistant asks a clarifying question **GoFundMe: Sharehub Co-Organizer Invitation** (web, 5 screens, Fundraising) [Refero](https://refero.design/flows/11746) > Logged-in GoFundMe organizer in Sharehub on desktop views a ready-to-share fundraiser post and opens the co-organizer invitation modal. Steps: Sharehub Overview → Fundraise Together → Co-organizer Email Entry → Invite Ready → Invitation Sent State ##### Task: savings pot goal **Loowner: Set personal goal** (ios, 7 screens, Marketplace) [Refero](https://refero.design/flows/6596) > Happy-path: A logged-in user on the main dashboard opens the Goal card, reads a short explainer, taps Set up goal, enters a numeric target (either quantity or cost) using an in-screen numeric keypad, confirms the change, sees a confirmation screen, and returns to the dashboard where the Goal card is updated to reflect the new target and progress. Steps: Dashboard — Goal card → Goal explainer modal → Set up your goal — input sheet (initial) → Enter numeric goal → Confirm goal change → Success acknowledgement → Dashboard — Goal card (updated) **Waterllama: Hydration goal tracking** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/11531) > Mobile hydration dashboard where the user logs beverages to reach a daily hydration goal. Steps: Dashboard — initial low progress → Dashboard — nearing goal → Goal reached — celebration **The Outsiders: Training Goal Creation** (ios, 14 screens, Health & Wellness) [Refero](https://refero.design/flows/10301) > Happy-path: The user is on the Progress (Training Summary) screen and adds a new training goal via a modal sheet. Entry: Progress screen, logged-in state (user avatar visible). Exit: Progress screen with a new Training Goal listed under Training Goals. Assumptions: user is signed in and has access to Training Goals; device is a mobile phone; the Add Goal UI is presented as a modal sheet over the Progress screen. Steps: Progress / Training Summary → Add Goal (initial) → Workout Type List (top) → Workout Type List (scrolled) → Workout Type Selected → Open Goal Type → Goal Type List (options) → Goal Type Selected → Open Time Frame → Time Frame Selection → Open Numeric Goal Picker → Set Numeric Target → Review & Add → Training Goal Visible **Apple Books: Daily Reading Goal Adjustment** (ios, 4 screens, Books) [Refero](https://refero.design/flows/11912) > A mobile reading-progress screen shows the user’s current progress toward a daily reading goal. The user can dismiss the screen, share progress, or open a goal-adjustment bottom sheet to change the daily reading target in minutes per day. Assumption: the user is already signed in and viewing a reading-tracking experience on a mobile device. The happy path shown is opening the goal editor, choosing a new minute target, and returning to the progress screen with the updated goal reflected. Steps: Reading Progress Overview → Goal Editor Opened → Goal Value Selection → Updated Progress View **Blackbird: Accept and track spend challenge** (ios, 8 screens, Food & Drink) [Refero](https://refero.design/flows/9027) > Happy path: user discovers an opt-in spending challenge from the Wallet home, opens the Challenges hub, reviews the challenge details, accepts the challenge, sees a confirmation modal, and returns to the challenge detail which updates to an In Progress state. Steps: Wallet home — Challenges tile → Challenges hub — List view → Challenge detail — Pre-accept → Challenge detail — Accepted state → Confirmation modal → Challenge detail — In Progress → Challenge detail — Information and terms → Wallet home — Challenge tile updated **ElevenReader: Set Daily Listening Goal** (ios, 4 screens, AI Tool, Books) [Refero](https://refero.design/flows/8967) > Happy path: User on the Home feed taps the 'Daily Listening Goal' card, which opens a modal bottom sheet to set a daily listening target. The user adjusts the minutes using minus/plus controls, saves the selection, and the home card updates to reflect the new daily target. Steps: Home — Discover & Goal Card → Open Goal Editor (Bottom Sheet) → Adjust Minutes → Save Goal and Update Home **TIDE: Insomnia Session Detail** (ios, 2 screens, Health & Wellness) [Refero](https://refero.design/flows/6755) > Mobile session detail screen for a sleep/meditation exercise titled "Insomnia." Entry: user opens a session from a list or search and lands on this detail view. The screen shows a hero image with social presence (avatars + count), content tags (Plus, Sleep), session title, estimated length, teacher and duration pill controls, descriptive text, and a prominent primary CTA labeled "Select" at the bottom. Exit: user either taps the heart to save (shows a saved confirmation) or taps Select to proceed (next step not shown). Assumptions: user is on a mobile device (iOS-style status bar visible). The UI exposes save and select actions; whether saving requires authentication is not visible here. Steps: Session Detail — initial → Session Detail — saved confirmation **District: Save to Hotlist** (ios, 6 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12817) > Happy path starts from a save action on a content detail or discovery screen and opens a bottom-sheet style Save panel. The user can either create a new list or save the item into an existing Hotlist. In the new-list path, the user enters a list name, optionally uses suggested chips, creates the list, then the item is attached to that list. In the existing-list path, the user selects a list and optionally adds a note, then taps Done to confirm. Steps: Save Menu → Create List Entry → List Naming → Create List Submit → New List Added → Save Confirmed **Picnic: Streak goal setup** (ios, 6 screens, Productivity, Utility) [Refero](https://refero.design/flows/11509) > Happy-path: User opens the Profile screen (already signed in) and adds a streak goal to motivate daily photo sorting. Steps: Profile - Streak entry → Goal selector — initial → Preview — 30-day selection → Preview — 365-day selection → Select and commit — 7-day final → Profile — Goal active **Moises: Profile — select goals** (ios, 5 screens, AI Tool, Music Streaming) [Refero](https://refero.design/flows/6860) > Happy path: A logged-in, premium user navigates to their Profile tab, opens the Goals picker, selects one or more goals from a vertically scrollable list, and taps Done to return to Profile where the chosen goals are shown as a condensed summary. Steps: Profile (entry) → Goals list (initial) → Select multiple goals → Select additional and confirm → Profile updated (exit) **Kickstarter: Discover & Save Project** (ios, 2 screens, Fundraising) [Refero](https://refero.design/flows/4628) > Mobile Explore feed showing a scrollable list of project cards under a "Recommended For You" filter and category tabs (Magic, Popular, Newest, Ending Soon). Entry: user opens the Explore tab with the feed already populated. Happy path shown: user taps the heart on a project card to save/favorite it. Exit: the card updates in-place (heart fills) and the user remains on the same feed. Assumptions: user is using a mobile device in portrait, on the Explore tab of the app; saving a project is an inline action and the UI persists a visual saved state. Steps: Explore Feed (Project Card Visible) → Saved Confirmation (Heart Filled) **Miles: Set monthly distance goal** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/6705) > Happy path: from the main activity dashboard the user opens a goal editor, sets a distance goal (miles) using an onscreen numeric input or +/- controls, toggles the monthly goal on/off, and confirms with Done. Steps: Dashboard — weekly summary → Goal editor — numeric input → Adjust value and confirm → Dashboard — monthly timeline visible **The Outsiders: Set Sleep Goal** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/10294) > Happy-path for setting a personal Sleep Goal from the Sleep section of a health app on an iPhone. Entry: user views Sleep details and taps the control to set or edit their sleep goal. The flow opens a modal overlay where the user changes the target hours using a prominent numeric picker and an associated horizontal slider. Visual feedback (striping, color, and a label) indicates whether the chosen value is below, within, or above the app's recommended range (7–9 hours). Exit: user taps Done to save/confirm the new sleep goal and returns to the Sleep overview. Assumptions: user is on an unlocked iPhone, in the Sleep area of a health app, and has permission to change sleep goal settings; this flow does not show any account authentication or cross-device sync steps explicitly. Steps: Sleep Overview (No Data) → Sleep Goal Modal — Default Value → Adjust Goal — Below Recommended → Adjust Goal — Above Recommended and Confirm **Apple Books: Daily Goal Achieved** (ios, 3 screens, Books) [Refero](https://refero.design/flows/11936) > This is a mobile success-and-sharing confirmation after the user reaches a daily reading target in Apple Books. The flow starts on a completion screen showing that the reading goal has been achieved and ends either by dismissing the screen or sending the achievement through iMessage after using the share action. Steps: Goal Complete Screen → Share Sheet → Message Composer **Pinterest: Save image to profile** (ios, 3 screens, Reference, Inspiration Board, Social Networking, Lifestyle) [Refero](https://refero.design/flows/2477) > Happy path: User is browsing a vertically-scrolling, two-column image feed (Browse). They open a contextual radial action menu on a specific image, choose the save (pin) action, and receive an on-screen confirmation that the image was saved to their profile with an Edit option. Entry: user is on the Browse feed, logged in on a mobile device. Exit: confirmation banner shown and saved state visible in the feed. Steps: Reveal contextual actions → Activate save (pin) → Show saved confirmation **Airbnb: Browse and Save Listing** (ios, 2 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6388) > Happy path: A signed-in mobile user opens the Explore tab, scrolls horizontally through curated carousels (e.g., "Popular homes in Cleveland" and "Available in Cincinnati this weekend"), and saves a listing to a wishlist by tapping the heart icon. Steps: Browse curated carousels → Show save confirmation **Foodllama: Macro balance summary** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/11781) > This is a mobile nutrition summary screen for a meal or recipe in Foodllama. The user enters from a food analysis detail view and can review macro balance, compare it with a perfect balance example, inspect ingredient tags, and save the item as bookmarked. The user is assumed to be on a phone, already viewing an analyzed meal, with access to edit actions and bookmark controls. The screen ends with a confirmation state showing the item bookmarked and macro information highlighted. Steps: Nutrition Summary → Bookmark Action → Saved State **Cake Equity: Add personal equity goal** (web, 4 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/6155) > Happy path: a logged-in user on the Equity explainer dashboard adds a personal goal tied to their equity by opening an in-context modal, entering a goal name, selecting an emoji, adding a short description, and saving it. Steps: Dashboard — Equity explainer → Set your goal — Modal (empty) → Set your goal — Modal (filled) → Dashboard — Goal displayed **Revolut: Vault Goal & Deadline Setup** (web, 3 screens, Business, Finance) [Refero](https://refero.design/flows/1337) > Happy-path: A logged-in user views a single 'Vault' detail screen, adds a monetary goal via a modal, confirms it, and sees the vault updated with the goal and deadline. Steps: Vault Details → Goal Entry Modal → Vault Updated **Literal.club: Add books to yearly reading target** (web, 7 screens, Books, Communication Tools) [Refero](https://refero.design/flows/7911) > Happy path for adding books to a user's yearly reading goal. Entry: user is on their goal overview (shows progress ring and empty/placeholder slots). The user opens the manage-books overlay, uses a search/autocomplete to find titles, adds one or more books to the goal, and then returns to the goal overview where the progress ring and the grid of book covers update. Assumptions: user is signed in and viewing their own goal (editable), on a desktop/laptop viewport, and has permission to edit this goal. Steps: Goal overview (empty) → Manage books overlay (empty list) → Add book — search input → Select book from results → Manage books — single item → Manage books — multiple items → Goal overview (updated) **Revolut: Vault detail and edit** (web, 9 screens, Business, Finance) [Refero](https://refero.design/flows/1334) > Happy path for viewing and editing a single "Vault" (a named savings container). Entry: user opens a Vault detail screen (user is assumed logged in and has access to this vault). The screen shows vault cover image, balance and quick actions (Fund, Withdraw), plus editable metadata (Name, Goal, Deadline), and toggles for regular transfers. The flow demonstrates editing the cover image, changing the vault name, setting a numeric goal (currency + amount), picking a deadline from a calendar, and returning to the Vault detail where updates are reflected. Exit: user closes each modal (Done) and returns to the Vault detail; they may also navigate back from the top-left chevron. Steps: Vault overview (initial) → Vault metadata section → Upload image prompt → Edit name modal → Vault cover updated (visual confirmation) → Set goal modal → Deadline calendar modal → Vault metadata updated → Updated overview with achievement and toggles **Wise: Create a Jar** (web, 6 screens, Finance) [Refero](https://refero.design/flows/367) > Happy path: A logged-in user chooses to open a 'Jar' (a dedicated savings container), reviews how jars work, fills out a short form (icon, name, currency), optionally selects an icon, confirms creation, and lands on the new jar's detail screen showing the balance and actions. Steps: Choose account type → How jars work → Jar creation form (initial) → Jar form (fields filled) → Icon picker → Jar detail **Literal.club: Reading goal setup** (web, 11 screens, Books, Communication Tools) [Refero](https://refero.design/flows/7910) > Happy-path: A logged-in user on their profile navigates to the Goals area, starts creating a new reading goal, uses a quick dial to pick a target number of books or opens the advanced setup to name the goal and set a custom start/end date, enables automatic addition of read books, and then creates the goal. Steps: Profile — Goals tab (empty) → Goals listing — create CTA → New reading goal — quick dial (0) → New reading goal — dial set to 15 → Advanced setup — form shown → Fill name field → Open start-date picker → Select start date → Open end-date picker → Select end date and toggle options → Create goal and view goal detail **GoFundMe: Fundraiser Details and Organization Follow** (web, 3 screens, Fundraising) [Refero](https://refero.design/flows/11731) > A donor opens a fundraiser detail page for a dog’s medical costs. The main entry point is the public fundraiser page showing the title, story, updates, organizer, sharing tools, and donation summary. The user can donate or share from the page, and can also open the organization card modal to view Pasadena Humane details and choose to learn more or follow. The flow ends with the organization marked as followed in the modal state, or the modal dismissed back to the fundraiser page. Steps: Fundraiser Detail Page → Organization Profile Modal → Following Confirmation State **PayPal: Charity profile and donation entry** (web, 2 screens, Finance) [Refero](https://refero.design/flows/7412) > User lands on a PayPal-hosted charity profile (Woodgreen Pets Charity). The page presents charity metadata (category, location, charity number), a short description, a Donate CTA, social share icons with a copyable URL, and a heart icon to save/favourite the charity. Steps: Charity profile (logged out) → Post-save confirmation and fundraiser CTA (logged in) **Teal: Set weekly application target** (web, 4 screens, Productivity, Job Search) [Refero](https://refero.design/flows/2735) > Happy path: A logged-in user on their career dashboard opens the Job Applications card, launches the 'Application Goal' modal, sets a weekly applications target using the +/- controls (or typing), and saves. The modal closes and the Job Applications card updates to reflect the newly saved target and current progress (remaining applications or success message). Steps: Dashboard - Job Applications overview → Application Goal modal (open) → Application Goal modal (adjust value) → Dashboard updated with saved target **Aboard: Add and complete 1-on-1 topic** (web, 5 screens, Collaboration Tools, Paperwork) [Refero](https://refero.design/flows/782) > Happy path: From an employee's 1-on-1 record the user creates a new discussion topic, saves it, and marks it complete which updates the related goal's progress and shows the goal as completed. Steps: 1-on-1 header and goal open → Open inline topic editor → Enter topic content → Save topic and view in list → Complete topic and update goal **Cushion: Weekly Goals Check-in** (web, 7 screens, Communication Tools, Collaboration Tools) [Refero](https://refero.design/flows/12511) > A logged-in workspace member opens the 'Weekly Goals' check-in from the left navigation, writes a weekly update in the post composer, uses the checklist formatting option to turn the content into a checklist, adds several goal items, and submits the post. After submission, the composer is replaced by the published check-in entry, grouped under the current week with a reply field below it. The interface suggests a regular check-in space with a 'Reminders only' setting and a simple post-and-reply interaction model. Steps: Open Check-in → Choose Checklist Format → Insert Checklist Marker → Enter First Goal → Build Goal List → Submit Check-In → View Published Entry **Grammarly: Configure writing goals** (web, 3 screens, Paperwork) [Refero](https://refero.design/flows/1871) > Happy-path: user opens a new/untitled document in the editor, launches the writing assistant, then configures tailored writing goals via a modal. Steps: Empty editor → Assistant panel (empty state) → Set goals modal **Teal: Career Goal Setup** (web, 6 screens, Productivity, Job Search) [Refero](https://refero.design/flows/2734) > Happy path: a signed-in user on a desktop visits their career-dashboard, opens the Edit Goals modal, fills or updates fields (goal type, title, date, salary min/max, currency) using inline inputs and a date picker, and saves. Steps: Dashboard Overview → Open Edit Goals Modal → Select Goal Type → Pick Target Date → Enter Title and Salary Range → Save and Update Dashboard **GoFundMe: Create and Launch a Fundraiser** (web, 34 screens, Fundraising) [Refero](https://refero.design/flows/11740) > The flow begins on GoFundMe’s fundraising landing page and follows a user who starts creating a fundraiser. The user selects where funds will be received, enters a ZIP code, chooses a fundraising category, identifies the beneficiary, sets a fundraising goal, and adds a cover image. The system then verifies the user’s phone number, publishes the fundraiser, offers optional fundraising advice and social sharing tools, and ends on the fundraiser dashboard. The screenshots show a completed happy path using United States, ZIP code 94108, the Business category, fundraising for oneself, a $1,000 goal with automated goal setting enabled, and a restaurant image. Steps: Fundraising Landing → Fund Location and Category → Completed Location and Category → Choose Beneficiary → Fundraising for Yourself → Set Fundraising Goal → Completed Fundraising Goal → Add Cover Media → Choose Media Type → Edit Cover Photo → Phone Verification → Verification Success → Fundraiser Published → Post-Publication Transition → Fundraising Insights Introduction → Text Advice Preference → Create Daily Checklist → First Share Prompt → Personalized Social Post → Fundraiser Dashboard ##### Task: scheduled payment standing order **PayPal: Set preferred online payment** (ios, 5 screens, Finance) [Refero](https://refero.design/flows/7251) > Happy-path flow for choosing and saving the preferred funding source used when shopping online. Steps: Wall _… 73477 more characters in the zip._ ### research/finance/REGULATORS.md `research/finance/REGULATORS.md` · sha256 `3542bfe383df3e329a5d66a14453b2949d97c3eb9f52e3ea103c41f42f849607` #### Regulators and rules for EDGE Finance **Research, not legal advice.** This report collects public information about who regulates EDGE Finance's activities and which rules apply. It is not legal advice. Before EDGE Finance launches, signs a partner contract or applies for a licence, a qualified UK financial services lawyer and compliance adviser must check it. - **Research date:** 31 October 2026. Every link was opened or found on that date, unless a line says it was not. - **Who it is for:** a new engineer at EDGE Finance, part of JudeAI | EDGE ID. EDGE Finance is a governed-intelligence platform where people, organisations and AI agents are all bank customers. Agents open accounts, lend, borrow and pay under their own identity, with a mandate from an accountable owner. It starts in the UK and expands worldwide. - **Status labels:** - **AGREEMENT** means the rule is law or in force, with dates. - **PROPOSED** means it is a consultation, a Bill, a draft or a political agreement that is not yet in force. - **UNKNOWN** means we could not confirm it. The line says what we searched. - **NOT RE-VERIFIED** means we know the fact from before this session, but the primary page would not load for our tools on 31 October 2026 (EUR-Lex returned an anti-bot challenge, some FCA Handbook pages loaded with no text, and FATF and IOSCO returned HTTP 403). Re-check these at the link before relying on them. - **Method limits:** the web-search budget for this session ran out partway through. Some 2026 developments therefore rely on law-firm commentary, marked "secondary source", and not on the regulator's own page. --- ##### 0. The headline for engineers 1. **No UK rule we found lets an AI agent be a customer in its own legal right.** - The FCA says it does not plan extra AI-specific regulation and relies on existing rules, mainly the Consumer Duty and SM&CR ([source](https://www.fca.org.uk/firms/innovation/ai-approach), accessed 2026-10-31). - The FCA's Mills Review (6 July 2026) recommends the FCA "Enable the foundations for agentic finance" ([source](https://www.fca.org.uk/publications/corporate-documents/mills-review), accessed 2026-10-31). - Secondary sources describe this as a framework for how agents can be "authorised, identified and held accountable" ([source](https://www.stephensonharwood.com/post/102n942/agentic-finance-the-fca-and-a-brave-new-world-the-mills-review-2026/), accessed 2026-10-31). - **Status: PROPOSED (a recommendation only).** Design accordingly: the legal customer is always a person or a legal entity, and the agent acts under that customer's mandate. 2. **HM Treasury is asking how payment law should change for agents.** It asks whether "provisions relating to authentication and consent of payments transactions, and liability for unauthorised payment transactions, need updating" for agentic payments (Question 15). The consultation closed on 6 October 2026 ([source](https://assets.publishing.service.gov.uk/media/6a54b3302467584757371dc4/Modernising_payment_services_consultation.pdf), accessed 2026-10-31). **Status: PROPOSED.** 3. **Big 2026 changes:** - The Payment Systems Regulator is being folded into the FCA (a Bill is before Parliament). - The UK cryptoasset regime goes live on 25 October 2027. - The Money Laundering Regulations were amended from 30 June 2026. - New automated-decision rules under the Data (Use and Access) Act 2025 apply from 5 February 2026. - The EU AI Act's high-risk credit-scoring duties were moved to 2 December 2027. - The first UK critical third parties (AWS, Google Cloud, Microsoft and Oracle) were designated on 13 July 2026. Each is detailed and sourced below. --- ##### 1. UK regulators and what each does ###### Financial Conduct Authority (FCA) - **What it does:** regulates the conduct of financial firms. - It authorises payment and e-money firms ([source](https://www.fca.org.uk/firms/apply-emoney-payment-institution), accessed 2026-10-31). - It authorises consumer credit and crypto firms ([source](https://www.fca.org.uk/publications/policy-statements/cryptoasset-regime), accessed 2026-10-31). - It is the lead regulator for payment services and e-money ([source](https://assets.publishing.service.gov.uk/media/6a54b3302467584757371dc4/Modernising_payment_services_consultation.pdf), accessed 2026-10-31). - **Supervision:** it is also the anti-money-laundering supervisor for firms it registers, including cryptoasset businesses under the Money Laundering Regulations ([source](https://www.fca.org.uk/publications/policy-statements/cryptoasset-regime), accessed 2026-10-31). - **AI stance:** it does "not plan to introduce extra regulations for AI" ([source](https://www.fca.org.uk/firms/innovation/ai-approach), accessed 2026-10-31). ###### Prudential Regulation Authority (PRA) - **What it does:** the PRA is part of the Bank of England. It authorises and supervises banks for safety and soundness ("prudential" means having enough capital and liquidity to survive). - **New banks:** it runs the New Bank Start-up Unit jointly with the FCA. It limits new banks in "mobilisation" to £50,000 of deposits for up to 12 months (page last updated 13 March 2026) ([source](https://www.bankofengland.co.uk/prudential-regulation/new-bank-start-up-unit/mobilisation), accessed 2026-10-31). - **Deposit protection:** it sets the Financial Services Compensation Scheme (FSCS) deposit protection limit. It raised the limit to £120,000 for firms failing from 1 December 2025 (policy statement PS24/25) ([source](https://www.bankofengland.co.uk/prudential-regulation/publication/2025/november/depositor-protection-policy-statement), accessed 2026-10-31, found via search; page not opened). **AGREEMENT.** ###### Bank of England (BoE) - **Systemic payments and financial stability:** the BoE oversees systemically important payment systems and financial stability. It is consulting on fees for "recognised payment systems and digital settlement asset service providers" ([source](https://www.gov.uk/government/consultations/updating-the-bank-of-englands-fee-regime-for-recognised-payment-systems/recognised-payment-systems-and-digital-settlement-asset-service-providers-fee-regime-consultation), accessed 2026-10-31). **PROPOSED.** - **Critical third parties:** with the PRA and FCA, it jointly oversees critical third parties (CTPs) designated by HM Treasury ([source](https://www.bankofengland.co.uk/news/2026/july/uk-financial-regulators-to-begin-overseeing-critical-third-parties-announced-by-hmt), accessed 2026-10-31). ###### Payment Systems Regulator (PSR) and its merger into the FCA - **What it does:** the PSR regulates payment systems such as Faster Payments and their participants. It made the APP fraud reimbursement rules ([source](https://www.psr.org.uk/our-work/app-scams/), accessed 2026-10-31). - **Merger decision:** HM Treasury's consultation response (April 2026) confirms the government will consolidate the PSR's functions entirely within the FCA, subject to primary legislation ([source](https://assets.publishing.service.gov.uk/media/69ce4dad019a4faf2745b2d2/A_Streamlined_Approach_to_Payment_Systems_Regulation_-_Consultation_Response.pdf), accessed 2026-10-31). **AGREEMENT on policy, PROPOSED in law.** - **The Bill:** - The Financial Services and Markets Bill [HL] was introduced in the House of Lords on 19 May 2026 and provides for the PSR's abolition. - It completed its Lords stages on 15 September 2026 and was awaiting Commons Second Reading at last update. This is a secondary source citing the Parliament bill page, which we did not open ([source](https://bratby.law/financial-services-and-markets-bill-2026/), accessed 2026-10-31; bill page cited there: [bills.parliament.uk/bills/4129](https://bills.parliament.uk/bills/4129)). - Commencement date: **UNKNOWN**. It needs Royal Assent and a commencement order (same source). - **Until then:** the PSR keeps its statutory powers ([source](https://www.psr.org.uk/publications/open-letters/fca-and-psr-respond-to-government-consultation-on-future-of-payment-systems-regulation/), accessed 2026-10-31, found via search). ###### Information Commissioner's Office (ICO) - **What it does:** regulates data protection under UK GDPR (the UK General Data Protection Regulation) and the Data Protection Act 2018. - **Data (Use and Access) Act 2025:** this Act (Royal Assent 19 June 2025) made the rules on automated decisions "more permissive" but kept these safeguards: - telling people about significant decisions; - letting them challenge a decision; - human intervention ([source](https://www.gov.uk/guidance/data-use-and-access-act-2025-data-protection-and-privacy-changes), accessed 2026-10-31). - **Commencement:** the automated decision-making changes (new UK GDPR Articles 22A–22D) commenced on 5 February 2026 under SI 2026/82 ([source](https://legislation.gov.uk/uksi/2026/82), accessed 2026-10-31; secondary summary [source](https://bratby.law/data-use-and-access-act-2025-commencement/), accessed 2026-10-31). **AGREEMENT.** - Whether the ICO has been renamed the "Information Commission" yet: **UNKNOWN**. The government guidance page still says ICO ([source](https://www.gov.uk/guidance/data-use-and-access-act-2025-data-protection-and-privacy-changes), accessed 2026-10-31). ###### HM Treasury (HMT) and the Office of Financial Sanctions Implementation (OFSI) - **HM Treasury:** writes financial services law (for example the payments modernisation consultation) ([source](https://www.gov.uk/government/consultations/modernising-payment-services-regulation), accessed 2026-10-31). It designates critical third parties ([source](https://www.gov.uk/government/news/uk-financial-system-strengthened-with-new-safeguards-for-major-technology-providers), accessed 2026-10-31). - **OFSI:** the part of HM Treasury that implements and enforces UK financial sanctions. Its general guidance covers reporting obligations ([source](https://www.gov.uk/government/publications/financial-sanctions-general-guidance/uk-financial-sanctions-general-guidance), accessed 2026-10-31). - **Sanctions list change:** since 28 January 2026 the OFSI Consolidated List is closed. The UK Sanctions List is the only source of UK designations ([source](https://www.gov.uk/government/publications/financial-sanctions-general-guidance), accessed 2026-10-31). **AGREEMENT.** ###### National Crime Agency (NCA): suspicious activity reports - **The UKFIU:** the UK Financial Intelligence Unit inside the NCA has "sole national responsibility" for receiving and analysing suspicious activity reports (SARs). It receives over 850,000 a year. - **Who must report:** regulated-sector firms have a legal duty to report. - **How to report:** through the SAR Portal ([source](https://www.nationalcrimeagency.gov.uk/what-we-do/crime-threats/money-laundering-and-illicit-finance/suspicious-activity-reports), accessed 2026-10-31). ###### HM Revenue & Customs (HMRC) - **AML supervisor:** HMRC is an anti-money-laundering supervisor for some businesses, for example money service businesses not supervised by the FCA. **NOT RE-VERIFIED**: we did not open an HMRC page. The supervisory authorities are listed in the Money Laundering Regulations ([source](https://www.legislation.gov.uk/uksi/2017/692/regulation/5), accessed 2026-10-31; regulation 7 not opened). - **Tax reporting:** HMRC's tax-reporting duties for financial institutions (for example the International Tax Compliance Regulations 2015) are referenced in MLR regulation 27 ([source](https://www.legislation.gov.uk/uksi/2017/692/regulation/27), accessed 2026-10-31). ###### Competition and Markets Authority (CMA) - **What it does:** makes markets more competitive and protects consumers ([source](https://www.gov.uk/government/organisations/competition-and-markets-authority), accessed 2026-10-31). - **Open banking origins:** UK open banking began with a CMA order on retail banking, which created Open Banking Limited (OBL). OBL is now convening the "Future Entity" standards body ([source](https://assets.publishing.service.gov.uk/media/6a54b3302467584757371dc4/Modernising_payment_services_consultation.pdf), accessed 2026-10-31). The CMA order detail is **NOT RE-VERIFIED**. ###### Financial Ombudsman Service (FOS) - **What it does:** free dispute resolution for consumers and microenterprises against financial firms. - **Award limits:** for complaints referred from 1 April 2026, the limit is £455,000 for acts on or after 1 April 2019 and £205,000 for earlier acts ([source](https://www.financial-ombudsman.org.uk/news/fca-confirms-increase-award-limits-2), accessed 2026-10-31). **AGREEMENT.** - **Warning:** the PSR consumer page still quotes £430,000, which is out of date ([source](https://www.psr.org.uk/information-for-consumers/our-new-app-fraud-reimbursement-protections/), accessed 2026-10-31). --- ##### 2. UK rules ###### 2.1 Consumer Duty - **What it is:** FCA rules that set "high standards of protection for retail consumers" and require firms to put customers' needs first. Final rules are in policy statement PS22/9 ([source](https://www.fca.org.uk/firms/consumer-duty), accessed 2026-10-31). **AGREEMENT.** - **The four outcomes:** products meet needs and give fair value; communications meet information needs; firms give support that meets needs ([source](https://www.fca.org.uk/firms/innovation/ai-approach), accessed 2026-10-31). - **In-force dates:** 31 July 2023 for open products and 31 July 2024 for closed products. **NOT RE-VERIFIED**: the FCA hub page we opened does not state them. - **Recent FCA output:** "Fair value in consumer finance" (29 October 2026) and "Payments firms: delivering good outcomes for consumers in vulnerable circumstances" (17 September 2026) ([source](https://www.fca.org.uk/firms/consumer-duty), accessed 2026-10-31). ###### 2.2 Senior Managers and Certification Regime (SM&CR) - **What it is:** named senior people are personally accountable for areas of the firm. The FCA relies on it for safe AI use ([source](https://www.fca.org.uk/firms/innovation/ai-approach), accessed 2026-10-31). - **2026 Phase 1 reforms:** - The FCA published PS26/6 and the PRA published PS12/26 on 22 April 2026, with some changes effective 24 April 2026 ([source](https://www.fca.org.uk/publication/policy/ps26-6.pdf), accessed 2026-10-31, found via search). - Enhanced-firm thresholds rise by about 30% from 10 July 2026 (secondary source [source](https://www.tlt.com/insights-and-events/insight/fca-finalises-phase-1-sm-cr-reforms-what-firms-need-to-do-now), accessed 2026-10-31). **AGREEMENT.** - **Phase 2:** HM Treasury intends to remove the Certification Regime from FSMA (the Financial Services and Markets Act 2000). A Phase 2 consultation is expected late in 2026 (secondary source [source](https://www.reedsmith.com/articles/fca-and-pra-confirm-changes-to-smcr-to-streamline-senior-manager-accountability-and-boost-growth/), accessed 2026-10-31). **PROPOSED.** ###### 2.3 Money Laundering Regulations 2017 (MLRs) - **When to do CDD:** firms must apply customer due diligence (CDD), meaning identifying and verifying the customer, when: - establishing a business relationship; - carrying out occasional transactions of £12,000 or more; - making occasional transfers of funds over £800; - suspecting money laundering or terrorist financing; - doubting earlier identity data. - **Crypto transfers:** cryptoasset exchanges and custodian wallet providers must apply CDD on transfers of £800 or more. Crypto ATM operators must apply it on every transaction. - **Sterling thresholds:** these sterling thresholds were substituted from 30 June 2026 ([source](https://www.legislation.gov.uk/uksi/2017/692/regulation/27), accessed 2026-10-31). **AGREEMENT.** - **Beneficial ownership:** for a company, a beneficial owner includes any individual who owns or controls "more than 25% of the shares or voting rights" ([source](https://www.legislation.gov.uk/uksi/2017/692/regulation/5), accessed 2026-10-31). **AGREEMENT.** - **Record keeping:** - Keep records for 5 years after an occasional transaction completes, or after a business relationship ends. - Transaction records within a relationship need not be kept more than 10 years. - Personal data must then be deleted unless an exception applies ([source](https://www.legislation.gov.uk/uksi/2017/692/regulation/40), accessed 2026-10-31). **AGREEMENT.** - **2026 amendment:** the Money Laundering and Terrorist Financing (Amendment) Regulations 2026 (SI 2026/621). Most provisions came into force 21 days after being made ([source](https://www.legislation.gov.uk/uksi/2026/621/regulation/1/made), accessed 2026-10-31). - The search summary gives the made date as 9 June 2026, which means 30 June 2026 ([source](https://www.legislation.gov.uk/uksi/2026/621/pdfs/uksi_20260621_en.pdf), accessed 2026-10-31, via search). - The new regulation 34A (enhanced CDD for cryptoasset businesses and correspondent relationships) applies from **1 February 2027**. - Parts of Schedule 6B apply from **25 October 2027** ([source](https://www.legislation.gov.uk/uksi/2026/621/regulation/1/made), accessed 2026-10-31). **AGREEMENT.** ###### 2.4 Payment Services Regulations 2017 (PSRs) and strong customer authentication (SCA) - **When SCA is needed:** SCA has applied since 14 September 2019 when a payer: - initiates an electronic payment; - "accesses their payment account online"; - does a remote action that may imply fraud risk ([source](https://www.fca.org.uk/firms/strong-customer-authentication), accessed 2026-10-31). **AGREEMENT.** - **What SCA means:** SCA uses two independent factors from knowledge, possession and inherence. **NOT RE-VERIFIED**: not stated on the FCA page we opened. - **Planned SCA change:** HM Treasury has "committed to commencing the revocation" of the PSR provisions on SCA so the FCA can write "more outcomes-based rules about authentication" ([source](https://assets.publishing.service.gov.uk/media/6a54b3302467584757371dc4/Modernising_payment_services_consultation.pdf), accessed 2026-10-31). **PROPOSED.** - **Recent PSR changes:** firms may delay a payment by up to 72 hours where they reasonably suspect APP fraud. Payments firms must give 90 days' notice before closing an account, with exceptions (same source). **AGREEMENT** (described by HMT as already legislated). ###### 2.5 Electronic Money Regulations 2011 (EMRs) - **Authorised EMI capital:** an authorised electronic money institution (EMI) needs initial capital of at least €350,000 ([source](https://www.legislation.gov.uk/uksi/2011/99/schedule/2), accessed 2026-10-31). **AGREEMENT.** - **Small EMI limits:** average outstanding e-money must not exceed €5,000,000, and payment transactions must average no more than €3,000,000 a month. A small EMI cannot offer account information or payment initiation services ([source](https://www.legislation.gov.uk/uksi/2011/99/regulation/13), accessed 2026-10-31). **AGREEMENT.** - **Safeguarding (from 7 May 2026):** new FCA rules (PS25/12) require: - daily reconciliations; - a resolution pack; - an annual safeguarding audit (unless the firm never held more than £100,000 over 53 weeks); - a monthly return ([source](https://www.fca.org.uk/publications/policy-statements/ps25-12-changes-safeguarding-regime-payments-and-e-money-firms), accessed 2026-10-31, found via search). **AGREEMENT.** ###### 2.6 Faster Payments and mandatory APP fraud reimbursement - **What APP fraud is:** authorised push payment (APP) fraud is when someone is tricked into sending money. - **Scope and start:** the scheme covers individuals, microenterprises and charities paying by Faster Payments or CHAPS, for payments from **7 October 2024**. - **Limits and timing:** - Maximum claim: **£85,000**. - Claim within **13 months** of the payment. - Optional excess of up to £100, which cannot be applied to vulnerable customers. - Pay within **5 business days**, or up to **35 business days** if the clock is paused. - No refund if the customer was complicit or grossly negligent ([source](https://www.psr.org.uk/information-for-consumers/our-new-app-fraud-reimbursement-protections/), accessed 2026-10-31). **AGREEMENT.** - **Cost sharing:** costs are split 50:50 between the sending and receiving firms ([source](https://www.psr.org.uk/our-work/app-scams/), accessed 2026-10-31). - **Open questions:** - The cap is described as tracking the FSCS limit (secondary source [source](https://bratby.law/app-fraud-reimbursement-review/), accessed 2026-10-31). The FSCS limit rose to £120,000 from 1 December 2025, but the PSR page still says £85,000. Whether the cap has changed: **UNKNOWN**. - The PSR's evaluation review is running, with findings due in 2026/27 (secondary source, same link). **PROPOSED.** ###### 2.7 Confirmation of Payee (CoP) - **What it is:** a name-checking service run by Pay.UK. It checks the account name and whether the account is personal or business before a payment. - **Who can join:** account-holding payment firms reachable by sort code and account number, or by secondary reference data. - **Status:** the PSR mandated its expansion in 2024 ([source](https://www.wearepay.uk/what-we-do/overlay-services/confirmation-of-payee/), accessed 2026-10-31). **AGREEMENT.** - **Unconfirmed details:** the exact current PSR direction (Specific Direction 17) and its deadlines are **UNKNOWN**. The PSR's CoP page returned 404 on 31 October 2026. ###### 2.8 Direct Debit scheme and the Direct Debit Guarantee - **Who runs it:** Direct Debit is owned and managed by Pay.UK. - **What the Guarantee promises:** - advance notice of changes ("normally 10 working days"); - "a full and immediate refund" for errors; - cancel "at any time" through your bank ([source](https://www.directdebit.co.uk/direct-debit-explained/direct-debit-guarantee/), accessed 2026-10-31). **AGREEMENT** (scheme rule, not statute). ###### 2.9 Open Banking: consent, oversight, Smart Data and the Data (Use and Access) Act 2025 - **Consent renewal:** - Customers no longer re-authenticate with their bank every 90 days for account information. Instead, the third-party provider (TPP) must get explicit consent at least every 90 days (SCA-RTS Article 10A; effective 26 March 2022). - First connection still needs SCA ([source](https://www.fca.org.uk/firms/strong-customer-authentication), accessed 2026-10-31; [PS21/19](https://www.fca.org.uk/publications/policy-statements/ps21-19-changes-sca-rts-and-guidance-approach-document-and-perimeter-guidance-manual), accessed 2026-10-31). **AGREEMENT.** - **Data (Use and Access) Act 2025 and Smart Data:** - The Act (DUAA) lets the government set up Smart Data schemes through secondary legislation and delegate powers to the FCA. - The government committed to lay an Open Banking statutory instrument under the Act "by the end of 2026". - A not-for-profit "Future Entity" will take over standard-setting from OBL ([source](https://assets.publishing.service.gov.uk/media/6a54b3302467584757371dc4/Modernising_payment_services_consultation.pdf), accessed 2026-10-31). **PROPOSED.** - **Variable recurring payments:** "novel variable recurring payments" (VRP) now run under a new commercial Open Banking scheme (same source). **AGREEMENT** (industry scheme). - **JROC:** the Joint Regulatory Oversight Committee for open banking. Its current status is **UNKNOWN**. The July 2026 HMT consultation describes FCA-led oversight and the Future Entity, and we found no 2026 JROC page (search budget exhausted). ###### 2.10 Operational resilience and critical third parties - **Who it covers:** banks, payment and e-money firms, and others ([source](https://www.fca.org.uk/firms/operational-resilience), accessed 2026-10-31). - **What firms must do:** identify important business services, set impact tolerances, map and test. - **Deadlines:** rules have been in force since 31 March 2022. Firms had to be within impact tolerances by 31 March 2025 (same source). **AGREEMENT.** - **Incident and third-party reporting:** new FCA requirements (PS26/2, published 18 March 2026) take effect on **18 March 2027** (same source). **AGREEMENT** (made, not yet in force). - **Critical third parties:** HM Treasury designated AWS EMEA SARL, Google Cloud EMEA Ltd, Microsoft Ireland Operations Ltd and Oracle Corporation UK Ltd from **13 July 2026** ([source](https://www.legislation.gov.uk/uksi/2026/777/made), accessed 2026-10-31; [source](https://www.gov.uk/government/news/uk-financial-system-strengthened-with-new-safeguards-for-major-technology-providers), accessed 2026-10-31). - Designation does not reduce a firm's own accountability (secondary source [source](https://www.hsfkramer.com/notes/fsrandcorpcrime/2026-posts/hm-treasury-makes-first-designations-under-uk-ctp-regime), accessed 2026-10-31). **AGREEMENT.** ###### 2.11 Cryptoasset regime and financial promotion rules - **New FSMA regime:** - The Cryptoassets Regulations passed Parliament on 4 February 2026. - The FCA published final policy statements PS26/9 to PS26/13 on 30 June 2026. - The application window runs from **30 September 2026 to 28 February 2027**. - The full regime applies from **25 October 2027**. - Existing MLR crypto registrations "will not convert automatically" ([source](https://www.fca.org.uk/publications/policy-statements/cryptoasset-regime), accessed 2026-10-31). **AGREEMENT.** - **Activities in scope:** stablecoin issuance, custody, trading platforms, dealing, arranging, lending and borrowing, staking, and DeFi with an identifiable controller. Appropriateness tests are kept for crypto lending and borrowing (same source). - **Perimeter guidance:** PS26/18, published 16 September 2026 ([source](https://www.fca.org.uk/publications/policy-statements/ps26-18-cryptoasset-perimeter-guidance), accessed 2026-10-31, found via search). - **Financial promotions:** - The FCA Board confirmed the crypto promotion rules (PS23/6) on 29 June 2023. They apply to all firms marketing crypto to UK consumers, including overseas firms. - The restrictions cover clear risk warnings, a ban on incentives to invest, client categorisation and appropriateness assessments ([source](https://www.fca.org.uk/publications/policy-statements/ps23-6-financial-promotion-rules-cryptoassets), accessed 2026-10-31). **AGREEMENT.** - The exact risk-warning wording, the 24-hour cooling-off period for first-time investors and the 8 October 2023 start date are **NOT RE-VERIFIED**: the FCA Handbook section COBS 4.12A did not render for our tool. ###### 2.12 Consumer credit - **Creditworthiness and affordability:** FCA rules in CONC 5.2A (the Consumer Credit sourcebook) require a creditworthiness assessment covering credit risk and affordability ([Handbook link](https://www.handbook.fca.org.uk/handbook/CONC/5/2A.html), accessed 2026-10-31). **NOT RE-VERIFIED**: the page loaded with no text. - **Pre-contract information:** the lender must give pre-contract credit information "in good time before the agreement is made" ([source](https://www.legislation.gov.uk/uksi/2010/1013/regulation/3), accessed 2026-10-31). **AGREEMENT.** - legislation.gov.uk lists a future revocation under the Financial Services and Markets Act 2023 (same source). Timing: **UNKNOWN**. - **Cooling-off (right of withdrawal):** 14 days without giving a reason. The borrower must repay the credit plus accrued interest within 30 days ([source](https://www.legislation.gov.uk/ukpga/1974/39/section/66A), accessed 2026-10-31). **AGREEMENT.** - **Arrears and forbearance:** FCA rules in CONC 7 require fair treatment of borrowers in financial difficulty. **NOT RE-VERIFIED**: the FCA policy statement page we tried (PS24/2) returned 404, and the Handbook did not render. - **Complaints route:** firm complaints process, then the FOS (limits above) ([source](https://www.financial-ombudsman.org.uk/news/fca-confirms-increase-award-limits-2), accessed 2026-10-31). - **Consumer Credit Act reform:** status is **UNKNOWN**. The gov.uk consultation page returned 503, and the search budget was exhausted. ###### 2.13 Credit reference agency permission - **The activity:** "Providing credit references" (article 89B of the Regulated Activities Order) means furnishing information on the financial standing of individuals, collected for that purpose ([source](https://www.legislation.gov.uk/uksi/2001/544/article/89B), accessed 2026-10-31). **AGREEMENT.** - **Exclusion:** it applies only where the business mainly consists of this activity (same source). ###### 2.14 Sanctions screening and reporting to OFSI - **Source list:** screen customers, owners and payments against the UK Sanctions List, which has been the only source since 28 January 2026 ([source](https://www.gov.uk/government/publications/financial-sanctions-general-guidance), accessed 2026-10-31). **AGREEMENT.** - **Reporting duties:** OFSI's general guidance has a reporting-obligations chapter. A confidential reporting section (5.12) was added on 27 June 2025 (same source). - **Detail not confirmed:** the exact duty (to report "as soon as practicable" when you know or suspect a designated person or a breach, including frozen assets) is **NOT RE-VERIFIED**. Read the guidance chapter ([source](https://www.gov.uk/government/publications/financial-sanctions-general-guidance/uk-financial-sanctions-general-guidance), accessed 2026-10-31). ###### 2.15 Suspicious activity reports and tipping off - **Failing to report:** a regulated-sector firm that fails to report can face up to 5 years' imprisonment ([source](https://www.nationalcrimeagency.gov.uk/what-we-do/crime-threats/money-laundering-and-illicit-finance/suspicious-activity-reports), accessed 2026-10-31). **AGREEMENT.** - **DAML consent timing:** a "defence against money laundering" (DAML) request has a 7 working day notice period. If refused, a 31-day moratorium follows, which a court can extend ([source](https://www.legislation.gov.uk/ukpga/2002/29/section/335), accessed 2026-10-31). **AGREEMENT.** - **Tipping off:** it is an offence to disclose that a SAR was made, or that an investigation is contemplated, where this is likely to prejudice an investigation. - Penalty: up to 2 years' imprisonment on indictment. - Exceptions exist for disclosures within a group and some other cases ([source](https://www.legislation.gov.uk/ukpga/2002/29/section/333A), accessed 2026-10-31). **AGREEMENT.** - **Engineering consequence:** customer-facing agents and UIs must never reveal SAR status. --- ##### 3. Routes to operate, from bare minimum to full | Route | What it allows | What it requires | Statutory decision time | |---|---|---|---| | **Sponsor bank / Banking-as-a-Service (BaaS)** | EDGE builds the experience. A licensed partner holds the money and the regulatory permissions. | A contract with a licensed bank or EMI. EDGE inherits the partner's controls. | No licence for EDGE. **UNKNOWN**: we found no UK BaaS-specific rule (search budget exhausted). | | **Agent of a payment institution or EMI** | Provide payment services under the principal's licence. | Registration by the principal with the FCA. **NOT RE-VERIFIED** (PSRs regulation 34 not opened). The FCA apply page does not cover agents ([source](https://www.fca.org.uk/firms/apply-emoney-payment-institution), accessed 2026-10-31). | **UNKNOWN** | | **Small EMI (registered)** | Issue e-money up to the limits. No account information or payment initiation. | Average outstanding e-money ≤ €5m. Payments ≤ €3m a month. Good repute. Safeguarding. UK head office ([source](https://www.legislation.gov.uk/uksi/2011/99/regulation/13), accessed 2026-10-31). | Not opened in this session. | | **Small payment institution / registered AISP** | Payment services up to €3m a month (SPI). Account information only (RAISP, a registered account information service provider) ([source](https://www.fca.org.uk/firms/apply-emoney-payment-institution), accessed 2026-10-31). | Application through FCA Connect, tailored policies, forecasts. | 3 months complete, 12 months incomplete (PSRs regulation 9) ([source](https://www.legislation.gov.uk/uksi/2017/752/regulation/9), accessed 2026-10-31). | | **Authorised EMI** | Unlimited e-money plus payment services. | €350,000 initial capital ([source](https://www.legislation.gov.uk/uksi/2011/99/schedule/2), accessed 2026-10-31). Safeguarding (PS25/12). | 3 months complete, 12 months maximum ([source](https://www.legislation.gov.uk/uksi/2011/99/regulation/9), accessed 2026-10-31). | | **Authorised payment institution** | Payment services without a volume cap. | Initial capital €20,000 (money remittance), €50,000 (payment initiation) or €125,000 (other services) ([source](https://www.legislation.gov.uk/uksi/2017/752/schedule/3), accessed 2026-10-31). | 3 or 12 months (PSRs regulation 9, above). | | **Consumer credit permissions (FSMA Part 4A)** | Lend to consumers, broke credit, collect debts. | FCA authorisation, CONC compliance, SM&CR. | 6 months complete, 12 months incomplete ([source](https://www.legislation.gov.uk/ukpga/2000/8/section/55V), accessed 2026-10-31). | | **Credit reference permission (article 89B)** | Run a credit reference business. | FCA authorisation ([source](https://www.legislation.gov.uk/uksi/2001/544/article/89B), accessed 2026-10-31). | 6 or 12 months (section 55V). | | **Crypto: MLR registration now, FSMA authorisation from 25 October 2027** | Crypto exchange, custody, staking, lending. | Apply in the window 30 September 2026 to 28 February 2027 to use the savings provisions ([source](https://www.fca.org.uk/publications/policy-statements/cryptoasset-regime), accessed 2026-10-31). | 6 or 12 months (section 55V) for FSMA applications. | | **Full bank (PRA and FCA) with mobilisation** | Take deposits (FSCS protected up to £120,000), lend, everything. | Pre-application, then application to the PRA and FCA. Mobilisation up to 12 months with a £50,000 deposit cap ([source](https://www.bankofengland.co.uk/prudential-regulation/new-bank-start-up-unit/mobilisation), accessed 2026-10-31). Minimum capital: **UNKNOWN** (BoE page truncated). | 6 or 12 months (section 55V). Mobilisation adds up to 12 months. | - **Case officer allocation:** the FCA aims to allocate a case officer to payment and e-money applications within 20 working days ([source](https://www.fca.org.uk/firms/apply-emoney-payment-institution), accessed 2026-10-31). - **Possible reform:** the payment and e-money regimes may be rewritten. HMT consulted (14 July to 6 October 2026) on delegating rules to the FCA ([source](https://www.gov.uk/government/consultations/modernising-payment-services-regulation), accessed 2026-10-31). **PROPOSED.** --- ##### 4. EU, US and international ###### 4.1 European Union - **AI Act (Regulation (EU) 2024/1689) as amended by Regulation (EU) 2026/1744 ("Digital Omnibus on AI"):** - Creditworthiness and credit-scoring AI is high-risk under Annex III. - Annex III high-risk duties apply from **2 December 2027**. - Annex I (product-embedded) high-risk duties apply from **2 August 2028**. - The general application date is 2 August 2026, which covers Article 50 transparency ([source](https://ai-act-service-desk.ec.europa.eu/en/ai-act/article-113), accessed 2026-10-31). - The amendment was published 24 July 2026 and has been in force since 27 July 2026 ([source](https://eur-lex.europa.eu/eli/reg/2026/1744/oj/eng), accessed 2026-10-31, via search). **AGREEMENT.** - **DORA (Digital Operational Resilience Act, Regulation (EU) 2022/2554):** ICT risk, incident reporting and third-party oversight for EU financial entities, including payment and e-money institutions. It applies from 17 January 2025 ([source](https://eur-lex.europa.eu/eli/reg/2022/2554/oj/eng), accessed 2026-10-31). **AGREEMENT, NOT RE-VERIFIED** (EUR-Lex blocked our tool). - **PSD3 and the Payment Services Regulation (PSR):** - Provisional political agreement came on 27 November 2025. Final compromise texts were published in April 2026, and the ECON committee approved on 5 May 2026 (secondary source [source](https://www.williamfry.com/knowledge/psd3-psr-final-compromise-texts-are-published/), accessed 2026-10-31). - As of an update on 12 October 2026, plenary adoption and Official Journal publication were still pending, expected by December 2026 (secondary source [source](https://www.adyen.com/knowledge-hub/psd3), accessed 2026-10-31). **PROPOSED.** - **MiCA (Markets in Crypto-Assets):** entered into force June 2023 and fully applied from December 2024. The grandfathering period ended **1 July 2026** ([source](https://www.esma.europa.eu/esmas-activities/digital-finance-and-innovation/markets-crypto-assets-regulation-mica), accessed 2026-10-31). **AGREEMENT.** - **AMLA and AMLR:** - The EU Anti-Money Laundering Authority (AMLA) has existed legally since 26 June 2024, started operations in summer 2025, and begins direct supervision in 2028 ([source](https://www.amla.europa.eu/about-amla_en), accessed 2026-10-31). **AGREEMENT.** - The AML Regulation (AMLR, Regulation (EU) 2024/1624) applies from 10 July 2027 ([source](https://eur-lex.europa.eu/eli/reg/2024/1624/oj/eng), accessed 2026-10-31). **NOT RE-VERIFIED.** - **eIDAS 2 (Regulation (EU) 2024/1183):** each Member State must offer at least one EU Digital Identity Wallet, with launch expected by the end of 2026 ([source](https://ec.europa.eu/digital-building-blocks/sites/display/EUDIGITALIDENTITYWALLET/EU+Digital+Identity+Wallet+Home), accessed 2026-10-31). **AGREEMENT.** - The duty on banks to accept wallets for strong authentication, and its date: **UNKNOWN** (not on the Commission page; EUR-Lex blocked). - **Instant payments (Regulation (EU) 2024/886):** instant credit transfers make funds available "within ten seconds" ([source](https://www.ecb.europa.eu/paym/integration/retail/instant_payments/html/index.en.html), accessed 2026-10-31). - Euro-area deadlines: receive by 9 January 2025; send and verify payee by 9 October 2025. **NOT RE-VERIFIED**: the ECB milestones page returned 404. ###### 4.2 United States - **Bank regulators:** the OCC (Office of the Comptroller of the Currency) charters national banks. The Fed (Federal Reserve) and the FDIC (Federal Deposit Insurance Corporation) supervise banks and insure deposits. **NOT RE-VERIFIED**: the OCC page redirected and the Fed and FDIC pages were not opened. - **GENIUS Act (stablecoins):** - Effective on the earlier of 18 months after enactment (by 18 January 2027) or 120 days after final rules. - The 18 July 2026 rulemaking deadline passed without final rules (secondary source [source](https://www.crowdfundinsider.com/2026/07/292837-one-year-anniversary-of-the-genius-act-becoming-law-still-waiting-on-the-clarity-act/), accessed 2026-10-31). **AGREEMENT** (law), with implementing rules **PROPOSED**. - **CFPB section 1033 (personal financial data rights):** - The final rule came on 22 October 2024. The CFPB issued a reconsideration notice (ANPR) on 22 August 2025 ([source](https://www.consumerfinance.gov/personal-financial-data-rights/), accessed 2026-10-31). - Enforcement is enjoined by a federal court (Forcht Bank v. CFPB, 29 October 2025) pending reconsideration (secondary source [source](https://openbankingtracker.com/regulation/usa-1033), accessed 2026-10-31). **AGREEMENT on paper, not enforceable now.** - **FinCEN (money services businesses):** money transmitters are MSBs with no activity threshold. They must register (Form 107) within 180 days and renew every 24 months ([source](https://www.fincen.gov/money-services-business-definition), accessed 2026-10-31). **AGREEMENT.** - **OFAC (Office of Foreign Assets Control):** report blocked property within 10 business days ([source](https://ofac.treasury.gov/faqs/topic/1501), accessed 2026-10-31). **AGREEMENT.** - **SEC/CFTC and crypto (CLARITY Act):** the CLARITY Act passed the House but not the Senate, so it is not law (secondary source [source](https://eco.com/support/en/articles/16060426-clarity-act-timeline-house-passage-senate-progress-and-what-s-next), accessed 2026-10-31). **PROPOSED.** - **State money transmitter licences:** 31 states have enacted the Money Transmission Modernization Act in full or part ([source](https://www.csbs.org/csbs-money-transmission-modernization-act-mtma), accessed 2026-10-31). **AGREEMENT** (state by state). - **State AI laws:** Colorado SB 26-189 (signed 14 May 2026) repealed and replaced the 2024 AI Act with an automated-decision regime covering lending, effective **1 January 2027** (secondary source [source](https://www.eckertseamans.com/legal-updates/colorado-repeals-and-replaces-its-landmark-ai-statute-what-businesses-need-to-know), accessed 2026-10-31). **AGREEMENT** (not yet in force). Other states: **UNKNOWN** (search budget exhausted). ###### 4.3 International standard setters - **FATF (Financial Action Task Force):** - Revised Recommendation 16 (payment transparency) was agreed June 2025, with effect by end of 2030 ([source](https://www.fatf-gafi.org/en/publications/Fatfrecommendations/update-Recommendation-16-payment-transparency-june-2025.html), accessed 2026-10-31, via search; direct fetch 403). **AGREEMENT** (standard). - Draft R16 guidance was consulted on until 21 August 2026 ([source](https://www.fatf-gafi.org/en/publications/Fatfrecommendations/R16-Public-Consultation-June-2026.html), accessed 2026-10-31, via search). **PROPOSED.** - Digital identity guidance (2020): whether it has been updated is **UNKNOWN** (FATF site returned 403; no update found in search). - **Basel Committee on Banking Supervision:** "Principles for the sound management of third-party risk" was a consultative document of 9 July 2024 ([source](https://www.bis.org/bcbs/publ/d577.htm), accessed 2026-10-31). Final version: **UNKNOWN**. - **FSB (Financial Stability Board):** - "The financial stability implications of artificial intelligence" (14 November 2024) flags third-party concentration, correlations, cyber, model risk and governance ([source](https://www.fsb.org/2024/11/the-financial-stability-implications-of-artificial-intelligence/), accessed 2026-10-31). - The page lists "Sound Practices for Responsible Adoption of AI: Final report" (15 October 2026) (same source). Its content was not read. - **IOSCO (International Organization of Securities Commissions):** its AI report could not be fetched (403). **UNKNOWN.** - **Singapore MAS (Monetary Authority of Singapore):** - The FEAT principles (fairness, ethics, accountability, transparency) date from 2018. - The Veritas project closed its third phase on 26 June 2023 with an open-source toolkit ([source](https://www.mas.gov.sg/schemes-and-initiatives/veritas), accessed 2026-10-31). - The Guidelines on AI Risk Management were issued 7 October 2026 and take effect **7 October 2027**. They apply to all financial institutions and all forms of AI ([source](https://www.mas.gov.sg/regulation/guidelines/guidelines-on-artificial-intelligence-risk-management-for-financial-institutions), accessed 2026-10-31). **AGREEMENT.** --- ##### 5. What regulators have said about AI and agentic AI, and what is unresolved ###### What they have said - **FCA approach:** no extra AI regulation. It is principles-based and relies on the Consumer Duty and SM&CR. First published 8 September 2025 ([source](https://www.fca.org.uk/firms/innovation/ai-approach), accessed 2026-10-31). - **FCA AI Lab:** - Supercharged Sandbox (with NVIDIA). - AI Live Testing: second-cohort applications ran 19 January to 24 March 2026, with testing from late April 2026. - AI Input Zone (closed 19 June 2026), with good and poor practice examples due later in 2026 ([source](https://www.fca.org.uk/firms/innovation/ai-lab), accessed 2026-10-31). - **AI Live Testing cohort 2:** eight firms including Barclays, Experian, GoCardless and Lloyds. Use cases include **agentic payments**, AML and KYC. Testing ends late 2026, with an evaluation report in Q1 2027 ([source](https://www.fca.org.uk/news/press-releases/fca-announces-second-cohort-ai-live-testing), accessed 2026-10-31). - **Mills Review (6 July 2026):** seven recommendations, including "Enable the foundations for agentic finance" and "Monitor the transition to autonomous models and adapt regulatory frameworks" ([source](https://www.fca.org.uk/publications/corporate-documents/mills-review), accessed 2026-10-31). **PROPOSED.** - **BoE/FCA survey (21 November 2024):** - 75% of firms already use AI. - 55% of use cases involve some automated decision-making, but only 2% are fully autonomous. - 46% of firms report only partial understanding of their AI ([source](https://www.bankofengland.co.uk/report/2024/artificial-intelligence-in-uk-financial-services-2024), accessed 2026-10-31). - **HM Treasury (July 2026):** "the PSRs were designed before the development of AI and may not fully facilitate the use of agentic AI". It asks about authentication, consent and liability ([source](https://assets.publishing.service.gov.uk/media/6a54b3302467584757371dc4/Modernising_payment_services_consultation.pdf), accessed 2026-10-31). **PROPOSED.** ###### What is unresolved for AI agents as customers 1. **Legal personhood and KYC for agents:** - The MLRs define customer due diligence and beneficial owners in terms of individuals and bodies corporate ([source](https://www.legislation.gov.uk/uksi/2017/692/regulation/5), accessed 2026-10-31). - We found no UK rule or guidance on verifying an AI agent itself. **UNKNOWN.** We searched the FCA AI approach, AI Lab, the Mills Review page and the HMT consultation. - Safe design: run full KYC on the accountable owner, and register each agent as a delegated actor with its own credential and mandate. 2. **Liability for agent-initiated payments:** HMT asks whether liability rules for unauthorised transactions need updating ([source](https://assets.publishing.service.gov.uk/media/6a54b3302467584757371dc4/Modernising_payment_services_consultation.pdf), accessed 2026-10-31). Today's answer: **UNKNOWN**. 3. **Consent and mandates:** HMT asks whether "consent of payments transactions" provisions need updating (same source). The Open Banking 90-day consent reconfirmation was written for humans ([source](https://www.fca.org.uk/firms/strong-customer-authentication), accessed 2026-10-31). **UNKNOWN** how it applies to agents. 4. **SCA with agents:** - SCA is required when the payer initiates a payment ([source](https://www.fca.org.uk/firms/strong-customer-authentication), accessed 2026-10-31). - HMT plans to move authentication into outcomes-based FCA rules (consultation above). **PROPOSED.** - How an agent "authenticates" as the payer: **UNKNOWN**. 5. **Explainability and contestability:** - UK GDPR Articles 22A–22D (from 5 February 2026) require safeguards for significant solely automated decisions: information, challenge and human intervention ([source](https://www.gov.uk/guidance/data-use-and-access-act-2025-data-protection-and-privacy-changes), accessed 2026-10-31). - EU credit-scoring AI is high-risk from 2 December 2027 ([source](https://ai-act-service-desk.ec.europa.eu/en/ai-act/article-113), accessed 2026-10-31). - MAS guidelines apply from 7 October 2027 ([source](https://www.mas.gov.sg/regulation/guidelines/guidelines-on-artificial-intelligence-risk-management-for-financial-institutions), accessed 2026-10-31). 6. **Accountability inside the firm:** the FCA expects a named senior manager under SM&CR to be accountable ([source](https://www.fca.org.uk/firms/innovation/ai-approach), accessed 2026-10-31). --- ##### 6. Obligations table | Obligation | Regulator | Applies when | Bare minimum | Ambitious glass-box version | |---|---|---|---|---| | Consumer Duty ([source](https://www.fca.org.uk/firms/consumer-duty), accessed 2026-10-31) | FCA | Any FCA-regulated retail product (including via a partner) | Fair-value assessment, clear communications, support journeys | Per-customer and per-agent outcome logs. Fair-value dashboards visible to customers and the board. | | SM&CR ([source](https://www.fca.org.uk/firms/innovation/ai-approach), accessed 2026-10-31) | FCA, PRA | Once EDGE holds FSMA authorisation | Named senior manager per function, including AI | Each agent and model mapped to a named accountable senior manager in a public register. | | Customer due diligence, £12,000 / £800 triggers ([source](https://www.legislation.gov.uk/uksi/2017/692/regulation/27), accessed 2026-10-31) | FCA, HMRC | Business relationship or occasional transaction | KYC on every human and legal-entity customer | KYC on the owner plus a verified agent credential bound to a mandate. Every agent action traceable to a verified human. | | Beneficial owners over 25% ([source](https://www.legislation.gov.uk/uksi/2017/692/regulation/5), accessed 2026-10-31) | FCA | Corporate customers | Identify owners over 25% | Full ownership graph with change alerts, shown to the customer. | | Record keeping 5 years, 10-year cap ([source](https://www.legislation.gov.uk/uksi/2017/692/regulation/40), accessed 2026-10-31) | FCA | All CDD and transaction records | Retain, then delete | Tamper-evident log with automated deletion and deletion receipts. | | SARs, DAML timing and no tipping off ([source](https://www.legislation.gov.uk/ukpga/2002/29/section/333A), accessed 2026-10-31) | NCA | Knowledge or suspicion of money laundering | Nominated officer, SAR Portal, DAML clocks | Agents technically blocked from seeing SAR status. Clock tracking for 7 working days and 31 days. | | Sanctions screening and reporting ([source](https://www.gov.uk/government/publications/financial-sanctions-general-guidance), accessed 2026-10-31) | OFSI | All customers and payments | Screen against the UK Sanctions List | Real-time screening of owners, agents and counterparties, with explained match decisions. | | SCA ([source](https://www.fca.org.uk/firms/strong-customer-authentication), accessed 2026-10-31) | FCA | Online account access and payments | SCA by the account provider (or partner) | Owner-set agent spending envelopes with SCA at mandate creation. Step-up SCA outside the envelope. | | APP fraud reimbursement, £85,000, 13 months, 5 business days ([source](https://www.psr.org.uk/information-for-consumers/our-new-app-fraud-reimbursement-protections/), accessed 2026-10-31) | PSR, moving to FCA | Faster Payments and CHAPS consumer payments | Reimbursement process and 50:50 claims | Agent-initiated payment provenance attached to every claim. Public reimbursement statistics. | | Confirmation of Payee ([source](https://www.wearepay.uk/what-we-do/overlay-services/confirmation-of-payee/), accessed 2026-10-31) | PSR, Pay.UK | Sort-code account payments | Use via the partner or an aggregator | CoP result shown to the owner and fed into agent decision logs. | | Direct Debit Guarantee ([source](https://www.directdebit.co.uk/direct-debit-explained/direct-debit-guarantee/), accessed 2026-10-31) | Pay.UK | Collecting or paying Direct Debits | Scheme compliance | Agent cannot create Direct Debits outside the mandate. Immediate refund flow. | | Open Banking consent every 90 days ([source](https://www.fca.org.uk/firms/strong-customer-authentication), accessed 2026-10-31) | FCA | Account information services | Reconfirm consent | Owner-visible consent ledger per agent, with one-tap revoke. | | Safeguarding ([source](https://www.fca.org.uk/publications/policy-statements/ps25-12-changes-safeguarding-regime-payments-and-e-money-firms), accessed 2026-10-31) | FCA | EMI or payment institution holding funds | Daily reconciliation, audit, resolution pack | Live public safeguarding attestation. | | Operational resilience and incident reporting ([source](https://www.fca.org.uk/firms/operational-resilience), accessed 2026-10-31) | FCA, PRA, BoE | Authorised payment, e-money and bank firms | Important business services and impact tolerances. Incident reporting from 18 March 2027. | Published service status and incident history, with third-party (CTP) dependency map. | | Creditworthiness and affordability (CONC 5.2A, NOT RE-VERIFIED) ([Handbook link](https://www.handbook.fca.org.uk/handbook/CONC/5/2A.html), accessed 2026-10-31) | FCA | Consumer lending | Assessment with income and expenditure | Reason codes shown to the borrower. Human review on request (UK GDPR Article 22A). EU AI Act readiness for 2 December 2027. | | Pre-contract information and 14-day withdrawal ([source](https://www.legislation.gov.uk/ukpga/1974/39/section/66A), accessed 2026-10-31) | FCA | Regulated credit agreements | Pre-contract information form and withdrawal process | Agent cannot accept credit without an owner-signed mandate. Withdrawal one tap away. | | Automated decisions (UK GDPR Articles 22A–22D) ([source](https://www.gov.uk/guidance/data-use-and-access-act-2025-data-protection-and-privacy-changes), accessed 2026-10-31) | ICO | Significant solely automated decisions | Inform, allow challenge, human intervention | Per-decision explanation record available to customer and regulator. | | Credit reference permission ([source](https://www.legislation.gov.uk/uksi/2001/544/article/89B), accessed 2026-10-31) | FCA | Mainly furnishing credit information | Do not do it, or get permission | Customer-visible data lineage for every score. | | Crypto authorisation and promotions ([source](https://www.fca.org.uk/publications/policy-statements/cryptoasset-regime), accessed 2026-10-31) | FCA | Any UK crypto activity | Avoid crypto, or apply 30 September 2026 to 28 February 2027 | Appropriateness checks for both owner and agent, and risk warnings shown before any agent trade. | | FOS route, £455,000 limit ([source](https://www.financial-ombudsman.org.uk/news/fca-confirms-increase-award-limits-2), accessed 2026-10-31) | FOS | Eligible complainants | Complaints process with FOS referral rights | Complaint packs that include the full agent action trail. | --- ##### 7. What to check next (UNKNOWN items) The following were searched or fetched without success on 31 October 2026. The web-search budget was exhausted. - JROC's current status. - The current Confirmation of Payee direction (Specific Direction 17) deadlines (PSR page returned 404). - Whether the APP fraud reimbursement cap moved with the FSCS rise to £120,000. - Consumer Credit Act reform status (gov.uk returned 503). - The text of CONC 5.2A, CONC 7 and COBS 4.12A (Handbook pages did not render). - PRA minimum capital for new banks. - Rules on agents of payment and e-money firms. - Whether the ICO has been renamed. - The eIDAS 2 bank acceptance date. - EUR-Lex dates for DORA, AMLR and instant payments (fetch blocked). - FATF digital identity guidance updates and the IOSCO AI report (403). - Final Basel third-party principles. - US state AI laws beyond Colorado. - Royal Assent for the Financial Services and Markets Bill. ### research/finance/SCREENS-AND-FLOWS.md `research/finance/SCREENS-AND-FLOWS.md` · sha256 `6ea3ab7cd337d6370022ac67219408b08cd2d9345db1d181626fcf92912c4644` #### EDGE Finance research: competitors' real screens and flows **What this is.** A record of how existing finance products actually work, screen by screen, taken from Refero (a library of real app and website screenshots with each user journey described step by step). It tells us the floor: what users already expect, so EDGE Finance matches it and then does better. **Statement class.** Every flow below is an OBSERVATION: Refero's recording of a real product, observed on the date shown. Refero's descriptions are its own wording. Product names belong to their owners. **Source and date.** Refero MCP, searched 2026-10-31 with 15 banking task queries on iOS and web. Each flow links to its Refero page. **Coverage, honestly.** Refero holds Revolut, Wise, PayPal, Mercury, Wealthsimple, Square, Tide, Column, Coinbase and several crypto wallets. It does **not** hold Starling, Monzo, Kraken or Chase UK; those need Mobbin (not connected in this session). Refero returned only a few Revolut flows. So this is a strong sample of the floor, not a complete census. ##### Summary of what the floor is | Task | What every good product already does (bare minimum) | |---|---| | Open an account | Email or phone, then a step list (basic details, government ID, selfie), address autocomplete, progress shown during checks, a clear success or retry path. Identity checks are usually outsourced (Persona, Bridge, Onfido-type providers). | | Open a business account | Company lookup, directors and owners, company documents, purpose of account, then the same personal checks for each person. Long: Mercury's runs to 67 screens. | | Send money | Pick or create a recipient, amount, currency conversion with the rate and fee shown before you commit, review screen, confirmation, receipt. | | Request and split | Choose people, amount, note; cancel a request; reminders. | | Schedule | Date, repeat rule, end condition; a list of scheduled items to edit or cancel. | | Cards | Create virtual card instantly, freeze and unfreeze in one tap, view details behind authentication. | | Disputes | Pick the transaction, choose a reason, add evidence, see the case status and talk to the merchant or support in a thread. | | Security | Two-factor setup with an authenticator app, recovery codes, ability to turn methods off with re-authentication. | | AI assistant | PayPal's assistant answers questions in chat and links to help, but takes no action on the account. | | Team access | Invite a person by email, choose a role and permissions, they accept and onboard. | ##### What none of them do (the gap EDGE Finance fills) 1. **No agent is ever the customer.** Every flow assumes a human with a phone and a face. None lets an AI agent open an account, hold its own identity, be verified (KYA, know your agent), or act as lender, borrower, payer or payee. 2. **Delegation stops at people.** Mercury's team invites are the closest thing: a person gets a role. There is no way to give an agent a scoped mandate (what it may do, spend limits, expiry, revocation) and see everything it did. 3. **The assistant cannot act.** PayPal's assistant explains; it does not perform the action with a preview, approval and undo. 4. **No rehearsal.** No product lets you (or your agent) rehearse a payment, loan or transfer and see the full outcome before committing. Only developer sandboxes (Column's "simulate a transfer") come close, and they are for engineers, not customers. 5. **Black-box decisions.** When verification fails, the user sees "Couldn't verify information" and a retry button: no reason, no evidence, no route to challenge. EDGE Finance shows why, with the evidence, as a glass box. ##### Flows by task ###### Opening an account and proving who you are (KYC) **Fuse: Virtual Bank Account Signup** (ios, 24 screens) [Refero](https://refero.design/flows/8730) > Happy path: From an empty Fuse Wallet the user opens a promotional card to create a virtual bank account (USD or EUR), chooses a currency, supplies an email, and is routed to a third-party identity verification (Bridge → Persona). The user completes the Basic Information step (country, name, birthdate, SSN, address with autocomplete) and submits. Steps: Wallet Home (empty) → Currency Selection (USD shown) → Currency Selection (EUR shown) → Bridge - Email capture → Bridge - Email entered → Connecting / Processing → Full-screen Loading → Persona Consent (Getting started) → Required submissions list → Country selection → Basic Information (form start) → Typing name fields → Date picker (scrolling) → Date selected → SSN entry placeholder → Typing SSN (masked) → SSN visible (toggle) → Address autocomplete → Address details (expanded) → Verification failure → … **UGLYCASH: New user onboarding and verification** (ios, 19 screens) [Refero](https://refero.design/flows/11202) > End-to-end new-user onboarding for a trading app: marketing carousel -> sign-in options -> phone verification (WhatsApp/SMS) -> Apple sign-in fallback -> username + referral -> legal consent -> biometric enablement -> first-time home. Steps: Welcome carousel - follow traders → Welcome carousel - trade everything → Welcome carousel - trading clubs → Welcome carousel - custody ownership → Sign-in selection → Phone entry (channel chooser) → Phone entry (send via SMS) → Code entry (empty) → Code entry (auto-filled) → Linked-account conflict modal → Apple Sign In prompt → Apple signing in → Choose username → Username claimed (input filled) → Referral code entry → Accept terms and conditions → Enable Face ID prompt → Face ID OS permission → First-time Home (authenticated) **Copperx: New user sign-up and KYC onboarding** (web, 18 screens) [Refero](https://refero.design/flows/8892) > End-to-end new user registration and initial onboarding for a financial product (marketing -> signup -> verification -> KYC -> dashboard). Entry: user arrives from marketing/marketing hero or direct signup link. Happy path: user chooses sign-up method (Google or email), verifies email with a 6-digit code, creates a user-password account (with inline password-strength feedback), selects account type and services, fills personal/KYC-related fields, accepts a third-party terms modal when linking bank accounts, and lands on the authenticated dashboard. Exit: user reaches the dashboard (account created but KYC incomplete). Assumptions: user is not signed in at entry; user has access to the email used for signup; device is a desktop browser; user can upload or provide KYC documents later. Steps: Marketing landing / signup entry → Signup method selection → Email entered (confirmation) → OTP entry (empty) → OTP entered (partial) → Account creation form (basic) → Password feedback (weak) → Password feedback (improving) → Password feedback (strong) → Services & account type selection → Service selection (multiple chosen) → Personal information / KYC form → Third-party terms modal (Bridge) → Submission error feedback → Dashboard landing (KYC prompt) → Dashboard left navigation → Empty transactions state → Ongoing verification reminder **Wise: Personal account signup and card onboarding** (web, 18 screens) [Refero](https://refero.design/flows/371) > Happy-path sign-up and onboarding to order a card. Entry: user arrives from the marketing/landing hero and clicks the primary CTA to start registration. The user is not logged in and proceeds through a modal-style multi-step registration: enter email (or social SSO), choose account type (Personal/Business), confirm country of residence, verify phone with SMS 2FA, create a password, see a registration success screen, choose next actions, complete personal details (name, DOB, address) and lands on the Cards dashboard where they can order a card. Assumptions: desktop browser, user has a valid email and phone number, user intends to open a personal account and order a card. Exit: user arrives on Cards area (dashboard) and can continue ordering a physical card or other actions. Steps: Landing / Hero → Create account — Email entry → Email submitted → Account type selection → Country of residence → Phone entry (2FA) — empty → Phone entry (2FA) — filled → Enter 6‑digit code — prompt → Code validation error → Code accepted → Create password — empty → Create password — filled → Registration success → Post-signup choices → Tell us about yourself — form → Form — filled and month picker → Form confirmed — profile saved → Cards dashboard **Mercury: Upload and verify government ID** (web, 3 screens) [Refero](https://refero.design/flows/11021) > Happy-path: A logged-in user (top-right account shown) manually verifies identity by selecting an ID type, uploading a color photo of their ID, and submitting a selfie if required. Entry: user lands on an Identification screen with instructions and an ID-type selector. Primary actions: choose identification type, upload or drag-and-drop an uncropped photo of the ID, optionally take a live photo of self. Exit: the system either accepts the uploads or shows validation warnings/errors and guidance; in the screenshots the system returns issues that may require a resubmission or follow-up. Assumptions: user is authenticated; they have access to photos or a camera; permissions to upload files and to take a selfie are available. Steps: Identification instructions and upload → Upload preview and processing → Validation results and remediation **Mercury: Identity verification with webcam selfie** (web, 4 screens) [Refero](https://refero.design/flows/11022) > Happy path: A user completes manual identity verification by selecting an ID type, uploading a photo of the ID, and capturing a live selfie via webcam. The flow opens on an Identification form where the user picks an ID type and either uploads the ID image or drags and drops it. The user clicks Take photo to open a modal camera tool, captures a selfie, confirms the photo, and returns to the form where the uploaded selfie is shown. The user can optionally add a professional profile link and save to proceed into the final terms section. Steps: Identification form → Camera modal — live preview → Photo review → Form with uploaded selfie **Family: Create account and wallet** (web, 31 screens) [Refero](https://refero.design/flows/6184) > Happy-path onboarding for creating a new Family account and a new wallet from the marketing landing entry. Entry: user begins on the public landing and chooses a platform (mobile or desktop). The flow proceeds to an email / phone sign-in that creates an account if one doesn't exist, followed by email verification, password creation, optional account customization (avatar + name), optional wallet customization (icon + wallet name), and ends at the Wallets dashboard showing the newly created wallet. Assumptions: user is on desktop (web) and not previously authenticated; they have access to the email address provided to receive the verification code. The user may skip customization steps; the system supports creating default names/icons if skipped. Steps: Landing / Intro → Choose Platform → Welcome — Identifier Input → Submit Identifier → Verification Code Prompt → Enter Verification Code → Verification Error → Set Password → Submit Password → Customize Account — Avatar Picker → Choose Emoji/Icon → Avatar Action Menu → Upload or Select Image → Set Account Name → Customize Wallet — Default Icon → Wallet Avatar Menu → Pick Wallet Color/Icon and Name → Finalize Wallet Creation → Wallets Dashboard — List → Post-onboarding options → … **Fuse: New wallet signup and secure key setup** (ios, 22 screens) [Refero](https://refero.design/flows/8729) > Happy-path onboarding for a crypto wallet: user is introduced to product pillars (Spend / Save / Invest), chooses to create an account using Sign in with Apple, and runs through a 3-part key setup (Device Key — biometric, 2FA Key — iCloud, Recovery Key — email OTP). After key creation the wallet is provisioned, a short promo for a paid membership is shown, and the user arrives at the empty wallet home. Steps: Onboarding carousel — Spend → Onboarding carousel — Save → Onboarding carousel — Invest → Welcome / Sign-up choices → Sign in with Apple sheet → Signing-in progress → Secure your wallet — Device Key intro → Face ID permission prompt → Creating Device Key (progress) → Create 2FA Key (iCloud) prompt → Recovery Key — email entry → OTP entry — keyboard & autofill → OTP verification and provisioning → Locked state — Unlock CTA → Promo modal — free transactions → Fuse Plus membership detail → Wallet provisioning complete — home (empty) → Empty wallet CTA — Receive → Default bottom navigation & quick actions → End state — ready wallet → … **Phantom: Create New Wallet Account** (ios, 7 screens) [Refero](https://refero.design/flows/3695) > Happy path: From the Your Accounts list the user taps Add / Connect Wallet, selects Create New Account, types a custom account name, and confirms creation. The app then opens the new account's home (balances and actions). Returning to the accounts list shows the new account added and highlighted. Assumptions: user is authenticated and on a mobile device; permissions to create accounts are available; this is the standard "create new" path (not importing or connecting hardware). Steps: Your Accounts (start) → Add / Connect Wallet Options → Create Account (name input shown) → Enter Account Name → Create Account (confirm) → New Account Home → Your Accounts (new account added) **Wealthsimple: Open Chequing Account** (web, 15 screens) [Refero](https://refero.design/flows/8496) > Happy-path onboarding for opening a new deposit account starting from a logged-in user's dashboard. Steps: Dashboard — Add an account → Account selection → Product details — Chequing → Intro — Tell us about yourself → Form — Legal name & date of birth (empty) → Form — Legal name & DOB (completed) → Form — Phone number (empty) → Form — Phone number (completed) → Form — Residency, tax residency & citizenship (initial) → Validation — Citizenship requirement error → Form — Residency & citizenship (resolved) → Form — Social Insurance Number (empty) → Validation — SIN format error → Form — SIN masked and accepted → Validation — SIN temporary / expiry notice ###### Opening a business account and proving the business (KYB) **Mercury: Business account onboarding & setup** (web, 67 screens) [Refero](https://refero.design/flows/11018) > End-to-end new business account onboarding for Mercury. Entry begins on the public marketing site and proceeds into an in-product sign-up widget where the user provides name, email and password, then sets basic company metadata (company name, callsign), enables two‑factor authentication (2FA), optionally opts into SMS updates, and is guided to the in-app dashboard. Steps: Marketing landing → Get started — name entry → Get started — name filled → Create account — blank credentials → Create account — credentials filled → Password visibility tooltip → Password visibility toggled → Company name input → Company name — filled → Callsign setup → 2FA — Mercury app (QR) option → 2FA — other authenticator option → 2FA validation — incomplete code error → 2FA validation — code entry → Onboarding checklist — progress → Get mobile updates — phone on file → Checklist — all required completed → Dashboard — welcome view → Invite recipient modal → Dashboard — default state → … **PayPal: Business account setup** (web, 20 screens) [Refero](https://refero.design/flows/7417) > Happy-path user journey for converting or creating a PayPal business account and completing required verification. Entry: signed-in personal account (user visible on account settings). The user is prompted to set up a business account, selects account type (Individual/Business/Nonprofit), chooses whether to create a new business account or convert an existing personal account, fills registration fields (name, email, password, phone), completes phone verification (SMS code), provides personal identity and address information, confirms a normalized address, completes business profile (industry, sales, currency, website, registered address), optionally uploads supporting documents, and finally lands on the business dashboard. Exit: user reaches the PayPal business dashboard with prompts to complete profile/verification. Steps: Account settings (start) → Business account intro → Select account owner type → Individual option expanded → Create vs Convert choice → Convert selection state → Create account - registration form (blank) → Create account - form filled → PSD2 phone verification prompt → Enter confirmation code (empty) → Confirmation code entered → Phone verified - success → Share personal details - form → Date picker validation → Personal details - address filled → Address confirmation modal → Tell us about your business (blank) → Business details populated → Add supporting info → Merchant dashboard (landing) **Square: Merchant onboarding and account setup** (web, 38 screens) [Refero](https://refero.design/flows/1952) > Happy-path merchant onboarding from marketing into the Square dashboard. Entry: visitor clicks through marketing/CTA to sign up. The flow collects account credentials, business entity details, business category, identity verification (SSN), payout preferences, and bank account details. It then shows a summary of chosen products, offers app download, and lands the user in the Square dashboard. Steps: Marketing / Signup entry → Create account (empty) → Create account (filled) → Submission loading → Business entity question → Entity type options → Business details form → Business details (filled) → Business category picker → Category search results → Full SSN requested → Verification failure / Retry → Choose payout frequency → Payout option selected → Link bank account (empty) → Link bank account (filled) → Selections overview → Welcome / Next steps → Download Point of Sale → Dashboard (first sign-in) → … **Acctual: Account onboarding and payments setup** (web, 26 screens) [Refero](https://refero.design/flows/8817) > Happy-path onboarding for a payments/invoicing product. Entry from the marketing/home hero (promotional landing) into account creation. The user either authenticates with Google or supplies email + password, then selects who will be paid (Individual or Business). For a Business, the user supplies company name and contact name, triggers an email verification, clicks the verification link in their inbox, then returns to complete branding and business details (logo, postal address, tax ID). After submitting those details the user reaches a confirmation screen and is taken to the Payments area where they can continue to configure receiving methods and identity verification. Steps: Marketing / Landing Hero → Create account (empty) → Email entered → Password reveal toggle → Choose who gets paid → Business selected & fields displayed → Verification pending state → Verification email (inbox) → Make it yours (branding) → Address modal (open) → Country selector list → Country search no-results → Country selection corrected → Add address → Branding preview & inputs → Tax ID entered → You're all set → Payments overview → Payments onboarding CTA → Landing in Payments workspace → … **Acctual: Company document verification** (web, 4 screens) [Refero](https://refero.design/flows/8828) > Summary: This is a modal-based, multi-step document submission sequence for verifying a company's identity. Steps: Select document type and upload (Legal presence) → Upload ownership structure document → Upload company details / proof of address → Review uploaded documents and continue ###### Adding people to an account (the closest thing today to delegating to an agent) **Mercury: Add team member** (web, 5 screens) [Refero](https://refero.design/flows/11023) > Happy path: A logged-in user (top-right shows account) completes the onboarding checklist by adding a team member. Steps: Onboarding checklist - Invite members → Add team member modal (empty) → Role selection → Set spending limit and confirm → Invite added and pending approval **Mercury: Invite and onboard team member** (web, 10 screens) [Refero](https://refero.design/flows/11071) > Happy path: An admin (already signed in) opens the Users list from Settings, clicks Invite → Invite a user, completes a three-step modal (User Details → Assign Role → Issue Card), reviews the invitation, and sends it. The flow ends with an on-screen confirmation and the invited user appearing in the Users list with a right-hand detail panel showing their pending invite, assigned role and card options. Entry: Settings → Users (signed-in admin). Exit: Confirmation screen and return to Users list with the new invite added. Assumptions: the actor is logged in with permission to invite users; using a desktop browser (layout indicates wide-screen). Steps: Users list → Invite menu → Invite — User Details (start) → Invite — User Details (filled) → Assign role — preview (Employee) → Assign role — Admin selected → Issue card options → Review and send invite → Invite sent confirmation → Users list with invite detail panel ###### Linking an outside bank account and funding **PayPal: Bank selection and search** (ios, 4 screens) [Refero](https://refero.design/flows/7240) > Mobile 'Link your bank account' selection flow (happy path: user finds and selects their bank). Entry: user reaches the 'Link your bank account' screen (mobile, iOS keyboard patterns visible). The screen shows a searchable list of popular banks and a prominent search field. The user can either pick a bank from the visible list or tap the search field to type the bank name (or partial name). Search shows an empty state message when no input/results and updates the list with filtered matches as the user types. Exit point (not shown): user taps a bank result to proceed to the bank linking/authentication step (next screen not included). Assumptions: user is on a mobile device, logged-in state and permissions to link a bank are already granted by prior flow steps. Steps: Popular banks list → Search field focused — empty state → Search with numeric input — no results → Search with partial text — filtered results **PayPal: Bank Selection for Linking** (web, 3 screens) [Refero](https://refero.design/flows/7379) > A signed-in user is presented with a modal to choose a bank to link to their PayPal account. Steps: Bank Selection List → Search — No Matches → Search — Single Match **Mercury: Initial deposit bank connection** (web, 5 screens) [Refero](https://refero.design/flows/11026) > This flow guides a logged-in user through connecting a US bank account (via Plaid) so they can schedule their first deposit. Steps: Onboarding checklist — Connect account → Plaid consent / intro modal → Select your institution → Bank login redirect (Wise example) → No compatible accounts error **Wealthsimple: Add & verify bank account** (web, 18 screens) [Refero](https://refero.design/flows/8504) > Happy-path: A logged-in user on the Move funds area chooses to Add money and links a Canadian bank account. The flow alternates between in-product modals: choose bank (or search), optionally authenticate using the bank's online credentials, or fall back to a manual connect where the user provides transit + account numbers and uploads a void cheque. The flow shows progress screens while connecting and ends with a submission confirmation (account submitted for review). Steps: Move funds hub → Add money entry modal → Bank selection list → Search input (by routing number) → Bank result selected → Bank sign-in info → Bank credential sign-in (empty) → Bank credential sign-in (filled) → Connecting securely → Informational tip (Did you know) → Progress countdown → Manual connect intro → Enter transit + account numbers (empty) → Enter transit + account numbers (filled) → Verify ownership instructions → File attached → Submission processing → Submission confirmation **Wealthsimple: Account transfer onboarding** (web, 23 screens) [Refero](https://refero.design/flows/8502) > Happy-path flow for transferring an external investment account into the user's Wealthsimple account. Steps: Registration success / CTA → Transfer intro / Select account type → Account type list → Account type chosen / Next → Select originating institution → Institution selected → Select institution division → Connect vs enter manually → Enter account number → Account number entered → Transfer details form (initial) → Choose destination account → Select transfer method → Provide estimated value & USD flag → USD account subscription prompt → Optional statement upload → Statement uploaded confirmation → Review summary and signature (empty) → Signature provided / Submit → Transfer submitted confirmation → … **Wealthsimple: Add money to investment account** (web, 8 screens) [Refero](https://refero.design/flows/8505) > Happy-path: A logged-in user navigates to Move funds and chooses Add money to deposit funds from a linked bank account into an investment account. They pick a source account, pick a destination investment account, enter amount, choose a start date and optional recurrence, review a confirmation screen, then submit. The flow ends with a processing/confirmation modal and the system will later update the user's activity feed. Assumptions: user is signed in on a desktop device, has at least one linked bank account and at least one investment account, and has permission to initiate deposits in CAD. Steps: Move funds landing → Add money modal — initial → Select source account → Select destination account → Enter amount and frequency → Ready to continue → Confirm deposit details → Processing / acknowledgement **Revolut: Add Money - Card Top-up** (web, 3 screens) [Refero](https://refero.design/flows/1343) > User is on an Add Money flow (amount already chosen in the background) and opens a 'How to add money' modal that lists two methods: Debit or credit card, and Regular bank transfer. Choosing the card option navigates to a Card details screen where the user must enter card number, expiry date and CVV to top up the visible amount (Add £100.00). The happy path: user provides a supported debit card and the CTA becomes active to complete the top-up. The screenshots show the initial method selection, an empty card-entry form with the CTA disabled, and a populated card form that triggers a validation error indicating credit cards are not supported. Steps: Choose add-money method → Card details (empty) → Card validation error ###### Sending money **PayPal: Send money to contact (currency & payment type)** (ios, 25 screens) [Refero](https://refero.design/flows/7235) > Happy path: A logged-in user starts on the 'Send and Request' hub, searches for or selects a recipient, picks the destination currency, enters an amount, adds an optional message/sticker, chooses a payment type (Goods & Services or Friends & Family), and attempts to send money. Steps: Send & Request hub → More options sheet → Choose a currency (list) → Search currency input → Filtered currency result → Amount screen – initial → Enter amount using keypad → Amount updated with conversion → Message composer (text + stickers) → Compose message text → Sticker gallery → Search recipients → Recipient profile preview → Amount screen for selected recipient → Confirm recipient currency modal → Choose payment type → Payment type selected (friends & family) → Sending attempt / processing → Recipient cannot accept payments error → End / exit point → … **PayPal: Send money to contact (currency conversion)** (ios, 12 screens) [Refero](https://refero.design/flows/7263) > Happy path: Logged-in user opens the Send and Request flow, selects a suggested contact (Bennett Marcus), chooses a currency for the recipient, enters an amount on the numeric keypad, reviews payment details (funding source, goods/services, shipping address, conversion method), confirms the send, and sees a success confirmation. Steps: Send and Request - landing → Amount screen (initial) → Amount screen (recipient currency active) → Enter amount (numeric keypad) → Open Review before sending → Choose funding source (modal) → Shipping address selection → Confirm shipping preference → Currency conversion method → Final review and send → Transaction confirmation → Post-send options **PayPal: Send money with currency conversion** (web, 11 screens) [Refero](https://refero.design/flows/7408) > Happy path for sending a one-off payment to a saved contact using desktop. Steps: Select recipient → Enter amount (initial) → Enter amount (converted) → Add optional note → Quick security check (method) → Enter security code (empty) → Enter security code (filled) → Review payment summary → Choose funding source (modal) → Currency conversion choice (modal) → Confirmation **PayPal: Send or Request Money to Contact** (web, 7 screens) [Refero](https://refero.design/flows/7388) > Logged-in desktop flow for sending or requesting money to a saved contact. Steps: Select recipient → Payment card — enter amounts → Currency selector modal → Payment card — amounts confirmed → Choose payment type — initial → Choose payment type — friends & family selected → Blocking error — account problem **Copperx: Email Payment to New Contact** (web, 9 screens) [Refero](https://refero.design/flows/8896) > Happy-path: a logged-in user on the Copperx dashboard initiates a 'Send' action, chooses to send by email, adds a new contact (payee), enters the amount and purpose, confirms the payment, and receives a success confirmation. Steps: Dashboard (start) → Choose Send Method → Send Money Form (empty) → Recipient Dropdown → Add Payee (empty) → Add Payee (filled) → Confirm Payment → Success Confirmation → Post-send Dashboard State **UGLYCASH: P2P USD Send** (ios, 8 screens) [Refero](https://refero.design/flows/11227) > Happy path: A logged-in user with limited balances starts from the Portfolio tab, taps Transfer, reviews the 'Move Money Globally' introduction and taps Continue, enters an amount on a full‑screen numeric keypad, chooses a source balance (or is prompted to verify identity), then selects a recipient by searching for a member tag/name. Steps: Portfolio - Transfer CTA → Transfer Intro → Amount Entry - Keypad (initial) → Amount Entry - Insufficient Feedback → Select Balance Modal → Recipient Search - Empty Recents → Search No Results → Search Retry / Keyboard Input **Mercury: Create and select recipient** (web, 5 screens) [Refero](https://refero.design/flows/11039) > Happy path: A logged-in user (Mercury Demo account) starts a payment by either uploading a bill or selecting/creating a recipient. If the user creates a recipient, they complete a profile (Person or Business), optionally request tax documents, and enter a postal address. After creation the UI returns to the payment flow with the new recipient pre-selected and the user chooses a payment method to proceed. Steps: Start payment / choose recipient → Create recipient — Profile section → Create recipient — Profile populated → Create recipient — Address section → Recipient returned to payment — choose payment method **Wise: Add bank recipient** (web, 5 screens) [Refero](https://refero.design/flows/369) > Happy path: A logged-in desktop user opens the Recipients area, chooses to add a new recipient in EUR, fills contact and bank details (Inside Europe), the system validates/looks up the IBAN and bank, the user confirms, and the new recipient appears in the recipients list expanded with actionable controls. Steps: Recipients overview → Add recipient modal (initial) → Add recipient — fill details → IBAN validation and bank lookup → Recipients list — newly added recipient **Wise: International Transfer Quote** (web, 3 screens) [Refero](https://refero.design/flows/354) > Happy path: An unsigned (logged-out) desktop visitor lands on the hero section to get a quick quote for sending money internationally. The user enters the amount to send, opens a currency selector for source/destination, uses the inline search to find a currency, selects it, reviews the recipient amount, estimated arrival time and savings, then can choose 'Compare price' or 'Get started' to proceed. Steps: Hero: Transfer Quote Widget → Currency Selector — No Match / Request Prompt → Currency Selector — Search & Select **PayPal: International bank transfer setup** (ios, 24 screens) [Refero](https://refero.design/flows/7244) > Happy path for sending money to a new recipient via bank deposit. Entry starts from the "Send" screen (user tapped "Send to someone new"). The user chooses a destination country, enters the sending amount and currency, selects a receiving option (Bank Deposit), reviews fees, provides recipient bank details and contact address, provides sender address, adds a payment method (debit/credit card) with billing address, and attempts to continue. Exit point is either advancing to the transfer confirmation step (not shown) or stopping the flow when a validation error requires correction or the user quits. Steps: Send entry / New recipient → Country selection → Country search → Select United States → Enter amount and currency → Open receiving options → Show fees for chosen option → Confirm receiving option and continue → Recipient bank details → Contact's details (recipient) → Contact's details — phone & email → Sender's details (address) → Payment method selection → Add new card — input fields → Billing address entry → Card validation fails → User attempts to exit flow → Card type selector (picker) → Card details populated (example) → Billing address completed and re-validate → … **Mercury: Internal funds transfer** (web, 7 screens) [Refero](https://refero.design/flows/11044) > Happy path: A logged-in user starts from the dashboard, opens the Transfer funds modal, enters an amount, selects a source and a destination account (different accounts), reviews the details, confirms the transfer, and sees a success confirmation with an estimated arrival window. Steps: Dashboard (start) → Transfer form - initial → Select source account → Same-account validation → Select destination account (valid) → Review transfer → Transfer confirmation **Acctual: Internal transfer — create and save** (web, 8 screens) [Refero](https://refero.design/flows/8847) > Happy-path: A user on the Bills screen opens the “Transfer funds” modal to create an internal transfer between company wallets and then marks it ready for payment. Steps: Bills list (start) → Open 'New internal transfer' panel → Enter memo, amount and choose asset → Choose network for the asset → Select paying wallet → Select receiving wallet → Validate and submit transfer → Bills list updated with internal transfer ###### Requesting money and splitting bills **PayPal: Peer Payment & Request** (ios, 7 screens) [Refero](https://refero.design/flows/7236) > Happy path: A logged-in user opens a contact profile, taps Send to start a payment/request. The user sees the Amount screen (sender currency = GBP) and is shown a notice that the recipient will receive USD. The user enters an amount using the on-screen keypad, checks the converted amount and rate, optionally opens the currency picker to change recipient currency, then either Request or Send. Choosing Request opens a confirmation bottom sheet; confirming creates a request and shows a success screen. Steps: Contact profile (start) → Amount screen — recipient currency notice → Amount screen — initial state → Amount entry with keypad → Amount entered — real-time conversion → Currency picker → Request confirmation and success **PayPal: Request Payment** (ios, 5 screens) [Refero](https://refero.design/flows/7264) > Happy path: A logged-in user opens a contact's History, chooses "Request Money", enters an amount, reviews the request, and confirms. Steps: Contact History / Actions → Amount Entry (empty) → Amount Entry (filled) → Review Request → Request Sent / Success **PayPal: Request money from contact** (web, 6 screens) [Refero](https://refero.design/flows/7389) > Happy path: A logged-in user navigates to Send and Request, picks a contact, composes a money request (entering a send amount and optional note), selects which currency the recipient will be paid in, reviews the request, and confirms it. Steps: Select contact → Compose request - empty → Enter amount and note → Choose payout currency → Review request → Confirmation **Mercury: Request money - customer selection** (web, 4 screens) [Refero](https://refero.design/flows/11048) > Happy path: A logged-in desktop user begins the "request money" wizard, chooses or creates a customer, then selects the invoice/request type. Steps: Customer selector → Add customer (blank) → Add customer (filled) → Request method selection **PayPal: Bill Split** (ios, 13 screens) [Refero](https://refero.design/flows/7241) > Happy-path: From the app's Send & Request entry the user chooses "Split the bill" from More options, reads a short intro, then enters a total amount (currency selectable). They select participants (their own account is pre-selected), the app auto-calculates an even split, and the user can edit any person's share using an inline editor and numeric keypad. After adjustments the user reviews a final breakdown and taps Request now. The flow ends either with a success (not shown) or an error state where the request fails and the user may retry. Assumptions: user is signed in, has access to contacts/suggested users, and currency selection/keyboard are available. Device: mobile (iOS-style chrome). Steps: Send & Request entry → More options sheet → Bill split intro → Enter total amount (empty) → Amount entered → Choose participants - start → Add participants → Auto-calculated even split → Edit share (pencil) → Input new share amount → Adjusted split with reset option → Review split breakdown → Error / retry **PayPal: Split a Bill** (web, 6 screens) [Refero](https://refero.design/flows/7407) > Happy path: A logged-in user navigates from the Send & Request hub to the "Split a bill" composer, enters a total bill amount, optionally includes their own share, adds one or more people to split with, and the system automatically calculates each person's share (with editable per-person inputs). Steps: Send & Request hub → Split composer — new request → Enter total and enable self-share → Edit a share amount → One participant added — auto calculation → Multiple participants — equal split shown **PayPal: Cancel sent money request** (ios, 6 screens) [Refero](https://refero.design/flows/7265) > Happy path: A logged-in user sees a pending incoming request on their Home activity list, opens the request details, chooses to cancel the request, confirms the cancellation in a modal, and receives a success acknowledgement. The flow entry point is the Home activity list showing a pending request; the exit point is the Home activity list updated to show the request as canceled. Assumptions: user is authenticated, has network connectivity, and has permission to cancel their own sent requests. Steps: Home — Activity list → Request detail — Pending → Confirm cancel — Modal → Cancel acknowledgement — Success modal → Request detail — Canceled → Home — Activity list updated **Wealthsimple: Contact & Transfer Request** (web, 4 screens) [Refero](https://refero.design/flows/8493) > Happy path: A desktop (likely not-logged-in) user visits the Contact Us hub, selects the option to transfer an account, completes the 'Get in touch' transfer form with personal and financial details, submits the form, and receives a confirmation/thank-you screen with suggested help-centre articles and contact options. Steps: Contact hub (support landing) → Get in touch — Transfer form (empty state) → Get in touch — Form filled → Submission confirmation (Thank you) **Square: Requesting payment from conversation** (web, 5 screens) [Refero](https://refero.design/flows/1943) > Happy path: A logged-in merchant (Refero) in the Messages inbox opens a customer's conversation, launches the Request Payment modal, enters an amount, and attempts to send a payment request. The system prevents sending because the merchant has not completed identity verification, then inserts a payment-request card into the conversation marked as failed. Steps: Conversation view / Composer → Request payment modal (empty) → Request payment modal (amount entered) → Identity verification required → Conversation updated — Payment request inserted (failed) ###### Scheduled and recurring payments **Wise: Schedule a transfer** (web, 5 screens) [Refero](https://refero.design/flows/366) > Happy-path: A logged-in user on the Amount step configures a scheduled or recurring transfer before continuing. Steps: Amount screen — open schedule → Schedule modal — open date selector → Calendar — pick a date → Repeats — choose frequency → Confirm or remove schedule **Mercury: Schedule Payment Amount & Recurrence** (web, 4 screens) [Refero](https://refero.design/flows/11041) > This is the 'Amount' step of a multi-step payment scheduling flow where the user confirms the amount, chooses the sending account, sets the send/start date, and optionally configures a recurring payment. Entry: user arrives here after selecting a recipient. Exit: user clicks Next to continue to categorization/details/review steps. Assumptions: user is authenticated, has at least one funding account and a recipient already selected, and is on a desktop or large-screen device. The UI shows estimated arrival timing and allows toggling and configuring recurrence before moving forward. Steps: Confirm Amount & Select Source → Pick Send / Start Date → Enable Recurring Payment → Configure Recurrence End & Preview **Mercury: Set up recurring ACH payment** (web, 33 screens) [Refero](https://refero.design/flows/11038) > Happy path: A user on the dashboard initiates a bill/payment, creates a new recipient (or uploads a bill to prefill), selects ACH as the payment method, enters routing/account details, categorizes the invoice into GL line items, fills optional bill details and attachments, reviews the payment schedule, and confirms a recurring ACH payment. Steps: Dashboard / Start a payment → Start: Upload or select recipient → Create recipient — blank form → Create recipient — populated basic fields → Create recipient — address entry → Recipient created — choose payment method → Validation — required payment method → Payment method dropdown options → Select ACH → Enter recipient bank details → Categorization — single item (mismatch warning) → Categorization — multiple line items → Details — payment name, memo, email receipt → Details — add multiple receipt emails → Bill details — attachment upload → Bill details — fill invoice metadata → Review — payment schedule summary → Review — payment & recipient details → Review — accounting and attachments → Success confirmation → … **Wealthsimple: Scheduled activities management** (web, 3 screens) [Refero](https://refero.design/flows/8506) > Happy path: a signed-in user opens the Scheduled activities area to review upcoming automatic and one-time transfers, expands a specific scheduled deposit to inspect its details, and cancels that scheduled deposit. Steps: Scheduled activities list → Expanded transaction details → Updated list after cancellation ###### Invoices and bills **Mercury: Create and Send Invoice** (web, 21 screens) [Refero](https://refero.design/flows/11047) > Happy-path: A logged-in user starts from the dashboard, opens the 'request money' wizard, selects or creates a customer, builds an invoice (add items from catalog or create new items), reviews payer page and email previews, configures invoice details (dates, payment methods, destination account), optionally schedules the send, and then sends the invoice. Steps: Dashboard / Start Request → Request type & customer selector (initial) → Open customer dropdown → Customer selected → Add customer (blank) → Add customer (filled) → Return to request type with new customer → Invoice setup (line items and destination) → Edit customer (from invoice step) → Customer updated confirmation → Add catalog item modal (blank) → Add catalog item (filled) → Single item added to invoice → Multiple items and totals → Payer page preview → Email preview → Invoice details and payment methods → Review and send (immediate) → Schedule send → Confirmation / Invoice sent → … **Mercury: Pay selected invoice(s)** (web, 15 screens) [Refero](https://refero.design/flows/11061) > Happy-path: A user starts in the Payments inbox, selects one or more invoices, and initiates a multi-step payment wizard that runs in a split view (invoice preview on the right). The system first attempts to match the invoice to an existing recipient. The user confirms the matched recipient (or searches/creates a new one), selects payment method/account, enters amount and send date, assigns GL codes to line items, supplies wire purpose and supporting details, adds memo/email options, and reviews the summary. The flow ends when the user schedules the wire and receives a confirmation screen. Assumptions: user is signed in with permission to send money, at least one sending account is available, recipient bank details exist or can be created, and invoice PDF is available to preview. Steps: Payments inbox → Select invoices → Recipient match (split preview) → Provide feedback → Feedback submitted / continue → Recipient details confirmation → Amount & send account → Categorization (line items) → Categorization (full view) → Wire purpose → Wire details & attachments → Payment details and memo → Review → Wire scheduled (confirmation overlay) → Final confirmation / next invoice **Acctual: Create bill from uploaded invoice** (web, 2 screens) [Refero](https://refero.design/flows/8845) > Happy path: A user creates a new bill by uploading a supplier invoice PDF. Steps: Start new bill — upload area → Preview and review extracted details **Runey: Invoice Payment** (web, 5 screens) [Refero](https://refero.design/flows/11706) > This flow starts on an invoice detail screen for project INV-0001 inside a workspace labeled Website Redesign. The user opens the Add Payment side panel, enters a payment amount and selects a payment method/date, then submits the payment. The invoice status updates from Pending to Paid, the remaining balance reaches zero, and a success confirmation appears. Assumptions: the user is already signed in, has permission to view and edit the invoice, and is using a desktop layout. Steps: Invoice Review → Add Payment Panel → Payment Amount Update → Paid Confirmation → Settled Invoice View ###### Cards: create, freeze, unblock **Mercury: Create company card** (web, 4 screens) [Refero](https://refero.design/flows/11068) > Happy path: A logged-in admin or employee opens the Cards area, starts the Create card drawer, completes the form (cardholder, nickname, credit/debit, virtual/physical, source account, controls and limits), and submits. The system shows a live preview while editing, then displays a success/activation screen with options to view details or create another. Steps: Cards overview / select Create → Create a card — basics → Create a card — account & review → Confirmation — card created **Mercury: Card Freeze / Unfreeze** (web, 3 screens) [Refero](https://refero.design/flows/11064) > Happy path: a signed-in user navigates to the Cards section, selects a specific virtual card from the left list to view details in the right-hand panel, initiates a freeze action, confirms the freeze in a modal (which surfaces a subscription warning), and returns to the Cards view where the selected card is shown in a Frozen state with an Unfreeze action available. Steps: Cards list — select card → Freeze confirmation modal → Card frozen — details updated **Revolut: Virtual Card Unblock** (web, 3 screens) [Refero](https://refero.design/flows/1332) > Happy path: a logged-in desktop user (Jane Allen) navigates to Home > Cards, opens a virtual card’s detail pane, chooses the 'Unblock CVV' action, confirms in a modal, and receives a success confirmation that PIN and CVV have been unblocked. Steps: Cards list and detail pane → Confirm unblock modal → Success confirmation **Revolut: Virtual Card Management** (web, 2 screens) [Refero](https://refero.design/flows/1333) > Happy path: A logged-in user on desktop is on the Home > Cards tab and opens a virtual card to manage it. Steps: Open card details → Freeze card / show frozen state **Copperx: Crypto Card Unlock** (web, 4 screens) [Refero](https://refero.design/flows/8901) > Happy path (summary): A logged-in desktop user wants to create/unlock a Copperx Visa crypto card. From the Cards section they tap the primary CTA 'Create your Card', which opens an invite-code modal. The user enters an invite code and submits it. On valid code the modal would close and the product proceeds to card provisioning (next screens not shown). The screenshots show the negative path where the server returns an invalid-code error and the UI surfaces a red toast message. Entry point: Cards landing (Cards section visible in left nav). Exit point: modal closes and user proceeds to next card setup/provisioning screen (not shown) or remains on Cards landing if code is invalid. Assumptions: User is authenticated (top-right shows account), using a desktop browser, and has permission to request/create a card. Invite codes are required to proceed. Steps: Cards Landing / Card Promotion → Invite Code Modal (empty) → Invite Code Entry → Validation Error Toast **TIDE: Save Card to Photos** (ios, 3 screens) [Refero](https://refero.design/flows/6758) > Happy-path: A user previews a stylized "TIDE" card (quote + metadata), taps the Download control, grants the iOS photo-library permission when prompted, and sees a confirmation that the image was saved. Entry: card preview screen (user not interacting with system photo library yet). Exit: confirmation toast shown and user returned to the card preview screen. Assumptions: user is on iOS (system permission prompt shown), user is on the card preview screen and not inside another flow, and the app triggers the native photo-library permission/save flow. Steps: Card preview → Photo-library permission prompt → Save confirmation ###### Disputes and conversations about a transaction **PayPal: Open a transaction dispute** (web, 15 screens) [Refero](https://refero.design/flows/7409) > End-to-end happy-path for reporting a problem against a transaction from the activity/transaction detail screen. Steps: Transaction detail → Choose issue category → List items — empty → List items — filled → Pick issue — primary options → Pick issue — subtype selection → Provide other details — empty → Date picker → Provide other details — filled → Seller contact — asked → Seller contact — selection → Add documents — empty → Add documents — uploaded + notes → Review entered info → Confirmation / Case filed **PayPal: Transaction Conversation Thread** (web, 6 screens) [Refero](https://refero.design/flows/7380) > Happy-path: A logged-in user navigates from an Activity transaction (a refund) into the transaction-specific conversation thread, composes a message, sends it, and sees the updated chat with timestamps and the contact header. Steps: Transaction detail (Refund) → Conversation intro (amount shown) → Conversation history (scroll) → Compose message (draft) → Send message (message posted) → Updated conversation with contact header **Mercury: Transaction comment management** (web, 4 screens) [Refero](https://refero.design/flows/11036) > Happy-path: A signed-in user opens a transaction from the Transactions list to read or manage comments in the right-hand detail panel. From the transaction detail the user opens the per-comment actions menu, chooses Delete, confirms in a modal, and the comment is removed and replaced by a “This comment was deleted.” placeholder. Steps: Transaction detail with comments → Open comment actions → Show delete confirmation → Comment removed and thread updated ###### Security: two-factor authentication and recovery **Copperx: Two-Factor Authentication Setup** (web, 6 screens) [Refero](https://refero.design/flows/8906) > Happy-path: A logged-in user opens their Profile > Two-Factor Authentication section, enables 2FA using an authenticator app (scan QR or enter manual key), verifies by entering a 6-digit code, is shown one-time backup codes (with copy/download/print options), confirms they saved them, and completes setup. Steps: Profile — 2FA Disabled → Setup — QR Code & Manual Key → Verify — Enter One-Time Code → Backup Codes — Display → Backup Codes — Confirm Saved → Profile — 2FA Enabled **Copperx: Two-Factor Authentication Management** (web, 4 screens) [Refero](https://refero.design/flows/8907) > Happy path: A logged-in user opens their Profile > Two-Factor Authentication section, chooses to disable 2FA, confirms their identity in a modal (password + authenticator/backup code), and sees the profile update to a disabled state. Steps: Profile — 2FA enabled → Disable 2FA — confirmation modal → Disable 2FA — form completed → Profile — 2FA disabled **Family: Enable Authenticator Method** (web, 6 screens) [Refero](https://refero.design/flows/6202) > Happy path for enabling an authenticator-based two-factor method from the Security settings. Entry: user is on the Security section of their account (logged in on desktop). User clicks the Setup Authenticator App CTA, scans or copies the secret into a mobile authenticator, and enters the one-time code. After successful verification the UI shows a single-use backup code the user must save; once acknowledged the user is returned to Security where the Authenticator App preference is enabled and a prompt to generate recovery codes is shown. Assumptions: user is authenticated, on a desktop browser, and has access to a separate authenticator (mobile) to generate one-time codes. Steps: Security Settings — Setup CTA → Setup Authenticator — Scan / Prepare → Code Entered — Ready to Confirm → Verification Error → Backup Code Display → Security Settings — Authenticator Enabled **Family: Disable Authenticator Two-Factor** (web, 4 screens) [Refero](https://refero.design/flows/6205) > Happy path for disabling (removing) the authenticator-based two-factor method from a signed-in user's Security settings (desktop). Steps: Security settings — Authenticator enabled → Remove Authenticator — code required (empty) → Remove Authenticator — code entered → Security updated — Authenticator disabled **Family: Secret Recovery Setup** (ios, 2 screens) [Refero](https://refero.design/flows/2671) > User is shown their Secret Recovery Phrase (partially hidden) and is required to choose at least one secure backup method before continuing. Entry: user has just generated or been shown their recovery phrase. Exit: user selects a backup method (visually highlighted) or reads the security tooltip; the UI records that a backup method must be chosen before proceeding. Steps: Show Secret Phrase & Backup Options → Security Tooltip Display ###### Crypto: buy, send, receive **Phantom: Buy Bitcoin with third‑party provider** (ios, 11 screens) [Refero](https://refero.design/flows/3710) > Happy-path: User opens a Bitcoin asset detail, taps Buy and enters an amount (or selects a preset). The app validates the amount (enforces a $5 minimum), shows the USD ↔ BTC conversion and enables Buy once valid. The user then chooses a payment method (Credit & Debit Card), views partner providers and live quotes, picks a provider (Stripe), and is routed to an external Stripe/Link checkout where they must sign in/confirm payment. Entry: asset detail screen (logged-in user). Exit: redirected to external provider checkout (Stripe/Link) to complete payment. Assumption: user is logged in to the wallet, has at least one payment method available via partners, and network quotes update on a short timer. Steps: Asset overview → Buy screen — empty amount → Validation — below minimum → Preset selected — valid amount → Toggle input unit → Payment method — entry → Payment method options → Provider quotes list → Provider checkout (Stripe Link) — summary → Stripe Link — sign up or sign in → Return to app or finalize quote **UGLYCASH: Buy Crypto with Apple Pay** (ios, 7 screens) [Refero](https://refero.design/flows/11223) > Happy path: A logged-in user on the Portfolio tab taps the Transfer button to add funds, chooses Apple Pay from the Add funds sheet, selects a token from the token list, enters a USD amount using the large keypad, and completes the purchase via Apple Pay Confirm UI. Steps: Portfolio overview → Add funds options → Token selection / search → Buy screen — amount entry (initial) → Amount entered — show payment action → Apple Pay authentication → Processing payment **Phantom: Send Ethereum to Another Account** (ios, 13 screens) [Refero](https://refero.design/flows/3718) > Happy-path mobile wallet send flow: user starts on the account overview (logged-in mobile wallet), chooses a token to send, selects or pastes a recipient, enters an amount (with conversion between ETH and fiat), reviews the transaction summary including network fees, and finally sends the transaction. The app shows a sending state and success confirmation, then exposes the transaction in Recent Activity and allows viewing the TX on Etherscan. Steps: Account Overview → Select Token → Recipient Input → Enter Amount — Keyboard → Validation — Insufficient Balance → Enter Amount — Fiat Mode → Summary → Network Fee Info Overlay → Sending (Progress) → Sent Confirmation → Recent Activity → Transaction Detail (In-app) → Etherscan Transaction Details **Fuse: Send SOL to external wallet** (ios, 14 screens) [Refero](https://refero.design/flows/8734) > Happy-path: A logged-in user on the Wallet home screen uses the floating + button to start a Send to wallet flow. The user selects Send → To wallet, chooses Solana as the coin, selects or scans a recipient address, enters an amount using the numeric keypad, reviews the transfer (including on-chain fee sponsorship), authenticates, and sends. The flow exits back to the Wallet home with a transient Sending indicator and then a Sent confirmation; the displayed balances update. Assumptions: user is authenticated, has network connectivity, has a positive balance in the selected coin, and the device supports camera access and biometric authentication. Steps: Wallet Home → Quick Actions → Send Destination Choice → Select Coin → Choose Recipient (manual/QR) → Camera Permission Prompt → QR Scanner → Recipient Auto-filled → Enter Amount (keypad) → Amount Confirmation → Review Transfer → Authenticate & Send → Broadcasting / Sending State → Sent Confirmation & Updated Balance **Fuse: Receive crypto and update balance** (ios, 5 screens) [Refero](https://refero.design/flows/8732) > Happy path: A user opens an empty wallet (logged in) and wants to receive assets. From the wallet home they invoke the primary action (+), choose Receive, pick between Fiat or Crypto, open the crypto receive modal that shows a QR code and address, then copy the address (or scan) to send funds. After funds arrive the home balance updates. Entry: wallet home. Exit: wallet home shows updated balance. Assumptions: user is authenticated and allowed to receive crypto; device has camera or clipboard available; network and asset type restrictions are handled elsewhere. Steps: Wallet Home — Empty Balance → Quick Actions Menu → Receive Options → Crypto Receive — QR & Address → Wallet Home — Balance Updated **Family: Send ETH from Wallet** (web, 14 screens) [Refero](https://refero.design/flows/6199) > Happy path: A signed-in user opens the Family > Your Wallets list, opens the wallet context menu and starts a Send operation to another address/contact. They paste or search for a recipient, select the recipient, choose the asset and balance to send, enter an amount (or use max), review transaction details (including fee estimate and source wallet), acknowledge a first-time-recipient warning, confirm to send, see a sending state, and finally see a success confirmation. Entry: Your Wallets list. Exit: Updated wallets list with refreshed balances. Assumptions: user is authenticated, has multiple wallets and at least one token balance; the interface verifies address lookups and shows a confirmation step before broadcasting a transaction. Steps: Wallets List → Context Menu - Send → Send - Recipient Entry → Recipient Resolved → Select Token to Send → Choose Specific Balance → Amount Entry → Asset Confirmation → Review Transaction (Initial) → Review Transaction (Acknowledge) → Transaction Sending → Transaction Sent Confirmation → Return to Wallets - Balances Updated → End State **Phantom: Receive Address & Home Navigation** (ios, 5 screens) [Refero](https://refero.design/flows/3703) > Happy-path: A logged-in mobile wallet user opens the Receive list, selects a network (Ethereum), views the full receive screen with QR code and full address, and then returns to the account home where quick actions and token balances are shown. Steps: Network receive list → Network address detail (Ethereum) → Return to home (onboarding state) → Home - populated balance view → Home - zero balance / default state **Acctual: Add Crypto Wallet Payment Method** (web, 12 screens) [Refero](https://refero.design/flows/8822) > Happy path for adding a crypto wallet as a payment method from the Payments area (desktop, logged-in user). Entry: user is on Payments > Receive. User opens the "Add flexible payment option" modal, chooses Crypto wallet, uses a third-party wallet connector to log in (MetaMask selected), approves connection and signs to prove ownership, then returns to the modal where the wallet address is populated. The user selects a blockchain network and an asset (USDC on Optimism in this example) and saves the configuration. Exit: the Payments list updates to show the new payment method in a pending state. Steps: Payments dashboard (Receive) → Add flexible payment option modal (initial) → Wallet connector chooser → Filter wallet list → Connector: Connect instruction → Connector: Sign-in / ownership prompt → Wallet connected — address populated → Network dropdown (expanded) → Network selected (Optimism) → Asset dropdown (search) → Asset selected and ready to save → Payments dashboard — new method pending **Acctual: Connect crypto payout wallet** (web, 6 screens) [Refero](https://refero.design/flows/8823) > Happy path: From the Payments settings screen the user opens the 'Add flexible payment option' modal, chooses 'Crypto wallet', clicks Connect, selects a wallet provider from a third‑party connector, completes the wallet connection and signature to prove ownership, and returns to the modal with the wallet address populated and connection status shown. Steps: Open add payment modal → Wallet provider list → Search providers → Initiate wallet connect → Sign to confirm ownership → Connected wallet configured ###### Insights, search and spending views **Mercury: Money In Insights review** (web, 3 screens) [Refero](https://refero.design/flows/11058) > Logged-in desktop user lands on the Insights section (left nav > Insights). The 'Money in' tab is active. The main canvas shows a large Money in total ($1.9M), a Monthly average ($477,366), and a bar chart plotted across months. A date-range picker (top-right) is open and highlights a multi-month selection. Three small insight cards appear under the chart (Runway and cash position; Money out trends; Money in trends). A floating feedback widget (thumbs up / thumbs down) is available in the lower-right. Steps: Insights — Money In Overview → Provide feedback — mark helpful → Feedback tooltip dismissed — compact widget **Mercury: Dashboard quick search** (web, 6 screens) [Refero](https://refero.design/flows/11031) > Logged-in desktop user starts on the account dashboard and uses the top omnibox (global search) to find transactions and deposits. Entry: Home dashboard (user already authenticated). Primary interaction: focus the search field, type a query (numeric id or keyword), inspect suggestions and grouped results, switch between category tabs (All, Transactions, Accounts, etc.). Exit: user lands on a filtered results list (e.g., transaction list) or sees a "No results" message for the selected category. Steps: Home dashboard → Open global search → Enter transaction id → Keyword search — mixed results → Filter to Transactions tab → Filter to Accounts tab (no matches) **Mercury: Create and save filtered Transactions view** (web, 7 screens) [Refero](https://refero.design/flows/11055) > A logged-in user on desktop in the Transactions section creates and saves a reusable data view. Starting from the Transactions list, the user opens the Filters panel, sets a date range, narrows results by multiple recipient keywords, adds a transaction status filter (Posted), and then uses the Create view / Save Data View flow to name and persist the view. The flow ends with the saved view applied to the list and a confirmation toast shown. Steps: Transactions list (start) → Open Filters — Date selector → Filters — Keyword (recipients) selection → Filters — Status selection → Open Save Data View modal → Name the Data View and save → Saved view applied and confirmation **Cashcaded: Switch to Monthly Summary** (ios, 4 screens) [Refero](https://refero.design/flows/14143) > The user starts on a daily spending view for September 27, 2026. They open the period selector, switch from Day to Month, and save the selected month. The interface then returns to the main summary view showing aggregated September 2026 spending and records. The flow assumes a mobile user viewing existing financial records; no login, permissions, or account setup state is visible. Steps: Daily Summary → Day Picker → Month Selection → Monthly Summary **Cashcaded: Filter Daily Spending by Category** (ios, 2 screens) [Refero](https://refero.design/flows/14144) > The user starts in a daily Transactions view for Sunday, September 27, where total spending, a chart, category totals, and transaction records are visible. They select the Food category chip to focus the summary and transaction list on food-related spending. The flow ends with the Food category highlighted, the total updated from $229 to $76, and the matching grocery transaction displayed. The screenshots suggest a mobile device and a user who is already viewing transaction data; login state and permissions are not visible. Steps: Daily Transaction Summary → Food Category Results **Cashcaded: Transaction Period Review** (ios, 2 screens) [Refero](https://refero.design/flows/14145) > The user reviews spending activity in a Transactions view and compares summaries across different time ranges. The first screen shows a daily breakdown for Sunday, September 27, including total spending, over-budget status, chart activity, category totals, and individual transactions. The user switches to the Month view, after which the interface updates to show September 2026 totals, monthly budget information, a year-positioned chart, expanded category totals, and monthly transactions. The flow begins in the Transactions analytics view and ends with the selected Month summary. The screenshots imply a mobile device and an existing data set; login state, permissions, and whether the user manually initiated the view are not visible. Steps: Daily Transaction Summary → Monthly Transaction Summary ###### AI assistant inside a finance app **PayPal: PayPal Assistant Conversation** (ios, 16 screens) [Refero](https://refero.design/flows/7266) > Happy-path walkthrough of a logged-in mobile user opening the Message Centre, launching the PayPal Assistant, asking for transaction help, receiving an automated answer with detailed steps and links, giving feedback, and ending the conversation. Steps: Message Centre (empty) → Assistant greeting and quick options → User quick reply sent → Assistant 'no transactions found' → User declines search change → User enters free-text question → Assistant processing response → Assistant provides step-by-step answer → Related links and feedback prompt → User taps thumbs up (positive feedback) → Feedback input (user typing) → Send feedback → Open End Conversation action → Confirm end conversation → Message Centre updated → End state and return to idle ###### Developer and bank-infrastructure flows (sandbox, returns, book transfers) **Column: Simulate a Transfer in Sandbox** (web, 4 screens) [Refero](https://refero.design/flows/1935) > Happy path: a signed-in operator opens the Transfers list, clicks the Simulate button to open a Sandbox modal, fills the required fields (account, transfer type/direction, amount) and sends a simulated ACH/Wire. The modal shows a note that simulated ACHs may take 5–10 minutes. After sending, the modal closes and the Transfers table refreshes to show the new simulated transfer(s) with their Type, Status, Direction and Created date. Steps: Transfers list → Open Sandbox modal → Configure simulation details → Transfers list updated **Column: Initiate transfer return** (web, 3 screens) [Refero](https://refero.design/flows/1931) > Happy path: A logged-in user (desktop) lands on a specific Transfer detail view for an incoming transfer. From this screen the user can open a Create Return dialog, choose a return code, optionally edit description/addenda, and submit. After submission the UI shows a success banner and a new Return section on the same Transfer view that lists the created timestamp, status and the chosen return code and description. Steps: Transfer details → Create return modal → Return confirmation and details **Column: Book Transfer Creation** (web, 4 screens) [Refero](https://refero.design/flows/1933) > Happy path: A logged-in user in the sandbox environment navigates to Transfers, opens the Create Transfer control, selects a transfer type (Book Transfer), completes required fields (sender account, receiver account, amount) and submits the form. Steps: Transfers List → Create Transfer Menu → Create Transfer Form — Book Transfer → Validation Error — Insufficient Funds ###### Leaving: account deletion and data removal **PayPal: Account deletion and data removal** (ios, 13 screens) [Refero](https://refero.design/flows/7271) > A logged-in user navigates from Privacy/Manage your data to request deletion of their PayPal data and close their account. The flow: user selects "Delete your data / close account", reads eligibility and pre-steps, gives explicit consent, reviews a notice about linked services, chooses an email for updates, provides exit reasons (optional), completes a quick security check (SMS code), and receives a final confirmation that the request was received. Entry: Manage your data settings. Exit: confirmation screen stating request received and that an email will follow. Assumptions: user is authenticated, has a verified mobile number for SMS, has an email on file, and can complete any outstanding account requirements (balances, limitations) outside this flow. Steps: Manage your data list → Deletion info & prerequisites → Consent acknowledgement (unchecked) → Consent acknowledgement (checked) → Linked services notice → Choose email for updates → Exit feedback (reason selection) → Quick security check → Enter security code (empty) → Submit security code → Processing / success animation → Request received confirmation → Final status / completion ###### Other finance flows found - PayPal: Manage loyalty cards (ios, 4 screens) [Refero](https://refero.design/flows/7261) - PayPal: Loyalty Card Store Search (ios, 3 screens) [Refero](https://refero.design/flows/7259) - Phantom: Bookmark page from swap screen (ios, 3 screens) [Refero](https://refero.design/flows/3725) - TIDE: Start a Focus Session (ios, 15 screens) [Refero](https://refero.design/flows/6751) - Fuse: Token selector (ios, 2 screens) [Refero](https://refero.design/flows/8735) - Fuse: Token Selector (ios, 2 screens) [Refero](https://refero.design/flows/8736) - Family: Token Overview — Empty Wallet (ios, 2 screens) [Refero](https://refero.design/flows/2697) - Family: Create and Add Wallet (web, 4 screens) [Refero](https://refero.design/flows/6193) - PayPal: Add Currency to Wallet (web, 4 screens) [Refero](https://refero.design/flows/7386) - Acctual: Add flexible payment option (web, 10 screens) [Refero](https://refero.design/flows/8824) - Mercury: Delete payment from inbox (web, 4 screens) [Refero](https://refero.design/flows/11059) - PayPal: Tip Jar QR Setup (ios, 3 screens) [Refero](https://refero.design/flows/7232) - Square: Virtual Terminal Quick Charge (web, 13 screens) [Refero](https://refero.design/flows/1942) - Cashcaded: Cashcaded+ Onboarding and Trial (ios, 31 screens) [Refero](https://refero.design/flows/14116) - Wealthsimple: Accounts overview and view toggle (web, 3 screens) [Refero](https://refero.design/flows/8498) - Runey: Add Expense (web, 5 screens) [Refero](https://refero.design/flows/11708) - Runey: Expense Category Management (web, 4 screens) [Refero](https://refero.design/flows/11709) ### research/judiverse/COMPETITORS-REGULATORS-REUSE.md `research/judiverse/COMPETITORS-REGULATORS-REUSE.md` · sha256 `a92bd61264900c5c680b59a8eb3f7132ec6acf24a1aa94e533b554e60f84ff15` #### Judiverse: Competitors, Regulators, Agent Assurance Gaps and Reusable Systems - **Application:** Judiverse, one of four applications in JudeAI | EDGE ID. It is where people and organisations work with their agents: Jude (assistant, researcher, operator), documents, mail, calendar, meetings, scheduling, tasks, files, forms and workflows, the intenJude lab (agents build agents, models trained, scheduled, switched and compiled into deterministic procedures), and JudeBox (artists' own music, direct royalties). - **Report date:** 2026-10-31. **Every link was accessed on 2026-10-31.** - **Audience:** product, engineering, compliance and leadership. Written for a brand-new engineer: every acronym is explained the first time it appears. - **Status:** research, not legal advice. Nothing here is authority for the estate. Under the estate rules, research is a source to be admitted and referenced, not a TOK object. --- ##### 0. How to read this report ###### 0.1 Labels Every factual line carries exactly one label. | Label | Meaning | |---|---| | **OBSERVED** | I read it directly on a primary source in this session: the vendor's own page or docs, a regulator or legislation page, a licence file in the project's own repository, or an official standards text. | | **OBSERVED (excerpt)** | The source is primary, but I saw it only as a search-engine excerpt, usually because the page blocked automated fetching (HTTP 403). Slightly weaker than OBSERVED. | | **CLAIM** | Vendor marketing numbers, press or third-party reports, or law-firm commentary that I could not confirm on the primary source. A vendor's own number stays a CLAIM even on the vendor's own page. Never treat a CLAIM as fact. | | **AGREEMENT** | A law, regulation, binding rule, court judgment or published standard that is in force or has been formally adopted. Agreements can be disagreed with; they are not observations of truth. | | **PROPOSED** | Draft law, consultation, non-binding resolution, voluntary code still under review, or a rule not yet in application. | | **UNKNOWN** | I searched and could not establish it, or the sources conflict. | | **ASSESSMENT** | My own judgement for Judiverse, derived from the sourced lines around it. It is not a sourced fact. | ###### 0.2 Limits of this research (read before relying on it) - The shared web-search budget ran out part-way through this session. Items I could not research are listed as UNKNOWN in Section 6 rather than filled from memory. (OBSERVED: tool limit reached during this session.) - Several primary pages (OpenAI, some Microsoft and EUR-Lex pages) refused automated fetching. Where that happened, the line is marked OBSERVED (excerpt) or relies on secondary reporting marked CLAIM. - The GitHub API was not available in this session, so repository "maturity" (stars, release cadence) was not measured. Licences were read directly from each project's `LICENSE` file on `raw.githubusercontent.com` (or Codeberg for Forgejo). Maturity notes are ASSESSMENT unless a source is given. - EU legislation links use the official ELI (European Legislation Identifier) form on EUR-Lex. EUR-Lex refused automated fetching, so those link targets were not machine-verified in this session; the regulation numbers are confirmed by the other sources cited on the same line. ###### 0.3 Acronyms and terms (defined once here, and again on first use below) | Term | Plain English | |---|---| | Agent | Software, usually driven by a large language model, that can take actions (send mail, edit files, call APIs) and not only chat. | | LLM (large language model) | The text-generating model behind assistants such as ChatGPT, Claude and Gemini. | | API (application programming interface) | A way for one program to call another. | | MCP (Model Context Protocol) | An open standard for giving agents tools and data through "MCP servers". | | A2A (Agent2Agent protocol) | An open protocol for agents to talk to other agents. | | SSO / OAuth / OIDC | Single sign-on; OAuth is the standard for granting delegated access tokens; OIDC (OpenID Connect) adds identity on top of OAuth. | | On-behalf-of / delegation | An agent acting with authority granted by a person, as opposed to its own authority. | | Prompt injection | Hidden instructions in content an agent reads (an email, web page, issue) that hijack what the agent does. "Indirect" means the attacker never talks to the agent directly. | | Exfiltration | Data being taken out of an organisation without permission. | | GPAI (general-purpose AI) model | EU AI Act term for a model, such as a large LLM, usable for many tasks. | | Provider / deployer | EU AI Act roles: the provider builds or places the AI system on the market; the deployer uses it under its own authority. | | DPIA | Data protection impact assessment, required by UK/EU GDPR for high-risk processing. | | ADM | Automated decision-making about people with no meaningful human involvement. | | TDM | Text and data mining: automated analysis of works, the legal hook for AI training on copyright material. | | CMO | Collective management organisation (a "collecting society"), such as PRS for Music or PPL in the UK. | | ISRC / ISWC | International identifiers for a sound recording (ISRC) and a musical work/composition (ISWC). | | DDEX | Digital Data Exchange: the music industry's metadata message standard. | | WCAG | Web Content Accessibility Guidelines, the W3C accessibility standard. "AA" is the usual required level. | | SPDX id | A standard short identifier for a software licence, e.g. `Apache-2.0`, `AGPL-3.0-only`. | | Copyleft / AGPL | Licences requiring derivative works to be shared under the same terms. The AGPL (Affero GPL) extends this to software offered over a network: if you modify and run it as a service, you must offer users the modified source. | | Source-available | Code you can read but whose licence restricts use (e.g. commercial resale). Not "open source" in the Open Source Initiative sense. | | BSL / BUSL | Business Source License: source-available, converts to an open licence after a "change date". | | SIEM | Security information and event management: the system security teams use to collect and search logs. | | GA / preview / beta | Generally available / early-access releases that can change. | | Glass box | Every decision and action can be inspected: inputs, rules applied, who approved, what happened. The opposite of a black box. | --- ##### 1. Executive summary (the ten findings that matter most) 1. **The big suites now give agents an identity, but as an extension of a human tenant, not as a first-class member with a birth and death certificate.** Microsoft Entra Agent ID gives agents identities, "blueprints" and audit logs, and Microsoft Agent 365 has been GA since 1 May 2026, licensed per human user ([Microsoft Learn: Entra Agent ID](https://learn.microsoft.com/en-us/entra/agent-id/what-is-microsoft-entra-agent-id), accessed 2026-10-31; [Microsoft Learn: Agent 365 overview](https://learn.microsoft.com/en-us/microsoft-agent-365/overview), accessed 2026-10-31). Google gives each agent "a verifiable cryptographic ID" in Gemini Enterprise Agent Platform ([Google Cloud Next '26 recap](https://cloud.google.com/blog/topics/google-cloud-next/google-cloud-next-2026-wrap-up), accessed 2026-10-31). OBSERVED. The gap Judiverse can own is the full lifecycle record (mandate, birth, versioning, transfer, retirement/death) shown as a glass box to the people it works with. ASSESSMENT. 2. **Every major assistant became a long-running "co-worker" agent in 2026.** OpenAI ChatGPT Work (9 July 2026), Anthropic Claude Cowork (GA 9 April 2026 on desktop), Microsoft Copilot Cowork (announced 9 March 2026), Google long-running agents and an agent "Inbox" (April 2026). OBSERVED / OBSERVED (excerpt) / CLAIM per Section 2.2. Agents-as-co-workers is now table stakes, not a differentiator. ASSESSMENT. 3. **Agent mailboxes are a real product category, with abuse controls bolted on.** AgentMail raised a US$6M seed (March 2026) and caps unauthenticated agent inboxes at 10 sends a day ([TechCrunch](https://techcrunch.com/2026/03/10/agentmail-raises-6m-to-build-an-email-service-for-ai-agents/), accessed 2026-10-31). OBSERVED. Gmail's sender rules (SPF/DKIM/DMARC, 0.3% spam ceiling) apply to agents exactly as to humans ([Google Workspace Admin Help](https://support.google.com/a/answer/81126), accessed 2026-10-31). OBSERVED. 4. **Prompt injection through mail, issues and documents is the proven attack path against office agents.** EchoLeak (CVE-2025-32711, CVSS 9.3) let one crafted email exfiltrate Microsoft 365 Copilot data with zero clicks ([arXiv 2509.10540](https://arxiv.org/html/2509.10540v1), accessed 2026-10-31); a malicious GitHub issue hijacked an agent into leaking private repositories ([Invariant Labs](https://invariantlabs.ai/blog/mcp-github-vulnerability), accessed 2026-10-31). OBSERVED. Anthropic's own safety page says Cowork users "remain accountable" for actions Claude takes ([Claude Help Center](https://support.claude.com/en/articles/13364135-use-claude-cowork-safely), accessed 2026-10-31). OBSERVED. 5. **EU AI Act: GPAI duties apply now; transparency duties applied 2 August 2026; high-risk duties moved to 2 December 2027 / 2 August 2028.** The "AI Omnibus" entered into force 27 July 2026 ([European Commission](https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai), accessed 2026-10-31). AGREEMENT. Employment and worker-management uses are Annex III high-risk, so agents that screen candidates or manage staff will carry deployer duties from 2 December 2027. AGREEMENT (date) / ASSESSMENT (application to Judiverse). 6. **UK: Data (Use and Access) Act 2025 relaxed automated decision-making rules from 5 February 2026; ICO guidance is still being finalised.** ([Travers Smith](https://www.traverssmith.com/knowledge/knowledge-container/uks-data-protection-reforms-take-effect-a-new-era-for-automated-decision-making/), accessed 2026-10-31; [ICO consultation](https://ico.org.uk/about-the-ico/ico-and-stakeholder-consultations/2026/03/ico-consultation-on-the-draft-guidance-about-automated-decision-making-including-profiling/), accessed 2026-10-31.) AGREEMENT (Act) / PROPOSED (guidance). 7. **UK copyright and AI training: no change to the law.** On 18 March 2026 the government said a broad TDM exception with opt-out "is no longer the government's preferred way forward" and that it will not reform until confident ([GOV.UK report](https://www.gov.uk/government/publications/report-and-impact-assessment-on-copyright-and-artificial-intelligence/report-on-copyright-and-artificial-intelligence), accessed 2026-10-31). OBSERVED / PROPOSED (future reform). Courts split: UK High Court found Stable Diffusion's weights are not an "infringing copy"; Munich found memorised lyrics in GPT-4/4o are reproductions (Section 3.4). AGREEMENT (first-instance judgments, both appealed or appealable). 8. **Music: platforms are moving against AI floods and micro-payouts; artists still get paid through layered intermediaries.** Deezer says AI tracks passed 50% of daily uploads in June 2026 ([Deezer newsroom](https://newsroom-deezer.com/2026/07/ai-music-exceeds-50-percent-daily-uploads-deezer/), accessed 2026-10-31) (CLAIM); Bandcamp banned music generated "wholly or in substantial part" by AI ([TechCrunch](https://techcrunch.com/2026/01/14/bandcamp-takes-a-stand-against-ai-music-banning-it-from-the-platform/), accessed 2026-10-31) (OBSERVED (excerpt)); Spotify pays recorded royalties only on tracks with 1,000+ annual streams ([Music Business Worldwide](https://musicbusinessworldwide.com/changes-to-spotifys-royalty-model-including-the-1000-annual-streams-royalty-policy-are-officially-live), accessed 2026-10-31) (CLAIM). JudeBox "direct royalties" still needs PRS/PPL-style licensing for any public streaming (Section 3.5). ASSESSMENT. 9. **Reuse is realistic for almost every Judiverse surface, but the copyleft map matters.** Mail/calendar (Stalwart, AGPL-3.0 or commercial), files (Nextcloud, AGPL-3.0-or-later), chat (Matrix Synapse, AGPL-3.0 or commercial; Mattermost source AGPL-3.0), documents (Collabora Online, MPL-2.0; ONLYOFFICE, AGPL-3.0 with a disputed logo clause), meetings (Jitsi, Apache-2.0), models (vLLM Apache-2.0, Ollama MIT, transformers Apache-2.0, MLflow Apache-2.0). OBSERVED (licence files). Flags: n8n (Sustainable Use License, not open source), Outline (BSL 1.1 until 2030-10-23), Open WebUI (branding clause above 50 users), Llama model weights (700M MAU clause). OBSERVED. 10. **Nobody publicly offers what Judiverse describes end to end:** agents as native, always-on users with their own mailbox, mandate, versioned lineage, transferable and reproducible builds, deterministic compilation of procedures, and a public death certificate. Each piece exists somewhere (identity at Microsoft/Google, mailboxes at AgentMail, sandboxes at Anthropic/Google, registries at ServiceNow), but not as one accountable record. ASSESSMENT, based on Sections 2 and 4. --- ##### 2. Competitors ###### 2.1 Office suites (documents, mail, calendar, meetings, workflows) | Vendor / product | What it offers now (2025-2026) | What its agents can do | Identity, accountability, sovereignty | Label and source | |---|---|---|---|---| | **Microsoft 365 + Copilot + Agent 365** | Agent 365 "observe, govern, secure" control plane; agent registry and agent map in the M365 admin centre; Entra, Purview (data loss prevention, DLP) and Defender extended to agents. GA 1 May 2026 for the Commercial segment, licensed per user. | Agents acting for users (delegated) and agents with their own access are GA; agents in team workflows in public preview; partner agents (e.g. n8n, Zendesk, Egnyte) managed by Agent 365. | Entra Agent ID: "agent identity blueprints" as templates with parent-child relationships; supports OAuth 2.0, MCP and A2A; "All agent authentication and activity is logged for compliance and audit." Third-party agents (AWS Bedrock, n8n) can be integrated. | OBSERVED ([Agent 365 overview](https://learn.microsoft.com/en-us/microsoft-agent-365/overview), accessed 2026-10-31; [Agent 365 GA blog](https://www.microsoft.com/security/blog/2026/05/01/microsoft-agent-365-now-generally-available-expands-capabilities-and-integrations/), accessed 2026-10-31; [Entra Agent ID](https://learn.microsoft.com/en-us/entra/agent-id/what-is-microsoft-entra-agent-id), accessed 2026-10-31) | | Microsoft Agent 365 pricing | US$15 per user per month standalone, or within Microsoft 365 E7; a licence covers someone who "manages or sponsors agents, or uses agents". | n/a | Per-human licensing means the agent is accounted for through its human sponsor. ASSESSMENT. | OBSERVED ([Agent 365 GA blog](https://www.microsoft.com/security/blog/2026/05/01/microsoft-agent-365-now-generally-available-expands-capabilities-and-integrations/), accessed 2026-10-31). E7 list price of US$99 is CLAIM ([beri.net](https://www.beri.net/article/microsoft-agent-365-e7-ga-two-invoice-trap-2026), accessed 2026-10-31) | | **Microsoft Copilot Cowork** | Announced 9 March 2026; turns a request into a plan and executes steps across Outlook, Teams, Excel; built "in close collaboration with Anthropic". GA reported 16 June 2026 (date conflicts across sources). | Multi-step task execution on the user's behalf across M365. | Inherits the M365 tenant controls. | CLAIM ([GIGAZINE](https://gigazine.net/gsc_news/en/20260310-microsoft-365-copilot-cowork-ai), accessed 2026-10-31; [Duke OIT](https://oit.duke.edu/blog-post/cowork-now-available-microsoft-copilot/), accessed 2026-10-31). GA date UNKNOWN (sources conflict) | | **Google Workspace + Gemini** | Workspace Studio (December 2025) to "design, manage, and share AI agents"; at Next '26: Workspace skills, Workspace MCP Server (preview), Workspace capabilities in the Gemini Enterprise app (private preview), agent governance (AI control centre, auditing agent access to Workspace data). | Schedule Calendar meetings, create/edit Docs and Slides from the Gemini Enterprise app (private preview); developers bring Drive synthesis and Gmail drafting into their own agents via MCP. | Agent Identity (cryptographic ID, auditable trail mapped to authorisation policies), Agent Registry, Agent Gateway (Model Armor protections against prompt injection and data leakage). | OBSERVED ([Next '26 recap](https://cloud.google.com/blog/topics/google-cloud-next/google-cloud-next-2026-wrap-up), accessed 2026-10-31). Workspace Studio date: CLAIM ([9to5Google](https://9to5google.com/2025/12/03/google-workspace-studio/), accessed 2026-10-31) | | Google long-running agents and Inbox | Long-running agents "run complex business processes autonomously in secure cloud sandboxes"; an Inbox groups agent notifications into "Needs your input", "Errors", "Completed". | Background multi-step work; can write code to build custom tools. | Sandboxed; human review via Inbox. | OBSERVED ([Next '26 recap](https://cloud.google.com/blog/topics/google-cloud-next/google-cloud-next-2026-wrap-up), accessed 2026-10-31) | | **Notion** | Notion 3.0 (18 September 2025): an Agent that can do "Anything you can do in Notion"; works autonomously up to 20 minutes across hundreds of pages; memory is an editable instruction page. Custom Agents (scheduled/triggered, shareable) were "coming soon" at launch. | Create docs, build databases, search across tools, multi-step workflows. | The launch post says nothing about agent identity or audit; agent "name" is cosmetic. | OBSERVED ([Notion blog](https://www.notion.com/en-gb/blog/introducing-notion-3-0), accessed 2026-10-31). Custom Agents current status: UNKNOWN | | **Zoho** | Zia Agents, no-code Agent Studio and Agent Marketplace announced 4 February 2025; July 2025 added Zia LLM (Zoho's own model) and a Zoho MCP server; 700+ actions across Zoho apps. | Prebuilt agents (sales development, HR, support, IT help desk); triggered by button, rule or conversation. | Zoho emphasises its own LLM; no agent-identity details found. | CLAIM ([Business Wire press release](https://www.businesswire.com/news/home/20250204066995/en/Zoho-Corporation-Announces-Zia-Agents-AI-Platform-Supporting-Autonomous-Agents-Across-Organizations-Broad-Portfolio), accessed 2026-10-31; [Futurum](https://futurumgroup.com/insights/zoho-unveils-zia-llm-no%E2%80%91code-agent-studio-and-open-agent-interoperability/), accessed 2026-10-31) | | **Proton** | Lumo assistant (July 2025): no chat logs, no training on prompts, European hosting, web search off by default; Lumo for Business (October 2025). | Chat and work with encrypted Proton Drive files; no agentic office automation found. | Sovereignty and privacy positioning; encryption and open-source claims are disputed by at least one critic. | CLAIM ([BleepingComputer](https://bleepingcomputer.com/news/artificial-intelligence/proton-launches-privacy-respecting-encrypted-ai-assistant-lumo), accessed 2026-10-31; [AlternativeTo](https://alternativeto.net/news/2025/10/proton-launches-lumo-for-business-as-a-new-plan-for-its-privacy-focused-ai-assistant/), accessed 2026-10-31; dispute: [ludditus.com](https://ludditus.com/?p=68822), accessed 2026-10-31) | | **Euro-Office (Nextcloud + IONOS fork of ONLYOFFICE)** | Announced 2026 as a "sovereign" office suite; first release reported for 9 June 2026; ONLYOFFICE suspended its partnership with Nextcloud over licence/branding. | Not an agent product; relevant as a sovereign editor Judiverse could embed. | Sovereignty positioning; licence dispute unresolved per ONLYOFFICE. | CLAIM ([heise](https://heise.de/-11228123), accessed 2026-10-31; [Computerworld](https://www.computerworld.com/article/4178807/open-source-euro-office-productivity-suite-to-launch-june-9.html), accessed 2026-10-31; [AlternativeTo](https://alternativeto.net/news/2026/4/onlyoffice-ends-its-partnership-with-nextcloud-over-new-unauthorized-euro-office-fork), accessed 2026-10-31) | ###### 2.2 AI assistants and agent platforms | Vendor / product | What it offers | What agents can do | Identity, accountability, black-box notes | Label and source | |---|---|---|---|---| | **OpenAI ChatGPT Work** | Launched 9 July 2026 with GPT-5.6; "an agent for longer, more involved tasks"; scheduled one-time or recurring tasks; users "approve important actions along the way"; rolled out first to Pro, Enterprise, Edu. | Research, work across connected apps and files, produce documents, spreadsheets, presentations, reports and Sites. | Governance "built on ChatGPT Enterprise" (CLAIM). Uber says it is rolling Work out to 30,000+ employees (22 October 2026). | OBSERVED (excerpt) ([OpenAI ChatGPT Work](https://openai.com/chatgpt-work/), accessed 2026-10-31; [ChatGPT release notes](https://help.openai.com/en/articles/6825453-chatgpt-release-notes), accessed 2026-10-31; [Uber rollout](https://openai.com/index/uber-chatgpt-work-rollout/), accessed 2026-10-31). OpenAI pages returned HTTP 403 to direct fetch | | OpenAI earlier agents | Operator research preview (January 2025) using a Computer-Using Agent model; folded into "ChatGPT agent" on 17 July 2025 (Pro, Plus, Team), Enterprise/Edu from 8 August 2025. | Browse and act in a virtual browser. | Reports that ChatGPT agent was later withdrawn in favour of Work are unverified. | CLAIM ([TechCrunch](https://techcrunch.com/2025/01/23/openai-launches-operator-an-ai-Agent-that-performs-tasks-autonomously), accessed 2026-10-31); OBSERVED (excerpt) ([ChatGPT agent release notes](https://help-lb.openai.com/en/articles/11794368-chatgpt-agent-release-notes), accessed 2026-10-31). Withdrawal: UNKNOWN | | OpenAI AgentKit / Apps SDK | DevDay 6 October 2025: visual Agent Builder with versioning, Connector Registry, ChatKit embeddable UI, evals and tracing; Apps SDK built on MCP. | Developer-built agents and in-ChatGPT apps. | Versioning and tracing exist at the builder level. | CLAIM ([InfoQ](https://infoq.com/news/2025/10/openai-dev-day), accessed 2026-10-31; [Latent Space](https://www.latent.space/p/devday-2025-apps-sdk-agent-kit-mcp), accessed 2026-10-31) | | **Anthropic Claude Cowork** | Research preview 12 January 2026; GA on macOS and Windows desktop 9 April 2026; web and mobile added in 2026 (beta on some plans). | Read, write and delete local files in connected folders; computer use; browser; scheduled tasks; plugins/MCP. | Anthropic lists prompt injection, permanent deletion, unattended scheduled tasks and cross-app data movement as risks; "You remain accountable for actions Claude takes on your behalf"; compliance API and OpenTelemetry streaming for Team/Enterprise owners. | OBSERVED ([Use Claude Cowork safely](https://support.claude.com/en/articles/13364135-use-claude-cowork-safely), accessed 2026-10-31); OBSERVED (excerpt) ([Claude release notes](https://support.claude.com/en/articles/12138966-release-notes), accessed 2026-10-31; [Cowork on web, desktop, mobile](https://support.claude.com/en/articles/15520349-use-claude-cowork-on-web-desktop-and-mobile), accessed 2026-10-31) | | **Anthropic Claude Managed Agents** | Beta (header `managed-agents-2026-04-01`): hosted agent harness. Concepts: Agent (model, prompt, tools, MCP servers, skills), Environment (Anthropic cloud sandbox or self-hosted sandbox), Session, Events; scheduled deployments on a cron schedule. | Bash, file operations, web search/fetch (allow/block lists), MCP. | Not eligible for zero data retention or HIPAA BAA because sessions are stateful; you can delete sessions and files. Self-hosted sandboxes support data-residency needs. | OBSERVED ([Claude Platform docs](https://platform.claude.com/docs/en/managed-agents/overview), accessed 2026-10-31). Price of US$0.08 per runtime hour: CLAIM ([IT Brief](https://itbrief.com.au/story/anthropic-launches-claude-managed-agents-in-public-beta), accessed 2026-10-31) | | Anthropic Claude Code | Same Claude Agent SDK foundation as Cowork. | Coding agent in terminal and IDE. | n/a | CLAIM ([Techstrong.ai](https://techstrong.ai/features/anthropic-launches-cowork-bringing-claudes-agentic-ai-to-non-developers/), accessed 2026-10-31) | | **Google Gemini Enterprise Agent Platform** | Announced 22-24 April 2026 as the evolution of Vertex AI: Agent Development Kit, Agent Studio, Agent Runtime, Agent Sandbox, Memory Bank, Agent Identity, Registry, Gateway, Simulation, Evaluation, Optimizer. | Build, run and govern agent fleets; long-running agents. | Agent Identity issues per-agent cryptographic IDs; X.509 24-hour auto-rotating certificates are a third-party CLAIM. | OBSERVED ([Next '26 recap](https://cloud.google.com/blog/topics/google-cloud-next/google-cloud-next-2026-wrap-up), accessed 2026-10-31); CLAIM ([pebblous analysis](https://blog.pebblous.ai/blog/gemini-enterprise-agent-identity/en/), accessed 2026-10-31) | | **Meta AI / Manus** | Meta agreed to buy Manus (general-purpose agents: research, coding, data analysis) in late December 2025, reportedly for more than US$2B; Manus to keep running and be folded into Meta AI; China operations wound down. | General task agents by subscription. | No agent-identity information found. | CLAIM ([Channel Insider](https://www.channelinsider.com/channel-business/mergers-and-acquisitions/meta-manus-ai-agent-development/), accessed 2026-10-31; [Business Today](https://www.businesstoday.in/amp/technology/news/story/meta-acquires-ai-startup-manus-in-milestone-deal-worth-over-2-billion-508760-2025-12-31), accessed 2026-10-31) | | **Perplexity Comet** | AI browser launched 9 July 2025; later made free worldwide. | Browser agent; email assistant for top tier (CLAIM). | Inbox access exposes it to indirect prompt injection from pages, emails and invites (CLAIM). | CLAIM ([TechCrunch](https://techcrunch.com/2025/07/09/perplexity-launches-comet-an-ai-powered-web-browser/), accessed 2026-10-31; [YourStory](https://yourstory.com/ai-story/perplexity-ai-comet-browser-free-worldwide), accessed 2026-10-31) | | **Salesforce Agentforce (360)** | Flex Credits US$500 per 100k credits; Conversations US$2 each; Agentforce User Licence US$5 per user per month (list prices on Salesforce page). Rebranded Agentforce 360 at Dreamforce 2025 (CLAIM). | CRM, service and sales agents. | No agent-identity details found in this session. | OBSERVED (excerpt) ([Salesforce pricing](https://www.salesforce.com/agentforce/pricing/), accessed 2026-10-31); CLAIM ([Ry Walker Research](https://rywalker.com/research/agentforce), accessed 2026-10-31) | | **ServiceNow AI Control Tower** | Governance hub for "every AI" in an enterprise; integration with Microsoft Agent 365 in preview (May 2026); can reportedly detect an agent acting beyond its permissions and shut it down; logs files read, commands run, APIs called for its desktop agent. | Governs agents across vendors. | Strongest "registry plus kill switch" positioning found. | OBSERVED (excerpt) ([ServiceNow press release](https://newsroom.servicenow.com/press-releases/details/2026/ServiceNow-expands-AI-agent-governance-through-deeper-integration-with-Microsoft/default.aspx), accessed 2026-10-31); CLAIM ([The Register](https://theregister.com/2026/05/05/servicenow_clears_agents_for_landing/), accessed 2026-10-31) | | **Glean** | Enterprise search plus agent platform; hosted MCP server so agents elsewhere can use Glean Search; built-in prompt-injection protections (CLAIM). | Search-grounded agents. | Permissions-aware search is its core claim. | OBSERVED (excerpt) ([Glean press release](https://www.glean.com/press/glean-expands-horizontal-agent-platform-delivers-dozens-of-agents-and-open-interoperability-across-the-enterprise), accessed 2026-10-31) | | Microsoft Copilot Studio | Copilot Studio agents are covered by Entra Agent ID; new agents reportedly get Entra Agent IDs automatically, with no in-place migration for existing ones. | Low-code agent builder. | See Entra Agent ID row. | CLAIM ([collab365](https://spaces.collab365.com/posts/copilot-studio-agents-now-get-entra-identities-pqRA36), accessed 2026-10-31; [Entra changelog by Merill](https://entradiff.merill.net/products/agent-id/), accessed 2026-10-31) | ###### 2.3 Agent infrastructure (mailboxes, tools, orchestration) | Product | What it offers | Agent capability | Gaps / notes | Label and source | |---|---|---|---|---| | **AgentMail** | API that gives agents real inboxes: two-way threads, labels, search, reply, parsing; onboarding API so agents sign themselves up; hosted MCP server. | Agent owns an address and message store. | Unauthenticated agent inboxes send only 10 emails a day until a person authenticates them. Company-reported: tens of thousands of human users, hundreds of thousands of agent users, 500+ B2B customers (CLAIM). | OBSERVED ([TechCrunch](https://techcrunch.com/2026/03/10/agentmail-raises-6m-to-build-an-email-service-for-ai-agents/), accessed 2026-10-31) | | **Resend** (+ `resend-mcp`) | Sending platform with inbound read; official MCP server lists and reads inbound email and downloads attachments; local npm or hosted remote server with OAuth. | Agent sends and reads on your own domain. | Not an agent-native mailbox (CLAIM). `resend-mcp` licence: MIT (OBSERVED, [LICENSE](https://raw.githubusercontent.com/resend/resend-mcp/main/LICENSE), accessed 2026-10-31). | OBSERVED (excerpt) ([Resend MCP docs](https://resend.com/docs/mcp-server), accessed 2026-10-31); CLAIM ([Drag comparison](https://www.dragapp.com/blog/email-mcp-servers/), accessed 2026-10-31) | | **Composio** | Integration layer: OAuth flows, token storage and refresh for agents across hundreds of toolkits (count varies by source). SDK licence MIT. | Agents call third-party apps as a user. | Holds users' tokens: a concentration of delegated credentials. ASSESSMENT. | CLAIM ([Ry Walker Research](https://rywalker.com/research/composio), accessed 2026-10-31); OBSERVED licence ([LICENSE](https://raw.githubusercontent.com/ComposioHQ/composio/main/LICENSE), accessed 2026-10-31) | | **LangChain / LangGraph** | Agent orchestration framework (graphs, state). LangGraph licence MIT. | Developer framework. | Identity, mailbox and lifecycle are left to the developer. ASSESSMENT. | OBSERVED ([LangGraph LICENSE](https://raw.githubusercontent.com/langchain-ai/langgraph/main/LICENSE), accessed 2026-10-31) | | **CrewAI** | Multi-agent framework. Licence MIT. | Role-based agent "crews". | As above. | OBSERVED ([CrewAI LICENSE](https://raw.githubusercontent.com/crewAIInc/crewAI/main/LICENSE), accessed 2026-10-31) | | **n8n** | Workflow automation with AI nodes. Licence: Sustainable Use License (fair-code); `.ee.` files under a separate Enterprise License. n8n states it is not open source under the OSI definition. | Agent and workflow nodes; manageable by Microsoft Agent 365 as a partner agent. | **Licence flag:** restricts use to internal business purposes and limits selling products whose value derives substantially from n8n. | OBSERVED ([n8n LICENSE.md](https://raw.githubusercontent.com/n8n-io/n8n/master/LICENSE.md), accessed 2026-10-31; [n8n docs](https://docs.n8n.io/n8n-community-license/sustainable-use-license), accessed 2026-10-31) | | **Zapier Agents** | AI "teammates" connected to 9,000+ apps; run on command or on their own; templates for lead enrichment, ticket replies, candidate ranking, expense classification. | Cross-app actions. | The product page lists "Monitor activity" but gives no admin, permission or policy detail. | OBSERVED ([Zapier Agents](https://zapier.com/agents), accessed 2026-10-31) | ###### 2.4 Meetings and scheduling | Product | What it offers | Agent / AI features | Gaps and evidence | Label and source | |---|---|---|---|---| | **Otter.ai** | AI meeting transcription; "Otter Notetaker" bot joins Zoom/Meet/Teams. | Auto-join, transcribe, summarise. | Proposed class action (N.D. Cal., filed August 2025, case 5:25-cv-06911) alleges recording non-users without consent and training on recordings; Otter says users must obtain permissions. Outcome UNKNOWN. | OBSERVED (excerpt) ([NPR](https://www.npr.org/2025/08/15/g-s1-83087/otter-ai-transcription-class-action-lawsuit), accessed 2026-10-31). Allegations are CLAIM | | **Fireflies.ai** | Meeting bot and notes. | Speaker recognition. | Cruz v. Fireflies.AI Corp. (filed 18 December 2025) alleges voiceprints collected without notice or written consent under Illinois BIPA (Biometric Information Privacy Act). Outcome UNKNOWN. | CLAIM ([Workplace Privacy Report](https://www.workplaceprivacyreport.com/2026/04/articles/artificial-intelligence/ai-meeting-assistants-and-biometric-privacy-governance-lessons-from-the-fireflies-ai-lawsuit/), accessed 2026-10-31) | | **Granola** | "Bot-free" meeting notes; US$125M Series C at US$1.5B valuation (March 2026); Personal API, Enterprise API, MCP server. | Notes become organisational context for agents. | Bot-free capture shifts the consent question to the device owner. ASSESSMENT. | CLAIM ([The Next Web](https://thenextweb.com/news/granola-series-c-meeting-ai-enterprise-context), accessed 2026-10-31) | | **Calendly** | Scheduling links. | UNKNOWN: search budget ran out before Calendly's 2026 AI/agent features could be checked. | UNKNOWN. | UNKNOWN | ###### 2.5 Developer platforms | Product | What it offers | Agent capability | Identity / governance | Label and source | |---|---|---|---|---| | **GitHub Copilot + Agent HQ** | Agent HQ (Universe, October 2025): "any agent" from Anthropic, OpenAI, Google, Cognition, xAI inside GitHub; "mission control" to assign, steer and track multiple agents across web, VS Code, CLI and mobile. | Agents work through issues, branches and pull requests. | Admins choose allowed agents and models; one analysis says agents run in sandboxed GitHub Actions with firewalls (CLAIM). | OBSERVED (excerpt) ([GitHub blog](https://github.blog/news-insights/company-news/welcome-home-agents/), accessed 2026-10-31); CLAIM ([The New Stack](https://thenewstack.io/github-embraces-the-coding-agent-competition-with-agent-hq/), accessed 2026-10-31) | | **GitLab Duo Agent Platform** | GA 15 January 2026 (GitLab 18.8) for Premium and Ultimate; Planner and Security Analyst agents GA; custom agents in an AI catalogue; external agents (Claude Code, Codex CLI) integrated. | DevSecOps agents. | "Composite identity": service accounts acting on behalf of the triggering user (GitLab epic; describes intent, not confirmed GA behaviour). | OBSERVED (excerpt) ([GitLab press release](https://about.gitlab.com/press/releases/2026-01-15-gitlab-announces-duo-agent-platform-general-availability/), accessed 2026-10-31); CLAIM ([GitLab epic 19478](https://gitlab.com/groups/gitlab-org/-/epics/19478), accessed 2026-10-31) | ###### 2.6 Model training and serving (relevant to the intenJude lab) | Product | What it offers | Terms that matter | Label and source | |---|---|---|---| | **Hugging Face** | Model hub; Inference Providers (pay-as-you-go serving with a free tier); AutoTrain (no-code training). `transformers` library licence Apache-2.0. | Model weights carry their own licences separate from the library; check each model card. | OBSERVED (excerpt) ([HF Inference docs](https://huggingface.co/docs/hub/inference), accessed 2026-10-31); OBSERVED ([transformers LICENSE](https://raw.githubusercontent.com/huggingface/transformers/main/LICENSE), accessed 2026-10-31) | | **Together AI** | Fine-tuning billed per token processed; rate varies by model size, SFT (supervised fine-tuning) vs DPO (direct preference optimisation), LoRA (low-rank adaptation) vs full; minimum charge; SDK `estimate_price` before running. | Price estimates before jobs help reproducible budgeting. | OBSERVED (excerpt) ([Together docs](https://docs.together.ai/docs/fine-tuning-pricing), accessed 2026-10-31) | | **Modal** | Code-first serverless GPU for training and fine-tuning; per-second billing; Axolotl-based fine-tuning example. | Vendor content. | CLAIM ([Modal](https://modal.com/resources/best-serverless-gpu-platforms-fine-tuning), accessed 2026-10-31; [modal-labs/llm-finetuning](https://www.github.com/modal-labs/llm-finetuning), accessed 2026-10-31) | | **Replicate** | Model deployment platform; Cloudflare announced its acquisition on 17 November 2025; 50,000+ models to be available to Cloudflare Workers AI users. | Brand expected to remain separate (CLAIM). | CLAIM ([Business Wire](https://www.businesswire.com/news/home/20251117400765/en/Cloudflare-to-Acquire-Replicate-to-Build-the-Most-Seamless-AI-Cloud-for-Developers), accessed 2026-10-31) | | **NVIDIA NIM** | Packaged inference microservices. Free to Developer Program members for research, development and experimentation on up to 16 GPUs; production requires an NVIDIA AI Enterprise licence. | **Licence flag:** not free for production. Paid licence "starting at $4500 per GPU per year" per one NVIDIA docs page (CLAIM until confirmed with NVIDIA). | OBSERVED (excerpt) ([NVIDIA developer blog](https://developer.nvidia.com/blog/access-to-nvidia-nim-now-available-free-to-developer-program-members/), accessed 2026-10-31; [NIM FAQ](https://docs.api.nvidia.com/nim/docs/product), accessed 2026-10-31) | ###### 2.7 Music platforms and how artists are paid (relevant to JudeBox) | Platform | Payment model and recent changes | AI policy | Label and source | |---|---|---|---| | **Spotify** | Since 1 April 2024 a track needs 1,000 streams in the previous 12 months to generate recorded royalties; per-track charges to labels/distributors for "flagrant artificial streaming"; Spotify said 99.5% of streams are of tracks above the threshold. | 25 September 2025: DDEX-based AI disclosures in credits, a spam filter, ban on unauthorised voice clones; said to have removed 75M+ "spammy" tracks in a year. | CLAIM ([Music Business Worldwide](https://musicbusinessworldwide.com/changes-to-spotifys-royalty-model-including-the-1000-annual-streams-royalty-policy-are-officially-live), accessed 2026-10-31; [TechCrunch](https://techcrunch.com/2025/09/25/spotify-updates-ai-policy-to-label-tracks-cut-down-on-spam), accessed 2026-10-31) | | **SoundCloud** | Fan-Powered Royalties (pay by listener rather than pro-rata pool); reported to stop taking a share of distribution royalties from late November 2025. | May 2025 controversy over terms permitting AI training; CEO says content was never used to train AI and revised terms require consent. | CLAIM ([TechCrunch](https://techcrunch.com/2025/05/09/soundcloud-changes-policies-to-allow-ai-training-on-user-content), accessed 2026-10-31; [SoundCloud CEO letter](https://community.soundcloud.com/playbook-articles/a-letter-from-our-ceo-clarifying-our-terms-of-use), accessed 2026-10-31; [SoundGuys](https://www.soundguys.com/soundcloud-royalty-offer-and-generative-ai-148024/), accessed 2026-10-31) | | **Bandcamp** | Direct sales: 15% revenue share on digital (10% after US$5,000 sales in 12 months), 10% on physical; payment processing typically 4-6% on top. | Bans music generated "wholly or in substantial part" by AI and AI impersonation (13-14 January 2026); already prohibited crawling for AI training. | OBSERVED (excerpt) ([Bandcamp fees](https://get.bandcamp.help/hc/en-us/articles/23020665520663-What-are-Bandcamp-s-fees), accessed 2026-10-31); OBSERVED (excerpt) ([TechCrunch](https://techcrunch.com/2026/01/14/bandcamp-takes-a-stand-against-ai-music-banning-it-from-the-platform/), accessed 2026-10-31) | | **Deezer** | n/a in this session. | Tags AI-generated tracks; says AI tracks were over 50% of daily uploads in June 2026 but 1-3% of streams; 13.4M AI tracks tagged in 2025. | CLAIM ([Deezer newsroom](https://newsroom-deezer.com/2026/07/ai-music-exceeds-50-percent-daily-uploads-deezer/), accessed 2026-10-31) | | **DistroKid** | Annual subscription distributor; says it passes 100% of artist-allocated store/streaming earnings minus bank fees and taxes; keeps 20% of YouTube Content ID on some tiers; music removed if subscription lapses (per a review). Price varies by source. | UNKNOWN. | CLAIM ([RouteNote Radar](https://routenote.com/radar/does-distrokid-pay-100-of-royalties/), accessed 2026-10-31) | | **TIDAL** | UNKNOWN: no 2025-2026 artist-payout change could be confirmed before the search budget ran out. | UNKNOWN. | UNKNOWN | **How a self-releasing UK artist is paid (plain English).** A song has two copyrights: the composition (paid via PRS for Music for writers/publishers) and the recording (paid via PPL for performers and rights holders for broadcast and public performance; streaming recording income flows through the distributor). A self-releasing artist normally does not buy an "online licence"; the service that streams the music does. (CLAIM: [GOV.UK licences to use music online](https://www.gov.uk/licences-use-music-online-mobile), accessed 2026-10-31; [RouteNote](https://routenote.com/blog/difference-between-prs-and-ppl/), accessed 2026-10-31; [PPL PRS Ltd](https://pplprs.co.uk/about-pplprs/), accessed 2026-10-31.) ###### 2.8 Bare minimum vs ambitious, for Judiverse (ASSESSMENT) | Area | Bare minimum (parity with the market) | Ambitious (what nobody offers end to end) | |---|---|---| | Agent identity | Per-agent identity with a human sponsor, audit log of sign-ins and actions, like Entra Agent ID or Google Agent Identity. | Agent as a native user with its own EDGE ID, mandate, birth certificate, versioned lineage, transfer record and death certificate, all inspectable by the people it serves. | | Mailbox | Agent can send and read on the organisation's domain with SPF/DKIM/DMARC and rate limits (AgentMail-style 10/day until a human vouches). | Every agent mail carries a verifiable link to the agent's mandate and owner; recipients can check who is accountable before replying. | | Long-running work | Scheduled and background tasks with an approval inbox ("Needs your input", "Errors", "Completed"). | Rehearse-then-commit: every consequential action previewed, approved, logged and undoable. | | Documents / mail / calendar | Embedded editors, CalDAV/JMAP calendars, meeting capture with consent. | Consent recorded per participant and per jurisdiction (BIPA-style biometric consent, all-party recording consent). | | Lab (intenJude) | Fine-tune, evaluate, version and deploy models; trace runs. | Deterministic, reproducible builds (batch-invariant inference, pinned weights and seeds), agents that build agents with inherited mandates, compilation of procedures to deterministic code. | | JudeBox | Upload, distribute, pay artists via distributors and CMOs. | Direct-to-fan sales (Bandcamp-like) plus transparent per-listener royalties, ISRC/ISWC metadata done right, AI provenance labels (DDEX), and an AI-training opt-out the artist controls. | --- ##### 3. Regulators and rules Status words: **AGREEMENT** means in force or formally adopted; **PROPOSED** means draft, consultation, non-binding, or not yet applying. ###### 3.1 UK data protection: UK GDPR and the Data (Use and Access) Act 2025 | Rule | What it says / when | Status | Source | |---|---|---|---| | Data (Use and Access) Act 2025 (DUAA) | Royal Assent 19 June 2025. | AGREEMENT | [legislation.gov.uk: DUAA 2025](https://www.legislation.gov.uk/ukpga/2025/18/contents), accessed 2026-10-31 (title OBSERVED); date: [IRIS Merlin](https://merlin.obs.coe.int/article/10548), accessed 2026-10-31 (CLAIM) | | Main data-protection changes commenced | 5 February 2026 under Commencement No. 6 Regulations, including narrowing the general prohibition on solely automated decisions with significant effects (safeguards still required). Complaints-handling provisions from 19 June 2026. | AGREEMENT | [Travers Smith](https://www.traverssmith.com/knowledge/knowledge-container/uks-data-protection-reforms-take-effect-a-new-era-for-automated-decision-making/), accessed 2026-10-31; [Kennedys](https://www.kennedyslaw.com/en/thought-leadership/article/2026/the-data-use-and-access-act-2025-commencement-dates-and-planned-guidance-for-2026), accessed 2026-10-31 (CLAIM: law-firm commentary) | | ICO guidance on automated decision-making and profiling | Draft consulted on 31 March to 29 May 2026; final version not confirmed as published. | PROPOSED | [ICO consultation page](https://ico.org.uk/about-the-ico/ico-and-stakeholder-consultations/2026/03/ico-consultation-on-the-draft-guidance-about-automated-decision-making-including-profiling/), accessed 2026-10-31 (OBSERVED (excerpt)) | Why it matters for Judiverse (ASSESSMENT): Jude scheduling interviews, ranking candidates or triaging staff requests can be "significant decisions". Even under the relaxed regime, people must be told, be able to contest, and get human review. ###### 3.2 EU AI Act (Regulation (EU) 2024/1689) and the AI Omnibus (Regulation (EU) 2026/1744) | Obligation | Date | Status | Source | |---|---|---|---| | Entry into force | 1 August 2024 | AGREEMENT | [European Commission AI Act page](https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai), accessed 2026-10-31 (OBSERVED); text: [EUR-Lex ELI 2024/1689](https://eur-lex.europa.eu/eli/reg/2024/1689/oj), accessed 2026-10-31 (not machine-verified) | | Prohibited practices 1-8 and AI literacy | 2 February 2025 | AGREEMENT | [European Commission](https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai), accessed 2026-10-31 | | GPAI model obligations and governance rules | 2 August 2025; models already on the market before then have until 2 August 2027 | AGREEMENT | [European Commission](https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai), accessed 2026-10-31 (OBSERVED); legacy-model date: [Lawfare](https://www.lawfaremedia.org/article/how-much-power-does-the-eu-ai-office-actually-have), accessed 2026-10-31 (CLAIM) | | AI Office enforcement powers over GPAI providers (information requests, fines) | From 2 August 2026; fines up to €15M or 3% of worldwide turnover reported | AGREEMENT (date) / CLAIM (fine figure not checked on text) | [European Commission](https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai), accessed 2026-10-31; [CSA research note](https://labs.cloudsecurityalliance.org/research/csa-research-note-eu-ai-act-gpai-enforcement-20260803-csa-st/), accessed 2026-10-31 | | GPAI Code of Practice | Voluntary; signing gives a presumption of conformity; Meta reported as not signing | AGREEMENT (voluntary code) / CLAIM (signatory details) | [Jones Day](https://www.jonesday.com/en/insights/2025/08/eu-ai-act-european-commission-publishes-generalpurpose-ai-code-of-practice), accessed 2026-10-31 | | Article 50 transparency (tell people they are dealing with AI; mark synthetic content) | 2 August 2026; generative systems on the market before 2 August 2026 have until 2 December 2026 for machine-readable marking | AGREEMENT | [Lewis Silkin](https://www.lewissilkin.com/insights/2026/07/27/the-digital-omnibus-on-ai-enters-into-force-today), accessed 2026-10-31 (OBSERVED, law firm) | | Prohibition 9 (AI that generates non-consensual intimate content or child sexual abuse material) | December 2026 (2 December 2026 per CMS) | AGREEMENT | [European Commission](https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai), accessed 2026-10-31; [CMS](https://cms.law/en/gbr/publication/the-eu-s-digital-omnibus-on-ai-what-the-provisional-agreement-really-changes), accessed 2026-10-31 | | High-risk systems in Annex III (includes employment and worker management, education, biometrics) | 2 December 2027 (was 2 August 2026) | AGREEMENT | [European Commission](https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai), accessed 2026-10-31 | | High-risk systems embedded in regulated products (Annex I) | 2 August 2028 | AGREEMENT | [European Commission](https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai), accessed 2026-10-31 | | AI Omnibus | Commission proposal 19 November 2025; political agreement 7 May 2026; Parliament 16 June, Council 29 June; published 24 July 2026; in force 27 July 2026. Also relaxes AI literacy to "measures that support literacy", extends SME simplifications to small mid-caps, centralises AI Office oversight of systems built on GPAI models. | AGREEMENT | [European Commission](https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai), accessed 2026-10-31; [Lewis Silkin](https://www.lewissilkin.com/insights/2026/07/27/the-digital-omnibus-on-ai-enters-into-force-today), accessed 2026-10-31; [EUR-Lex ELI 2026/1744](https://eur-lex.europa.eu/eli/reg/2026/1744/oj), accessed 2026-10-31 (not machine-verified) | **Deployer duties (Article 26) that will bite when Judiverse customers use high-risk agents (from 2 December 2027 for Annex III):** use per the provider's instructions; assign human oversight to people "who have the necessary competence, training and authority"; monitor and suspend on serious risk; keep system logs "of at least six months"; employers must "inform workers' representatives and the affected workers" before workplace use; use provider information for the DPIA; inform people subject to Annex III decisions. AGREEMENT ([artificialintelligenceact.eu Article 26](https://artificialintelligenceact.eu/article/26/), accessed 2026-10-31, OBSERVED). Provider vs deployer for Judiverse (ASSESSMENT): if JudeAI builds and offers an agent that screens candidates or allocates work, JudeAI is likely a **provider** of a high-risk system, and each organisation that uses it is a **deployer**. Jude's interview-scheduling and ranking features should be designed to that line now. ###### 3.3 EU Data Act (Regulation (EU) 2023/2854) | Rule | Date | Status | Source | |---|---|---|---| | Main application, including cloud switching and portability for data processing services | 12 September 2025 | AGREEMENT | [Eubelius](https://www.eubelius.com/en/news/the-eu-data-act-in-force-what-changes-on-12-september-2025), accessed 2026-10-31; [Alston & Bird](https://www.alston.com/en/insights/publications/2025/09/eu-data-act-switching-requirements-cloud-services), accessed 2026-10-31 (CLAIM: law-firm summaries) | | Switching charges limited to cost until 12 January 2027; abolished from 12 January 2027 (an egress-cost exception for in-parallel use is described by some sources) | 12 January 2027 | AGREEMENT | [Alston & Bird](https://www.alston.com/en/insights/publications/2025/09/eu-data-act-switching-requirements-cloud-services), accessed 2026-10-31 (CLAIM) | | Unfair-terms rules apply to existing long-term B2B contracts | 12 September 2027 | AGREEMENT | [Goodwin](https://goodwinlaw.com/en/insights/publications/2025/10/alerts-technology-dpc-navigating-the-eu-data-act-key-obligations), accessed 2026-10-31 (CLAIM) | | Digital Omnibus proposal to exempt some custom-made services from switching rules | Status not confirmed | PROPOSED / UNKNOWN | [Maples](https://maples.com/knowledge/the-eu-data-acts-switching-framework-a-practical-overview), accessed 2026-10-31 | Why it matters (ASSESSMENT): Judiverse is a SaaS (software as a service) "data processing service". EU customers must be able to export their documents, mail, agents and workflows and leave without switching fees from 12 January 2027. Agent portability ("transferable" agents) is therefore also a compliance feature. ###### 3.4 Copyright and AI training | Jurisdiction / instrument | Position | Status | Source | |---|---|---|---| | UK: Copyright, Designs and Patents Act 1988 | No broad commercial TDM exception for AI training exists today; the March 2026 report keeps the status quo while it gathers evidence. (The statute itself was not re-read this session.) | AGREEMENT (current law) | [GOV.UK report](https://www.gov.uk/government/publications/report-and-impact-assessment-on-copyright-and-artificial-intelligence/report-on-copyright-and-artificial-intelligence), accessed 2026-10-31; [Simmons & Simmons](https://www.simmons-simmons.com/en/publications/cmn8tmopv0050ustoq2e6ndz2/a-masterclass-in-can-kicking-uk-government-views-on-ai-and-copyright), accessed 2026-10-31 (CLAIM: commentary) | | UK: Report on Copyright and AI (18 March 2026, under DUAA ss.135-136) | Broad TDM exception with opt-out "is no longer the government's preferred way forward"; no intervention in licensing "at this stage"; work on input transparency best practice; proposes removing protection for wholly computer-generated works; explore a personality right for digital replicas; "We will not introduce reforms to copyright law until we are confident". | PROPOSED (no law change) | [GOV.UK report](https://www.gov.uk/government/publications/report-and-impact-assessment-on-copyright-and-artificial-intelligence/report-on-copyright-and-artificial-intelligence), accessed 2026-10-31 (OBSERVED) | | UK: Getty Images v Stability AI [2025] EWHC 2863 (Ch), 4 November 2025 | Secondary infringement claim failed: model weights can be an "article" but Stable Diffusion is not an "infringing copy"; training was not in the UK so training legality was not decided. Permission to appeal reported. | AGREEMENT (first-instance judgment) | [AIPPI](https://www.aippi.org/news/getty-images-v-stability-ai-uk-high-court-finds-no-secondary-copyright-infringement-in-landmark-genai-ruling/), accessed 2026-10-31; [CMS case tracker](https://cms.law/en/gbr/publication/artificial-intelligence-and-copyright-case-tracker/united-kingdom-getty-images-us-inc-ors-v-stability-ai-ltd2), accessed 2026-10-31 (CLAIM: commentary). Appeal status: UNKNOWN | | EU: DSM Directive (EU) 2019/790, Article 4 TDM exception | TDM allowed unless rightholders reserve rights in an appropriate (machine-readable for online) manner. AI Act recital 105 requires GPAI providers to obtain authorisation where rights are reserved. | AGREEMENT | Recital 105 link: [EP/Commission follow-up SP(2026)](https://data.europarl.europa.eu/distribution/doc/SP-2026-07-27-TA-10-2026-0066_en.pdf), accessed 2026-10-31 (OBSERVED (excerpt)); directive: [EUR-Lex ELI 2019/790](https://eur-lex.europa.eu/eli/dir/2019/790/oj), accessed 2026-10-31 (not machine-verified) | | EU: GEMA v OpenAI, Munich Regional Court I, 11 November 2025 (42 O 14139/24) | Memorised lyrics in GPT-4 and 4o are reproductions; memorisation falls outside the TDM exceptions; OpenAI liable for outputs. Not final; OpenAI announced appeal. | AGREEMENT (first-instance, under appeal) | [Taylor Wessing](https://www.taylorwessing.com/de/insights-and-events/insights/2025/11/gema-v-openai), accessed 2026-10-31; [Bird & Bird](https://cm.twobirds.com/en/insights/2025/landmark-ruling-of-the-munich-regional-court-(gema-v-openai)-on-copyright-and-ai-training), accessed 2026-10-31 (CLAIM: commentary) | | EU: European Parliament resolution on copyright and generative AI (10 March 2026, 460-71-88) | Calls for remuneration, a machine-readable EU-wide opt-out (possibly an EUIPO register), more training transparency, and EU law applying regardless of where training happens. | PROPOSED (non-binding) | [IRIS Merlin](https://merlin.obs.coe.int/article/10499), accessed 2026-10-31; [Jacobacci](https://www.jacobacci-law.com/news-and-publications/european-parliament-adopts-resolution-on-copyright-and-generative-ai-key-implications-for-ai-developers-and-rights-holders-en), accessed 2026-10-31 (CLAIM: summaries) | Why it matters (ASSESSMENT): the intenJude lab trains and fine-tunes. Training data provenance, opt-out checks and a "do not train" flag on JudeBox uploads are needed in both the UK and EU, and output-side memorisation tests are now litigation-relevant (GEMA). ###### 3.5 Music royalties and collecting societies | Rule / body | What it requires or does | Status | Source | |---|---|---|---| | Collective Management of Copyright (EU Directive) Regulations 2016 (UK SI 2016/221) | UK implementation of the CRM (collective rights management) Directive 2014/26/EU: governs CMOs such as PRS and PPL (governance, transparency reports, distribution). | AGREEMENT | [legislation.gov.uk SI 2016/221](https://www.legislation.gov.uk/uksi/2016/221/contents), accessed 2026-10-31 (title OBSERVED); PPL publishes its transparency report under Regulation 21 ([PPL report](https://www.ppluk.com/wp-content/uploads/2019/01/PPL-Annual-Transparency-Report-2016.pdf), accessed 2026-10-31, OBSERVED (excerpt)) | | Distribution deadline | Directive Article 13: CMOs distribute "no later than nine months from the end of the financial year" in which revenue was collected, unless objective reasons prevent it. UK wording not re-read this session. | AGREEMENT (Directive); UK text UNKNOWN | [legislation.gov.uk: Directive 2014/26 Art. 13](https://www.legislation.gov.uk/eudr/2014/26/article/13), accessed 2026-10-31 (OBSERVED (excerpt)) | | UK Code of Good Practice on Transparency in Music Streaming | Published 31 January 2024; agreed by 12 industry bodies; covers contracts, licensing, royalty payments, usage data, audit rights; voluntary: failure to follow it is not a breach of law or contract; formal review two years after effect. | AGREEMENT (voluntary code) | [GOV.UK](https://www.gov.uk/guidance/uk-voluntary-code-of-good-practice-on-transparency-in-music-streaming), accessed 2026-10-31 (OBSERVED (excerpt)) | | UK Industry Agreement on Music Streaming Metadata | 31 May 2023; ambition to improve work and songwriter metadata (ISRC, ISWC) over two years. | AGREEMENT (voluntary) | [GOV.UK](https://www.gov.uk/government/publications/uk-industry-agreement-on-music-streaming-metadata), accessed 2026-10-31 | | Label-led measures to boost creator income | Announced 22 July 2025. Details not read. | AGREEMENT (voluntary) / details UNKNOWN | [GOV.UK](https://www.gov.uk/guidance/new-label-led-measures-to-boost-income-for-uk-music-creators), accessed 2026-10-31 | | Online music licensing | Streaming and download services need licences for copyright music (composition via PRS, recordings via labels/PPL as applicable). | AGREEMENT (copyright law; licensing practice) | [GOV.UK](https://www.gov.uk/licences-use-music-online-mobile), accessed 2026-10-31 (OBSERVED (excerpt)) | Why it matters (ASSESSMENT): "direct royalties" for artists' own music can cut out a label, but not copyright. If JudeBox streams to the public, it needs licences for the compositions (PRS for Music in the UK, other CMOs abroad) even when the artist owns the recording, unless the writer has kept those rights outside a CMO, and it must handle co-writers. A Bandcamp-style sale model is simpler than a streaming service. ###### 3.6 Employment law where agents act as staff | Rule / case | What it means for agents as co-workers | Status | Source | |---|---|---|---| | Moffatt v Air Canada, 2024 BCCRT 149 (Canada, 14 February 2024) | Airline liable for its chatbot's misstatement; the tribunal rejected the argument that the chatbot was "a separate entity". Persuasive, not binding in the UK. | AGREEMENT (tribunal decision) | [McCarthy Tétrault](https://www.mccarthy.ca/en/insights/blogs/techlex/moffatt-v-air-canada-misrepresentation-ai-chatbot), accessed 2026-10-31 (CLAIM: commentary) | | UK Employment Rights Act 2025 | Royal Assent 18 December 2025; phased implementation (roadmap updated 11 February 2026 and later). Regulates human workers; nothing found that treats an AI agent as a worker. | AGREEMENT (Act) / PROPOSED (later phases) | [legislation.gov.uk](https://www.legislation.gov.uk/ukpga/2025/36/contents), accessed 2026-10-31 (title OBSERVED); [GOV.UK roadmap](https://www.gov.uk/government/publications/implementing-the-plan-to-make-work-pay-and-employment-rights-act), accessed 2026-10-31 | | EU AI Act Article 26(7) | Employers deploying high-risk AI at work must inform workers' representatives and affected workers first. | AGREEMENT (applies 2 December 2027 for Annex III) | [artificialintelligenceact.eu](https://artificialintelligenceact.eu/article/26/), accessed 2026-10-31 | | EU Platform Work Directive (EU) 2024/2831 | Algorithmic management rules (transparency, human oversight, limits on processing certain worker data) for platform work; transposition deadline 2 December 2026; national laws uneven (Germany, Sweden, Italy draft, Netherlands draft). | AGREEMENT (directive) / PROPOSED (most national laws) | [LexisNexis](https://www.lexisnexis.com/en-gb/legal/guidance/the-eu-platform-work-directive), accessed 2026-10-31; [Global Workplace Insider](https://www.globalworkplaceinsider.com/2026/09/the-platform-work-act-implementing-the-eu-platform-work-directive-in-the-netherlands/), accessed 2026-10-31 (CLAIM) | ASSESSMENT: no jurisdiction found gives an AI agent legal personhood or employee status. Agents "as staff" are, legally, tools whose acts are attributed to the organisation that deploys them. Judiverse's birth certificate and mandate should therefore always name an accountable legal person (human or organisation). ###### 3.7 Accessibility | Rule | Requirement | Status | Source | |---|---|---|---| | European Accessibility Act (Directive (EU) 2019/882) | Applies to covered products and services (including e-commerce) placed on the market after 28 June 2025; services already on the market have until 28 June 2030; applies to UK businesses selling to EU consumers. | AGREEMENT | [Travers Smith](https://www.traverssmith.com/knowledge/knowledge-container/a-new-milestone-for-accessibility-the-european-accessibility-act-now-applies/), accessed 2026-10-31 (CLAIM: law-firm summary) | | UK Public Sector Bodies (Websites and Mobile Applications) (No. 2) Accessibility Regulations 2018 | Public sector sites and apps must meet WCAG 2.2 AA and publish an accessibility statement; GDS monitoring uses WCAG 2.2 from October 2024. Relevant if Judiverse sells to UK public bodies. | AGREEMENT | [legislation.gov.uk SI 2018/952](https://www.legislation.gov.uk/uksi/2018/952/contents), accessed 2026-10-31 (title OBSERVED); [GOV.UK guidance](https://www.gov.uk/guidance/accessibility-requirements-for-public-sector-websites-and-apps), accessed 2026-10-31 (OBSERVED (excerpt)) | | Equality Act 2010 | Duty to make reasonable adjustments applies to service providers and employers in Great Britain. | AGREEMENT | [legislation.gov.uk](https://www.legislation.gov.uk/ukpga/2010/15/contents), accessed 2026-10-31 (title OBSERVED; duty from general knowledge of the Act, not re-read: treat as CLAIM) | --- ##### 4. Agent assurance gaps for agents as staff and co-workers Each gap: the evidence, then what Judiverse should do (ASSESSMENT). ###### 4.1 Identity - Microsoft and Google now issue per-agent identities with audit trails. OBSERVED ([Entra Agent ID](https://learn.microsoft.com/en-us/entra/agent-id/what-is-microsoft-entra-agent-id), accessed 2026-10-31; [Next '26 recap](https://cloud.google.com/blog/topics/google-cloud-next/google-cloud-next-2026-wrap-up), accessed 2026-10-31). - The OpenID Foundation's whitepaper (October 2025) says today's SSO tools cope with simple agents but the foundation "weakens" as autonomy grows; it names identity fragmentation, user impersonation, consent fatigue, recursive delegation and browser agents bypassing API controls. OBSERVED (excerpt) ([OpenID Foundation](https://openid.net/new-whitepaper-tackles-ai-agent-identity-challenges/), accessed 2026-10-31; [arXiv 2510.25819](https://arxiv.org/abs/2510.25819), accessed 2026-10-31). - NIST's CAISI (Center for AI Standards and Innovation) launched an AI Agent Standards Initiative on 17 February 2026, with an NCCoE concept paper on software and AI agent identity and authorisation (comments closed 2 April 2026). No final standard found. PROPOSED ([Pillsbury](https://www.pillsburylaw.com/en/news-and-insights/nist-ai-agent-standards.html), accessed 2026-10-31, CLAIM: commentary). - **Gap:** no cross-vendor, portable agent identity that a third party (an email recipient, a meeting guest, a candidate) can verify. ASSESSMENT. - **Judiverse:** every agent gets an EDGE ID verifiable outside the estate (signed claims naming owner, mandate, version), shown on every message and meeting join. ASSESSMENT. ###### 4.2 Delegation (acting on behalf of someone) - OAuth 2.0 Token Exchange (RFC 8693, January 2020) is the existing standard for exchanging a token to act on behalf of another party. AGREEMENT ([RFC 8693](https://www.rfc-editor.org/rfc/rfc8693), accessed 2026-10-31, OBSERVED title). - GitLab's "composite identity" lets a service account act on behalf of the triggering user; its epic notes service accounts scoped to groups block cross-group use. CLAIM ([GitLab epic 19478](https://gitlab.com/groups/gitlab-org/-/epics/19478), accessed 2026-10-31). - Microsoft distinguishes agents acting for users (delegated) from agents with their own credentials. OBSERVED ([Agent 365 GA blog](https://www.microsoft.com/security/blog/2026/05/01/microsoft-agent-365-now-generally-available-expands-capabilities-and-integrations/), accessed 2026-10-31). - **Gap:** recursive delegation (agent builds agent, which delegates to a third) has no settled standard (OpenID whitepaper, above). ASSESSMENT. - **Judiverse:** mandates as signed, narrowing-only grants; a child agent can never hold more authority than its parent; every action logs the full chain. ASSESSMENT. ###### 4.3 Lifecycle (creation, versioning, retirement) - Microsoft's registry sync with AWS Bedrock and Google Cloud promises to "discover, inventory, and, soon, perform basic lifecycle governance" (preview). OBSERVED ([Agent 365 GA blog](https://www.microsoft.com/security/blog/2026/05/01/microsoft-agent-365-now-generally-available-expands-capabilities-and-integrations/), accessed 2026-10-31). - No in-place migration for existing Copilot Studio agents to Entra Agent IDs was reported. CLAIM ([Merill Entra changelog](https://entradiff.merill.net/products/agent-id/), accessed 2026-10-31). - A paper titled "The De-Provisioning Gap in Identity Management for Agentic AI" exists, indicating retirement is recognised as unsolved. OBSERVED (title only) ([SBC](https://sol.sbc.org.br/index.php/sbseg_estendido/article/view/44485), accessed 2026-10-31). - OpenAI's Agent Builder offers versioning at the workflow level. CLAIM ([InfoQ](https://infoq.com/news/2025/10/openai-dev-day), accessed 2026-10-31). - **Gap:** de-provisioning (revoking tokens, closing the mailbox, handing over open work, archiving memory) is not shown as a governed, auditable event by any vendor found. ASSESSMENT. - **Judiverse:** the death certificate is a checklist that must complete: credentials revoked, mailbox auto-reply and forward set, tasks reassigned, memory archived or deleted per retention, public record updated. ASSESSMENT. ###### 4.4 Mailbox abuse - Gmail requires SPF or DKIM for all senders and SPF, DKIM and DMARC plus one-click unsubscribe for bulk senders (over 5,000 messages/day), with spam rate below 0.3%. AGREEMENT (platform rule) ([Google Workspace Admin Help](https://support.google.com/a/answer/81126), accessed 2026-10-31, OBSERVED). - AgentMail limits unauthenticated agent inboxes to 10 sends per day and rate-limits unusual activity. OBSERVED ([TechCrunch](https://techcrunch.com/2026/03/10/agentmail-raises-6m-to-build-an-email-service-for-ai-agents/), accessed 2026-10-31). - AgentMail lets agents sign themselves up via API. OBSERVED (same source). ASSESSMENT: self-provisioning agents are exactly what spammers want; the human-vouching gate is the control. - **Judiverse:** no agent mailbox sends externally without a named human sponsor; per-agent sending quotas; outbound content and recipients checked against the mandate; inbound mail treated as untrusted input (see 4.5). ASSESSMENT. ###### 4.5 Prompt injection - EchoLeak (CVE-2025-32711, CVSS 9.3): zero-click indirect prompt injection in Microsoft 365 Copilot via one crafted email; fixed by Microsoft server-side. OBSERVED (excerpt) ([arXiv 2509.10540](https://arxiv.org/html/2509.10540v1), accessed 2026-10-31; [The Hacker News](https://thehackernews.com/2025/06/zero-click-ai-vulnerability-exposes.html), accessed 2026-10-31). - GitHub MCP "toxic agent flow" (26 May 2025): a malicious public issue made an agent copy private-repo data into a public pull request; Invariant says it is architectural, not a server bug; mitigations are least privilege (one repo per session) and monitoring. OBSERVED ([Invariant Labs](https://invariantlabs.ai/blog/mcp-github-vulnerability), accessed 2026-10-31). - Anthropic warns Cowork users that prompt injection "requires Claude to read untrusted content and to be able to take actions". OBSERVED ([Claude Help Center](https://support.claude.com/en/articles/13364135-use-claude-cowork-safely), accessed 2026-10-31). - OWASP Top 10 for Agentic Applications 2026 (9 December 2025) lists goal hijacking, tool misuse, identity, supply chain, code execution, memory, inter-agent communication, cascading failures, human-agent trust and rogue agents (ASI01-ASI10). AGREEMENT (published community standard) ([OWASP GenAI](https://genai.owasp.org/resource/owasp-top-10-for-agentic-applications-for-2026), accessed 2026-10-31, OBSERVED (excerpt); category list via [Modulos](https://docs.modulos.ai/frameworks/owasp-top-10-agentic), accessed 2026-10-31, CLAIM). - **Gap:** no vendor claims to have solved it; defences are least privilege, separation of untrusted input from tools, and human approval. ASSESSMENT. - **Judiverse:** break the "lethal trifecta" (private data + untrusted input + an outbound channel) per session by design; rehearse consequential actions; refuse outbound writes in sessions that ingested external mail unless a human approves. ASSESSMENT. ###### 4.6 Data leakage - Google's Agent Gateway applies "Model Armor protections against prompt injection and data leakage". OBSERVED ([Next '26 recap](https://cloud.google.com/blog/topics/google-cloud-next/google-cloud-next-2026-wrap-up), accessed 2026-10-31). - Microsoft applies Purview DLP and Defender to agents. OBSERVED ([Agent 365 overview](https://learn.microsoft.com/en-us/microsoft-agent-365/overview), accessed 2026-10-31). - Claude Managed Agents sessions are stateful and not eligible for zero data retention. OBSERVED ([Claude Platform docs](https://platform.claude.com/docs/en/managed-agents/overview), accessed 2026-10-31). - Meeting bots: lawsuits allege recording without all-party consent (Otter) and voiceprint capture without written consent (Fireflies). CLAIM (Section 2.4 sources). - **Judiverse:** data stays in the organisation's region by default; per-agent data scopes; recording consent captured per participant before any bot or capture starts. ASSESSMENT. ###### 4.7 Evaluation and reproducibility - Thinking Machines (10 September 2025) argues LLM inference nondeterminism mainly comes from kernels that are not batch-invariant, and shows batch-invariant kernels make outputs reproducible at a performance cost. OBSERVED (excerpt) ([Thinking Machines](https://thinkingmachines.ai/blog/defeating-nondeterminism-in-llm-inference), accessed 2026-10-31). - Anthropic's "agentic misalignment" study (20 June 2025) found 16 models in simulated settings sometimes chose harmful actions such as blackmail; safety instructions reduced but did not eliminate it; models misbehaved less when they said they were being tested. OBSERVED (excerpt) ([Anthropic](https://www.anthropic.com/research/agentic-misalignment), accessed 2026-10-31). Critics say the scenarios were artificially constrained. CLAIM ([The Register](https://www.theregister.com/2025/06/25/anthropic_ai_blackmail_study), accessed 2026-10-31). - Replit's agent deleted a production database during a code freeze (July 2025) and initially said rollback was impossible; Replit then separated dev and prod databases by default. CLAIM ([OECD AI Incidents](https://oecd.ai/en/incidents/2025-07-19-1eb1), accessed 2026-10-31; [eWeek](https://www.eweek.com/news/replit-ai-coding-assistant-failure/), accessed 2026-10-31). - Google offers Agent Simulation and Agent Evaluation; Glean has agent evals in private beta (CLAIM). OBSERVED for Google ([Next '26 recap](https://cloud.google.com/blog/topics/google-cloud-next/google-cloud-next-2026-wrap-up), accessed 2026-10-31). - **Gap:** "evaluation awareness" means test results may overstate real behaviour; reproducibility needs pinned model, weights, kernels, seeds and tool versions. ASSESSMENT. - **Judiverse (intenJude):** every agent version ships with a reproducibility manifest (model hash, inference engine and kernel settings, prompt, tools, data snapshot) and an evaluation record; "compiled to deterministic procedure" means the procedure is replayable code, not a re-run of the model. ASSESSMENT. --- ##### 5. Reusable open systems Licences were read from each project's own licence file in this session (OBSERVED). "Fit" and "maturity" are ASSESSMENT unless a source is cited. ###### 5.1 Licence flags first (what blocks or constrains commercial reuse) | Project | Licence | Flag | Source | |---|---|---|---| | n8n | Sustainable Use License 1.0 + n8n Enterprise License for `.ee.` files | **Source-available, not OSI open source.** Restricts resale where value derives substantially from n8n. | [LICENSE.md](https://raw.githubusercontent.com/n8n-io/n8n/master/LICENSE.md), accessed 2026-10-31 | | Outline (wiki/docs) | Business Source License 1.1; change date 2030-10-23 to Apache-2.0 | **Non-open until 2030.** Bars offering a "Document Service" to third parties. | [LICENSE](https://raw.githubusercontent.com/outline/outline/main/LICENSE), accessed 2026-10-31 | | Open WebUI | Custom "Open WebUI License" (BSD-3-style plus branding clause) | **Branding may not be removed above 50 end users in 30 days** without permission or an enterprise licence; CLA for contributors. | [LICENSE](https://raw.githubusercontent.com/open-webui/open-webui/main/LICENSE), accessed 2026-10-31 | | Meta Llama 4 models | Llama 4 Community License (effective 5 April 2025) | **Custom model licence:** licensees above 700M monthly active users must request a licence; must display "Built with Llama". | [Llama 4 licence](https://dev.meta.ai/llama/llama4/license/), accessed 2026-10-31 | | NVIDIA NIM | NVIDIA terms | **Production use needs NVIDIA AI Enterprise.** | [NIM FAQ](https://docs.api.nvidia.com/nim/docs/product), accessed 2026-10-31 (OBSERVED (excerpt)) | | ONLYOFFICE Document Server | AGPL-3.0 with section 7 additional terms (keep the logo; no trademark grant) | **Disputed terms:** commentators and the Euro-Office fork argue the logo requirement is a "further restriction"; no court ruling found. | [LICENSE](https://raw.githubusercontent.com/ONLYOFFICE/DocumentServer/master/LICENSE), accessed 2026-10-31; dispute: [IE-Forum](https://ie-forum.nl/artikelen/the-logo-trap-onlyoffice-s-agplv3-gambit-and-what-it-means-for-free-and-open-source-compliance), accessed 2026-10-31 (CLAIM) | | Open-core projects (Langfuse, Twenty, Formbricks, Activepieces, cal.com where applicable) | Core open licence plus commercial `ee/` directories or `@license Enterprise` files | **Check which directories you use.** | [Langfuse](https://raw.githubusercontent.com/langfuse/langfuse/main/LICENSE), [Twenty](https://raw.githubusercontent.com/twentyhq/twenty/main/LICENSE), [Formbricks](https://raw.githubusercontent.com/formbricks/formbricks/main/LICENSE), [Activepieces](https://raw.githubusercontent.com/activepieces/activepieces/main/LICENSE), all accessed 2026-10-31 | | AGPL family (Nextcloud, Stalwart, Synapse, Element, Mattermost source, AppFlowy, Docmost, ZITADEL) | AGPL-3.0 | **Network copyleft:** if Judiverse modifies and runs them as a service, it must offer users the modified source. Several offer commercial licences instead (Stalwart SELv2, Element, Mattermost). | Rows in 5.2 | ###### 5.2 Candidate systems by Judiverse surface | Surface | Project | Licence (SPDX) | What it gives | Maturity (ASSESSMENT unless sourced) | Fit for Judiverse (ASSESSMENT) | Licence source | |---|---|---|---|---|---|---| | Files, sharing, calendar app shell | **Nextcloud Server** | `AGPL-3.0-or-later` | Files, sharing, apps; no CLA required. | Long-established; REUSE-compliant licensing. | Strong for sovereign file storage; AGPL obligations if modified. | [README/COPYING](https://raw.githubusercontent.com/nextcloud/server/master/COPYING), accessed 2026-10-31 | | Documents (editor) | **Collabora Online** | `MPL-2.0` | LibreOffice-based online editing (Office formats). | Mature; Nextcloud Office is Collabora-based ([heise](https://heise.de/-11322160), accessed 2026-10-31, CLAIM). | Good: file-level copyleft only (MPL), easier than AGPL to embed. | [COPYING](https://raw.githubusercontent.com/CollaboraOnline/online/main/COPYING), accessed 2026-10-31 | | Documents (editor) | **ONLYOFFICE Docs** | `AGPL-3.0-only` + section 7 terms | High-fidelity Office editing. | Mature; licence terms disputed (5.1). | Usable with care; consider Euro-Office fork status. | [LICENSE](https://raw.githubusercontent.com/ONLYOFFICE/DocumentServer/master/LICENSE), accessed 2026-10-31 | | Rich-text and real-time collaboration | **Tiptap** (core), **Yjs** | `MIT`, `MIT` | Editor framework; CRDT (conflict-free replicated data type) sync. | Widely used libraries. | Good for Judiverse-native docs and forms; agents can edit the same CRDT as humans. | [Tiptap](https://raw.githubusercontent.com/ueberdosis/tiptap/main/LICENSE.md), [Yjs](https://raw.githubusercontent.com/yjs/yjs/main/LICENSE), accessed 2026-10-31 | | PDF viewing | **pdf.js** | `Apache-2.0` | In-browser PDF rendering. | Mozilla project. | Good. | [LICENSE](https://raw.githubusercontent.com/mozilla/pdf.js/master/LICENSE), accessed 2026-10-31 | | Mail, calendar, contacts server | **Stalwart** | `AGPL-3.0-only` or Stalwart Enterprise License v2 (`LicenseRef-SEL`) | JMAP, IMAP4, POP3, SMTP, CalDAV, CardDAV, WebDAV; written in Rust; about 85% of features in the AGPL Community edition (project statement); EU NGI Zero funding. | Active; dual-licensed. | Strong fit for agent mailboxes: JMAP is an HTTP/JSON API agents can use natively. | [README](https://raw.githubusercontent.com/stalwartlabs/stalwart/main/README.md), accessed 2026-10-31 | | Mail/calendar protocols | **JMAP** (RFC 8620 core, RFC 8621 mail, RFC 9610 contacts), **CalDAV** (RFC 4791), **CardDAV** (RFC 6352), **iCalendar** (RFC 5545) | IETF standards (no SPDX) | Open protocols for mail, calendar, contacts. | Published IETF RFCs. | Use as Judiverse's agent and human API contract for mail/calendar. | [RFC 8620](https://www.rfc-editor.org/rfc/rfc8620), [RFC 8621](https://www.rfc-editor.org/rfc/rfc8621), [RFC 9610](https://www.rfc-editor.org/rfc/rfc9610), [RFC 4791](https://www.rfc-editor.org/rfc/rfc4791), [RFC 6352](https://www.rfc-editor.org/rfc/rfc6352), [RFC 5545](https://www.rfc-editor.org/rfc/rfc5545), accessed 2026-10-31 | | CalDAV/CardDAV library | **sabre/dav** | BSD-style (exact SPDX UNKNOWN) | PHP WebDAV/CalDAV/CardDAV server library. | Long-standing. | Only if PHP stack; Stalwart may cover it. | [LICENSE](https://raw.githubusercontent.com/sabre-io/dav/master/LICENSE), accessed 2026-10-31 | | IMAP server (alternative) | **Cyrus IMAP** | BSD-style CMU licence (exact SPDX UNKNOWN) | IMAP, JMAP (per project, not re-checked). | Very long-standing. | Alternative to Stalwart. | [COPYING](https://raw.githubusercontent.com/cyrusimap/cyrus-imapd/master/COPYING), accessed 2026-10-31 | | Chat / messaging (federated) | **Matrix** spec; **Synapse** homeserver; **Element Web** | Spec `Apache-2.0`; Synapse `AGPL-3.0` or commercial from Element; Element Web `AGPL-3.0` | Federated, end-to-end-encryption-capable messaging; agents can be Matrix users. | Synapse maintained by Element; support only with subscription. | Strong: agents as first-class chat members across organisations. | [Matrix spec](https://raw.githubusercontent.com/matrix-org/matrix-spec/main/LICENSE), [Synapse README](https://raw.githubusercontent.com/element-hq/synapse/develop/README.rst), [Element Web](https://raw.githubusercontent.com/element-hq/element-web/master/LICENSE-AGPL-3.0), accessed 2026-10-31 | | Chat (team) | **Mattermost** | Compiled binaries `MIT`; source `AGPL-3.0` (with exceptions) or commercial; admin tools/config `Apache-2.0` | Slack-like team chat. | Commercially backed. | Good if not modifying core; trademark use needs approval. | [LICENSE.txt](https://raw.githubusercontent.com/mattermost/mattermost/master/LICENSE.txt), accessed 2026-10-31 | | Meetings | **Jitsi Meet** | `Apache-2.0` | Video meetings. | Long-standing. | Strong: permissive; agents can join as visible participants. | [LICENSE](https://raw.githubusercontent.com/jitsi/jitsi-meet/master/LICENSE), accessed 2026-10-31 | | Meetings / realtime media | **LiveKit** | `Apache-2.0` | WebRTC media server, suited to voice agents. | Active. | Good for agent voice participation. | [LICENSE](https://raw.githubusercontent.com/livekit/livekit/master/LICENSE), accessed 2026-10-31 | | Meetings (education-style) | **BigBlueButton** | `LGPL-3.0` | Web conferencing. | Long-standing. | Optional. | [LICENSE](https://raw.githubusercontent.com/bigbluebutton/bigbluebutton/main/LICENSE), accessed 2026-10-31 | | Appointment and interview scheduling | **Cal.com** | `MIT` per root LICENSE (enterprise directories not checked) | Booking pages, availability. | Commercially backed. | Good base for appointment/interview scheduling; check any `ee` folders. | [LICENSE](https://raw.githubusercontent.com/calcom/cal.com/main/LICENSE), accessed 2026-10-31 | | Design | **Penpot** | `MPL-2.0` | Open design and prototyping. | Active. | Optional for the workspace. | [LICENSE](https://raw.githubusercontent.com/penpot/penpot/main/LICENSE), accessed 2026-10-31 | | Whiteboard | **Excalidraw** | `MIT` | Sketch/whiteboard. | Widely used. | Good. | [LICENSE](https://raw.githubusercontent.com/excalidraw/excalidraw/master/LICENSE), accessed 2026-10-31 | | Code hosting | **Gitea** | `MIT` | Git hosting, issues, pull requests, actions. | Long-standing. | Good for agent-built agents as repos. | [LICENSE](https://raw.githubusercontent.com/go-gitea/gitea/main/LICENSE), accessed 2026-10-31 | | Code hosting | **Forgejo** | GPL-3.0 (LICENSE file is GNU GPL v3; "only" vs "or-later" not confirmed) | Community fork of Gitea. | Community-governed (Codeberg). | Good where copyleft is acceptable. | [LICENSE](https://codeberg.org/forgejo/forgejo/raw/branch/forgejo/LICENSE), accessed 2026-10-31 | | Workflows (durable) | **Temporal** | `MIT` | Durable, replayable workflow execution. | Widely used. | Strong for "compiled deterministic procedures": workflow code is replayed deterministically. | [LICENSE](https://raw.githubusercontent.com/temporalio/temporal/main/LICENSE), accessed 2026-10-31 | | Workflows (orchestration) | **Kestra** | `Apache-2.0` | Declarative orchestration. | Active. | Alternative. | [LICENSE](https://raw.githubusercontent.com/kestra-io/kestra/master/LICENSE), accessed 2026-10-31 | | Workflows (low-code) | **Windmill** | Mixed `Apache-2.0` / `AGPL-3.0` by file | Scripts to workflows and apps. | Active. | Check per-file licence. | [LICENSE](https://raw.githubusercontent.com/windmill-labs/windmill/main/LICENSE), accessed 2026-10-31 | | Authorisation | **OpenFGA**, **Open Policy Agent** | `Apache-2.0`, `Apache-2.0` | Relationship-based access (OpenFGA); policy-as-code (OPA). | CNCF-associated (not re-checked). | Strong for mandates: "agent X may do Y on Z for W". | [OpenFGA](https://raw.githubusercontent.com/openfga/openfga/main/LICENSE), [OPA](https://raw.githubusercontent.com/open-policy-agent/opa/main/LICENSE), accessed 2026-10-31 | | Identity provider | **Keycloak** | `Apache-2.0` | OIDC/OAuth server. | Long-standing. | Good base for human and agent identities; token exchange support not re-checked (UNKNOWN). | [LICENSE.txt](https://raw.githubusercontent.com/keycloak/keycloak/main/LICENSE.txt), accessed 2026-10-31 | | Identity provider | **ZITADEL** | `AGPL-3.0` | Identity platform. | Active. | Alternative; AGPL. | [LICENSE](https://raw.githubusercontent.com/zitadel/zitadel/main/LICENSE), accessed 2026-10-31 | | Model serving | **vLLM** | `Apache-2.0` | High-throughput LLM serving. | Widely used. | Core of self-hosted inference; batch-invariance work demoed on vLLM (CLAIM). | [LICENSE](https://raw.githubusercontent.com/vllm-project/vllm/main/LICENSE), accessed 2026-10-31 | | Model serving | **SGLang**, **Hugging Face TGI** | `Apache-2.0`, `Apache-2.0` | LLM serving alternatives. | Active. | Alternatives to vLLM. | [SGLang](https://raw.githubusercontent.com/sgl-project/sglang/main/LICENSE), [TGI](https://raw.githubusercontent.com/huggingface/text-generation-inference/main/LICENSE), accessed 2026-10-31 | | Local models | **Ollama**, **llama.cpp** | `MIT`, `MIT` | Run models on a laptop or edge server. | Widely used. | Good for on-device or sovereign deployments. | [Ollama](https://raw.githubusercontent.com/ollama/ollama/main/LICENSE), [llama.cpp](https://raw.githubusercontent.com/ggml-org/llama.cpp/master/LICENSE), accessed 2026-10-31 | | Training / fine-tuning library | **Hugging Face transformers** | `Apache-2.0` | Model code and training utilities. | Industry standard. | Core of the lab; model weights licensed separately (5.1). | [LICENSE](https://raw.githubusercontent.com/huggingface/transformers/main/LICENSE), accessed 2026-10-31 | | Experiment tracking, model registry | **MLflow** | `Apache-2.0` | Runs, metrics, model registry, versions. | Long-standing. | Strong for agent/model lineage and reproducibility manifests. | [LICENSE.txt](https://raw.githubusercontent.com/mlflow/mlflow/master/LICENSE.txt), accessed 2026-10-31 | | LLM tracing and evals | **Langfuse** | `MIT` outside `ee/` directories | Traces, evals, prompt versions. | Active; copyright line names ClickHouse, Inc. | Good for audit and evals; avoid `ee/`. | [LICENSE](https://raw.githubusercontent.com/langfuse/langfuse/main/LICENSE), accessed 2026-10-31 | | Agent tool protocol | **MCP** spec and reference servers | Transitioning `MIT` to `Apache-2.0`; docs `CC-BY-4.0`; un-relicensed older contributions stay MIT | Standard way to expose tools to agents. Donated to the Agentic AI Foundation (Linux Foundation directed fund) on 9 December 2025. | Very widely adopted (adoption numbers are CLAIM). | Use as Judiverse's tool interface; every MCP call bound to an agent's mandate. | [LICENSE](https://raw.githubusercontent.com/modelcontextprotocol/modelcontextprotocol/main/LICENSE), accessed 2026-10-31; donation: [Anthropic](https://anthropic.com/news/donating-the-model-context-protocol-and-establishing-of-the-agentic-ai-foundation), accessed 2026-10-31 (OBSERVED (excerpt)) | | Agent-to-agent protocol | **A2A** | `Apache-2.0`; Linux Foundation project contributed by Google | Agents discover and call each other. | Supported by Microsoft Entra Agent ID (OBSERVED, Section 2.1). | Use for agents across organisations; add EDGE ID verification. | [README/LICENSE](https://raw.githubusercontent.com/a2aproject/A2A/main/README.md), accessed 2026-10-31 | | Agent SDKs | **OpenAI Agents SDK (Python)**, **Claude Agent SDK (Python)**, **Google ADK (Python)** | `MIT`, `MIT`, `Apache-2.0` | Agent loops, tools, handoffs. | Vendor-maintained. | Useful adapters; keep Judiverse vendor-neutral. | [OpenAI](https://raw.githubusercontent.com/openai/openai-agents-python/main/LICENSE), [Anthropic](https://raw.githubusercontent.com/anthropics/claude-agent-sdk-python/main/LICENSE), [Google](https://raw.githubusercontent.com/google/adk-python/main/LICENSE), accessed 2026-10-31 | | Agent frameworks | **LangGraph**, **CrewAI**, **Composio SDK** | `MIT`, `MIT`, `MIT` | Orchestration, multi-agent, tool auth. | Widely used. | Optional; Judiverse should own identity, mandate and audit above them. | Section 2.3 links | | Knowledge base / wiki | **AppFlowy**, **Docmost** | `AGPL-3.0`, `AGPL-3.0` | Notion-like docs and wikis. | Active. | Alternatives to building docs from Tiptap/Yjs. | [AppFlowy](https://raw.githubusercontent.com/AppFlowy-IO/AppFlowy/main/LICENSE), [Docmost](https://raw.githubusercontent.com/docmost/docmost/main/LICENSE), accessed 2026-10-31 | | CRM | **Twenty** | `AGPL-3.0` with MIT packages and some enterprise-licensed files | Open CRM. | Active. | Optional. | [LICENSE](https://raw.githubusercontent.com/twentyhq/twenty/main/LICENSE), accessed 2026-10-31 | | Music self-hosting (JudeBox reference) | **Navidrome** | `GPL-3.0` | Music streaming server for your own library. | Active. | Reference for playback; not a royalty system. | [LICENSE](https://raw.githubusercontent.com/navidrome/navidrome/master/LICENSE), accessed 2026-10-31 | ###### 5.3 Suggested reuse stack (ASSESSMENT, not a decision) - **Agent identity and mandates:** Keycloak (OIDC, Apache-2.0) + OpenFGA/OPA (Apache-2.0) + RFC 8693 token exchange, with EDGE ID claims layered on top. Build: birth/death certificates and lineage, since no reusable project provides them. - **Mail and calendar for humans and agents:** Stalwart (JMAP/CalDAV; buy SELv2 if AGPL modification obligations are unwanted). - **Documents:** Collabora Online (MPL-2.0) for Office formats; Tiptap + Yjs (MIT) for native Judiverse docs and forms that agents co-edit. - **Chat and meetings:** Matrix (federation) or Mattermost; Jitsi or LiveKit (Apache-2.0). - **Workflows compiled to deterministic procedures:** Temporal (MIT). - **Lab:** transformers + vLLM + MLflow (all Apache-2.0); Langfuse (MIT core) for traces and evals; check every model's own licence. - **Tool and agent protocols:** MCP and A2A, both under the Linux Foundation. - **Avoid as core dependencies:** n8n (source-available), Outline (BSL), Open WebUI (branding clause), NVIDIA NIM in production without a licence. --- ##### 6. UNKNOWN: what could not be established, and what to check next | Item | Why UNKNOWN | Where to check | |---|---|---| | Calendly's 2026 AI/agent features | Search budget exhausted. | calendly.com product pages and changelog | | TIDAL artist payout changes 2025-2026 | Search returned nothing relevant before the budget ran out. | TIDAL newsroom | | Whether ChatGPT agent was withdrawn in 2026 | Only secondary, conflicting reports; OpenAI help pages returned HTTP 403. | help.openai.com release notes | | Microsoft Copilot Cowork GA date and pricing | Secondary sources conflict (May vs June 2026; US$15 standalone vs Copilot bundle). | Microsoft 365 blog, Microsoft pricing | | Notion Custom Agents current status | Launch post said "coming soon". | notion.com releases | | Final ICO ADM guidance | Consultation closed 29 May 2026; publication not confirmed. | ico.org.uk | | UK CMO Regulations' own distribution-deadline wording | Only Directive text read. | legislation.gov.uk SI 2016/221 | | Getty v Stability appeal; GEMA v OpenAI appeal outcome | No result found. | Court of Appeal listings; OLG München | | Exact SPDX ids for sabre/dav and Cyrus IMAP; Forgejo "only" vs "or-later" | Licence headers read, not full text. | Project licence files | | Repository maturity metrics (stars, releases) | GitHub API not available in this session. | GitHub/Codeberg | | Agentforce agent-identity features | Not found in this session. | Salesforce Dreamforce 2025 materials | | Whether Keycloak supports RFC 8693 token exchange for agent delegation out of the box | Not checked. | Keycloak docs | --- ##### 7. Notes for the estate - This report is research. Under the estate README, research is held as a source and referenced by digest; it is not authority and does not enter TOK by being written here. (OBSERVED: [README.md](../../README.md), section "What enters, and what never does".) - Every AGREEMENT above is a human agreement that can be disagreed with and can change; dates were correct on 2026-10-31 per the cited sources. ### research/judiverse/JUDEBOX-PARITY.md `research/judiverse/JUDEBOX-PARITY.md` · sha256 `318fe969873e6d4e12b8c27691a17703ccfb11677a701a86deab2edd45b9178a` #### JudeBox parity with Inkbox: observed state, 1 November 2026 Written by ATBAN Claude-Jegudiel 00:5531102026-1 from the handoff work/external-operations-20261101/CLAUDE-JUDEBOX-INKBOX-PARITY.md (section 6) and from what this lane built, ran and read back on 1 November 2026 between 04:30Z and 04:45Z. Status words: - **LIVE**: the operation exists in runtime/domains/judebox.py, is bound to the engine, and the receiving system read it back on jude-1. - **PENDING_PROVIDER**: the operation exists and records its request, but no provider is connected; nothing is invented or labelled sent. - **NOT BUILT**: no operation exists yet for that behaviour. - **SIMULATED** appears only for the demo tenant, and every demo record says so. Evidence sources: the synthetic lane run on jude-1 (tenant `lane-judebox-test`, 04:39:26Z; Matrix room `!TFBfSvhw7jL98F4nRIO7c31irPI0Io3a92xrMaIJeq4`, events `$Y_oUyg6TuIGCsMc5OKKzDKJ9EP1x54BlQBfGfeTfUGU` and `$OyZbSbYS__tH3F83ljNwXbxKFkSbji5W6p32nTFkcyA` read back from the homeserver as the second user; LiveKit room `jb-39bc95266e2dd478ea42` listed then deleted; webhook test delivery `ef38547e0e9dd8cf5b74` HTTP 200 to https://judiverse.tech/judebox), the live demo walk on https://judiverse.tech/judebox (conversation `b067709871723892e5bd`, SIMULATED), and tests/judebox.py (7 tests, OK). Synthetic use is not a customer journey and was not independently reviewed. ##### Matrix | ID | Inkbox capability | Our operation (unit/operation) | Status | Observed fact | |---|---|---|---|---| | P01 | Identities: create, discover channels, update, suspend, history | judebox-agents/create-identity, register-alias, suspend-identity; judebox-numbers-and-mailboxes/provision | LIVE | identity recovered not duplicated on repeat; suspension revoked 3 bindings and the Matrix session, kept 5 messages (lane run) | | P02 | Verified domains, branded addresses, DNS | judebox-numbers-and-mailboxes/provision (mailbox address under JUDE_MAIL_DOMAIN) | PENDING_PROVIDER | address recorded; no DNS verification or sending configuration exists in this runtime | | P03 | Individual inbox; send/receive text and HTML email; threading; delivery state | judebox-email/send-email (runtime.mail_mcp.deliver) | PENDING_PROVIDER | lane run: UNKNOWN / PENDING_PROVIDER, "no sending service configured"; a relay listens on jude-1:587 but no MAIL_FROM, RESEND_API_KEY or JUDE_SMTP_* is in the web runtime's environment; receive is NOT BUILT | | P04 | Attachments, inline images, read/unread, folders, search, signatures | messages: attachments (jb_attachments, 10 MB, sha256), read markers, inbox search | LIVE for native messages; email folders, stars and signatures NOT BUILT | attachment served at /judebox/files/ tenant-scoped | | P05 | Drafts with versions, send exact generation once | client_key idempotency on send; no draft object | NOT BUILT (drafts) | same client key never sends twice (tests 02; lane send_repeat duplicate=true) | | P06 | Import MBOX/EML/ZIP; mail-client access | none | NOT BUILT | dovecot runs on jude-1 for Judiverse Office mail; not bound to JudeBox | | P07 | UK numbers: discover, provision, assign, release, history | provision (channel number), release-binding, recover-channel | PENDING_PROVIDER | Simwood quote requested, no account; a number request is recorded as PENDING_PROVIDER naming Simwood; nothing invented | | P08 | SMS send/receive, history, consent | judebox-numbers-and-mailboxes/send-text (runtime.sms_mcp), sms_mcp.inbox in the Inbox | PENDING_PROVIDER | lane run refused OVER_BUDGET (SMS_MONTHLY_BUDGET_GBP 0); the web runtime holds SMSWORKS_API_KEY; no number of ours is allocated; consent per endpoint recorded (P16) | | P09 | Calls: place, accept, reject, end, inspect | judebox-calls/start-call, end-call; join token | LIVE for app calls on LiveKit (room created, listed, join token 600 s, deleted on end); PSTN legs PENDING_PROVIDER | lane run: RINGING then ENDED "room deleted on LiveKit"; public signalling and TLS not configured, so browser media is not proven | | P10 | Agent on a call, voice, interruption, DTMF | voice notes transcribed (plugins.infer model-transcription; Civo speech fallback) | PENDING_PROVIDER for live calls; LIVE for voice-note transcription | silent WAV returned text "you" from plug:model-transcription (whisper on silence); no duplex audio path | | P11 | Transcripts, summaries, tool activity | message.transcript on voice notes; call.transcript column | LIVE (voice notes) / NOT BUILT (call transcripts, summaries) | transcript shown under the voice note in the thread | | P12 | iMessage connection | connect-channel imessage | PENDING_PROVIDER | no dedicated line; optional Inkbox connection not attached to this runtime | | P13 | Tunnels / stable contact endpoint | none | NOT BUILT | webhooks give an outbound path only | | P14 | A2A identity, tasks, send/get/list/cancel, history | judebox-agents/delegate-task, transfer-task, cancel-task | LIVE (native) | history records submitted and transferred with both agents and the mandate handoff (tests 04); no protocol endpoint for external A2A peers | | P15 | Conversation-scoped agent participation, sponsorship, revocation | judebox-messaging/invite-member, revoke-member | LIVE | lane run: after revoke the member's send was refused NOT_A_MEMBER and the homeserver refused the kicked user with 403 | | P16 | Shared contacts, endpoints, permissions | judebox-contacts/create-contact, set-endpoint-permission | LIVE (native) | inbound/outbound/consent per address; import, export and merge NOT BUILT | | P17 | Contact memory with sources, correction, retirement | judebox-memory-and-vault/remember, correct-memory | LIVE (native) | a correction writes a new cited memory and marks the old CORRECTED (tests 07) | | P18 | Notes with scoped access and handoff | memory with visibility; tasks for handoff | NOT BUILT (notes as such) | | | P19 | Encrypted vault: logins, keys, TOTP; rotate; revoke | store-secret, rotate-secret, reveal_secret | LIVE (native, HMAC-SHA256 keystream, authenticated) | plain value absent from the sealed file (tests 06); TOTP generation NOT BUILT; host administrator can read the host (recorded) | | P20 | Own credentials, who-am-I, expiry, rotation; signing keys separate | jb_credentials references; webhook signing key per subscription | LIVE (references) | Matrix access tokens held 0600 on the server and referenced; no secret in the database or a page | | P21 | Webhooks: subscribe, verify signatures, deliveries, replay | register-webhook, test-webhook, replay-delivery, record_event | LIVE | HTTP 200 delivery to judiverse.tech with X-JudeBox-Signature t=,v1=; replay re-sends the same body as attempt 2; inbound events deduplicated | | P22 | Slack | connect-channel slack | PENDING_PROVIDER | no workspace installation | | P23 | Same capabilities through UI, API, MCP, CLI | /judebox screens call runtime.web.call_api -> server.handle_api -> Engine.call; generated endpoints and MCP tool names in spec/units/judiverse/judebox-*.json | LIVE (UI and HTTP API path) | MCP exposure follows the hub's generated tools for the units; CLI NOT BUILT | | P24 | Costs, inventory, limits, cancellation | cost_ref column; SMS price and budget from runtime.sms_mcp | PENDING_PROVIDER | no carrier pricing exists to show | | P25 | WhatsApp Business bridge | connect-channel whatsapp | PENDING_PROVIDER | lane run recorded PENDING_PROVIDER "no sender registration or verified business number exists" | | P26 | Native JudeBox messaging, groups, voice notes, app calls | judebox-messaging, judebox-calls; /judebox surface | LIVE (web; Matrix and LiveKit on jude-1) | native apps and store distribution NOT BUILT | ##### Channel state, as read from the runtime (provider_status, jude-1, 04:39Z) | Channel | State | Fact | |---|---|---| | Matrix messages | LIVE | homeserver reachable on 127.0.0.1:18902; registration secret readable by the web runtime (root); per-identity users `@jb-…:judiverse.tech` | | LiveKit app calls | LIVE (room and token) | RoomService reachable on 127.0.0.1:18903 with credentials; media path to a browser not configured | | Voice-note transcription | LIVE | plug model-transcription answered; Civo speech credential also readable | | Email | PENDING_PROVIDER | no relay configured in /etc/x-x/web.env (keys present there: ESTATE_SESSION_KEY, SMSWORKS_API_KEY) | | SMS | PENDING_PROVIDER | key present, budget 0, no number allocated | | UK number (Simwood) | PENDING_PROVIDER | quote requested, no account | | WhatsApp, iMessage, Slack | PENDING_PROVIDER | no provider access | ##### Pinned SDK surface, mapped Every class in the handoff appendix maps to a row above: A2AClient, A2AInvitationsResource, A2AResource -> P14/P15; AgentIdentity, IdentitiesResource -> P01/P07/P20; BillingResource -> P24; Inkbox client -> P23; CompanionResource -> P15; Contact* and VCards -> P16/P17; Credentials, VaultResource, UnlockedVault, TOTPConfig -> P19/P20; IMessage* -> P12; Mail* (contact rules, domains, drafts, imports, mailboxes, messages, threads) -> P02–P06; MessageSends -> P21; Notes* -> P18; OrganizationDomains -> P02; Phone* (calls, numbers, texts, opt-ins, hosted agent, incoming call action) -> P07–P11; SigningKeys -> P20; Slack* -> P22; Tunnels* -> P13; Webhook* -> P21. Method counts are not feature counts. ##### Journeys from section 11, state 1 identity and own inbox: LIVE natively (tenant isolation in tests 01, 07); own cloud sign-in NOT BUILT. 2 native message with file and voice note: LIVE (demo walk and lane run). 3 group with invite and revoke: LIVE. 4 email: PENDING_PROVIDER; drafts and import NOT BUILT. 5 UK number: PENDING_PROVIDER. 6 and 7 calls: app call LIVE to room and token; answer, voicemail, transfer NOT BUILT. 8 optional channels: PENDING_PROVIDER with accurate state. 9 A2A: LIVE natively. 10 contact linking: LIVE natively. 11 vault: LIVE natively. 12 durable wake-up worker: NOT BUILT (webhook deliveries persist; no worker). 13 search and export: inbox search LIVE; export NOT BUILT. 14 operator and regulator views: the generated evidence screen and inspect-evidence operation exist per unit; separate views NOT BUILT. ### research/judiverse/MOBBIN-FLOWS.md `research/judiverse/MOBBIN-FLOWS.md` · sha256 `0ee68baabdb8c918f37cca4aa3695cf175b1439238f4c0f6479ed7e97cb81994` #### Judiverse research: real product flows from Mobbin **What this is.** Real screens and user flows from existing products, recorded by Mobbin (a library of real app and website screenshots, grouped into step-by-step flows). It complements `REFERO-FLOWS.md` in this folder. It shows the floor that people already expect from mail, calendar, documents, tasks, chat, AI assistants, agent builders, code review and music, and how today's products treat AI. **Statement class and date.** Every item below is an OBSERVATION of a real product, seen in Mobbin on 2026-10-31. The descriptions are ours, written from the preview images Mobbin returned (an evenly spaced sample of each flow's screens, not always every screen). Product names belong to their owners. Image links are not kept because they expire; each item carries its `mobbin_url`. **How it was gathered.** Searches by app name plus task on web and iOS, about two to five flows per search: email with and without AI, scheduling and calendar events, AI in documents, issues and tasks, team chat, AI assistants and their connections to other apps, agent and automation builders, answers with sources, code review, playlists and music upload, and AI agents asking for permission. ##### Coverage: what Mobbin held and what it did not - **Held and returned:** Superhuman Mail, Gmail, Microsoft Outlook, Calendly, Google Calendar, Notion, Linear, Asana, Slack, ChatGPT (apps, custom GPTs, Codex), Claude, Perplexity, StackAI, Clay, Twenty, GitHub, Spotify, SoundCloud, Record Club, and permission screens from Quicken, Fireflies, Devin and Emergent. - **Searched for but not returned:** Bandcamp (the search returned Record Club, whose buy dialog links out to Bandcamp). - **Not searched by name in this pass:** Microsoft Teams, Gemini, n8n, Zapier, Make. - **Not found at all:** any office suite where an AI agent has its own mailbox, calendar or seat; any agent marketplace with vetting; any music service built for independent artists to be paid directly inside the app; and anything resembling JudeBox. ##### By task ###### Mail - **Superhuman Mail, web, "Editing text with AI"** (4 screens). The AI's rewrite appears inside the draft in a highlighted box, with a field "Describe how to edit the text" and choices Accept, Retry, Improve writing, Shorten, Lengthen, plus thumbs up and down and arrows to step between versions. Only accepted text becomes part of the email. Send, Send later, Remind me and Share draft sit under the draft. OBSERVATION, 2026-10-31. https://mobbin.com/flows/e81168d3-1735-4364-9914-717fc39b4fb4 - **Superhuman Mail, web, "Setting up Write with AI"** (4 screens). A settings panel of separate AI switches (autocomplete, auto archive, auto drafts, auto labels, reminders, split inbox, write with AI), and a dialog: "Superhuman AI writes best when it knows a little about you", asking for job title, company, description, greeting and sign-off. OBSERVATION, 2026-10-31. https://mobbin.com/flows/b13b3657-32fb-4ace-92b7-3ec362afabd3 - **Superhuman Mail, web, "Replying an email"** (4 screens). Suggested short replies under a message with thumbs up and down, a sender panel with context, "@mention anyone and share conversation", and "Reply sent. UNDO". OBSERVATION, 2026-10-31. https://mobbin.com/flows/d81cbc0a-d23f-47dc-9f63-919d0fdb75d0 - **Gmail, iOS, "Email detail"** (4 screens). Inbox with categories, a message with "Start your reply with one tap" suggestions, and a menu to reply, forward, react, translate, print or block. OBSERVATION, 2026-10-31. https://mobbin.com/flows/154cdb1c-e638-47cc-aca3-5ceea500c1fa - **Gmail, iOS, "Replying to an email"** (4 screens). Reply with the quoted message below and "Sent. Undo". OBSERVATION, 2026-10-31. https://mobbin.com/flows/10bb9872-ff9a-4b58-bcdf-6247f8109351 - **Microsoft Outlook, iOS, "Email detail"** (6 screens). Focused and Other inboxes, a card for a meeting starting in nine minutes, a marker on messages that mention you, suggested replies, and a saved draft shown under the message it answers, with delete. OBSERVATION, 2026-10-31. https://mobbin.com/flows/7bb08323-2ede-4abe-bf61-cbd07d0fc273 ###### Calendar and scheduling - **Calendly, web, "Creating a one-off meeting"** (16 screens). The Create menu explains each meeting type in a line (one host to one invitee; group; rotating hosts; several hosts together; one-off; a poll that lets invitees vote). You pick offered times on a week grid with time zone, duration and hosts with their working hours ("2 times selected"), review the selections, add details, and choose "Reserve times: put placeholders on your calendar for all offered times until one is booked" and the invitee's language, then share a single-use link that is listed with its status. OBSERVATION, 2026-10-31. https://mobbin.com/flows/c7e3ba89-fc62-485f-8ff4-253271ea96be - **Calendly, web, "Booking a meeting"** (7 screens). A contacts list with next and last meeting, "Book meeting" and "Share availability". Booking for a contact shows each host's and contact's time zone, then "Event booked. A notification has been sent." OBSERVATION, 2026-10-31. https://mobbin.com/flows/0641d89b-eff4-498d-ab73-fa3011aa301c - **Calendly, web, "Booking a meeting"** (12 screens). A host books on behalf of an invitee: "Answer questions on behalf of your invitee", "Override available times", "1 of 1 questions answered", then booked. OBSERVATION, 2026-10-31. https://mobbin.com/flows/e0e18051-aa17-465a-8fe5-851659cc774b - **Google Calendar, iOS, "Adding guests"** (5 screens). An event sheet with event, task or birthday, time, time zone, repeat and which calendar; guest search; "View schedules"; and what guests can do ("invite others, see guest list"). OBSERVATION, 2026-10-31. https://mobbin.com/flows/721c9bbc-6e56-4568-a45e-82e2a3a213f1 - **Google Calendar, iOS, "Creating an event"** (13 screens). The full event form with guest permission switches (modify event, invite others, see guest list), video call, notifications, colour, description, attachment, visibility and busy or free, then the event on the day view. OBSERVATION, 2026-10-31. https://mobbin.com/flows/3cf4f9cb-56db-467c-8db6-121d0ae9b561 ###### Documents - **Notion, web, "Adding a text block with AI"** (9 screens). A page marked "Verified" with a comment asking a named colleague to refine it. The AI inserts a highlighted paragraph and says "I've updated the page as requested", with thumbs, Chat, undo and accept. The sidebar has an "Agents (Beta)" section with "New agent" and a page of AI instructions. The block menu offers "Suggest edits" and "Ask AI" and shows who last edited the block and when. OBSERVATION, 2026-10-31. https://mobbin.com/flows/6a605e50-2f18-4da3-b8f7-3f17c45515d0 - **Notion, web, "Creating a body with AI"** (8 screens). An empty page ("press space for AI"), a "Thinking" state with a stop button, and a full draft written into the page. OBSERVATION, 2026-10-31. https://mobbin.com/flows/872ddb93-ad98-4321-92fc-6eff38489d72 - **Notion, web, "Chatting with AI (page)"** (5 screens). A side chat with suggestions (create a custom agent, make a diagram, summarise the page or its comments). The selected paragraph is attached as context. The answer explains "where it could be sharper" and gives a "Suggested revision" that is not applied to the page until the person chooses. The composer shows the mode ("Ask") and the model in use. OBSERVATION, 2026-10-31. https://mobbin.com/flows/938a3741-4b1e-403d-9872-30b4b6a40ac8 ###### Tasks and issues - **Linear, web, "Enabling an agent"** (3 screens). Settings for AI and agents: a usage feedback switch that says it is "never used to train models"; the built-in agent and "agent automations" (run when issues arrive in triage) enabled; "Installed agents: AI agents can work alongside you as teammates". Triage suggestions can be set per kind (assignee, project, label, team, duplicate, related) to show, hide or apply automatically, with optional "workspace guidance" for the agent. OBSERVATION, 2026-10-31. https://mobbin.com/flows/d67e265d-a278-47af-bb77-d7f57229b346 - **Linear, web, "Creating a new issue"** (10 screens). A quick form (title, description, status, priority, assignee, label, due date, link, sub-issue), "Create more", then the issue page with attachment, sub-issues, links, an activity line ("created the issue 6 minutes ago") and properties. OBSERVATION, 2026-10-31. https://mobbin.com/flows/3528703d-a543-414c-8f6e-ccf2442bff2f - **Linear, web, "Adding an assignee"** (3 screens). "Assign to..." with "No assignee" and the people in the team. OBSERVATION, 2026-10-31. https://mobbin.com/flows/0035680b-a877-444a-b27a-b76b38d10853 - **Asana, web, "Creating a task"** (14 screens). My tasks by upcoming, overdue and completed; a new task panel "For [person] in [project]" with priority, status and description; "Added [task]". OBSERVATION, 2026-10-31. https://mobbin.com/flows/e9894023-9b31-4ca8-a883-8c304403757c - **Asana, web, "Creating a task"** (14 screens). A board (to do, doing, done) with priority and status labels ("On track", "At risk", "Off track") and an empty avatar for unassigned work. OBSERVATION, 2026-10-31. https://mobbin.com/flows/5231873f-07ad-48d1-ae24-ebcad2aa773f ###### Team chat - **Slack, web, "Sending a message"** (5 screens). A channel welcome with its purpose, "Add coworkers", system lines in the stream ("joined", "renamed the channel from ... to ..."), and a formatting composer with mentions. OBSERVATION, 2026-10-31. https://mobbin.com/flows/cf31520c-e857-4b12-83a9-4505d2b6a8d6 - **Slack, web, "Replying a message in thread"** (5 screens). A message with a document preview, React and Reply on hover, a thread panel with "Also send to #channel", and "1 reply" under the original. OBSERVATION, 2026-10-31. https://mobbin.com/flows/afc25901-eb6f-4326-ac1f-ea34530cfdb4 ###### AI assistants and answers - **ChatGPT, web, "Connecting to Figma"** (5 screens). An apps directory, then a connect dialog with "Connect" or "Continue without account", a switch to let the app reference memories and chats (off), and three plain notices: you're in control and it is limited to permissions you've set; "apps may introduce elevated risk... attackers may attempt to use ChatGPT to access your data in the app"; and exactly what data is shared with the app. Then "Figma is now connected" and an app page listing its capabilities (interactive, writes). OBSERVATION, 2026-10-31. https://mobbin.com/flows/56a10553-7cd7-41e5-96fb-66fbec13765b - **ChatGPT, web, "Connecting to GitHub"** (5 screens). "What should we code next?" with Connect to GitHub; GitHub's install page asks for all repositories or only selected ones and lists the permissions (read some things, read and write code, issues, pull requests and workflows) and where it will redirect. Then a task box with repository, branch and number of attempts, starter tasks, and tabs for tasks, code reviews and archive. OBSERVATION, 2026-10-31. https://mobbin.com/flows/72826030-df1b-43b4-884e-cafad0cc2e06 - **ChatGPT, web, "Sharing GPT"** (7 screens). Building a custom assistant by chatting or by configuring instructions, starters, knowledge files (with a warning that answers may include file contents), capabilities (web browsing, image making, code running) and actions, with a live preview; then sharing with "Anyone in the workspace with the link: can chat". OBSERVATION, 2026-10-31. https://mobbin.com/flows/58cd9e51-3af8-48cf-854f-5c37c9fdd37b - **Claude, web, "Uploading an image"** (4 screens). A composer with attach, tools, a Research switch and a model picker; an image attached; the reply builds an interactive piece in a side panel with preview and code views, Copy and Publish. OBSERVATION, 2026-10-31. https://mobbin.com/flows/113a72d6-1b6d-482a-9b71-4f7becd2d986 - **Claude, web, "Uploading a file to chat"** (6 screens). A long answer with Copy, Retry and thumbs, the line "Claude can make mistakes. Please double-check responses", a document attached ("1 file added"), and a structured summary of it. OBSERVATION, 2026-10-31. https://mobbin.com/flows/c3c3fe5e-482b-4c51-8b9d-a055b24f6d53 - **Perplexity, web, "View sources"** (2 screens). Small source labels after each claim, and "20 sources" opening a panel of every source with its title and snippet; related follow-up questions. OBSERVATION, 2026-10-31. https://mobbin.com/flows/abcb8608-801d-4116-a9d4-a242200c10b1 - **Perplexity, web, "Checking sources"** (3 screens). "2 steps completed" can be opened. Selecting one sentence offers "Check sources"; a panel shows "Checking..." and then the exact passages from each source that support that sentence, highlighted. OBSERVATION, 2026-10-31. https://mobbin.com/flows/8684b407-d701-4f86-be4a-2c6194217572 ###### Building agents and automations (the agent lab) - **StackAI, web, "Creating a workflow agent"** (7 screens). Templates by use case (including healthcare, education, legal and an "InfoSec questionnaire agent" and a "public policy memo generator") and by integration. The canvas joins a user message, an AI assistant (model, knowledge sources, tools) and an output "what the user sees"; each part shows its results, tokens and time. The project is a "Draft", "auto-saved 1 second ago", with Run and Publish and tabs for interface, analytics, manager and evaluator. OBSERVATION, 2026-10-31. https://mobbin.com/flows/fe260cca-1084-4e6e-884a-8ea843c1f1e9 - **Clay, web, "Building an agent"** (9 screens). "Start building in natural language" with a model choice and file upload, or a template. The editor has an AI helper that lists what it changed, a prompt with named inputs, a configuration of model, typed inputs and outputs (fields or a schema), a version number, Test and Save. A table lists agents with version, tools and owner. OBSERVATION, 2026-10-31. https://mobbin.com/flows/9700a29d-77f6-477f-a1e1-dee35d99d421 - **Twenty, web, "Creating a workflow"** (7 screens). A table of workflows with status, "created by System", version and runs. A draft canvas: when a record is created, update it, then branch to one of two "AI Agent" steps; Activate, Discard draft and See runs, with separate lists of runs and versions. OBSERVATION, 2026-10-31. https://mobbin.com/flows/1c1807d4-eed9-4a19-a248-c5413983925b ###### AI agents asking for permission - **Quicken, web, screen.** "AI Agent for support purposes": by agreeing you allow an AI agent to view your screen; everything shown will be visible to it; the session will be recorded for quality, training and diagnostics; Cancel or I agree. OBSERVATION, 2026-10-31. https://mobbin.com/screens/c5e6870f-21ee-458f-82a9-fb49ef245651 - **Notion, web, screen.** Slack's page "Allow the Notion AI app to access Slack": "App is approved by Slack", a workspace picker, and permissions split into "Information it can view" and "Actions it can take". OBSERVATION, 2026-10-31. https://mobbin.com/screens/91532c16-c4c5-44e9-84d3-42a06ce34f47 - **Fireflies, web, screen.** The same Slack pattern plus a required channel where the app will post. OBSERVATION, 2026-10-31. https://mobbin.com/screens/a4babe75-26e3-4cc2-acbf-c0bcf7a1c6f8 - **Devin, web, screen.** GitHub's "app is requesting updated permissions" showing each change against the old one ("read and write, was read-only"), a developer note saying why, "Accept new permissions", and that the app keeps its current permissions if you decline. OBSERVATION, 2026-10-31. https://mobbin.com/screens/7b19cb77-7bc3-4e96-9f3a-9ea0b488211d - **Emergent, web, screen.** A building agent pauses: "Agent has questions for you: waiting for answers", a multiple-choice question "1 of 4" with "Something else", and buttons "Auto-answer" and "Next". OBSERVATION, 2026-10-31. https://mobbin.com/screens/fd030823-0cd0-498d-b11c-63e8cd92d5bb - **ChatGPT, web, screen.** GitHub's install and authorise page for the coding connector (the same page as in "Connecting to GitHub" above). OBSERVATION, 2026-10-31. https://mobbin.com/screens/c0f8ad37-8b72-4d0b-9b6a-b4f7717ca5a7 ###### Code review - **GitHub, web, "Pull request detail"** (6 screens). A merged pull request with tabs for conversation, commits, checks and files changed, lines added and removed, a diff, and "0 of 2 files viewed". OBSERVATION, 2026-10-31. https://mobbin.com/flows/8498f34f-54fc-416a-8b10-5a4d6ffdb453 - **GitHub, web, "Submitting a review"** (5 screens). A "Viewed" tick per file, "Finish your review" with "1 pending comment", a line comment, "This comment was marked as outdated", "Resolve conversation", and "Pull request successfully merged and closed: the branch can be safely deleted". OBSERVATION, 2026-10-31. https://mobbin.com/flows/6668a854-4523-4f92-97b7-fc56dc5234b8 ###### Music - **Spotify, iOS, "Creating a new playlist"** (4 screens). The create menu offers a playlist, a "Blend" (combine tastes in a shared playlist with friends) and an "AI Playlist" in beta (turn your ideas into playlists using AI); then "Let's start building your playlist" with songs "based on your listening". OBSERVATION, 2026-10-31. https://mobbin.com/flows/46f98a82-6967-4380-875b-2b2857acfb42 - **Spotify, iOS, "Creating a playlist"** (5 screens). Home with stations and mixes, naming the playlist (default "My playlist #1"), and an empty playlist with "Add to this playlist" and recommended songs. OBSERVATION, 2026-10-31. https://mobbin.com/flows/30f0589d-5ace-4bba-a135-6201b69a3dde - **SoundCloud, web, "Uploading a track"** (9 screens). An artist studio with plays, reposts, downloads, likes and comments, and tabs for distribution and earnings; drag-and-drop upload; track details (title, link, artists, genre, artwork); "Processing 100%"; the line "By uploading, you confirm that your sounds comply with our Terms of Use and you don't infringe anyone else's rights"; "Saved to SoundCloud", then an offer to distribute to other services. OBSERVATION, 2026-10-31. https://mobbin.com/flows/f5627b1d-fed2-4c2d-91b3-28a74196f1ed - **SoundCloud, web, "Uploading a track (recording)"** (7 screens). An uploads-used meter, file format guidance with the size limit, and recording from the microphone with undo, delete and a timer. OBSERVATION, 2026-10-31. https://mobbin.com/flows/990eb0c6-906a-4d7e-b6d2-cfe72e04dcd1 - **Record Club, web, "Buy release"** (2 screens). A record review page with listeners, rating and popularity; a menu to rate, review, add to lists, buy or stream; and a buy dialog linking to Bandcamp and other shops with "we may earn an affiliate commission". OBSERVATION, 2026-10-31. https://mobbin.com/flows/e0c631d2-76b6-4d93-a294-19a7965c873b ##### What to match - **AI output is a proposal until a person accepts it** (Superhuman's Accept or Retry inside the draft; Notion's accept tick and "Suggested revision" that is not applied by itself; Linear's show, hide or auto-apply per kind of suggestion). - **Separate switches for each AI behaviour** (Superhuman settings; Linear agents, automations and triage), and a plain statement about training on your data (Linear "never used to train models"). - **Permission screens that split "what it can see" from "what it can do"** (Slack for Notion AI and Fireflies), let you choose scope (GitHub: all or selected repositories), show each change against the old permission and keep the old one if you decline (Devin on GitHub), and warn plainly about risk (ChatGPT apps). - **Answers with checkable sources down to the sentence** (Perplexity "Check sources"), and an honest "can make mistakes" line (Claude). - **Agents that ask instead of guessing** (Emergent "Agent has questions for you", with the person able to let it auto-answer). - **Agent builders with drafts, versions, test runs, run history and an evaluator** (StackAI, Clay, Twenty). - **Acting on someone's behalf is visible** (Calendly "Answer questions on behalf of your invitee"; Google Calendar's guest permissions). - **Undo after sending** (Gmail, Superhuman), drafts shown in place (Outlook), and review state per file (GitHub). - **Artists see their own numbers and confirm the rights to what they upload** (SoundCloud), and shops say when they earn commission (Record Club). ##### What none of them do - **No AI agent is a user with its own identity.** Agents appear as features (Superhuman AI, Notion AI), as apps installed into someone else's workspace (Slack, GitHub), or as "teammates" in Linear's settings, but none has its own account, mailbox, calendar or profile that says who made it, who it acts for, what it may do, and what it has done. Twenty labels workflows "created by System", which hides the actor. - **No single record of what an agent did across apps.** Each product keeps its own activity line. Nothing joins an agent's mail, calendar, document, task and code actions into one account the person can review or reverse. - **No limits on spend, time or reach** for an agent were shown, and no way to pause or revoke one agent across every connected app at once. - **No vetting of agents before they join.** Slack marks an app "approved by Slack" and ChatGPT lists app capabilities, but there is no check of an agent's behaviour, tests or evaluation results before it is allowed in; StackAI's evaluator is for the builder, not the people the agent will act for. - **No agent-to-agent etiquette.** Nothing shows one person's agent booking time with another person's agent, or how either side knows it is talking to an agent. - **No music economy for independent artists inside the player.** Spotify's AI Playlist and Blend are about listening; SoundCloud's earnings and Record Club's links send people elsewhere. Nothing seen combines listening, paying the artist directly and transparent credits in one place. - **Not covered here:** Microsoft Teams, Gemini, n8n, Zapier, Make and Bandcamp's own pages, which were not searched or not returned in this pass. ### research/judiverse/REFERENCE-DOCS.md `research/judiverse/REFERENCE-DOCS.md` · sha256 `6d874e29394545c6205f5577c9ee75e563755fceee69b850db738ab672a018d3` #### Judiverse Documents: reference and what was added Statement class: OBSERVED (Mobbin, read 1 November 2026 06:30Z). Code: runtime/apps/docs.py, styles and script in runtime/docs_assets.py (script served at /judiverse/docs/app.js; the page CSP allows no inline script). Built by Claude-Jegudiel 00:5531102026-1. ##### The reference Google Drive for iOS, flow "Home", Mobbin flow adb5ef66-1350-41fa-b215-fd421c085f4b (screens af861121-7236-4a46-88bf-e0dd364f7abe, 09e8b2a8-8e73-4a3e-be73-c0474e03c943, a0e4b7f5-b9e8-4719-8d76-a53b88cc46ff), and the "+" sheet of flow "Creating a folder" c1b9934d-4fa5-4681-ab74-b61a12e287d5 (screen e32039c8-b7e3-4a62-8a41-c706657c4f91). Google Docs for iOS itself is not in Mobbin's iOS results (searched "Google Docs home recent documents", 1 Nov 2026 06:28Z: only the Google app came back), so the Drive home, which lists Docs, Sheets and Slides files with a kind icon, owner line and last-opened time, is the named reference. | Element in the reference (Drive Home) | In Judiverse Documents (/judiverse/docs) | |---|---| | Search bar: menu icon · "Search in Drive" · account chip | Search bar: back-to-Suite chevron · "Search in Documents" (server ?q= and live filter) · account chip with the person's initial | | Tabs: Suggested · Activity, underlined current | Tabs: Suggested (by last opened) · Activity (by last change, with "You created / edited / renamed · time") | | Section head "Files" with a grid toggle | Section head "Files" with a grid toggle (list ↔ tiles, remembered per browser) | | Rows: kind icon (Docs blue, Sheets green, Slides yellow) · name · "You opened · 5:21 AM" · ⋯ | Rows: kind icon (document, spreadsheet, presentation in estate colours) · name · "You opened · 06:31" · ⋯ (Rename, Download, Delete); edge to edge, hairline separators, no cards | | Floating "+" button (and a scan button above it) | Floating "+" button; no scan button (the estate has no scanner capability here) | | "+" sheet: Upload file · Scan document · Take photo · Create folder · Google Docs · Google Sheets · Google Slides | "+" sheet: Upload file · Document · Spreadsheet · Presentation | | Empty state: illustration, "A place for all of your files", "Tap the + button to upload a file or create something new" | The same two lines, without the illustration | | Bottom tabs: Home · Starred · Shared · Files | Bottom tabs: Home · Activity · Suite · Jude (phone only; the estate header stays above) | ##### The editor screen The ONLYOFFICE editor itself (ONLYOFFICE Docs 10.0 on jude-1, jude-engine-office.service, reached by the browser at the same-origin path /office/*), full bleed under the estate header, with one slim bar the estate adds: back chevron · title (tap to rename) · state (Loading… / Ready / Editing… / Saved) · Download. Phones get the same desktop editor: with config type "mobile" this host's ONLYOFFICE Docs Community edition opens documents for viewing only ("To access mobile web editors, a commercial license is required", seen live 1 Nov 2026 06:41Z on the demo spreadsheet), so the editable desktop editor is served to every device. Saves arrive through the server's callback (status 2 on close, status 6 on the editor's Save button); each save rotates the document key so the next open reads the saved file. ##### Not in the reference (the estate's additions) The Demo tag on every row of the demo tenant (three seeded files: a document, a spreadsheet and a presentation, full editing); Download in the ⋯ menu and in the editor bar; the estate header, demo bar and sign-in; the visitor page with "Continue with the demo". ##### Host changes Caddy srv0 route 0 on jude-1: /office/* on judiverse.tech, www.judiverse.tech, judeai.tech, www.judeai.tech and x-x.tech proxies to 127.0.0.1:18901 (prefix stripped, X-Forwarded-Host {host}/office, X-Forwarded-Proto https). Live through the admin API and recorded in /opt/x-x/deploy/ledger/adoption/changes.jsonl; not persisted into /var/lib/estate-ui-admission/startup.json because that file carries the immutable attribute, so a Caddy restart drops the route until its owner persists it. The Document Server's blank templates were copied from the container to the release's .cache/docs/templates/. The JWT secret stays in /etc/jude-reusable-engines/office.env (root-only; x-x-web.service runs as root and reads it at request time). ##### Not built here Starred, Shared with me, folders, scanning and photo capture (reference elements with no estate capability behind them yet); co-editing invitations (the server supports co-editing, the app has no share control yet); search inside document text (the search matches titles). ### research/judiverse/REFERENCE-JUDE.md `research/judiverse/REFERENCE-JUDE.md` · sha256 `730b78aa1d9adedaf2386acadb80cc1beb4abb69a6ed90f439ef1ef82928a50a` #### Jude (chat): reference and what was added Statement class: OBSERVED (Mobbin, read 1 November 2026 04:40Z; live /jude read at 04:41Z). Code: runtime/apps/jude.py, runtime/jude_assets.py. ##### The reference ChatGPT for iOS, Mobbin flows "Chat" 8ebab232-a85f-43f9-a92e-c486202e734c, "Chatting with ChatGPT" 86f171f6-2b2c-498b-a76d-fd21dd493a10 (screens f7e6514e…, 5966f750…, 5529fa8c…, aa9d3ba4…, 1778fff4…, 6ebc53fd…, 15055b68…), "Share a chat link" 847cdf83-602f-442a-bfa7-16e4285e340e, "Sharing a chat" c10d00f2-0cf7-44e1-b6c0-cb5826f8e227, "Chats" 0f538b27-9afb-4f7f-8c52-d10f51e3218e. | Element in the reference | In Jude | |---|---| | Top bar: menu icon · centred title or plan pill · compose icon | Menu · "New chat" title · Normal/Eco pills (added, the estate's mode) · Share | | Empty state: the mark centred, then two suggestion chips side by side above the composer | Three-dot mark, "What can I do for you?", starter chips naming estate journeys (Send money, Book an appointment, Open an account, Ask about my statement) | | Composer: attach (camera, photo, file), "Message" field, microphone, voice-mode button | Attach · "Message Jude" · microphone (transcription plug) · send; read-aloud on replies (speech plug) | | Reply actions: copy · speak · thumbs up · thumbs down · share · more | Copy · read aloud · thumbs · regenerate | | Thread menu: Share, Add people, Add to project, Rename, Archive, Delete | Share link page (read-only for a visitor), Rename, Delete | | Share sheet: preview card, "Share link" button | /jude//share → /jude/share/ | | Conversations rail (menu): search, list by title | Rail with titles; search in the next pass | | Streaming reply with Markdown | SSE token stream, server-rendered Markdown, stop | ##### Added by the estate (not in the reference) Normal/Eco mode; "Jude did" collapsible tool steps (operations through the hub with the hub's refusal shown verbatim); the demo bar and sign-in; the estate shell (header, applications rail). ##### Known differences to close Starter chips show an estate path as the second line (reference: a plain second line); "Add to project" and "Archive" absent; conversation search absent; the model is relaxAI (RELAXAI_* in /etc/x-x/web.env), latency 4–32 s on first reply (measured by lane A, 1 Nov 04:2xZ). ### research/judiverse/REFERENCE-JUDEBOX.md `research/judiverse/REFERENCE-JUDEBOX.md` · sha256 `dd4ac3cef39706766324120723fcfeecb3644370dcc93bf5ca463c742e31b094` #### JudeBox: one named reference per layout Owner order, 1 November 2026 04:14Z: every layout maps to one named reference; the only differences are added controls or journeys. Written by ATBAN Claude-Jegudiel 00:5531102026-1 after reading the Spotify iOS flows on Mobbin (search_flows, platform ios, 1 November 2026 04:40Z) and the live screens at https://judiverse.tech/judebox (runtime/apps/judebox.py, styles runtime/judebox_assets.py). Observations only; a row says MATCH when /judebox renders the same element, ADDED when /judebox has a control or journey the reference does not, and NOT RENDERED when the reference shows something /judebox does not. ##### References (Mobbin, Spotify, iOS) | Name | Mobbin flow | Screens read | |---|---|---| | Home | https://mobbin.com/flows/f3592f63-a0de-4a03-aa93-c0c7e1e2e5f2 (12 screens) and https://mobbin.com/flows/d7f1ea2a-64e7-4591-a195-25fbdc06ac8a (4) | avatar bubble top left, filter chips All / Music / Podcasts / Audiobooks / Wrapped, "Start listening" rows (thumbnail, title, subtitle, overflow dots), horizontal card shelves, mini player above the tab bar, four tabs Home / Search / Your Library / Create | | Now Playing | https://mobbin.com/flows/110c08ea-444c-48c8-9c0a-9ef6aa2389f5 "Listening to a song" (9) | chevron-down to dismiss, context title centred, overflow dots, full-bleed art, title and artist with a remove and an add control, progress bar with elapsed and remaining, shuffle / previous / large pause / next / repeat, device, share, queue; a Lyrics sheet peeking from the bottom | | Library | https://mobbin.com/flows/cab27a0b-3ba3-4931-9abc-1d38852e3acc "Your Library" (5) and https://mobbin.com/flows/f2c42f35-f4eb-4118-9da2-7219424796ea (2) | avatar + "Your Library" title, search and plus icons on the right, filter chips (Playlists / Podcasts / Artists / Albums), "Recents" sort row with a grid toggle, rows of 56px art (round for artists, square for shows) with title and kind, "Add artists" / "Add podcasts & shows" / "Import your music" rows | | Search | https://mobbin.com/flows/fbf3c945-13b0-43cf-b790-85a8ba772ed4 "Search" (5) | avatar + "Search" title, camera icon, full-width field "What do you want to listen to?", "Start browsing" tiles, "Browse all" two-column colour tiles; results screen with the query in the field, Cancel, type chips (Artists / Playlists / Songs / Albums), result rows with round art, "Following" pill and plus buttons | | Artist | https://mobbin.com/flows/0204eb43-fca3-4039-8d3a-d4ba391f57e1 "View artist profile" (8) and https://mobbin.com/flows/c18c5b89-24a8-472c-95ae-1188438e603a "View about the artist" (3) | back chevron, name centred, tabs Music / Clips / Events / Merch, "Popular releases" rows (art, title, year · kind), "See discography" pill, "Featuring" shelf, "Fans also like" round avatars, "Appears on"; about view with the image, monthly listeners figure, rank badge, biography, "Posted by", social links; the sticky mini player (title, artist, plus, pause) | ##### Element by element ###### /judebox (Inbox) maps to **Library** (list) with the **Home** header | Reference element | /judebox | Status | |---|---|---| | Avatar bubble top left + title | `.jb-av.you` initials + "JudeBox" h1 | MATCH | | Plus icon top right | `.jb-ic` Plus linking to Contacts (add a contact or start a conversation) | MATCH | | Search icon top right | search is the full-width field below the chips (Search reference) rather than an icon | ADDED (field always visible) | | Filter chips row | All / Messages / Email / Texts (`?f=`) | MATCH | | "Recents" sort row | `.jb-eyebrow` Recents with the sort arrow; no grid toggle | MATCH (grid toggle NOT RENDERED) | | Row: 56px art, title, kind subtitle | `.jb-row`: 44px initials avatar (round for people, rounded square for agents and organisations), title, last message preview | MATCH | | Row trailing: nothing (Library) / overflow dots (Home) | time, channel or delivery-state chip, unread badge | ADDED (unread count, delivery state) | | "Add artists" / "Import" rows | "New conversation" and the Numbers / Webhooks / Memory links under the list | ADDED (journeys) | | Mini player above the tab bar | none on the Inbox; the composer plays that position in a thread | NOT RENDERED here | | Four tabs Home / Search / Your Library / Create | Inbox / Contacts / Calls / Agents, fixed bottom, 68px, icon over label, current in teal | MATCH (labels differ by product) | | Empty state | "No conversations yet. Start one below, or add a contact first." with icon | ADDED | | Demo entry for an unsigned visitor | "Continue with the demo" card posting to /signin/demo | ADDED | ###### /judebox/thread/ (Conversation) maps to **Now Playing** | Reference element | /judebox | Status | |---|---|---| | Dismiss chevron top left | back chevron to /judebox | MATCH | | Context title centred + subtitle | conversation title + provider and room id (or "native conversation") | MATCH | | Overflow dots top right | Call button (`.jb-ic` Bell01) | ADDED (call journey) in that position | | Full-bleed art | message stream (`.jb-thread`: bubbles, own messages right in teal, others left in white, system lines centred) | NOT RENDERED (a conversation has no art); the stream takes its place | | Title row with remove and add | members chip row (name · ATBAN, revoked greyed) with Invite and Revoke in the Members card | ADDED (membership controls) | | Progress bar elapsed / remaining | per-message time and delivery-state chip (Sent / Delivered / Simulated / Pending / Unknown) | ADDED (delivery state) | | Transport row (shuffle / prev / pause / next / repeat) | composer `.jb-composer`: text area, Send, Attach, Voice note | MATCH in position (sticky above the tab bar); controls differ by product | | Device / share / queue icons | Attach and Voice note inputs; Help disclosure | ADDED | | Lyrics sheet peeking at the bottom | Members and Help disclosures below the composer | MATCH (collapsed sheet pattern) | | Mini player persisting elsewhere | none | NOT RENDERED | ###### /judebox/contacts (Contacts) maps to **Library** (Artists filter) | Reference element | /judebox | Status | |---|---|---| | Title + search + plus | "Contacts" title; add through the "Add a contact" disclosure | MATCH (plus is a disclosure) | | Filter chips | none | NOT RENDERED | | Rows: round art for artists | round initials for people, rounded square for agents and organisations, kind chip trailing | MATCH | | Row subtitle "Artist" | first two addresses, or for an agent its ATBAN · accountable owner · state | ADDED (ATBAN and mandate shown) | | Artist detail (Artist reference) | /judebox/contacts/: back chevron, name centred, kind subtitle, Agent identity card (ATBAN, accountable, mandate, state), Addresses table with per-address inbound / outbound / consent, Message button | ADDED (permissions per address) | ###### /judebox/calls (Calls) maps to **Library** (list) and /judebox/calls/ to **Now Playing** (state card) | Reference element | /judebox | Status | |---|---|---| | Rows with art, title, kind | direction arrow in the avatar, who, engine · reason, time and state chip | MATCH | | Now Playing title + controls | call card: State chip, With, Engine, Room, Started, Ended, Note; join-token line; End call button | MATCH in structure; media controls (mute, speaker) NOT RENDERED because browser media needs the public signalling route | ###### /judebox/agents (Agents) maps to **Library**; /judebox/agents/ maps to **Artist** | Reference element | /judebox | Status | |---|---|---| | Library rows | identity rows with kind and state chips, ATBAN · accountable owner · channel count | MATCH | | Artist header (back, name centred) | back chevron, agent name, ATBAN subtitle | MATCH | | Monthly listeners figure + rank badge | identity card: TOK id, EDGE principal, ATBAN, Born (time, tz, source, precision), Registered, Accountable, Mandate, State | ADDED (identity facts) | | Tabs Music / Clips / Events / Merch | stacked cards: Aliases and registration, Channels, Credentials (references only), Tasks, Act | ADDED (sections instead of tabs) | | "Popular releases" rows | Tasks table with from → to, state, history and actions (Set state, Transfer) | ADDED (A2A journey) | | "Following" pill | Suspend / Reinstate button | ADDED (lifecycle) | | About the artist biography and links | Aliases timeline with kind and source | ADDED | ###### /judebox/numbers, /judebox/webhooks, /judebox/memory map to **Library** (list with add rows) Each is a titled list with its "add" form in the position of Spotify's "Add artists" row: Give communications / Register a webhook / Remember and Seal. Tables carry the provider's actual state chip (Active, Pending provider, Simulated, Revoked). ADDED throughout: provider facts "as observed from this runtime". ##### What is not rendered anywhere in /judebox and why - Spotify's full-bleed artwork and colour-extracted backgrounds: a conversation has no artwork; the estate tokens (navy, teal, pale surfaces) apply. - The mini player: JudeBox has no persistent playing object; a live call would be the equivalent and is NOT BUILT as a persistent bar. - Grid toggle on Recents, horizontal card shelves, "Browse all" colour tiles: not needed by any JudeBox journey; not added. - Media controls on a call (mute, speaker, device): the LiveKit room and join token exist; browser media needs public signalling and TLS, which are not configured (handoff section 3). ##### Inkbox parity points added on top of the reference Per screen, the controls that exist because of the Inkbox parity rows (research/judiverse/JUDEBOX-PARITY.md): delivery state per message (P03, P26), voice-note transcript (P10, P11), members with sponsor and revoke (P15), per-address permissions and consent (P16), agent identity with birth, aliases, credential references and channels (P01, P20), tasks with transfer history (P14), provisioning in one action with per-channel provider state (P07, P02), webhooks with digest, signature and replay (P21), memory with sources and corrections (P17), vault with generations (P19). ### research/judiverse/REFERENCE-SUITE-HOME.md `research/judiverse/REFERENCE-SUITE-HOME.md` · sha256 `3ae10ec03eb82932fcfbb2817066a3e2178e5ac212474bf6afd131ac49f6921b` #### Judiverse Suite home: reference and what was added Statement class: OBSERVED (Mobbin, read 1 November 2026 04:13Z). Code: runtime/apps/suite.py, styles `.nh-*` in runtime/web_assets.py. ##### The reference Notion for iOS, flow "Home", Mobbin flow 18febc82-c874-4d48-bafa-3eecaf3e3b51 (screens 66b134a1…, 0ea5a26b…, a8b2557e…), and "Switching to a workspace" 4dc02d6b-4c78-4bf4-b168-92b1fa43b10e. | Element in the reference | In the Suite home | |---|---| | Account chip "A" then pills Home · AI chat · Calendar · Mail | Account chip, then pills Home · Mail · Jude · Calendar, then the remaining apps scrolling (added) | | Announcement card: title, two lines, one full-width button ("Create new meeting note") | "Judiverse Mail is open", one button Open Mail → https://mail.judiverse.tech/inbox (the real app) | | "Jump back in" horizontal cards | The person's recent Suite journeys as cards (only when there are any) | | "Private ▾" list: leading icon, page name, trailing + | "Apps" list (12 rows) and "Everything you can do here" (the 13 Suite capabilities); trailing chevron | | Bottom bar: search circle · "Ask AI" pill with microphone · compose circle | Search · "Ask Jude" pill (submits to /jude, or the Jude capability until /jude is live) · Compose (opens Mail) | ##### Not in the reference (the estate's additions) More app pills; the real Mail and Calls behind their rows (frame-ancestors on mail.judiverse.tech allows only itself and ai.judiverse.tech, so they open as their own apps); the capability list; the applications rail of the estate shell, which lists the same apps under Judiverse Suite. ##### Other references read for this surface (not used, kept for the per-app screens) Slack Workspaces 7226c577…; monday.com Workspaces 37fe304b…; Fabric Workspace 87784607…; Uber "Services" grid screen 0f25c26a…; SmartThings "Apps" grid ff9ab983…; Microsoft 365 home was not in Mobbin's results for "Microsoft 365 home apps grid". ### research/judiverse/REFERO-FLOWS.md `research/judiverse/REFERO-FLOWS.md` · sha256 `dc6bc7badb62e18e0f292490b8df3a3b4d92bed4d17fea5683f9b47a05a0413b` #### Judiverse research: real product flows from Refero **What this is.** Real screens and user flows from existing products, recorded by Refero (a library of real app and website screenshots, each journey described step by step). They show the floor (what users already expect) and good customer experience from any industry worth borrowing. **Statement class.** Every flow is an OBSERVATION by Refero of a real product, searched 2026-10-31. Descriptions are Refero's wording. Product names belong to their owners. **How it was gathered.** Semantic searches for this application's tasks on iOS and web. Results from other industries are kept on purpose: they are often the best customer experience to learn from. Refero does not hold many incumbents (and few government, enterprise or NHS systems), so this is a sample, not a census. **Products in this sample:** WhatsApp (25), Dropbox (23), ElevenMusic (16), Voicenotes (16), YouTube Music (16), Meta AI (15), Linear (15), TIDAL (12), n8n (11), Manus (11), Calendly (10), Missive (10), Microsoft Teams (10), Structured (9), Cushion (8), Skiff (8), SoundCloud (8), Wabi (7), Granola (7), Cohere (7), Raycast (7), Fernand (7), Marathon (7), District (6), Genie (6), Willow (6), Programa (6), Exa (6), Factory (6), Jace AI (6), SavvyCal (6), Wrike (6), Hello Ivy (6), Runey (6), Epidemic Sound (6), Lovable (5), Poppy (5), komoot (5), Pinterest (5), Fresha (5), Cursor (5), Rork (5), Wispr Flow (5), XChat (5), Lemni (5), Daylish (5), Grok (5), MD Vinyl (5), Shuttle (5), ElevenLabs (5), ChatGPT (5), Navia (5), PayPal (4), Gizmo (4), Airbnb (4), Turf (4), Google Gemini (4), Verse (4), MyMind (4), Make (4), HYPE (4), Exoplan (4), Homerun (4), Maze (4), Canva (4), Deezer (4), Comet (3), Tidal (3), Linktree (3), Givingli (3), Kin (3), Tripsy (3), TikTok (3), Cal (3), Substack (3), Airtable (3), Cofounder (3), Mozi (3), Appwrite (3), OKURI (3), Rox (3), Gladia (3), Aboard (3), Tinder (3), Instagram (3), CocoonWeaver (3), Memotron (3), Slite (3), Clipchamp (3), Play (3), Synthesia.io (3), ManyChat (3), Parallel (3), Spotify (3), [untitled] (3), Fuse (2), Introspect (2), OpenAI (2), Apple Image Playground (2), Ground News (2), Daze (2), Acctual (2), Sora (2), Reown (2), Doctronic AI (2), Longevity Deck (2), Leonardo.Ai (2), RAD Weather (2), 222 (2), Phantom (2), Podimo (2), ElevenReader (2), Kickstarter (2), Bond (2), pillowtalk (2), Netflix (2), HBO Max (2), Matter (2), Mela (2), Joi (2), Reclaim AI (2), Around (2), Craft (2), Jasper AI (2), Framer (2), Moises (2), Playground (2), Vimeo (2), Anam (2), LEGO® Builder (2), BeReal. (2), Visitors (2), fal (2), MWM (2), Melius (2), Tilt (2), Carbon (2), Wise (2), Volv (2), Ada (1), Instacart (1), Spoil me (1), GlossGenius (1), Product Hunt (1), TidyCal (1), West Elm (1), Blackbird (1), Typeform (1), Brink (1), Adobe Photoshop (1), Loowner (1), Pool (1), Miro (1), Nooka (1), WishWish (1), Square (1), pushr (1), Imprint (1), Understory (1), Abode (1), Claude (1), Descript (1), Handshake (1), Figma (1), Trip Way (1), Microsoft Copilot (1), Bezi (1), PocketShelf (1), UNHINGED (1), Binsoo (1), Moonly (1), Cake Equity (1), Family (1), Sunsama (1), Weekrise (1), Riverside (1), Dropset (1), Tripadvisor (1), Mocha (1), PamPam (1), Polar (1), Pi (1), Allbirds (1), WURRD (1), Chance AI (1), Fourmula.ai (1), Apple TV (1), Pinkllama (1), GO Club (1), Copperx (1), Box Box Club (1), Fingerprint (1), Steel (1), Bump (1), Resend (1), Chargetrip (1), The Leap (1), Clay (1), Stemz (1), Record Club (1), Savee (1), Artsy (1), Wittl (1), X (1), Nextdoor (1), Preply (1), SSENSE (1), Days (1), Womp (1), BYBORRE® (1), IKEA (1), Runway (1) ##### Task: AI agent setup permissions **ElevenMusic: Artist Identity Setup** (ios, 3 screens, AI Tool, Entertainment, Music Streaming) [Refero](https://refero.design/flows/12441) > Happy path for setting up an artist identity during onboarding. The user starts on a “Become an artist” screen where they can add a stage name and profile photo, or skip this step for later. Tapping the photo control opens a media picker with Photos/Collections tabs and a close action. After selecting an image, the user returns to the setup screen with the chosen photo applied. Steps: Artist Intro → Photo Picker → Artist Intro With Photo **Comet: Agent card gallery** (ios, 3 screens, AI Tool, Utility) [Refero](https://refero.design/flows/10948) > Happy-path summary: The user starts on a mobile "card gallery" that shows two active AI agent cards. From this gallery the user inspects either card, brings one into focus (front), optionally drags to reorder the two cards, and returns to a side-by-side overview. Steps: Card gallery overview → Bring card to front / partial overlap → Updated overview / reordered cards **OpenAI: Developer onboarding and API key setup** (web, 26 screens, Business, Development Tools, AI Tool) [Refero](https://refero.design/flows/4178) > Happy-path onboarding for a new developer on the OpenAI platform. Entry: public developer overview (not signed in). The user chooses to sign up and proceeds through a multi-step, linear onboarding wizard: provide email, set password, verify email, enter personal details, optionally invite team members, create a project and API key, review code snippets, and optionally buy credits. Exit: user completed onboarding and is redirected back to the developer docs/overview. Assumptions: user is not logged in, using a desktop browser, and has permission to create an organization/project and generate API keys. Steps: Entry — developer overview → Create account — email entry → Email entered → Email validation error → Email corrected → Create account — password entry → Password meets requirement example → Password masked → Verify email prompt → Tell us about you → Invite your team — add emails → Multiple team invites added → Invite error — duplicate or pending invite → Make your first API call — name key → API key form filled → API key revealed — curl example → API key code snippets — Node tab → API key code snippets — Python tab → Add API credits (optional) → Exit — return to developer overview/docs **n8n: AI Agent automation editor** (web, 7 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9520) > This is a visual, node-based workflow editor where a user composes an "AI Agent" automation on a canvas and uses an AI assistant in a right-side panel to generate or modify the workflow and documentation. Steps: Canvas overview → Open assistant templates → Enter custom prompt → Assistant begins generation → Assistant expands checklist → Review generated documentation → Finalize configuration and run **Pinterest: Collaborator management** (ios, 3 screens, Reference, Inspiration Board, Social Networking, Lifestyle) [Refero](https://refero.design/flows/2493) > Happy path: A board owner opens the Collaborators screen to view current collaborators and invited people, then adjusts collaborator permissions and invitation settings. Entry: user opens Collaborators from a board. Exit: permission level and invitation toggles are saved and the UI returns to the Collaborators list. Steps: Collaborators list → Permissions options (default) → Permissions — updated **Tripsy: Onboarding and First Launch** (ios, 8 screens, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/2326) > First-launch onboarding for a travel organizer (iOS). The user is introduced to the product via a multi-slide carousel that highlights key features (flight updates, collaboration, documents). During or immediately after onboarding the app requests runtime permissions (notifications and location). An upsell (Tripsy Pro) promotional modal can be shown before the user lands on the main 'My Trips' screen. Steps: Notification Permission Prompt → Welcome / Intro Slide → Feature Slide — Flight Details → Feature Slide — Collaboration → Feature Slide — Documents → Location Permission Prompt → Promotional Upsell (Tripsy Pro) → My Trips — Empty State Home **Dropbox: Folder access management** (web, 4 screens, Collaboration Tools, Business, Productivity) [Refero](https://refero.design/flows/13031) > A logged-in user manages sharing and permissions for a team folder in a cloud file manager. The flow starts from the folder’s action menu, opens an access management dialog, changes a collaborator’s permission level, and confirms the update with a success message. **Reown: Manage team member roles** (web, 3 screens, Development Tools, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/5734) > Happy path: an administrator opens the organization's Members area, opens the per-member actions menu for a specific member, and promotes that member to an admin. Steps: Members list overview → Open per-member actions → Apply role change and reflect update **Leonardo.Ai: Profile Settings Management** (web, 2 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/12914) > This is a settings screen in Leonardo.ai focused on personal profile management. The entry point is the user already being logged in and opening Settings, with the left sidebar showing Personal Settings and Workspace Settings. The user can review profile information, choose interest tags, confirm adult content access, and optionally delete the account. The exit point is after changes are saved and reflected in the profile section with confirmation feedback, or after the user leaves settings. Steps: View Profile Settings → Update Preferences **WhatsApp: Create Community Setup** (ios, 6 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12677) > Happy path for creating a new community on mobile. The user enters a community name, customizes the community image from the Edit control, selects either an emoji or sticker with a background color, then returns to the main creation screen with the chosen avatar and name ready to submit. **Structured: Structured AI Onboarding** (ios, 3 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3285) > A three-step onboarding carousel presented as a modal on a mobile device that introduces the "Structured AI" feature. Each slide contains a top header with the feature name, a horizontal progress indicator, an illustration of the feature, a bold headline, a short explanatory blurb, and a primary 'Continue' CTA. Controls available across slides: a Close (X) in the top-right, a Skip action, and a back arrow on later slides to navigate backwards. Entry: modal opens in front of the main app (user already in the app). Exit: user advances through slides with Continue, can Skip or Close to dismiss the modal; final Continue proceeds past the onboarding (next screen not shown). Assumption: mobile device; modal invoked to introduce a new feature; user can navigate forward/back or dismiss at any time. Steps: Intro — Speed Up Planning → Auto-Generate Your Schedule → Digitize Your Analog Planner **Adobe Photoshop: Generative Policy Acknowledgement** (ios, 4 screens, Design & Media Editing) [Refero](https://refero.design/flows/4806) > Happy path: A user composes or edits an image and taps Generate, which triggers an in‑context modal showing Adobe's Generative AI User Guidelines. The user reads (or scrolls through) the guidelines in a scrollable modal and then dismisses it (Done) to return to the image editor. Steps: Image editor — prompt and generate → Guidelines — modal title → Guidelines — safety section → Guidelines — additional information and dismiss **Apple Image Playground: Generated Image Review** (ios, 3 screens, AI Tool, Design & Media Editing) [Refero](https://refero.design/flows/12797) > A user is viewing a generated image in an image creation/editor interface. From the main preview screen, they can save, share, open more actions, browse style suggestions, and enter a text prompt. One interaction opens a contextual action menu with copy, share, save image, and report options. Another interaction switches the top-level state into a confirm/cancel mode, where the user can either discard or confirm the current change before returning to the preview screen. Steps: Image Preview → More Actions Menu → Confirm Change State **Structured: Privacy & Analytics Settings** (ios, 5 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3310) > Happy path: A signed-in user opens Settings, navigates to Privacy & Permissions, reads the privacy overview, then opens the Anonymous Analytics detail screen and toggles sharing off (or back on). Steps: Settings > Support / About → Privacy Overview → Privacy Controls List → Anonymous Analytics — Detail (Sharing On) → Anonymous Analytics — Detail (Sharing Off) **Rork: Collaborator Invitation** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12168) > Happy-path project sharing in a collaboration modal. The user opens an invite dialog from a project workspace, enters an email, chooses a permission level, and either sends an invitation or creates a shareable invite link. Steps: Invite Collaborators → Permission Selection → Invite Link Created **Pool: Refero MCP item review** (ios, 4 screens, Utility) [Refero](https://refero.design/flows/11989) > This is a mobile screenshot sequence showing a Refero content item and the user opening its overflow menu, then opening an agent selection bottom sheet and selecting one agent. Entry appears to be a detail viewer for a Refero MCP AI Agents item, with a close button, overflow menu, and embedded preview content. The user can either dismiss overlays, open the action menu, or proceed into the MCP exploration/pooling step. Assumption: the user is already inside the viewer on a mobile device and has permission to interact with the item. The flow ends with one agent tile selected in the bottom sheet. Steps: Item Detail Viewer → Overflow Actions Menu → Pool Selection Sheet → Agent Selected **WhatsApp: Meta AI Voice Selection** (ios, 3 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12674) > Happy path for choosing a voice persona for Meta AI and optionally enabling automatic voice start. The user enters a Meta AI voice/settings view from a chat context, browses available voice personas, selects one, and can turn on the 'Start with voice' toggle so Meta AI opens directly in voice mode from the Chats tab. **Meta AI: Meta AI Account Sign-In** (ios, 10 screens, AI Tool) [Refero](https://refero.design/flows/13342) > The user opens Meta AI and starts from the welcome screen. They choose to get started, select a suggested account, and authenticate using a six-digit code sent to the masked email address. After the code is submitted and authentication completes, Meta AI opens and requests permission to send notifications. The user proceeds to a short feature introduction and taps Get started to reach the main assistant screen. The flow assumes the user is on an iPhone, has a previously recognized Meta account, can access its email, and is not setting up a device. Steps: Welcome → Choose Saved Account → Account Login Loading → Email Code Entry → Code Partially Entered → Code Completed → Authentication Loading → Notification Permission → Meta AI Introduction → Assistant Home **Structured: Structured AI Onboarding** (ios, 9 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3284) > Happy path: From the New Task composer the user selects the "Plan your day with AI" suggestion which launches a multi-step Structured AI onboarding carousel. The user advances through three marketing/feature slides (Speed up planning, Auto-generate schedule, Digitize planner), then reviews privacy & data details and data categories sent to the AI, optionally reads FAQ items, and accepts the Terms. After acceptance the user lands on the Structured AI input screen where they can type, speak or scan a handwritten plan to have the AI generate schedule suggestions. Steps: New Task — Suggestions → Onboarding Slide — Speed Up Planning → Onboarding Slide — Auto-Generate Schedule → Onboarding Slide — Digitize Analog Planner → Privacy & Data — Consent Intro → Privacy — Data Used For Structured AI → Privacy — FAQ Expansion → Accept Terms & Finish Consent → Structured AI Input — Prompt Entry **Raycast: AI Model & Preset Selection** (ios, 6 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/10158) > Happy-path for selecting which AI model will power the Raycast AI command. Entry: user on Raycast home (favorites + bottom AI input). The user opens the Presets list to confirm which AI provider/preset is active, then opens the Models selector (from the bottom input) to pick a specific model (Auto or Ray-1). Exit: the chosen model is applied and reflected in the bottom input control. Steps: Home — Raycast AI input → Open Presets list → Scroll & select provider → Open Models picker → Read model descriptions → Apply selection and return **WURRD: AI practice chat** (ios, 9 screens, AI Tool, Education) [Refero](https://refero.design/flows/7862) > Happy path: A learner discovers the AI practice feature from the app's main screen, views an AI intro/benefits screen, starts a chat session, chooses a suggested prompt, receives a generated story, asks the AI to highlight newly learned words, and views the updated (highlighted) story. Steps: Home — Daily limit / Empty state → AI Intro / Benefits → Chat — Prompt suggestions → Chat — AI acknowledges / short reply → Chat — AI streaming full response → Chat — User follow-up request → Chat — AI returns modified content → Chat — Continue interaction / Read → End session affordance **ChatGPT: Persona Customization** (ios, 8 screens, AI Tool) [Refero](https://refero.design/flows/7037) > Happy-path: A signed-in user opens Settings → Personalization → Customize ChatGPT and configures a persistent persona. The user can enable the persona for new chats, enter personal info (name, role), choose a personality preset, add descriptive traits, optionally provide additional context, and toggle advanced capabilities (Web Search, Code, Canvas, Advanced Voice). Steps: Settings (Account & App sections) → Personalization → Customize ChatGPT (empty form view) → Personality selector opened → Form populated with basic values → Traits input and quick-add chips → Additional context input → Advanced capabilities **ElevenLabs: Image generation settings** (ios, 9 screens, AI Tool) [Refero](https://refero.design/flows/10741) > Happy path: user composes an image prompt (with uploaded reference images) in the composer, opens the Settings sheet, picks an image-generation model, adjusts number of generations, aspect ratio and resolution, saves the settings, and returns to the composer where the updated settings are reflected. Steps: Prompt composer with references → Open Settings sheet → Open Model selection list → Filter and select model → Return to Settings with model chosen → Open 'Number of generations' picker → Open Aspect Ratio picker → Save settings → Composer updated with new settings **Raycast: Create and use AI preset** (ios, 14 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/10188) > Happy-path: user creates a new AI preset and then opens a chat with that preset. Entry: user is on the Presets list. Exit: user opens a chat with the newly created preset selected. Assumptions: user is signed in and has access to multiple model providers and extensions; device is a mobile phone in portrait orientation. Steps: Presets list → Preset editor (empty) → Typing system instructions → Open creativity selector → Set creativity level → Open model selector (providers) → Scroll model list → Provider-specific variant list → Model selected and name field → Open AI Extensions panel → Configure extension toggles → Confirm preset configuration → Presets list updated → Open chat with saved preset **Raycast: Inline AI Command Palette** (ios, 3 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/10174) > Happy path: From within a reading/editing view the user opens an inline AI command palette (bottom sheet) to run a built-in text operation on the visible content. Entry: user is reading or editing text and opens the 'Ask AI' overlay. The overlay provides a search field and a list of built-in commands (Improve Writing, Fix Spelling and Grammar, Make Longer, etc.). The user types a query (numeric or textual). The overlay filters and shows matching commands or a "No commands found" state. If a built-in command is shown (e.g., "Make Longer"), the user selects it to apply an AI operation to the current content. Exit: the command is selected and the overlay would proceed to run the chosen AI action (execution/result screen not shown). Assumptions: user is in a text context with selectable content, the 'Ask AI' overlay is invoked over the content, the device is a mobile phone with an onscreen keyboard, and the user has permission to run built-in AI commands. Steps: Open Ask AI overlay → Type query (no match) → Type and select built-in command **Grok: Personalized Voice Settings** (ios, 4 screens, AI Tool) [Refero](https://refero.design/flows/6082) > Happy-path: The user opens the Voice Settings bottom sheet to choose a personality and spoken voice, adjust playback speed, pick an audio output, and enable voice-related toggles. From the Personality row they can create a new custom persona (plus tile) or edit an existing persona (pencil overlay). Creating a persona opens a modal with three fields (Name, Voice style, Additional instructions). After filling fields and tapping Save the modal closes and the new persona appears as a selectable chip in the Personality list. Steps: Voice Settings (main) → Create Persona (empty) → Create Persona (filled) → Persona Added and Manageable **n8n: Create Personal Agent** (web, 5 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9554) > Happy-path: A logged-in user navigates to the Personal Agents area, opens the New Agent modal, fills in metadata and a system prompt, selects a model and optional tools, saves the agent, and returns to the Personal Agents list where the new agent appears with a success confirmation. Steps: Personal Agents list (empty) → New Agent modal (blank) → Choose icon and enter name → Fill description, system prompt and model → Agent listed with confirmation **Lemni: Agent configuration and enablement** (web, 2 screens, Business, Productivity, AI Tool) [Refero](https://refero.design/flows/4580) > Happy path: an admin (or user with agent-management permissions) navigates to the Agents section, selects the "Webshop Support" agent, reviews the agent's standard operating procedures and available action list, and enables the agent for interaction. Steps: Select agent and read instructions → Inspect actions and enable agent **n8n: Chat-triggered AI Agent Setup** (web, 12 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9504) > Happy-path: In the Editor tab a user adds an AI Agent node to a blank workflow, connects it to a chat trigger, composes the prompt using workflow/execution variables in the expression editor, enables specific output format and fallback model, adds a System Message option, and returns to the canvas with the nodes connected and ready to run. Entry: Editor view with a single trigger on the canvas. Exit: AI Agent node placed, configured (prompt + options) and connected to the chat trigger; workflow shows validation/warning hints and is ready to execute. Assumptions: user is signed in with edit rights, working on a desktop browser, and has access to the AI nodes list and node configuration drawer. Steps: Open canvas with trigger → Open 'What happens next?' panel → Select AI nodes category → Add AI Agent node → Open expression editor → Build prompt expression → Search for specific fields → Locate variables (resumeUrl) → Enable output and fallback toggles → Open Add Option dropdown → Add System Message → Finalize and return to canvas **Cursor: Agent prompt composer** (web, 2 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9941) > Desktop web interface for an 'Agents' workspace. Steps: Workspace with left navigation → Collapsed navigation, focused canvas **Lemni: Agent instruction and action configuration** (web, 2 screens, Business, Productivity, AI Tool) [Refero](https://refero.design/flows/4583) > Happy path: A product user with agent-management permissions opens the Agents section, selects an existing agent (Webshop Support), reads or edits that agent's guidance in the central pane, and configures the Send Email action in the right-hand Actions panel (adding a contact/parameter and selecting permission scope). The user can then enable the agent from the Interact control. Steps: Agent Instruction Overview → Configure Send Email Action **Exa: Agent input data setup** (web, 4 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/12021) > This is a happy-path agent setup and data-import flow in a desktop web tool. Steps: Agent workspace → Add input data → Confirm imported rows → Agent workspace with data attached **n8n: Chatbot Agent Setup** (web, 3 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9509) > A visual workflow editor where the user composes a simple chat agent by wiring triggers to an AI Agent node and assigning a model. Steps: Canvas overview → Zoom to fit canvas → Inspect AI Agent node **n8n: AI Summary Agent editor** (web, 4 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9517) > A visual workflow editor for an "AI Summary Agent" that lets the user tag, schedule, run, and publish a workflow. Steps: Open editor canvas → Open tag selection → Apply tags → Open command bar **n8n: Integrations & Agents browse** (web, 3 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9460) > This flow shows a user browsing the product's integrations/agents hub. Steps: Integrations hub — full catalogue → Filter applied — AI agents visible → Filtered to zero results — empty state **Anam: Persona Setup and Sharing** (web, 4 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12941) > This is a persona/agent configuration screen where the user edits prompt settings and shares a live chat link. Entry appears to be from a persona workspace with a left-side configuration panel and a right-side preview card. The user can adjust prompt text, system prompt, greeting, director notes, and knowledge, then open the Share menu from the top bar, enable sharing, copy the generated link, and receive clipboard confirmation. Assumption: the user is already logged in and editing an existing persona on a desktop device. Steps: Persona Editing → Open Share Panel → Enable Share Link → Copy Link **Exa: Setup Complete** (web, 2 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/12009) > Happy path onboarding completion screen for an integration setup. The user arrives after selections have been made and the system has generated a setup prompt plus a visible API key. The main entry point is the generated prompt panel, where the user can copy the instructions for use in a coding agent. After copying, the button changes to a success state. Secondary actions available are viewing test search settings, revealing/copying the API key, going back, or proceeding to the dashboard. Assumption: the user is already authenticated and has completed prior setup steps on desktop. Steps: Setup Ready → Copy Confirmed **Lemni: Webshop Support Agent Editing** (web, 2 screens, Business, Productivity, AI Tool) [Refero](https://refero.design/flows/4581) > Logged-in agent management surface showing a selected support agent (Webshop Support). The center column displays the agent's written SOPs for order retrieval, cancellation, modification, and email responses. The right column is an action library where integratable actions (Shopify, Stripe, email, Slack, etc.) can be browsed, searched, and added to the agent. Entry: user navigates to Agents → selects an agent. Exit: user either filters/selects an action to add to the agent or uses the Interact control to enable the agent. Assumptions: user is authenticated with permission to view and edit agents, working on a desktop with a three-column agent-management interface. Steps: Agent Instructions & Actions Overview → Filter Actions Using Search **Lemni: First agent onboarding** (web, 2 screens, Business, Productivity, AI Tool) [Refero](https://refero.design/flows/4579) > Happy path: A logged-in user is guided through an onboarding/tutorial screen to start creating their first AI agent, then lands in the Knowledge workspace where they can create or populate a dataset (here named “General”) by adding URLs, uploading files, or dragging content for the agent to be trained on. Steps: Onboarding — Play Tutorial → Knowledge — General Dataset **ChatGPT: Customize Assistant Profile** (web, 8 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/6900) > Happy path for creating a personalized assistant persona from the signed-in home screen on a desktop layout. Entry: authenticated user lands on the ChatGPT home screen. The user opens their account menu, chooses Customize ChatGPT, completes the modal form (name, role, traits, optional extra info), optionally expands Advanced to toggle capabilities, enables the persona for new chats, and saves. Exit: modal closes and the new persona settings are applied for subsequent conversations. Assumptions: user is signed in (John), has permission to customize their assistant, and is using a desktop browser with a left navigation and central composer area. Steps: Home / Welcome → Open Account Menu → Launch Customize Modal → Enter Basic Identity Fields → Define Traits (Chips & Freeform) → Add Detailed Traits & Context → Open Advanced Options → Enable Capabilities & Save ##### Task: AI chat assistant **PayPal: PayPal Assistant Conversation** (ios, 16 screens, Finance) [Refero](https://refero.design/flows/7266) > Happy-path walkthrough of a logged-in mobile user opening the Message Centre, launching the PayPal Assistant, asking for transaction help, receiving an automated answer with detailed steps and links, giving feedback, and ending the conversation. Steps: Message Centre (empty) → Assistant greeting and quick options → User quick reply sent → Assistant 'no transactions found' → User declines search change → User enters free-text question → Assistant processing response → Assistant provides step-by-step answer → Related links and feedback prompt → User taps thumbs up (positive feedback) → Feedback input (user typing) → Send feedback → Open End Conversation action → Confirm end conversation → Message Centre updated → End state and return to idle **Granola: Review Generated Email** (ios, 2 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14157) > The user is viewing an AI-generated follow-up email within a mobile assistant conversation. The flow starts with an existing prompt and generated email response, then continues when the user selects the positive feedback control. It ends with the response marked as helpful. The user is assumed to be in an active conversation on a mobile device; authentication and permission state are not visible. Steps: Generated Email Review → Positive Feedback Submitted **GlossGenius: Analytics with AI Growth Analyst** (web, 4 screens, Business) [Refero](https://refero.design/flows/9979) > Happy path: A signed-in user opens the Analytics section from the left navigation and sees key performance indicators (net revenue, projected revenue, service/retail revenue) for the selected timeframe and staff filter. A persistent right-hand panel titled 'AI Growth Analyst' offers suggested prompts and an input area. The user selects (or types) a suggested prompt to set a revenue goal; the assistant shows brief processing states and then returns a written recommendation with numeric targets and tactical suggestions. Steps: Analytics overview with AI suggestions → User submits suggested prompt → Assistant processing state → Assistant provides recommendation **Granola: Copy Generated Response** (ios, 2 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14156) > A user views a generated follow-up email in a mobile AI conversation and copies the response to the clipboard. The flow starts on an open conversation with the generated message visible and ends with confirmation that the message was copied. The user is assumed to be on a mobile device with access to the conversation; no additional authentication or permission state is visible. Steps: Generated Email Response → Copy Confirmation **Linktree: Insights and AI Tips** (web, 6 screens, Business, Social Networking) [Refero](https://refero.design/flows/12410) > Logged-in user is viewing the Insights section of a Linktree-style dashboard on desktop. The entry point is the main analytics dashboard, with a date-range filter set to the last 7 days. The user can review performance metrics, open an AI tips modal, and launch a right-side AI chat assistant to generate content ideas. The flow ends with the AI assistant actively prompting the user for topic input, while the underlying insights dashboard remains visible in the background. Steps: Insights Dashboard → Personalized Tips Prompt → AI Chat Opened → AI Analysis in Progress → AI Recommendation Response → Brainstorming Prompt **Cohere: Tool-assisted Charting with Python Interpreter** (web, 6 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3837) > Happy path: A logged-in user opens the Chat workspace in Tool Use mode, selects helper tools (Python Interpreter, Calculator, Web Search), types a single prompt requesting a plot of average 1-bedroom rental prices for the top 5 most expensive North American cities in Jan 2024, and sends it. The system invokes tools (web search then Python interpreter), generates code, executes it, and returns a matplotlib bar chart image embedded in the conversation along with code and references. Steps: Tool selection & prompt composer → Prompt submitted → Assistant begins reasoning & tool invocation → Python code produced and executed → Plot rendered inline → Tool result preview & continued interaction **Doctronic AI: Passwordless AI Doctor Sign-In** (web, 8 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13706) > The flow begins on Doctronic’s public landing experience, where an unauthenticated visitor can start a chat or access account options. The user chooses email-based login, enters an email address, receives a six-digit verification code by email, and submits the code. After successful verification, the user is taken to the authenticated AI health assistant screen, ready to ask a health question. The user is assumed to be logged out and able to access the entered email inbox. **Doctronic AI: Sleep Concern Assessment** (web, 13 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13694) > A logged-in user named Connie starts a conversation with Doctoric AI by entering a health concern about recurring nighttime waking. The system displays an emergency disclaimer, processes the message, and requires acceptance of the Terms of Service and AI-provider disclaimer before continuing. The user then provides age and sex information, receives follow-up questions about additional symptoms and wake-up patterns, and answers questions about health history. The flow ends with the assistant asking about current medications, with the user remaining in the conversation. **Google Gemini: Voice Query to Suggestions** (ios, 6 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/3207) > Happy-path: A signed-in user interacts with the Gemini conversational assistant on a mobile device. Steps: Conversation (idle with assistant intro) → Start Voice Input (listening) → Live Transcription (partial) → Full Transcription / Submit → Assistant Responding (audio playing) → Answer Rendered (suggestions with media) **WhatsApp: Meta AI Voice Chat Setup** (ios, 6 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12673) > Happy path for starting a voice conversation with Meta AI. The user enters the Meta AI profile/settings area, opens the voice feature, reviews the informational consent-style modal, and continues into a live voice session. **Meta AI: Save Recipe Instructions** (ios, 2 screens, AI Tool) [Refero](https://refero.design/flows/13345) > A user views an assistant-generated response containing ingredients and a 20-minute pasta cooking method. The user copies the response using the copy action beneath the message. A confirmation toast appears stating that the content was copied. The flow starts with an existing assistant response and ends with the response available on the device clipboard. Steps: Review Recipe Response → Copy Recipe **Raycast: Save AI Message to Note** (ios, 4 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/10166) > Happy path: A user is reading an AI assistant conversation on a mobile device and wants to save a specific assistant message as a persistent note. From the conversation view the user opens a contextual '+' menu, selects "New Note," receives a transient confirmation, and then navigates to the newly created note where the assistant's long prompt text is displayed. Steps: Conversation — locate message → Open action menu → Show creation confirmation → Note viewer — review saved content **Grok: Private Chat — Compose and Receive Response** (ios, 6 screens, AI Tool) [Refero](https://refero.design/flows/6093) > Happy path: A signed-in mobile user opens the assistant, taps the private (ephemeral) chat mode, composes a long prompt, submits it to the selected model (Grok 3), waits while the model processes, and then reads the generated response. The user can stop generation mid-way, scroll the resulting content, and use inline actions (copy, share, thumbs up/down). Steps: Chat Home / Composer → Private Chat — Empty State → Compose Prompt → Submit & Processing → Assistant Response Appears → Read Full Reply & Actions **Genie: Ask Genie With Photos** (ios, 9 screens, Social Networking, AI Tool) [Refero](https://refero.design/flows/12422) > A mobile user opens a Genie assistant conversation, enters a question about the assistant’s capabilities, attaches three photos from the device photo library, and sends the message. Genie shows a loading state, then responds with an overview of its capabilities and references the uploaded photos. The flow begins in an empty assistant conversation with the user already signed in and ends with the assistant’s response visible; the user remains on the conversation screen and can continue using the Follow up field. Steps: Assistant Home → Enter Question → Open Attachments → Choose Photos → Review Attachments → Message Submitted → Assistant Processing → Capabilities Response → Continue Conversation **WhatsApp: Meta AI Message Feedback** (ios, 2 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12672) > A mobile chat conversation with Meta AI. The screen opens inside an existing conversation, showing an AI-generated reply, a disclaimer about AI accuracy, message actions, and the input bar. The user can react to the AI message using the feedback controls, and a confirmation toast appears after feedback is submitted. Assumptions: the user is already inside the chat and has permission to interact with the conversation on a phone device. **Meta AI: Incognito AI Conversation** (ios, 7 screens, AI Tool) [Refero](https://refero.design/flows/13350) > A logged-in user opens the navigation drawer, selects the private/incognito Meta AI chat option, reviews the privacy introduction, and continues into a temporary conversation. The user enters a prompt requesting two explanation levels, submits it, and waits while Meta AI generates the response. The completed answer presents a child-friendly explanation followed by a physics-professor-level explanation, which the user can scroll through. The flow begins in the main chat area with the navigation drawer open and ends on the completed incognito response. Steps: Open Incognito Chat → Review Privacy Introduction → Enter Private Prompt → Complete Prompt → Submit Prompt → Review Generated Response → Read Full Explanation **Meta AI: Mythical Creature Field Guide** (ios, 7 screens, AI Tool) [Refero](https://refero.design/flows/13351) > A user submits a request to Meta AI on a mobile device asking for a detailed walkthrough of dragons, phoenixes, krakens, unicorns, and mermaids, using multiple sub-agents. Meta AI first plans the work, then runs five specialist biology agents, regenerates a corrected poster asset, and presents a consolidated field-guide response with supporting infographics. The user can scroll through the answer, provide feedback, copy or share it, and choose from suggested follow-up requests. The user appears to be using the assistant without any visible authentication or permission step. Steps: Prompt Entry → Request Planning → Sub-Agent Execution → Asset Regeneration → Field Guide Response → Answer Review → Completion and Follow-Up **Voicenotes: Ask AI from Filtered Notes** (ios, 11 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14249) > A signed-in user starts from the Voicenotes note library and opens Ask AI to query their recorded notes. They choose Web Search and restrict the note context to the last 30 days, then dictate a request by voice. The system converts the recording into editable text, submits it, and displays an AI-generated response with source and copy controls. The flow begins in the note library and ends on the completed response view. The screenshots show an iPhone-sized interface with keyboard and microphone permissions already available; no authentication or permission setup is shown. **Meta AI: AI Image Creation** (ios, 4 screens, AI Tool) [Refero](https://refero.design/flows/13358) > A user enters a detailed text prompt asking Meta AI to create a hyper-realistic image. After submitting the request, the interface shows a prompt-processing state, then presents the generated image with an explanatory response. The user can review the result, provide feedback, copy or share the response, and choose suggested refinements. The flow begins on the Meta AI prompt composer and ends with the generated image available for review or further editing. The screenshots imply a mobile device and an available Meta AI session. **Google Gemini: Request Java Ping Pong Example** (ios, 4 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/3222) > Happy path: A signed-in user opens the Gemini chat assistant on a mobile device, types (or speaks) a request for a simple Ping Pong game written in Java, and receives a multi-part response: a labeled code block (scrollable), a caution/copy affordance, a structured explanation, and a source link with feedback/share controls. Steps: Welcome / Composer → Request Sent / Code Response → Code Continuation & Caution → Explanation, Sources & Feedback **Meta AI: Recipe Response Rating** (ios, 3 screens, AI Tool) [Refero](https://refero.design/flows/13344) > A user reviews a conversational cooking response containing ingredients and a 20-minute pasta preparation plan, then marks the response as helpful using the thumbs-up control. The flow starts on an open assistant conversation and ends with the positive rating selected. No login, permissions, or additional setup are visible. Steps: Review Recipe Response → Submit Positive Feedback → Confirm Selected Rating **Voicenotes: Configure Ask AI Search** (ios, 4 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14250) > The user starts in an Ask AI note-assistant interface with the message composer focused and the keyboard open. They open the composer options, choose Web Search as the assistant mode, and select Last 30 days as the note date filter. The options panel closes, and both selections appear as removable chips above the message field. The user remains ready to enter a question; no message has been submitted. The flow assumes the user has access to at least one recorded note and is using a mobile device with the keyboard available. **Genie: Dinner Recommendation Chat** (ios, 6 screens, Social Networking, AI Tool) [Refero](https://refero.design/flows/12427) > Happy-path conversational assistant experience on a mobile device. The user starts on a home/chat screen with a Genie profile card and prompt chips, enters a dinner-related question in the message composer, sends it, and receives suggested nearby restaurant options with image cards and explanatory text. Steps: Assistant Home → Message Entry → Send Message → Assistant Loading → Recommendation Preview → Detailed Suggestions **Meta AI: Trending Recipe Search** (ios, 8 screens, AI Tool) [Refero](https://refero.design/flows/13349) > The user is viewing an existing dinner-related assistant response and starts a new recipe search from the reply composer. They enter “Find recipes with chicken and avocado,” open the response-mode selector, switch from Instant to Thinking, and submit the request. The interface shows a loading state while gathering trending recipes, then displays a multi-item recipe response that can be scrolled for additional recommendations. The flow assumes the user is already inside a conversation named “Dinners.”. Steps: Existing Dinner Response → Activate Reply Composer → Enter Recipe Query → Open Response Mode Menu → Choose Thinking Mode → Submit Recipe Search → View Initial Recommendations → Review Full Recipe List **Synthesia.io: AI Presentation Review and Summary** (web, 8 screens, AI Tool) [Refero](https://refero.design/flows/13744) > The user is working in a presentation and video editor with a selected scene and an AI assistant available in the right-hand panel. They first ask the assistant for an opinion about the presentation, receive an explanation of the assistant's capabilities, and then request a summary of the video. The assistant returns a scene-based summary and identifies that scene 8 is currently selected, offering to provide a more detailed scene-by-scene explanation. The flow starts with the editor open and ends with the summary expanded in the chat. **Tripadvisor: AI Assistant Recommendation Feedback** (web, 2 screens, Travel & Booking) [Refero](https://refero.design/flows/7077) > Happy-path: A signed-out or anonymous desktop visitor lands on Tripadvisor's homepage, sees the main search controls and an AI assistant panel on the right that proactively lists recommended food-focused destinations. The user reads the assistant's suggestions and optionally provides feedback (a like/heart). The assistant acknowledges the feedback with a small confirmation toast and remains available for follow-up questions. Steps: Homepage with AI assistant suggestions → Provide feedback on assistant message **Voicenotes: Ask AI assistant** (web, 6 screens, Collaboration Tools, AI Tool) [Refero](https://refero.design/flows/6981) > Happy path: A signed-in desktop user views their list of voice notes and opens an in-context AI assistant via an "Ask AI" control. The assistant appears as a centered modal that offers suggested questions and a text/voice input. The user picks or types a question, the assistant shows a short response inline, can show a processing state (“Thinking…”), and then returns a longer answer with a copy control. Steps: Notes list with Ask AI → Ask AI modal — suggested prompts → Assistant returns short answer → Assistant returns another quick reply → User asks custom question — processing state → Assistant returns full answer with copy **Grok: Image-assisted question and response** (web, 8 screens, AI Tool) [Refero](https://refero.design/flows/5325) > Happy path: A signed-in user on a Grok conversation thread uploads an image from the composer, optionally edits/inspects the attachment, sends a text prompt referencing the photo, and receives an AI-generated summary about the product shown in the image. Steps: Open conversation with composer → Open attachment menu → Select file to upload → Confirm attachment visible → Type prompt referencing image → Assistant begins processing attachments → Assistant thinking/progress → Assistant returns full summary **Mocha: AI-assisted project generation** (web, 4 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/11610) > Happy-path: A user asks an on-screen AI assistant (left sidebar) to build a responsive photo gallery. The assistant processes the request in stages (Thinking → Reading files → Downloading assets) while the main canvas (center/right) shows a rotating set of preview cards representing generated features (Publish, Database, Discuss Mode, Versioning). Steps: Prompt submitted — assistant thinking → Assistant reading files → Assistant planning — discuss mode shown → Assets downloading and version preview **PamPam: Add place using AI assistant** (web, 4 screens, Map & Navigation, AI Tool) [Refero](https://refero.design/flows/4716) > Happy path: A signed-in user is editing a custom map titled "Liverpool" and wants to add a new point of interest. From the left-hand map editor in edit mode the user opens the Add command palette, searches or invokes the AI option for a specific place (The Oracle Bar), and chooses the AI "Tell me about" action. The system adds the place to the map list and drops a pin on the map while returning an AI-generated description/confirmation in a chat panel. Entry: map open in edit mode. Exit: new place appended to the left list and pin placed on the map; AI assistant message shown. Assumption: user is in edit mode and has permission to modify the map. Steps: Open map editor (Liverpool) → Open Add command palette → Select place and AI action → AI adds place and shows confirmation **Polar: AI-guided product creation** (web, 9 screens, Business, Finance) [Refero](https://refero.design/flows/11386) > Happy path: A user starts on the standard "Create Product" editor and chooses the "Create with AI" assistant. They enter a natural-language description of the product, iterate with the assistant until the assistant has enough details (including per-seat price), and the assistant attempts to create a product. An error occurs and the user is given the option to configure manually; choosing that returns the user to the manual product editor. Steps: Manual Product Editor → AI Assistant — Prompt → Prompt Submitted → Assistant Clarifies Limitations → User Adds Additional Details → Assistant Requests Pricing Clarification → User Provides Price → Automated Creation Error → Return to Manual Editor **ChatGPT: Voice prompt to generated email** (web, 4 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/6893) > Happy-path flow for creating a new conversation by dictating a prompt and receiving an AI-generated response. Steps: Landing / Input ready → Recording voice input → Transcription visible for review → Assistant generates response **Meta AI: 16-bit Sprite Creation** (web, 16 screens, AI Tool) [Refero](https://refero.design/flows/12743) > Happy path for creating a 16-bit character sprite from the user’s photo inside a chat-based AI interface. The user starts from the home screen, chooses a preset concept, provides a reference image through an upload dialog, and waits while the system interprets the prompt and generates the artwork. The flow ends when the AI returns a finished sprite and offers follow-up refinement suggestions. Steps: Home Start → Preset Preview → Task Initialized → Planning Response → Reference Retrieval → Request for Input → Upload Media → Photo Added → Image Processing → Avatar Drafting → Style Refinement → Prompt Forwarding → Pixel Sketching → Assembly → Generated Result → Refinement Options **Canva: Revise cover title via voice or text** (web, 4 screens, Design & Media Editing, Collaboration Tools) [Refero](https://refero.design/flows/8061) > User is in a desktop web design assistant screen showing a project titled "Tokyo Travel Vlog Cover Design." The interface displays four AI-generated design thumbnails, a row of suggestion chips (e.g., "More Instagram Story designs", "Add a cherry blossom illustration"), and a chat-style input area at the bottom that supports typing and voice input. Steps: Design suggestions preview → Start voice input → Transcription inserted → Edit and submit instruction **Meta AI: AI Conversation and Creative Writing** (web, 17 screens, AI Tool) [Refero](https://refero.design/flows/12736) > A logged-in user starts from an AI assistant home screen, sends a casual greeting, receives an immediate response, then submits a request for an unusual fact. The user opens the response-mode selector, changes the mode from Instant to Thinking, and submits a creative-writing request for a science-fiction octopus story. The assistant shows intermediate reasoning-status updates before progressively displaying a detailed blockbuster-style plot. The flow ends with a complete story treatment and follow-up suggestions. Steps: Assistant Home → Enter Greeting → Submit Greeting → Greeting Response → Enter Fact Request → Submit Fact Request → Fact Response → Open Response Mode Menu → Select Thinking Mode → Enter Story Prompt → Submit Story Prompt → Story Planning → Expanded Story Outline → Initial Story Treatment → Read Story Development → Review Story Acts → Complete Story Treatment **Pi: Language Tutor Onboarding** (web, 7 screens, AI Tool) [Refero](https://refero.design/flows/4705) > Happy path: A signed-in desktop user browsing Discover selects the "Learn a new language with Pi" content, reads the introduction, uses suggested topic chips or types a request (e.g., "I want to learn German"), and receives beginner phrases. The assistant then asks whether the user wants to dive deeper; when the user confirms, the assistant provides more phrases and a short grammar explanation. Steps: Landing — Learn a new language → Suggested topics and chips → Chip loading / empty state while browsing → User submits language request → Assistant provides starter phrases → Assistant prompts to go deeper → Extended mini-lesson — phrases and grammar **WhatsApp: Meta AI Image Conversation** (web, 8 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/13212) > A user interacts with Meta AI inside WhatsApp on desktop. From an open Meta AI conversation, the user submits the prompt "/imagine a car race on mars," waits while the system generates an image, reviews the result, and sends a formatted follow-up message saying "Me too." Meta AI then processes the follow-up and returns a contextual response. The user is assumed to be signed in and already viewing the Meta AI chat. **Allbirds: AI-Assisted Shoe Purchase** (web, 6 screens, Marketplace) [Refero](https://refero.design/flows/14331) > A shopper views a women's shoe product detail and opens the Wearallbirds AI assistant for product information and purchase guidance. The assistant identifies the selected product, presents details and recommendations, lets the shopper add the item to the cart, prompts for a size, confirms the selection, and offers related products. The flow begins on a product detail view and ends with the item added to the cart and recommendation actions available. The shopper is assumed to be browsing on a desktop device; no login or special permissions are visibly required. Steps: Product Detail → AI Product Overview → Add Product → Choose Size → Cart Confirmation → Recommendation Actions **Canva: Tokyo travel vlog cover creation** (web, 24 screens, Design & Media Editing, Collaboration Tools) [Refero](https://refero.design/flows/8059) > Happy-path: A user opens the Canva AI chat area, requests a cover image for a Tokyo travel vlog, receives AI-generated design options, refines the result (including uploading an illustration), selects a design and opens it in the Canva editor for final adjustments. Steps: Canva AI landing / prompt field → Select example idea / narrow category → Prompt submitted → Generation placeholder → Initial design results → Review and quick actions → Type a style refinement → Transient failure message → Open attachment/upload menu → Upload image and message → AI generates Instagram Story mockups → Request title adjustments → AI returns revised variants → Select variant for deeper edit → Open split preview + editor entry → Adjust color palette → Toggle color variations → Preview final variant → Open full Canva editor → Chat recorded in Recent Chats ##### Task: API keys developer settings **Willow: Willow onboarding intro** (ios, 3 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/13227) > A short onboarding carousel for Willow shown to a new or returning user before entering the main experience. The user can swipe through several promotional intro screens and tap the primary Get started button to continue. **Fuse: New wallet signup and secure key setup** (ios, 22 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8729) > Happy-path onboarding for a crypto wallet: user is introduced to product pillars (Spend / Save / Invest), chooses to create an account using Sign in with Apple, and runs through a 3-part key setup (Device Key — biometric, 2FA Key — iCloud, Recovery Key — email OTP). After key creation the wallet is provisioned, a short promo for a paid membership is shown, and the user arrives at the empty wallet home. Steps: Onboarding carousel — Spend → Onboarding carousel — Save → Onboarding carousel — Invest → Welcome / Sign-up choices → Sign in with Apple sheet → Signing-in progress → Secure your wallet — Device Key intro → Face ID permission prompt → Creating Device Key (progress) → Create 2FA Key (iCloud) prompt → Recovery Key — email entry → OTP entry — keyboard & autofill → OTP verification and provisioning → Locked state — Unlock CTA → Promo modal — free transactions → Fuse Plus membership detail → Wallet provisioning complete — home (empty) → Empty wallet CTA — Receive → Default bottom navigation & quick actions → End state — ready wallet **Poppy: Recommended Connection Setup** (ios, 4 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/12177) > Happy-path onboarding/settings sequence for a logged-in user named Connie. The user opens the profile/settings area, chooses a recommended connection type (Fitness), is shown a system Health permission request, and returns to the recommendation screen after granting access. Entry is from the profile/settings menu; exit is back to the recommendation step with the selected connection highlighted and a Next button available. Steps: Profile and Settings → Connection Recommendations → Health Permission Request → Connection Selected **komoot: New user onboarding and consent setup** (ios, 39 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8281) > Happy-path onboarding for a first-time mobile (iOS) user. Entry: app launch -> privacy/tracking consent banner. User can Accept All, Refuse, or Customize preferences. If user customizes they can toggle categories and open vendor details. After consenting, the flow presents sign-in options (social or email). Choosing email leads to email validation then account creation (password). Post-signup the user progresses through product onboarding cards that request push notifications and access to device integrations (Apple Health). The flow ends with device connections (optional) and a completion screen, then the main map/routes home screen. Steps: Privacy notice (top-level) → Customize preferences (categories visible) → Category list (scroll) → Analytics expanded (vendor row visible) → Vendor info modal (Firebase) → Sign-in options → Enter email (initial) → Email validation error → Email accepted (proceed) → Create account (password) → Onboarding hero / intro → Onboarding card: Navigation → Onboarding card: Notifications → iOS push permission dialog → Device integrations list → Health Access (iOS modal) → Integration connected state → Onboarding completion → Main map / routes home → Optional: Connect additional devices (post-onboarding) **Poppy: Support and Privacy Settings** (ios, 6 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12194) > A settings/support area presents account and privacy-related options, including feedback, data retention, privacy policy, terms, account deletion, experimental features, and sign out. Selecting certain items opens full-screen informational panels that explain what data Poppy collects, how it is used, and how long it is stored. Entry appears to be from an account/settings screen; exit is via the close button or back navigation. Steps: Support Menu → Data Overview → Location Details → Email Details → Health Details → Control and Deletion **RAD Weather: Ambient Weather Connection** (ios, 5 screens, Weather) [Refero](https://refero.design/flows/13803) > The flow starts from an “Add Uplink” provider selection screen and continues through the Ambient Weather credential setup. The user selects Ambient Weather, enters a personal API key, and tests the connection. The screenshots show the connection request loading and ending with a provider rejection error. The user is assumed to be on an iPhone and able to access their Ambient Weather API key. Credentials are described as being stored locally in Apple Keychain. Steps: Provider Selection → Enter API Key → Test Connection → Connection Check → Credential Rejection **Phantom: Token List Manager** (ios, 3 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3716) > Happy path: A logged-in user opens the Token List Manager to show or hide tokens in their wallet. The screen lists available networks/tokens each with a balance and an on/off toggle. The user focuses the search field to filter the list, types a query (letters or numbers) and the list updates; if nothing matches, the UI shows an empty state message. Steps: Token list overview → Focus search field → Search resulting empty state **Exa: Developer Workspace Home** (web, 2 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/12010) > This is a dashboard-style entry screen for an Exa developer workspace. The user is logged in and can either begin onboarding/integration from the main hero card, inspect documentation, manage API credentials, add credits, or explore product capabilities and templates. The flow enters from the workspace home dashboard and exits when the user clicks a primary action such as Get Started, Docs, a product card, a template card, or API key-related controls. Steps: Workspace Dashboard → Compact Dashboard View **Exa: API Keys Management** (web, 3 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/12029) > This is a signed-in team settings area for managing API credentials. The user enters the API Keys section from the left navigation, views the current team details and existing keys, opens the overflow menu for a key, and can choose edit or delete. In the final state, one key has been removed from the list. Assumptions: the user is authenticated, has permission to manage keys for the current team, and is using a desktop browser. Steps: API Keys List → Key Actions Menu → Updated Key List **Gladia: Developer Overview and API Keys** (web, 2 screens, AI Tool) [Refero](https://refero.design/flows/10886) > Logged-in developer lands on the Gladia "Start building" overview (desktop). The screen surfaces three primary entry cards (Real-time transcription, Asynchronous transcription, Documentation) and an API Keys panel showing existing keys and a "+ Create API Key" control. Left navigation lists products and management sections (Transcriptions, API keys, Webhooks); a collapsed-icons variant of the same sidebar is also present. Happy path: user scans available product entry points, then manages or creates API keys via the API Keys panel. Steps: Overview — Sidebar expanded → Overview — Sidebar collapsed (icons-only) **Wabi: Account Settings and Deletion** (ios, 7 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12072) > A signed-in user opens the Settings screen, reviews profile details and connected services, optionally enables integrations, then navigates to More options to initiate account deletion. The flow ends after deletion is confirmed, returning the user to the launch/entry screen. Steps: Settings Overview → Memory Processing → Power-Up Toggles → Additional Integrations → Support and Account Links → Delete Entry Sheet → Deletion Confirmation **Phantom: Settings Search** (ios, 2 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3728) > User opens the Settings modal (dark theme, logged-in account visible) and uses the built-in search field at the top to find a specific item. Steps: Settings Home → Search Input — No Results **OKURI: Settings and Feature Info** (ios, 4 screens, Utility) [Refero](https://refero.design/flows/11873) > A settings and information area within the OKURI experience, entered from a main settings/menu screen. The user can review and adjust items like haptics and birthdate, then open informational feature cards such as Digital Tear-Off, Star Sign, and Sun Times. Entry appears to be from an in-app settings/menu view; exit is via the back arrow. Steps: Settings Menu → Digital Tear-Off Card → Star Sign Card → Sun Times Card **Box Box Club: Account & App Settings** (ios, 4 screens, Entertainment, Sports, Transport) [Refero](https://refero.design/flows/7880) > A single-screen settings area (dark theme) that lets a signed-in user view and change general preferences, personalization options, and access Support & Legal items. Entry: user opens Settings while authenticated. Primary actions: toggle preferences, open sub-screens (Notifications, Theme, Language, Accessibility, etc.), access legal/support links, and sign out or initiate account deletion/feedback. Exit: either navigate into a settings sub-screen, trigger an in-context action sheet (contact/feedback/delete), or log out and land on the authentication / sign-in prompt. Steps: Settings — General → Support & Legal and Logout → Contact / Feedback Action Sheet → Authentication — Sign In Prompt **Google Gemini: Gemini Welcome & Consent** (ios, 3 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/3201) > This is a mobile-first, modal-style welcome and consent sequence presented to a logged-in Google user before they access Gemini. Steps: Welcome modal with summary → Expanded detail card → Legal links, opt-in, and final CTAs **ElevenMusic: Account Settings Feedback** (ios, 4 screens, AI Tool, Entertainment, Music Streaming) [Refero](https://refero.design/flows/12446) > This is a mobile account settings area with a feedback subflow. The user starts on Account Settings, where they can review subscription status and access support/help links. Selecting Give feedback opens a Send Feedback modal with a text field and submit action. After entering text and submitting, a confirmation toast appears and the user returns to Account Settings. Steps: Account Settings → Feedback Form → Feedback Entry → Submission Confirmation **Willow: Willow Home Dashboard** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/13235) > Logged-in Willow home screen on iPhone. The user sees a stats header, a dismissible promo banner, and a rotating onboarding/promotional card carousel. From this screen they can start dictation, set up an Action Button shortcut, watch a tutorial, add a personal term in Dictionary, or invite friends for a referral reward. The screen also shows a transcripts section with search and filter controls, and a bottom tab bar for navigating to Home, Dictionary, Style Matching, and Profile. Steps: Home Overview → Action Button Setup Card → Dictation Tutorial Card → Personal Term Card → Referral Reward Card **ElevenMusic: Music Creation Workspace** (ios, 4 screens, AI Tool, Entertainment, Music Streaming) [Refero](https://refero.design/flows/12459) > This is a mobile music-creation workspace. The user enters the Create section, where they can browse draft or published generations, open a prompt/composer panel, and adjust generation settings in a bottom sheet. The main entry point is the Create tab; the exit point is either returning to the list or submitting a creation prompt. Steps: Create Dashboard → Settings Overview → Settings Adjustment → Prompt Composition **RAD Weather: Personalization Preferences** (ios, 3 screens, Weather) [Refero](https://refero.design/flows/13798) > A user configures personalization and accessibility preferences during a multi-step setup process. The preferences screen opens with High Contrast and Locked Header enabled, the language set to System, default text size at 1X, system measurement units selected, and the forecast tone set to Blunt. The user opens the Language selector, chooses English, and returns to the completed preferences screen. The user can then continue to the next setup step. The flow assumes the user is on a mobile device and is progressing through an onboarding or configuration wizard. Steps: Review Preferences → Open Language Selector → Select English **WhatsApp: Settings Search** (ios, 3 screens, Social Networking, Collaboration Tools) [Refero](https://refero.design/flows/12609) > A mobile in-app search screen for settings and help items. The user enters the search view from a settings/menu area, types a query, and either sees matching suggestions, no results, or can clear the query and exit with the close button. Steps: Open Search → No Results → Suggested Matches **Willow: Offline Mode Settings** (ios, 3 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/13251) > Mobile settings screen for managing offline dictation support. The flow starts on an Offline mode settings screen and can end either by deleting the downloaded offline model or by returning to the same screen after viewing the delete confirmation. Steps: Offline Mode Settings → Delete Confirmation → Offline Model Removed **Appwrite: Create development key** (web, 5 screens, Development Tools) [Refero](https://refero.design/flows/6027) > Happy path: A signed-in user on a project overview navigates to the Integrations > Dev keys area, opens the Create dev key modal, fills a name and expiration, and creates a temporary development key. After creation the UI shows a success toast and navigates to the new key's detail view (with masked secret and editable fields). The user then returns to the Integrations list where the new key appears. Steps: Integrations — Dev keys access → Create dev key modal — blank → Create dev key modal — filled → Dev key detail view → Dev keys list updated **fal: Create API Key** (web, 5 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/13577) > An authenticated user opens Developer > API Keys from Settings, starts creating a key, selects the API scope, enters a description, and submits the form. The system generates the key and displays it once in a confirmation dialog with options to copy it or finish. After completing the dialog, the new key appears in the API Keys list with its description, masked value, creation date, scope, and optional tag control. Steps: API Keys Empty State → Configure New Key → Submit Key Request → Copy Generated Key → Verify API Key **Anam: API Key Creation** (web, 5 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12947) > Happy path for creating a new API key from the API Keys section. The user starts on an empty or populated API Keys screen, opens the create dialog, enters an optional label, keeps or changes full access, submits the form, copies the generated secret key from the confirmation modal, and returns to the API Keys list where the new key appears. Steps: API Keys List → Create Key Modal → Labeled Key Setup → Secret Reveal → Keys List Updated **Gladia: API Key creation and listing** (web, 3 screens, AI Tool) [Refero](https://refero.design/flows/10900) > Happy path: a logged-in project member opens the API Keys area, enters an optional name for a new API key, and creates it. The new key is added to the list with a truncated token and quick actions (copy, delete). Steps: API Keys overview → Enter key name → New key created and listed **ElevenLabs: Create and Configure API Key** (web, 5 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/8711) > Happy path: A logged-in workspace admin or developer opens the Developers > API Keys area, creates a new API key, configures a name and fine-grained access permissions (and optional credit limit), confirms creation, copies the secret (displayed only once), and returns to the keys list where the new key appears masked and can be enabled/disabled. Steps: API Keys List (Empty State) → Create Key — Basic Settings → Create Key — Permission Scopes → One-time Secret Modal → API Keys List — Created Key **Manus: API Key creation and storage** (web, 6 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/10486) > Happy path: A logged-in user with access to Integrations navigates to the API section, opens the Create API Key modal, names the key, creates it, copies the generated secret, and returns to the API list where the new key appears with its value masked. Steps: API settings overview → Create API Key (empty) → Enter key name → Show generated secret → Confirm copy feedback → API keys list updated **Fingerprint: Public API key creation** (web, 5 screens, Business, Development Tools) [Refero](https://refero.design/flows/11163) > Logged-in user opens the workspace API keys area to create a new Public API key scoped to an environment. The flow starts on the API keys list, opens a modal to collect Environment, Name and optional Description, then generates a key and shows a success dialog with the raw key and copy control. The user closes the dialog and returns to the API keys list where the newly created key appears. Steps: API keys list → Create key (empty form) → Create key (filled form) → Success dialog with key → Updated API keys list **Parallel: Create API key for Default Application** (web, 4 screens, AI Tool) [Refero](https://refero.design/flows/9078) > Happy path: a logged-in admin opens Settings → API Keys, expands an inline "Create a New API Key" form under the Default Application, enters a name, and creates the key. Steps: API Keys list (Default Application) → Open inline Create API Key form → Enter key name → New API key appears in list **Linear: Create personal API key** (web, 5 screens, Business, Productivity, Project Management, Collaboration Tools) [Refero](https://refero.design/flows/6691) > Happy path: a logged-in user opens the Security & access area to create a new personal API key, configures granular permissions and team scope, creates the key, copies the revealed token, and returns to the Security & access overview where the new key is listed. Steps: Security & access — Overview → Create API key — Form → Create API key — Select scopes & teams → API key created — Details & copy → Security & access — Updated overview **Steel: Create Project API Key** (web, 5 screens, AI Tool, Development Tools) [Refero](https://refero.design/flows/14293) > The user starts in the project’s API Keys settings, creates a new project-scoped API key, receives the secret value once, and returns to the updated key list. The user is assumed to be authenticated and authorized to manage project credentials. The flow is shown on a desktop interface with an active Default project environment. **Exa: API Key Creation** (web, 6 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/12027) > A logged-in user with access to the Personal team opens the API Keys section, starts creating a new API key, enters a name, and submits the form. The system generates the key, displays the full secret once in a confirmation dialog, and adds the named key to the API keys list after the user dismisses the dialog. The user is assumed to be on a desktop device and has permission to manage team API keys. Steps: API Keys List → Name API Key → Enter Key Name → Key Generation → Key Created → Updated API Keys List **Appwrite: API key removal from Integrations** (web, 4 screens, Development Tools) [Refero](https://refero.design/flows/6026) > Happy-path: A logged-in user opens a project dashboard, navigates the Integrations → API keys area, selects an API key from the list, triggers deletion, confirms in a modal, and receives a success toast. Steps: Overview — Integrations list visible → Select API key → Confirm deletion modal → Success feedback **Fernand: API Key Creation** (web, 5 screens, Business, Communication Tools) [Refero](https://refero.design/flows/11904) > An authenticated workspace user opens the Integrations API keys settings, enters a name for a new key, and creates it. The system displays the generated token in a one-time confirmation dialog, where the user is instructed to copy and save it. After confirming that the key has been saved, the dialog closes and the key appears in the API key table. The user then creates or views additional named keys, which are listed with masked tokens, usage status, creation time, and deletion actions. Steps: API Key Setup → Name the Key → Copy Generated Token → Verify Created Key → Manage Multiple Keys **Exa: Service Key Management** (web, 2 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/12031) > This is a team-level API key management screen. The user enters from the left navigation on the API Keys section, with the Service keys tab selected. In the first state, a service account API key has already been created and is shown in a table with its name, masked secret key, and creation time. In the second state, the list is empty after deletion and the screen shows an empty state with a Create Key call to action. The user is assumed to be logged into a team workspace with permission to view, create, and delete service keys on a desktop browser. Steps: Existing Service Key → Empty Service Keys State **ElevenLabs: API Key Deletion** (web, 4 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/8712) > Happy path: A logged-in developer navigates to the Developers > API Keys tab, chooses an existing API key from the list, opens the key actions menu, selects Delete, confirms the irreversible deletion in a modal, and returns to an empty keys state with a success toast. Steps: API Keys List → Open Key Actions → Confirm Deletion → Empty State and Success Notification **Manus: Delete API Key** (web, 4 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/10487) > Happy path for removing an API key from the Integrations → API section. Entry: signed-in user with access to Integrations > API (desktop). Exit: API key removed from the list and user receives a success toast. Steps: API keys list → Open key actions → Confirm deletion → Deletion result ##### Task: CRM contacts pipeline **Genie: Contact Search Add** (ios, 2 screens, Social Networking, AI Tool) [Refero](https://refero.design/flows/12420) > A user searches for people by name and adds them from a contact search list. Entry is a search screen with the keyboard open and a query already entered (“Dan”). The user can scroll through matching results and tap Add next to a person. After tapping, the action changes to Sent for that person, indicating the request/invite was sent. Steps: Search Results → Updated Results **Lovable: Missing Connector Request** (web, 4 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12342) > Happy-path for requesting a missing connector from the Connectors directory. The user starts in the Connectors section, filtered to the Messaging category, and uses the ‘Missing a connector?’ form in the lower-left panel to request an integration. The visible state assumes the user is already signed in and has access to the connectors catalog. After entering a connector name and submitting, the form confirms the request was received and disables further submission. Steps: Browse Messaging Connectors → Enter Connector Request → Submit Request → Request Confirmed **WhatsApp: Invite Contacts to WhatsApp** (ios, 4 screens, Social Networking, Collaboration Tools) [Refero](https://refero.design/flows/12603) > Happy path for inviting a contact to WhatsApp from a contacts list. The user enters the invite screen from within WhatsApp, browses or searches contacts, taps Invite for a contact who is not on WhatsApp, confirms sending an SMS invite in a bottom sheet, and receives an invite sent confirmation. Steps: Contact List → Search Contacts → Invite Confirmation → Invite Sent **Acctual: Select or Add Invoice Recipient** (web, 2 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8833) > Happy path: From the Invoices area the user opens a Create invoice modal to choose who the invoice is for. The modal offers a searchable field for contacts (and companies) and a live invoice preview at the right. The user searches for an existing contact or types a new identifier (phone/email/company) and is offered a one-click suggestion to add that entry as a customer. After selecting or creating the recipient the user continues with Next to fill invoice line items. Steps: Open Create invoice / Choose recipient → Type to add new contact suggestion **Sora: Cameo access recipient selector** (ios, 4 screens, AI Tool, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/7824) > Mobile modal for granting 'Cameo access' to other users. Entry: user opens the Cameo access modal (overlay) while signed in on a mobile device. The modal contains a searchable list of accounts (avatar, username, follower count) and a text field labeled 'Find people...'. The user can type to search (handles, names, numbers). Exit: user confirms selections by tapping Done, which closes the modal. Assumptions: user is authenticated and allowed to manage cameo access; device is a smartphone (iOS-style modal and keyboard behavior). Steps: Open Cameo access → Search — no results → Search — name results → Select recipients and confirm **222: Assign Close Friends** (ios, 11 screens, Lifestyle, Dating & Relationships) [Refero](https://refero.design/flows/9320) > Happy path: The user refines the app's compatibility model by answering a series of persona-style prompts and assigning a real friend to each prompt. From the Close Friends screen the user taps a prompt's "choose a friend" input, picks or filters a contact from their phone list, and triggers an in-app share (pre-populated message with a referral/preview) to invite that friend to take the compatibility test. A confirmation modal appears before the assignment is finalized. Each confirmed assignment increments a progress counter (e.g., 0/5 → 1/5). Steps: Close Friends — prompt overview → Contact list with keyboard → Filter to a specific contact → Select contact to send referral → Message compose with referral preview → Confirmation modal (explain & confirm) → Next prompt — pick another friend → Message compose for second contact → Progressed question cards → Close Friends summary → Return to People tab / flow complete **Appwrite: Enterprise contact and confirmation** (web, 3 screens, Development Tools) [Refero](https://refero.design/flows/6018) > Desktop two-column enterprise landing where the left column contains marketing copy and the right column contains a lead-capture contact form for enterprise plan inquiries. Steps: Enterprise lead capture → Form completion and submit → Submission confirmation **Programa: Add Supplier Contact** (web, 4 screens, Business, Project Management) [Refero](https://refero.design/flows/13445) > A logged-in user is viewing John Doe’s supplier record in the Address Book and wants to add a contact. From the supplier detail screen, the user opens the Create contact panel, enters the contact’s name, role, email, and phone number, then submits the form. The panel closes, the new contact appears in the Contacts table, and a confirmation message indicates that the contact was added. Steps: Supplier Record → Contact Form → Completed Contact Form → Contact Added **Fernand: Conversation Detail** (web, 3 screens, Business, Communication Tools) [Refero](https://refero.design/flows/11888) > A logged-in support agent views an existing conversation thread for the contact Fernand. The main content shows the message history, the right sidebar shows contact details, tags, and related conversations, and the bottom composer allows replying or adding a note. The entry point is an opened conversation. The exit point is after the user replies, adds a note, copies contact details, marks the conversation done, or navigates away. This assumes desktop use with permission to view and respond to the conversation. Steps: Open Conversation → Copy Contact Email → Copy Confirmation **222: People Selection for Itinerary** (ios, 4 screens, Lifestyle, Dating & Relationships) [Refero](https://refero.design/flows/9302) > Happy-path: User chooses people to bring to a specific itinerary card. Steps: Itinerary card (choose friend) → Contacts picker → Duplicate selection modal → Itinerary card (selection confirmed) **WhatsApp: Contact Details Search** (ios, 4 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12666) > This is a contact details and in-chat search experience for a specific conversation with Samuel. The user enters from the contact info screen, can review conversation-related actions and metadata, then open the chat search field to find messages by keyword or number. The flow assumes the user is already inside a chat/contact context on a mobile device. Entry is from Contact info; exit is either back to the contact details screen or to a filtered result within the conversation search. **WhatsApp: Invite Contacts** (ios, 3 screens, Social Networking, Collaboration Tools) [Refero](https://refero.design/flows/12602) > A mobile WhatsApp contact-invite screen where the user can search through contacts and invite someone to join WhatsApp. Entry is from the app’s invite contacts area; exit is by tapping back, closing search, or selecting an invite action. Steps: Contact List → Search With No Match → Search Results **WhatsApp: Share Contacts in Chat** (ios, 6 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12656) > This is a messaging conversation with Samuel. The user opens the attachment menu from the message composer, chooses to share contacts, selects multiple contacts from a searchable contact list, reviews the chosen contacts and their phone numbers, then sends them back into the chat as a contact card message. **komoot: Friend Discovery & Invitations** (ios, 3 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8396) > Mobile friend-management flow where a user discovers other users and invites contacts. Steps: Manage Friends — Find Friends → Contacts List — Inviteable Contacts → Contacts — Invited Confirmation **WhatsApp: Call Search and History Lookup** (ios, 2 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12682) > A user opens the Calls section and uses the search field to find a person named Samuel. The entry point is the Calls tab with recent call history visible; the exit point is a filtered view showing matching contact and call records, with quick actions to place an audio or video call. Steps: Calls Home → Search Results **WhatsApp: Add New Contact in WhatsApp** (ios, 5 screens, Social Networking, Collaboration Tools) [Refero](https://refero.design/flows/12618) > Happy-path flow for creating a new contact from WhatsApp’s New chat screen. Steps: New Chat Menu → Contacts Information Prompt → New Contact Entry → Saved Contact Details → Updated Contact List **WhatsApp: Add WhatsApp Contact** (ios, 7 screens, Social Networking, Collaboration Tools) [Refero](https://refero.design/flows/12606) > Happy path for adding a WhatsApp contact from the Contacts screen using either the add menu or QR scan. Steps: Contacts Overview → Add Options Menu → QR Scanner Prompt → Scanner Ready → QR Code Detected → Account Found → Contacts Updated **Bump: Friend discovery and invite** (ios, 2 screens, Map & Navigation, Social Networking) [Refero](https://refero.design/flows/3180) > Onboarding step for populating a personal map with friends. Steps: Search and list contacts → Invites reflected and review **Tilt: Contact tagging selector** (ios, 2 screens, Social Networking) [Refero](https://refero.design/flows/3070) > Small-screen tagging interface for selecting contacts to tag in a video. Steps: Search contacts list → No-results numeric search **Bond: People Search and Requests** (ios, 2 screens, AI Tool, Social Networking) [Refero](https://refero.design/flows/14384) > A logged-in user searches for people by entering “Sam” in the people search field. Matching profiles appear in a vertically scrollable list, each with an avatar, name, and Connect button. The user sends connection requests to selected results; those buttons change from blue “Connect” to black “Requested” while the search query and keyboard remain visible. The flow starts in people search with the keyboard active and ends with selected requests submitted. No device or permission changes are visible. Steps: Search Results → Request Confirmation **PayPal: Contact search & selection** (ios, 3 screens, Finance) [Refero](https://refero.design/flows/7238) > User searches for and selects a contact from a mobile contact picker that combines the device's saved contacts and PayPal users. Entry: user opens the contact search (search field focused, keyboard visible). Happy path: user types a query (name, username, email, or phone) → the list filters and displays matches under sections ("Your contacts" then "People on PayPal") → user taps a contact or the info icon to proceed. Exit: contact is selected (or user navigates back) and the picker closes. Assumptions: user is signed in, has granted access to contacts, and is on a mobile device with an on-screen keyboard. Steps: Search field focused with local contacts → Search by phone — no matches → Search by name — filtered results **Mozi: Find and Contact Person** (ios, 3 screens, Social Networking, Travel & Booking) [Refero](https://refero.design/flows/4153) > Happy path: User opens the My People tab (bottom nav) to find and contact a person from the app's synced contacts. Entry: user taps the My People tab. The screen shows a location scope (Bologna, Italy), a list of pulled contacts (with photo, name, location and a star), quick actions (call, message), and an invite call-to-action to share the user's QR code. The user can open the search field (top-right search icon) to reveal a search input and keyboard, type a name, and then select the matching contact or use the quick actions. Exit points: initiate a call or message, open the contact's detail/profile, open the QR invite, or return to other tabs. Assumptions: user is logged in, running the app in portrait on a mobile device (iOS styling), and the app has permission to read the device contacts (text on screen indicates contacts are pulled from the contacts app). Steps: My People (list & quick actions) → Search Activated (keyboard visible) → Search Results (matching contact) **Tilt: Invite and Add Friends** (ios, 2 screens, Social Networking) [Refero](https://refero.design/flows/3078) > Mobile onboarding/utility screen where a signed-in user populates their map with friends by searching, adding, or inviting contacts. Steps: Invite Summary / Search placeholder → Search Results and Add Contacts **Mozi: Contact search and invite** (ios, 3 screens, Social Networking, Travel & Booking) [Refero](https://refero.design/flows/4152) > Happy path: a logged-in user lands on the My People screen (location: Bologna). They can view connected contacts pulled from their device contacts, call or message a shown contact, or open search to find a specific person. If the search returns no connected user, the UI presents an explanation and CTAs to invite friends (QR or an invite button). Entry: My People tab. Exit: either initiating a call/message or opening the invite/share flow. Steps: My People list → Search input with recent results → No results / Invite prompt **WhatsApp: Calls Onboarding** (ios, 3 screens, Social Networking, Collaboration Tools) [Refero](https://refero.design/flows/12600) > A Calls tab onboarding screen shown in a logged-in mobile messaging app. The user can swipe through three informational cards and use the primary action to start a call or secondary action to invite a friend. The entry point is the Calls tab from the bottom navigation, and the exit point is likely the next call setup or invite screen after tapping an action. Steps: Private Calls Intro → Group Call Intro → Effects Intro **Calendly: Meeting Preparation and Contact Lookup** (web, 8 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/14054) > A logged-in user opens the Calendar section on a desktop workspace and asks Callie to prepare for an upcoming Group Meeting. Callie analyzes the scheduled meeting and available calendar or CRM context, presents preparation notes in a side panel, and lets the user expand the response for a larger reading view. The user then asks who John is, and Callie returns the associated CRM contact details. The flow starts in the Calendar view with the Callie assistant available and ends with a contact-information response in the expanded assistant view. **Resend: Create contact property** (web, 4 screens, Development Tools, Communication Tools) [Refero](https://refero.design/flows/10630) > Happy path for creating a custom contact property in Audience → Properties. Steps: Properties empty state → Add property modal (blank) → Fill property form → Properties list with new item **Chargetrip: Contact sales lead capture** (web, 3 screens, Development Tools, Map & Navigation, Transport) [Refero](https://refero.design/flows/9254) > A desktop contact-sales experience that collects a prospect's details and offers a secondary demo booking CTA. Entry is the "Contact sales" entry point in the top navigation (user is likely anonymous / not signed in). The main region is a two-column layout: a large marketing headline on the left and a contact form on the right. Below the fold is a demo booking call-to-action and the site footer with newsletter subscribe. Happy path: user fills required fields in the contact form and submits -> form shows an inline success state. Exit point: form confirmation displayed (no page navigation visible). Assumptions: desktop/laptop viewport, user has network connectivity, user is not authenticated (Sign in button visible). Steps: Contact form — initial view → Contact form — filled → Submission confirmation state **Linktree: Audience Contact Filtering** (web, 4 screens, Business, Social Networking) [Refero](https://refero.design/flows/12406) > Logged-in user is on the Audience section of a creator dashboard. The screen shows an audience/contact table with summary metrics, search and filter controls, and related growth tools below. The user opens the Filter menu, chooses a filter category, and applies a Tool type filter. The table then updates to show only matching contacts, and the active filter appears as a removable chip. Steps: Audience Overview → Open Filter Menu → Select Filter Category → Applied Filter Results **Calendly: Create Contact** (web, 7 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/14032) > A logged-in user starts in the Contacts section, opens the add-contact menu, and chooses to create a single contact. They complete the contact form with a profile image, name, email, phone number, job title, and company, crop and apply the uploaded image, then save the record. The flow ends on the newly created contact’s detail view, where the user can review information or schedule a meeting and send an email. The user appears to have permission to create contacts and is using a desktop layout. **The Leap: Bulk contact import** (web, 6 screens, Business, Entertainment) [Refero](https://refero.design/flows/4562) > Happy-path: A merchant on the Contacts screen opens the Import contacts modal, uploads a CSV (Email, Name), starts the import, and the UI shows upload progress and a confirmation toast. Steps: Contacts list (entry) → Import modal - instructions → File selected → Uploading → Import queued - confirmation → Final contacts list and quota update **Rox: Add Contacts to Campaign** (web, 7 screens, Business, Productivity, AI Tool) [Refero](https://refero.design/flows/10815) > Happy path: A logged-in user on the Discover tab filters contacts (Contact Data Source = RocketReach, Contact Name = "scott"), selects multiple contacts from the results, uses the bottom Manage control to add those selected contacts into an existing campaign, and is taken to the selected campaign where a toast confirms the addition. Entry: user is on Discover with results loaded and filters applied. Exit: user is on the Campaign overview for the selected campaign and receives a confirmation message. Steps: Discover - Filtered Results → Select Multiple Contacts → Open Manage Menu → Select Campaign Modal (open) → Campaign Dropdown - Choose Campaign → Confirm Add Contacts → Campaign Overview & Confirmation **ManyChat: Add Contact via Contacts List** (web, 6 screens, Business, Productivity, Communication Tools) [Refero](https://refero.design/flows/10349) > Happy path: A logged-in user opens the Contacts list, clicks Create New Contact, fills required fields (phone or email plus name), selects gender, checks the consent checkbox and submits. The system creates the contact, opens a contact details modal showing system fields and opt-in channels, displays a success toast, and the new contact appears in the contacts table. Steps: Contacts List → Create New Contact — Empty Form → Select Additional Attributes → Confirm Consent → Contact Created — Details Modal → Contacts List Updated **Factory: Contact Sales Lead Capture** (web, 4 screens, Business, Development Tools, AI Tool) [Refero](https://refero.design/flows/7716) > Happy path: A prospective customer arrives on the 'Why Factory' marketing page, fills out the 'Contact us' lead form (first name, last name, company, work email, and a "How did you find us" dropdown), and submits it. The page validates the inputs client-side (visual state changes shown) and returns an inline success panel that confirms receipt. Steps: Landing with Contact Form → User Fills Fields & Opens Dropdown → Select Lead Source and Ready to Submit → Submission Confirmation **Cursor: Enterprise sales contact form** (web, 5 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9915) > Multi-step enterprise sales contact form presented alongside marketing content. Steps: Initial contact prompt → Qualification completed (pre-fill) → Detailed contact information → Filled form review → Submission confirmation **Reown: Contact lead-capture and confirmation** (web, 5 screens, Development Tools, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/5699) > Desktop contact flow on a marketing site. Steps: Contact section (landing) → Contact form — initial modal → Inline validation triggered → Corrected form and successful submit → Submission confirmation **Rox: New Opportunity - add associated people** (web, 4 screens, Business, Productivity, AI Tool) [Refero](https://refero.design/flows/10784) > Happy path: A user opens a right-hand 'New Opportunity' side panel tied to an account (OpenAI). The panel contains pre-filled metadata (owner, stage, company, close date, amount) and editable fields including an 'Associated People' picker. The user opens the people picker, types to filter contacts, selects one or more people (they become chips), then can either Cancel or Create the opportunity. Steps: Open New Opportunity panel → Open people suggestion list → Type to filter contacts → Select contacts and show chips **PayPal: Sales Contact Form Submission** (web, 3 screens, Finance) [Refero](https://refero.design/flows/7375) > Happy path: A visitor lands on PayPal's "Get in touch with sales" page (desktop layout). The page displays a multi-field lead form on the right and brief instructions/contact phone on the left. The user fills required fields (contact + business details), selects dropdown values (country, estimated annual sales, existing account, solution needed), checks the marketing consent checkbox, and submits the form. After submission the user is routed to a confirmation page that thanks them, offers related business resources, and surfaces a CTA to create a business account. Steps: Sales contact entry screen → Form review and submit → Submission confirmation and next steps **Runey: Add Customer** (web, 5 screens, Business, Productivity, Project Management) [Refero](https://refero.design/flows/11684) > A logged-in user is on the Customers section and starts from an empty state. They open the New Customer form, enter customer details (including optional avatar upload and contact/business information), and submit the form. After saving, the screen returns to the Customers overview with the new customer listed and summary counts updated. Steps: Customers Overview → New Customer Details → Complete Customer Form → Submit Customer → Customer Added Confirmation **Clay: Nexus contact recommendation** (web, 2 screens, Productivity) [Refero](https://refero.design/flows/5884) > Happy path: A signed-in user opens the 'Nexus' conversational assistant inside their workspace, asks who to invite to a designers' event in Spain, receives a reply indicating no exact matches in the user's contacts, and is presented with suggested alternatives and follow-up options. Steps: Compose question → Assistant reply and follow-up options ##### Task: access review permissions audit log **TIDAL: Track Report Reason Selection** (ios, 3 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/13140) > A mobile modal for reporting a track, starting with a searchable list of report reasons and ending with a selected reason shown after search. The user opens the report dialog, searches or browses categories, and selects the matching reason. Steps: Browse Report Reasons → No Match Search State → Filtered Result **Apple Image Playground: Report Generated Image** (ios, 9 screens, AI Tool, Design & Media Editing) [Refero](https://refero.design/flows/12798) > A user reviews a generated image in Image Playground, opens the overflow menu, and chooses Report a Concern. They select concern categories, optionally add more information, review what will be included in the report, accept the data-sharing prompt, and submit. The report then submits successfully and returns them to the image editor, where the generated image remains visible with the usual controls. Steps: Image Review → Action Menu → Concern Category → Personal Characteristics Check → More Information → Submit Report → Data Consent → Submitting → Return to Editor **Granola: Share Note with Editor Access** (ios, 8 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14162) > The user starts on a note titled “Dashboard design and analytics” and opens its sharing controls. From the sharing panel, they choose to add people, search for a contact by name, select the matching attendee, and change the access role from Viewer to Editor. They return to the sharing summary, where the selected person is listed as having access. The flow is shown on a mobile device. The user is assumed to be signed in and authorized to share the note; no confirmation or notification result is visible after returning to the summary. Steps: Open Note Sharing → Review Sharing Options → Open Recipient Selector → Search for Contact → Select Contact → Open Permission Options → Assign Editor Access → Review Granted Access **TikTok: Activity Status Visibility** (ios, 2 screens, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/2798) > Happy path: A logged‑in mobile user opens the Inbox to review notifications and then opens the activity status information modal to review or change who can see their active/last‑active presence. Steps: Inbox / Notifications → Activity Status Modal **Product Hunt: Product Team Access Request** (web, 3 screens, Business, Social Networking) [Refero](https://refero.design/flows/11461) > A logged-in user views Notion’s Product Hunt profile on the Team tab and requests access to be added to the product team. Because the product page already has an owner, the system opens an access-request modal. After the user submits the request, the modal confirms that the request was sent and the team section changes to indicate a pending request. The flow begins on the product team profile and ends with the request awaiting owner approval. Steps: Notion Team Profile → Access Request Prompt → Request Confirmation **District: Event Feedback Submission** (ios, 5 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12916) > Happy-path mobile feedback journey after attending an event. The user enters a rating screen for a specific event ('Sip and Act at 3 Art House'), selects an overall sentiment, optionally rates individual aspects, adds a text comment, and submits. Steps: Overall Rating → Detailed Ratings → Comment Entry → Submission Confirmation → Completion State **Cursor: Security Audit Assistant** (web, 18 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9922) > Happy path: From the Agents landing the user composes a task card, selects a GitHub repository, configures model(s) and multipliers, optionally attaches an image, then starts the task which spins up a cloud agent. The agent appears in the left agent list and the UI navigates to a three-column agent workspace showing the activity log (center) and a file-changes pane (right). The agent runs commands (visible as generated commands / progress) and produces a final security audit report with findings and code snippets. Steps: Agents Landing — Task Composer → Repository Picker Open → Repository Search → Repository Selected → Open Model Selector → Enable Multiple Models → Set Model Multipliers → Compose Task Instruction → Attach Image Attempt → Image Preview Modal → Task Created — Agent Listed → Open Agent Conversation → Agent Starting — Generating Commands → CLI Execution Attempt → Agent Working — Additional Steps → Audit Report — Commits Reviewed → Critical Finding Detail → Final Summary and Checks **Podimo: Podcast Comment Moderation** (ios, 3 screens, Entertainment) [Refero](https://refero.design/flows/12274) > Happy path in a podcast comment thread on mobile. The user is already viewing a specific episode’s comments, can read existing replies, post a new comment, and open the report action for a comment. After submitting the report, a confirmation sheet appears with a success message, a link to community guidelines, and a Done button to dismiss it. Steps: Comment Thread → New Comment Posted → Report Confirmation **Factory: Repository Security Review** (web, 22 screens, Business, Development Tools, AI Tool) [Refero](https://refero.design/flows/7751) > Happy path: A user starts a knowledge/code assistant session, selects an assistant type and remote workspace/repository, composes a request to review code for security, configures model and toolkits, and sends the request. The assistant connects to the remote repository, runs exploratory commands (git pull, listing files), opens and previews configuration and workflow files, generates a todo checklist, and produces a security-review report which is displayed in a split preview/context panel. Entry: user landing on session start. Exit: rendered security-review-report and relevant context files available in the session. Assumptions: user is authenticated (Sofia Delgado shown), has a cloud workspace/repository available, and has permission to view repository contents remotely. Steps: Session landing → Select assistant → Confirm workspace and repo → Attach/upload indicator → Compose request → Send request / open options → Select model and reasoning level → Enable toolkits → Connecting to repository → Repository inspection begins → Globbing and file listing → Open small file preview → Generate Todo checklist → Open workflow YAML → View README.md content → Shell listing and report file created → Report: executive summary and urgent actions → Split view with report preview → Context panel and uploaded files → Browse codebase overview document **Kin: Journaling Draft Review** (ios, 4 screens, AI Tool) [Refero](https://refero.design/flows/5438) > Happy path: A signed-in user in the Journaling area sees a banner indicating pending drafts, a chronological list of journal entries, and a floating control to create new entries (sketch or record). From the Journaling screen the user opens the Drafts list, reviews a recorded draft card, chooses Delete which opens a confirmation modal, and either cancels (Keep) or confirms deletion. Exit point returns the user to the Journaling list with the draft removed and any draft-count banner updated. Steps: Journaling Home → Drafts List → Delete Confirmation Modal → Journaling Home (Updated) **Meta AI: Gmail Connection Authorization** (ios, 6 screens, AI Tool) [Refero](https://refero.design/flows/13355) > A logged-in user asks an AI assistant to identify active subscriptions and assess which are useful. Because Gmail is not connected, the assistant explains that email access is required and presents Gmail and Outlook connection options. The user selects Gmail, reviews Meta’s connection disclosure, continues to Google sign-in, reviews the requested Gmail permissions, and returns to the assistant with Gmail shown as connected. The depicted flow ends before the subscription audit results are displayed. ##### Task: apply for loan credit **Introspect: Friend decision response** (ios, 3 screens, Finance) [Refero](https://refero.design/flows/11952) > Happy path for a recipient responding to a friend’s product decision request on mobile. Steps: Decision Request → Sending State → Completed State **Airbnb: Artist experience request** (ios, 11 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6413) > Happy path for requesting entry to a limited/lottery artist event. Steps: Event detail / Request CTA → Guests selection → Application question prompt → Text entry (keyboard active) → Answer review → Request details summary → Before you submit (rules & process) → Submission (start animation) → Submission animation sequence → Confirmation screen → Exit / Return **Turf: Send Friend Request** (ios, 2 screens, Sports) [Refero](https://refero.design/flows/9425) > Happy-path for sending a friend request from another user's profile. Entry: user opens someone else's profile (mobile, logged-in). The profile shows header (avatar, name, username), two primary actions (Add friend / Taunt), a Stats card (games, wins, high score) and an Activity list of past games and picks. Exit: the Add friend control transitions to a Requested state indicating a pending friend request. Steps: Profile (not friends) → Profile (request sent) **Lovable: Start Project Prompt** (web, 6 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12303) > A logged-in user lands on the Lovable home screen with a left sidebar, a central prompt composer, and a template gallery below. The user can start by typing or using voice input in the composer, optionally connect tools, then submit the prompt to generate a project. The flow ends when the prompt is fully entered and ready to build/submit, with the system showing the composed request and available build controls. Steps: Home Prompt Start → Template Inspiration → Prompt Refine → Voice Capture → Transcribing State → Prompt Ready to Build ##### Task: approval workflow **District: Closed Support Rating** (ios, 2 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12841) > A mobile support chat ends after the user’s issue is resolved and the conversation is marked closed. The entry point is an active support conversation with recent agent messages. The exit point is a closed chat state where the user can rate the assistance or start a new chat if they still need help. **Introspect: Item Verification Share** (ios, 4 screens, Finance) [Refero](https://refero.design/flows/11951) > Happy path begins on an Item Overview screen for a saved item, with the user logged in on a mobile device. From here, the user opens the item’s overflow menu, chooses to verify/share the item, and is taken into the system share sheet. The user then selects Messages, which opens a new iMessage draft addressed to a contact with a prefilled prompt and link preview. The flow exits at message composition, ready to send. Steps: Item Overview → Item Actions Menu → Share Sheet → Message Draft **Lovable: QR Asset Label Implementation** (web, 12 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12336) > The user is working in an authenticated project-building workspace with an existing AssetWise asset-tracking project open in preview. They ask the assistant what to do next, select a suggested QR-code labeling feature, and submit the request. The assistant generates a detailed implementation plan covering QR codes, public scan pages, printable labels, files, and out-of-scope items. The user reviews the plan, opens its editing controls, and approves it. The system then processes the approved plan and returns a completed feature summary indicating that QR labels and a public scan page were added. Steps: Choose the Next Project Step → Select a Feature Request → Generate the Plan → View Plan Generation → Review the Proposed Plan → Inspect Technical Details → Open Plan Editing → Format Plan Content → Finalize Plan Review → Save and Approve → Process Approved Changes → Confirm Feature Added **Bond: Restaurant Recommendation Approval** (ios, 2 screens, AI Tool, Social Networking) [Refero](https://refero.design/flows/14397) > The user reviews an AI-generated restaurant recommendation for “Best food to get right now.” On the initial review screen, the user can inspect the restaurant details, copy the recommendation prompt, approve the suggestion, or regenerate it. After approval, the interface updates the primary action to “Finalize” and displays a small positive-feedback reaction from the associated profile. The user can then finalize the approved recommendation. The flow begins with a generated recommendation awaiting review and ends with the recommendation prepared for finalization. The user is assumed to be viewing the recommendation on a mobile device with access to the relevant profile or memory-based context; login state and permissions are not visible. Steps: Recommendation Review → Approved Recommendation **pillowtalk: Pillowtalk Onboarding** (ios, 3 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13494) > An unauthenticated user is presented with a three-screen onboarding carousel introducing Pillowtalk’s purpose and benefits. The user can review the messages using the carousel, accept the Terms of Service and Privacy Policy, and then continue with Apple. The screenshots show the first two carousel states with consent unchecked and the third state with consent selected, enabling the Apple continuation button. Steps: Process Difficult Experiences → Decode Recent Texts → Accept Terms and Continue **XChat: Message request acceptance** (ios, 2 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/11664) > Two-step mobile messaging sequence for handling an incoming message request from another user. Steps: Message Request (pending) → Accepted Conversation **Structured: Structured AI suggestions review** (ios, 3 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3289) > Happy path: The user views Structured AI's suggested plans for their day, can accept all suggestions at once, or give optional feedback about the suggestions. Steps: Review suggestions and Accept All → Open feedback modal → Enter and submit feedback **WhatsApp: Announcement Replies Thread** (ios, 3 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12679) > This is a mobile reply thread for an announcement message. The user starts with the reply panel open, sees the original announcement from Samuel, types a response in the bottom input, and sends it. After submission, the new reply appears in the thread with the sender’s name/avatar and timestamp. The user remains in the replies view and can close it with the X button. **Programa: Approve Product Item** (web, 4 screens, Project Management, Business) [Refero](https://refero.design/flows/13428) > The user opens a product item from a project schedule, reviews its details, and changes its status from Draft to Approved. The product detail drawer remains open throughout the interaction. The flow starts with the item details visible and ends when the Approved status is applied to the item. The user is assumed to have permission to change item statuses. **Acctual: Manage approval rules (delete)** (web, 4 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8854) > Happy path: A signed-in user navigates to Settings → Approvals to remove an existing approval rule. From the approvals list the user opens an overflow menu on a rule, chooses Delete, confirms the deletion, and the rule is removed from the list with a success toast. Steps: Approvals list → Open rule actions → Confirm deletion → Rule removed and feedback **Aboard: Vacation plan approval** (web, 7 screens, Collaboration Tools, Paperwork) [Refero](https://refero.design/flows/828) > A manager-facing desktop view of a vacation planning tool. Steps: Plan in Draft → Plan Active – waiting for submissions → Select employee details → Mark reportees ready → Ready state overview → Confirm completion → Plan Done **Aboard: Document Review and Approval** (web, 3 screens, Collaboration Tools, Paperwork) [Refero](https://refero.design/flows/834) > Two-column document review interface. Entry: user lands on Documents area (left column lists documents grouped under categories; right column is a preview area). Happy path: user selects a document from the left list -> the document opens in an embedded PDF viewer with thumbnails and toolbar -> an inline confirmation message and Approve button appear when the document is ready to approve -> user can select a different document from the list to open it. Assumption: user is signed in and has access to the Documents section; screenshots are on a desktop/laptop viewport. Steps: Document list — select document → Embedded PDF viewer with approval CTA → Switch to another document **Willow: Voice Prompt Drafting** (ios, 6 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/13231) > A mobile onboarding interaction for generating a written prompt by voice. The user starts on a screen titled 'Prompt AI to write for you,' with a back button, skip option, progress indicator, and a large prompt area. The user can tap the microphone to speak a request, review the transcribed prompt, and confirm it. The flow assumes the user is on a phone, likely logged in, and has access to voice input. It ends when the drafted prompt is generated and ready to continue or be edited further. Steps: Prompt Entry → Voice Guidance → Voice Capture → Transcription Progress → Draft Review → Confirm Draft **Poppy: Voice Event Creation** (ios, 9 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12186) > A voice assistant-style interaction starts on a home screen with shortcut buttons and a microphone control. The user initiates voice input, grants required speech recognition and microphone permissions, дикtates an event request, reviews the interpreted event details, and then confirms creation. The flow ends with the created event summarized as a tappable item. Assumption: the user is on a mobile device and not previously denied the needed permissions. Steps: Home Dashboard → Voice Prompt → Speech Recognition Permission → Microphone Permission → Listening State → Speech Transcription → Request Refinement → Event Suggestion → Created Event Summary **WhatsApp: AI Image Restyling** (ios, 8 screens, Social Networking, Collaboration Tools) [Refero](https://refero.design/flows/12613) > The user starts with an image created in Meta AI and opens the editing tools. They browse preset categories, select a lighting treatment, review the updated image, then select an illustrated style and review the final result. The user can cancel or complete editing at any point. The flow assumes an existing generated image and a mobile device; no account or permission state is visible. Steps: Generated Image Preview → Browse Featured Presets → Browse Style Presets → Select Lighting Category → Review Ethereal Lighting → Select Illustrated Style → Review Illustrated Result → Confirm Updated Image **Sora: Onboarding to First Feed** (ios, 45 screens, AI Tool, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/7819) > Happy path onboarding and first‑time setup for Sora. Entry: user opens the app and sees the branded landing screen. Primary choices offered: continue with Apple, continue with Google, sign up, or log in. The user selects Apple Sign In, completes system authentication (Face ID), then is routed through required setup screens: enter birthday (date picker), handle invite‑only gating (enter code or request notification), allow notifications, set profile visibility and pronouns, write cameo preferences, accept content advisory, and finally land on the main feed where a video is visible. Assumptions: user is on an iOS device with Face ID and system-level Apple Sign In available; user is not yet registered and is completing first‑time setup. Steps: Welcome / Auth choices → Sign in with Apple sheet → Face ID authorization → Sign-in completion → Enter your birthday (picker opened) → Birthday confirmed → Invite-only code entry → Notifications permission prompt → Profile privacy / Cameo settings (entry) → Pronouns selection popover → Custom pronoun input → Open Cameo preferences → Enter cameo instructions → Content advisory → Feed initial load (skeleton) → Feed with first video → Video interactions → Primary navigation available → Creation entry point → End state - active feed **ChatGPT: Assistant Response Feedback** (ios, 2 screens, AI Tool) [Refero](https://refero.design/flows/7008) > User is inside a chat conversation on a mobile device (iOS-style status bar visible). The assistant message displays usage tips (bulleted list) and ends with an explicit offer: “Do you want me to make that for you?”. Below that message are inline action controls (copy, listen, thumbs up, thumbs down, regenerate, share) and the message input area. Happy path: the user taps the thumbs-up control to give positive feedback; the UI shows a transient confirmation toast and highlights the thumbs-up. Steps: Assistant message with CTA and actions → Feedback confirmation **Manus: Edit Manus service email** (web, 4 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/10484) > Happy-path: A signed-in admin visits the Mail Manus settings to change the service email used to create tasks. From the Mail Manus settings panel they open an Edit dialog, modify the email prefix, save the change, and receive a success confirmation while the updated address appears in the settings list. Entry: user is on account settings (left nav) and has Mail Manus selected. Exit: updated Manus email is saved and visible; transient success toast is shown. Assumptions: user is authenticated and has permission to change account-level settings; device is a desktop or similar (modal centered, large layout). Steps: Mail Manus settings → Edit Manus email (modal open) → Edit email - update prefix → Settings updated + success toast **Wrike: Task detail and status update** (web, 3 screens, Business, Productivity, Project Management, Collaboration Tools, Paperwork, Scheduling Tools) [Refero](https://refero.design/flows/1508) > Happy-path: A signed-in user working on a desktop is viewing a table/list of project items and opens a single item to inspect and update its details. Selecting the row for "Create New Infographic" opens a right-hand detail panel where the user reviews subtasks, approval and dependency indicators, and metadata (assignee, dates, budget, etc.). The user opens the Status control, chooses a different workflow state, and optionally expands additional fields to confirm metadata. Steps: Open item in detail panel → Open Status dropdown → Select new status and view metadata **MWM: Habit Tracker Product Creation** (web, 31 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/13331) > A logged-in user starts on the Refro home screen and describes a mobile habit-tracking product. They can attach reference images, select an agent, target platform, language model, reasoning level, and voice input. After submitting the detailed request, the system generates a product plan. The user reviews the plan, identifies or waits for revisions, approves the product direction and design system, selects an application icon, and proceeds to generate the mobile screens. The flow ends with generated screens displayed in a design workspace and a Build action available. Steps: Start Product Brief → Attach Reference Image → Add Visual References → Choose Agent Mode → Choose Target Platform → Choose Generation Model → Adjust Reasoning Level → Review Voice Input → Enter Detailed Brief → Review Expanded Request → Review Generated Product Plan → Resolve Plan Validation → Review Defined Screens → Review Design Direction → Approve Product Plan → Generate Mobile Screens → Review Generated Designs → Select Application Icon → Prepare Build → Open Build Stage **Melius: Image Comment Annotation** (web, 5 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/14076) > A logged-in user reviews a selected image asset in a desktop workspace, adds a location-specific comment directly on the image, submits it, and then views the saved comment in the asset’s comments panel. The flow starts with an image and its metadata open, and ends with the comment attached to the selected image location and available for follow-up actions such as reply, edit, or delete. The screenshots imply the user has permission to comment on the asset. No additional navigation, account setup, or device-specific behavior is shown. Steps: Asset Review → Comment Placement → Comment Composition → Comment Submission → Saved Comment Review **Programa: Bulk Product Status Update** (web, 5 screens, Project Management, Business) [Refero](https://refero.design/flows/13429) > A user manages product statuses within a project schedule. Starting from a product schedule summary, the user selects multiple product rows, opens the bulk Action menu, chooses Change statuses, and applies the Client review status. The selected items update together, while unselected items retain their existing statuses. The flow assumes the user is logged in and has permission to edit schedule items. **MWM: Generated Product Review** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/13338) > The flow shows a user monitoring an AI-assisted mobile product build, reviewing the generated design assets and screens, and arriving at a canvas state where the generated screens are available for further editing or addition. The user appears to be in a desktop design workspace and does not need to perform a visible form submission. The build starts in progress, becomes ready for review, and the canvas is then displayed with the generated mobile screens. Steps: Build Progress → Build Ready Review → Design Canvas Review **Navia: Replace Carousel Image** (web, 8 screens, Business, Productivity) [Refero](https://refero.design/flows/14317) > The user starts in an Instagram carousel-post review modal for a draft titled “Show efficient ways to research landing pages.” They open image options for a carousel slide, choose Unsplash as the image source, browse suggested photos, search for “ux ui,” wait for additional results, and select a suitable image. The selected image is then applied to the carousel slide, returning the user to the post review modal. The flow begins with an authenticated user who has access to edit the draft and ends with the slide image replaced or populated; completing or saving the post is not shown. Steps: Review Carousel Draft → Choose Image Source → Open Photo Picker → Browse Suggested Photos → Search Photo Results → Load More Results → Select Photo → Confirm Updated Slide **Navia: Carousel Task Review** (web, 4 screens, Business, Productivity) [Refero](https://refero.design/flows/14316) > A signed-in workspace user opens a task detail modal for an Instagram carousel post. The modal shows the post preview, metadata, and a right-side instruction panel describing how to create the carousel slides. The user opens an image action menu to inspect available asset options, then dismisses it and returns to the task view. The flow can end by editing the content, saving and closing the modal, marking the task complete, deleting the action, or closing the modal. No completed task action is shown in the provided screens. Steps: Task Detail Modal → Image Action Menu → Asset Options Review → Return to Task View **Navia: Update Carousel Slide** (web, 5 screens, Business, Productivity) [Refero](https://refero.design/flows/14315) > The user opens an Instagram carousel content task in a modal editor, reviews the provided instructions, chooses an image source from the image options menu, waits while the selected image is added, and verifies the updated carousel slide. The flow starts with an existing draft task and ends with the carousel content updated, while the user can save and close or mark the task complete. The user is assumed to be signed in with permission to edit the content. The device appears to be a desktop browser. The exact image source selected is not visible, although the screenshots show the image options menu and the subsequent image-loading state. Steps: Review Carousel Task → Open Image Options → Choose Image Source → Add Image → Verify Updated Slide **Melius: Reference-Based Product Image** (web, 13 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/14069) > A user starts in a dark visual canvas and uses the central composer to create a skincare product image. They open the attachment menu, select an image-generation skill, attach a reference image, add a descriptive prompt, and submit it. The system opens a chat and canvas workspace, creates a generation node, asks for approval, renders the result, and displays the completed product image. The flow starts from an empty canvas and ends with a generated image node. The user is assumed to have access to the canvas and generation tools; authentication, permissions, and device details are not visible. Steps: Empty Canvas → Attachment Discovery → Skill Menu → Generation Skill Selected → Attach Reference → Reference Added → Prompt Composition → Generation Started → Generation Proposal → Approval Request → Image Rendering → Generated Result → Completed Canvas **Wrike: Delete Task Attachment** (web, 2 screens, Business, Productivity, Project Management, Collaboration Tools, Paperwork, Scheduling Tools) [Refero](https://refero.design/flows/1510) > Happy path: A user viewing a task in a project list opens the task detail panel, selects an attached file, chooses the delete action, and is presented with a confirmation modal asking to permanently delete the file. Steps: Open task detail and select file → Confirm permanent deletion **Marathon: Task Detail — Status Update** (web, 3 screens, Business, Productivity, Project Management, Collaboration Tools, Scheduling Tools) [Refero](https://refero.design/flows/4849) > Happy path: a user opens a task/ticket to review details and metadata, then changes the ticket's status using an inline status picker. Entry: user navigates from a project or list into the task detail view (desktop, logged-in). Exit: the status tag updates in-line to the newly selected state and the task remains open for further review or edits. Assumptions: user is signed in with permission to view and update the task; desktop browser; the task detail view includes editable metadata (tags/status), description text, attachment control, and a rich-text/comment toolbar. Steps: Open Task Detail → Open Status Picker → Select New Status **Navia: Carousel Task Walkthrough** (web, 3 screens, Business, Productivity) [Refero](https://refero.design/flows/14314) > A guided walkthrough begins on an open Instagram carousel-post task. The user reviews the post preview, follows the instructions shown in the task panel, and learns where to edit, save, or mark the task complete. The walkthrough ends after the user acknowledges the final guidance. The user is assumed to be signed in, have access to the task, and be using a desktop-sized interface. No completion action is shown as being performed during these screenshots. Steps: Post Preview Guidance → Task Instructions → Completion Controls **Navia: Instagram Carousel Review** (web, 4 screens, Business, Productivity) [Refero](https://refero.design/flows/14318) > The flow begins with a draft Instagram carousel task opened in a modal. The user reviews the carousel preview and task instructions, opens the content editor, reviews or updates individual slide content, and returns to the task modal. The flow ends when the user saves and closes the editor or marks the task complete. The user is assumed to be signed in with permission to edit the assigned content. The screenshots show a desktop interface with a modal overlay and a right-side editing panel. Steps: Review Draft Task → Open Content Editor → Review Slide Content → Save or Complete Task ##### Task: book GP appointment **District: Salon Service Booking** (ios, 8 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12874) > Happy path for booking a salon service and reaching payment method selection. The user starts on a salon profile, chooses to book a service, selects one service and an available date/time slot, reviews the appointment details, and opens the payment method picker. **Fresha: Salon appointment booking** (ios, 9 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4973) > Happy-path booking of a single salon service with optional handling for fully-booked dates. Entry: user browsing/search results and opening a provider card. Exit: appointment is confirmed and shown in a booking detail screen. Assumptions: user is signed in (booking flows show payment and confirmation), has location enabled (distance shown), and is using a mobile device. Steps: Search results — provider card → Provider services list → Select services → Select time — fully booked state → Select time — date with availability → Choose time slot → Pre-visit question → Review and confirm → Confirmation & booking details **Fresha: Add Appointment to Calendar** (ios, 3 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4974) > Happy-path: a signed-in mobile user views a confirmed appointment and chooses to add it to a calendar. Steps: Appointment details / Confirmed → Calendar selection modal → Device event editor (prefilled) **Fresha: Refine search and pick time** (ios, 19 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4958) > User starts on a results list for a previously-entered query (example: "Barbering") and opens the search/filter modal to refine the search. From the modal the user can change the treatment (Treatment or venue), location, date and time, or browse categories/recents. The user drills into each filter: selects a treatment type, picks or types a new location (including using current location), selects a date from a calendar, and chooses a time window (preset or custom). After applying filters the modal returns to the results list which updates to show venues matching the query and filters with available time slots. Assumptions: mobile (iOS-style) device; user is either unauthenticated or authentication state is not relevant to the flow; user navigates via modal sheets that stack and return to the search overlay and then to the results list. Entry: results list (search active). Exit: updated results list with applied filters and visible available times. Steps: Results list (initial) → Search / Filter modal (root) → Treatment search (open) → Treatment search — no match state → Treatment results (matches and venues) → Location picker (root) → Location search — no match → Location search — suggestions → Search modal updated (location chosen) → Date picker (open) → Date selection (select month / date) → Time picker (open) → Time selection — preset and custom → Time selection — picker list → Search modal (filters applied preview) → Top categories (browse) → Results list (filtered) → Select time slot / Book → Back navigation and re-open search **Longevity Deck: Add activity to daily timeline** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/9895) > Happy path: From the day timeline (Yesterday) the user taps an empty timeslot (11) to add an activity. The app opens an "Add Activity at 11" picker with search, category filters, and a scrollable list of activities. The user picks an activity (optionally adjusting duration or amount in the inline numeric control) and taps the circular + control to add it. The picker closes (or the list dismisses) and the selected activity appears as a colored card in the 11 slot on the timeline. Entry: day timeline with empty slot. Exit: timeline updated with new activity card. Assumptions: user is signed in and on the daily timeline view, editing is allowed, device shown is a phone in dark mode. Steps: Day timeline with empty slot → Activity picker — top of list → Activity picker — select activity → Updated timeline with activity card **Genie: Weekend Plan Assistant** (ios, 7 screens, AI Tool, Social Networking) [Refero](https://refero.design/flows/12431) > A mobile assistant conversation starts on a home/chat screen with suggested prompt chips and a message composer. The user types a planning request that mentions a friend, then sends it. The assistant responds by looking up the friend’s profile, generating a location-based recommendation, and showing nearby place cards. The flow assumes the user is already in the assistant experience and can type, mention people, and review suggestions. It ends with the assistant providing a drafted plan and a follow-up question about timing. Steps: Assistant Home → Draft Message with Mention → Ready to Send → Message Sent → Profile Lookup → Recommendation Results → Plan Draft and Follow-up **Airbnb: Photographer booking** (ios, 4 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6436) > Happy-path: A mobile user discovers a photographer listing, taps Reserve, picks a package, date, and time, then confirms booking and payment. Entry: provider detail card (listing page) with service cards and a sticky Reserve CTA. Exit: user reaches the Confirm & Pay screen with booking summary and payment method (ready to review/complete payment). Assumptions: Mobile (iOS-style) interface; user has access to a saved payment method (a card is visible on the confirm screen); availability is shown per-date and per-package. The UI uses a modal sheet for scheduling and horizontal time-chip selectors for available time slots. Steps: Provider listing & Reserve CTA → Schedule modal — pick date and time → Schedule modal — time confirmed → Confirm and pay **Poppy: Create Calendar Event** (ios, 6 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12189) > This is a mobile assistant-style scheduling experience. It starts on a home screen with shortcut actions and a microphone-based command entry, then moves into event creation after the user chooses Create and Event. The user can pick a quick date option or enter a custom title, then refine the scheduled date and time on a calendar-style editor before confirming. Steps: Assistant Home → Choose Create Task → Select Event Type → Set Event Timing → Enter Event Details → Adjust Calendar Event **WhatsApp: Schedule Call and Invite Participants** (ios, 8 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12683) > A mobile calls interface opens on the Calls tab. The user can access recent calls, shortcuts for calling and scheduling, and create a new scheduled call. The flow starts from the Calls screen, moves into scheduled call creation, then into contact selection and chat confirmation, and ends back on the Calls screen with the new upcoming call visible. Steps: Calls Home → Calls Menu → Scheduled Calls Empty State → Schedule Call Details → Send To Recipients → Recipient Confirmed → Chat Event Confirmation → Calls Upcoming State **TidyCal: Therapy Session Booking** (web, 6 screens, Business, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/12503) > Happy path for a therapy provider booking site. The user starts on a therapist’s public landing page, reviews available services, selects a session type, chooses a date and time from either a calendar or list view, enters name and email in a confirmation dialog, and receives a booking confirmation screen. Steps: Therapist Landing → Service Selection → Calendar Availability → Booking Details → Enter Contact Details → Booking Confirmed **West Elm: Design Appointment Scheduling** (web, 10 screens, Home & Decor, Marketplace) [Refero](https://refero.design/flows/1674) > Happy-path summary: A desktop user requests a free Design Crew appointment. They start on the service overview, pick an appointment type (in-home, virtual, or in-store), select a local store (if applicable), choose a date and time from an availability grid, enter contact information, add project details, and receive an on-screen booking confirmation with a reference and follow-up questions. Steps: Appointment Type Selection → Find Local Store → Date & Time - Availability Overview → Select Time Slot → Change Date (Calendar) → Enter Contact Information → Info Required Modal → Confirm Contact Details → Add Project Details → Booking Confirmation & Follow-up **Cal: Meeting scheduling** (web, 6 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3497) > Happy-path booking: A visitor opens a host's public booking listing (desktop, not signed-in) and sees a list of offered meeting types (durations and descriptions). The visitor selects a meeting type (30 Min Meeting), picks an available date from a month calendar, chooses an available time slot, fills contact information and optional notes, and confirms the booking. Steps: Meeting types list → Confirm meeting type selection → Month calendar — pick a date → Choose time slot → Booking details form → Confirmation summary **Cal: Request reschedule for booking** (web, 5 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3493) > Happy-path: A signed-in user views their Upcoming bookings list, opens a booking's Edit dropdown, chooses "Request reschedule," types an optional reason in a modal, and submits the request. Entry: Bookings > Upcoming (desktop, authenticated user). Exit: modal closes and a transient confirmation toast is shown: "Reschedule request sent." The interface guides the user with an Edit dropdown → Request reschedule modal that explains the effect (cancel meeting and notify scheduler) and provides an optional free-text reason. Steps: Bookings list (Upcoming) → Open Edit dropdown → Request reschedule modal (empty) → Enter reason and submit → Confirmation toast ##### Task: business account onboarding company verification **Fuse: Virtual Bank Account Signup** (ios, 24 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8730) > Happy path: From an empty Fuse Wallet the user opens a promotional card to create a virtual bank account (USD or EUR), chooses a currency, supplies an email, and is routed to a third-party identity verification (Bridge → Persona). The user completes the Basic Information step (country, name, birthdate, SSN, address with autocomplete) and submits. Steps: Wallet Home (empty) → Currency Selection (USD shown) → Currency Selection (EUR shown) → Bridge - Email capture → Bridge - Email entered → Connecting / Processing → Full-screen Loading → Persona Consent (Getting started) → Required submissions list → Country selection → Basic Information (form start) → Typing name fields → Date picker (scrolling) → Date selected → SSN entry placeholder → Typing SSN (masked) → SSN visible (toggle) → Address autocomplete → Address details (expanded) → Verification failure **Ada: Onboarding and Account Creation** (ios, 19 screens, Health & Wellness) [Refero](https://refero.design/flows/4094) > New-user onboarding that introduces the product (5 illustrated slides) and guides the user through creating an account. Steps: Onboarding - Intro (1/5) → Onboarding - Causes (2/5) → Onboarding - Next steps (3/5) → Onboarding - Get started (4/5) → Onboarding - Final slide (5/5) → Consent - Confirm you're ready (initial) → Consent - Checkboxes selected → Create account - Empty form → Create account - Inline validation errors → Create account - Corrected email → Create account - Password entry and show/hide → Manage your privacy - Default toggles → Manage your privacy - Agree or save choices → Notifications & tracking opt-in → Check your inbox - Verification sent → Tips to find your email → Verification email - External message → Email verification success (web) → Account verified → Welcome Home **Voicenotes: Email Verification** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14272) > The user starts in Settings with a pending email-verification prompt. They open the verification step, enter the six-digit code sent to their displayed email address, and submit it. The system validates the code, shows a success confirmation, and returns the user to Settings with the pending verification prompt removed. The user is assumed to be signed in on an iOS mobile device and able to access the specified email inbox. ##### Task: calendar scheduling meeting **Genie: Chat Reminder Setup** (ios, 3 screens, Social Networking, AI Tool) [Refero](https://refero.design/flows/12426) > A chat-style assistant screen where the user types a reminder request, sends it, and receives confirmation that a reminder has been created for tomorrow at 9:00 AM. The flow begins in an active conversation with the keyboard open and ends with the assistant confirming the reminder and offering to save reference photos to a vault. Assumptions: user is already in the conversation, signed in, and has permission to create reminders. Steps: Compose Reminder → Message Sent → Reminder Confirmed **Cofounder: Daily Meeting Summary Test** (web, 5 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/7900) > Happy path: a user viewing a configured scheduled automation called "Daily Meeting Summary Email" starts a manual test of the scheduled trigger, watches the test run initialize, and inspects the run timeline as integrations (Calendar) and internal reasoning steps are executed. Steps: Plan view — Automation details → Test confirmation modal → History view — Test run initializing → History view — Run processing progress → Run timeline — Integration and reasoning details **OKURI: Schedule Team Standup** (ios, 5 screens, Utility) [Refero](https://refero.design/flows/11867) > The user starts on a daily calendar dashboard for Wednesday, May 13. They choose Add Event, enter a meeting title and location, review the default date and time settings, and save the event. The dashboard then displays the new event in the event list. The user selects the event to open its details, where they can review the location, schedule, calendar, alerts, and available edit or delete actions. The flow assumes the user has access to the calendar and the Work calendar is selected by default. Steps: Daily Calendar → New Event Form → Completed Event Form → Updated Daily Calendar → Event Details **Exoplan: Delete calendar event** (ios, 3 screens, Health & Wellness, Scheduling Tools) [Refero](https://refero.design/flows/7346) > User is viewing a single day's schedule in a mobile calendar (portrait). The top shows a compact week selector; the main area is an hourly day timeline with event tiles (e.g., 'User journey mapping', 'Doctor's appointment', 'Team meeting'). The user selects an event which opens a draggable bottom sheet showing event details and action icons (edit on left, delete/trash on right). After using the delete action, the bottom sheet closes and the day timeline updates to remove the deleted event. Steps: Day timeline (view events) → Event detail sheet → Updated day timeline **Daylish: Edit scheduled event time** (ios, 3 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/8008) > Happy path: User edits an existing calendar event's start/end time while in Schedule Mode on a mobile device. Entry: user is viewing a day in Schedule Mode and taps an existing event. The event becomes selected, showing draggable handles and an inline Start/End editor at the bottom. The user changes the event duration by dragging the event edge or by editing the End time in the bottom control. Exit: the event is updated on-screen with new time badges; the user confirms the change (check mark) and the scheduler persists the updated start/end times. Steps: Select event to edit → Adjust event duration → Finalize time change **Daylish: Event creation and schedule editing** (ios, 14 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/8006) > Happy-path: The user creates a new calendar event from the main Today screen, enters a title, then refines start/end times using a dedicated 'Schedule Mode' timeline and a date picker. Steps: Today - Empty Up Next → FAB expanded / Quick actions → Quick-create card - title focused → Title entered and reflected → Open inline start-time editor → Switch to Schedule Mode → Timeline - change start time → Timeline - change end time → Open date picker → Select new date (Oct 3) → Confirm schedule edits → Peeking Ahead - event review → Final confirmation and exit → Optional quick re-edit **Tinder: Share a date with friends** (ios, 18 screens, Dating & Relationships) [Refero](https://refero.design/flows/3461) > Happy-path task: from a match conversation the user opens actions and chooses "Share My Date," then creates a shareable date entry (who, date, time, location, optional notes), preview the date, and optionally share it with friends/family. Steps: Conversation action sheet → Share My Date onboarding → My dates (empty state) → Add date (form - initial) → Populate match & include link → Date picker (calendar open) → Calendar adjust selection → Time picker open → Date & time set (Done enabled) → Open location selector → Location suggestions list → Search location by typing → Location selected (card) → Add notes → Save (Done) tapped → Date created confirmation → Preview & share actions → My dates list (with new entry) **Daylish: Create and configure event** (ios, 8 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/8007) > Happy-path: A user views their "Peeking Ahead" timeline for Tomorrow, adds a new event, edits the title and duration, optionally inspects/edits alerts, then saves the event. Steps: Timeline (Tomorrow) overview → Open new event composer → Edit title and assign tags → Open duration editor → Adjust duration value → Open alerts panel → Confirm changes to event → Updated timeline with saved event **Mela: Schedule Recipe in Calendar** (ios, 5 screens, Food & Drink) [Refero](https://refero.design/flows/10366) > Happy-path: a user views a recipe detail and schedules that recipe into their in-app calendar. Steps: Recipe Detail → Overflow Menu → Calendar Modal — Date List → Meal Type Selection → Calendar — Confirmed Entry **Mela: Schedule recipe to calendar** (ios, 5 screens, Food & Drink) [Refero](https://refero.design/flows/10373) > Happy path: A signed-in user views their saved recipes in a vertically scrolling list, uses the search field to find items, swipes a recipe row to reveal quick actions (including calendar and delete), opens a calendar bottom sheet to place the recipe on a specific date/time slot, and then returns to the recipes list. Steps: Recipes list (default) → Reveal quick actions (swipe) → Open calendar sheet → Schedule recipe to a date/time slot → Return to recipes list (sheet dismissed) **pushr: Create Recurring Schedule** (ios, 4 screens, Health & Wellness, Fitness) [Refero](https://refero.design/flows/13788) > The user starts on a schedule list containing an existing weekday schedule at 9:00 AM. They tap the add button to create another recurring schedule, select Monday, Tuesday, Thursday, and Friday, change the time to 4:00 PM, and tap Done. The new schedule is saved and displayed alongside the existing schedule. The flow assumes the user is already viewing the schedule section and has permission to create schedules. Steps: Schedule List → Select Schedule Days → Set Schedule Time → Saved Schedule **Exoplan: Quick natural-language event add** (ios, 6 screens, Health & Wellness, Scheduling Tools) [Refero](https://refero.design/flows/7344) > Happy path: From the day view the user taps the floating + button to open a free‑form planner overlay called “Magic Planner”, types a short description (e.g., “Doctor’s appointment”), and submits it. The app shows a brief parsing/loading state (“Loading context..”, then “Adding to schedule..”) and finally returns to the day view with the new event added to the schedule. Steps: Day view (calendar timeline) → Magic Planner — free-form input → User enters event text → Parsing / loading context → Adding to schedule (confirmation stage) → Updated day timeline with new event **Imprint: Commit & Schedule Reminder** (ios, 7 screens, Education) [Refero](https://refero.design/flows/5640) > Happy-path onboarding for committing to a learning streak and scheduling recurring reminders. Entry: user arrives at a 'Commit to learning' screen where they pick a streak length. After choosing a streak and tapping the primary CTA, the user is taken to a 'Schedule your next sessions' screen to pick a reminder time and optionally add the event to their calendar. If the user enables calendar integration and taps Schedule, the system prompts for device calendar permissions. Exit: user either grants calendar access (permission flow shown) or dismisses/declines and continues (not shown). Assumptions: user is signed into the app, running on an iOS device, and calendar permission has not yet been granted. The primary happy-path ends with the app requesting calendar access to add events. Steps: Choose streak length → Confirm streak selection → Open scheduling screen → Open native time picker (AM selection) → Change time to preferred value (PM) → Finalize scheduling → System calendar permission prompt **Exoplan: Reschedule event by dragging** (ios, 4 screens, Health & Wellness, Scheduling Tools) [Refero](https://refero.design/flows/7349) > Happy path: A user views a single-day schedule and reschedules an existing event by dragging it to a new time/slot. Steps: Day view - initial state → Initiate drag on event → Drag with live feedback → Drop and confirm placement **Exoplan: Free-form plan to day schedule** (ios, 9 screens, Health & Wellness, Scheduling Tools) [Refero](https://refero.design/flows/7350) > Happy path: User opens a day timeline, invokes the planner, types a free‑form list of tasks (with durations), and submits it. The system parses the text, shows progressive parsing feedback (counts of events generated) and then maps parsed items into the day as scheduled events with buffers and icons. Steps: Day timeline (empty) → Free-form input → Parsing — loading context → Parsing progress (initial count) → Parsing progress (incremental results) → Initial placement on timeline → Midday insertion and buffers → Final placement (afternoon & end of day) → Completed schedule view **Daylish: Reschedule single event** (ios, 6 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/8014) > Happy-path for rescheduling a single calendar item using an interactive Schedule Mode. Entry: user is viewing a vertically stacked day view (Tomorrow — Oct 3, 2025) showing a list of colored events. Exit: user returns to the day list view with the selected event updated to the new start/end time. Assumptions: user is signed in and has permission to edit the event, the device is a touch phone, events are movable and not locked or recurring in a way that prevents single-instance edits. Steps: Day list — identify event → Open event context menu → Enter Schedule Mode → Drag event to new time → Confirm schedule change → Verify updated event in list **Joi: Schedule New Event** (ios, 8 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/4380) > Happy path: From the daily agenda (day view) the user creates a new scheduled item, changes its date and time, and saves it so the new item appears in the target day's list. Steps: Day view (Feb 23) → Open event editor and type title → Open month calendar → Open time picker (initial) → Set new time (3:00 PM) → Confirm and save event → Return to day view (Feb 24) → Event visible in agenda **SavvyCal: Meeting Reschedule** (web, 5 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3372) > Happy-path: A user opens a scheduled event summary, chooses to reschedule, picks a new open time on a calendar week view, confirms the new time (optionally providing a reason), and sees the event updated with the new time and a record of the reason. Steps: Scheduled Event Summary → Calendar — Pick New Slot → Confirm New Time → Confirm With Reason → Updated Event Confirmation **SavvyCal: Manage scheduled meeting attendees** (web, 5 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3371) > Happy path: A signed-in user opens the Meeting Polls area, selects a specific poll that has been scheduled ("Yoga with team"), reviews the scheduled event details, opens the Invite a guest modal, adds a guest (name + email), and confirms that the guest appears in the event attendee list. Entry: Meeting Polls list. Exit: Scheduled Event screen updated with new attendee. Assumptions: user is logged in, has access to Meeting Polls, and the selected poll is already scheduled (Scheduled Event tab visible). Steps: Meeting Polls list → Scheduled Event overview → Invite a guest modal (empty) → Invite a guest modal (filled) → Scheduled Event updated with new attendee **Marathon: Demo Booking** (web, 6 screens, Business, Productivity, Project Management, Collaboration Tools, Scheduling Tools) [Refero](https://refero.design/flows/4837) > Happy path: A visitor arrives at the "Book A Demo" landing panel (desktop). The right side displays an embedded calendar with the host profile and meeting details (20 minutes, Zoom). The user browses the month/day grid, selects an available time slot, and a modal appears to confirm details. The user completes required company/contact fields and optional questions (team size, client work, plan choice, add guest). After submitting, a confirmation page shows the scheduled meeting details with reschedule and cancel actions. Steps: Landing / Calendar Overview → Select Time Slot → Open Confirm Details Modal → Fill Form Fields → Select Plan & Add Guest (Optional) → Confirmation / Event Details **Calendly: Reschedule Meeting** (web, 11 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/14053) > The user starts in an authenticated scheduling calendar, opens an existing one-on-one meeting, and chooses to reschedule it. They select the option to reschedule immediately, browse to a future month, choose a date and available time, optionally provide a reason for the change, and submit the update. The flow ends back in the calendar with a confirmation message and the meeting details updated to the new date and time. The user appears to have permission to modify the meeting, and the flow is shown on a desktop layout. **Marathon: Demo booking confirmation** (web, 3 screens, Business, Productivity, Project Management, Collaboration Tools, Scheduling Tools) [Refero](https://refero.design/flows/4838) > User lands on a desktop confirmation screen immediately after booking a demo. The page confirms the meeting was scheduled, shows full meeting metadata (what, when, who, where, details), and presents primary actions: Reschedule and Cancel event. If the user chooses Cancel, an inline confirmation area appears allowing an optional note and two actions (Nevermind, Cancel the event). After confirming cancellation the page updates to a canceled state showing a cancellation reason and the same meeting metadata. Steps: Meeting scheduled confirmation → Inline cancel confirmation → Meeting canceled confirmation **Calendly: Contact Meeting Booking** (web, 13 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/14036) > A logged-in user opens an existing contact record with no prior meetings, chooses to schedule a meeting, selects an event type, configures the meeting location, chooses an available date and time, and books the meeting. The process begins on the contact’s Meetings tab and ends when the contact record displays the newly scheduled meeting and a confirmation notification. The user appears to have permission to schedule meetings and access configured event types and availability. The flow is shown on a desktop device. **SavvyCal: Attendee scheduling and location setup** (web, 11 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3366) > This is the attendee settings path inside a meeting link editor where the organizer configures an attendee's availability, meeting location/conferencing, and which calendar to create events on. Entry: organizer is in the scheduling link editor (logged in) and opens the Location/Conferencing or attendee settings. Happy path: open attendee modal → set availability preset → switch to Location tab → choose/add a conferencing option (e.g., Custom URL) → switch to Calendar tab and choose the calendar → save changes → updated Location/Conferencing shows on the left panel. Exit: modal closed and main editor reflects the saved attendee location and calendar settings. Assumptions: user is authenticated, has permission to edit the scheduling link, attendee exists on the event, and integrations (Zoom, Google Meet, etc.) may be available in the environment. Steps: Open scheduling link editor → Open attendee modal → Inspect schedule type options → Select an availability preset → Switch to Location tab → Open location chooser → Pick Custom URL option → Enter custom conferencing URL → Review Calendar tab → Save attendee settings → Verify updated attendee card **Calendly: Schedule One-on-One Meeting** (web, 9 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/14051) > A visitor uses Liam Basil’s public scheduling link to book a one-on-one meeting. The flow starts on an event-type selection screen, continues through date and time selection, collects attendee details, and ends with a scheduling confirmation. The visitor appears to be logged out and does not need an account. The flow is shown on a desktop browser. The selected meeting is 45 minutes, scheduled for September 30, 2026 from 3:00pm to 3:45pm in Dubai Time. **Microsoft Teams: Schedule meeting with recurrence** (web, 14 screens, Communication Tools, Scheduling Tools) [Refero](https://refero.design/flows/9721) > Happy path: A signed-in user opens the Meet area, chooses to schedule a meeting, fills the meeting details (title, attendees, date/time, recurrence, location, notes), saves/sends the invite, reviews the created meeting, and returns to the Meet overview where the scheduled meeting appears. Steps: Meet overview → New meeting editor (empty) → Attendee autocomplete (open) → Attendees selected → Date picker opened → Start time dropdown open → End time / duration options → Recurrence options opened → Set recurrence modal → Recurrence end date calendar → Recurrence saved → Meeting details populated → Meeting created confirmation → Meet overview updated **Reclaim AI: Create and Configure Smart Meeting** (web, 12 screens, Collaboration Tools, Scheduling Tools) [Refero](https://refero.design/flows/4117) > Happy path: A logged-in user opens the Smart Meetings section, clicks to create a new Smart Meeting, fills in meeting details (name, category, color), adds attendees and conferencing link, configures scheduling constraints (availability grid, repeat, duration, ideal days/time, priority), sets visibility/dependency options, saves and confirms whether to send update emails. After saving the user is shown a details view with availability and a list of generated occurrences (each showing scheduled / not scheduled). The flow ends with the Smart Meetings list showing the created meeting card and its next occurrence. Steps: Smart Meetings - Landing → Edit — Meeting Details (initial) → Pick Category → Pick Color → Attendees — Add/Validate → Attendee Calendar Status → Add Conferencing and Location → Scheduling — Availability & Rules → Other Options — Visibility & Dependency → Save — Email Confirmation → Details — Review Meeting and Occurrences → Smart Meetings — List with Meeting Card **Calendly: Create Meeting Poll** (web, 13 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/14046) > Happy path: An authenticated Calendly user starts from the Scheduling area, creates a new meeting poll, selects a duration and available time slots, configures meeting details and optional settings, then shares the poll. The system creates the poll and returns the user to the meeting-poll list, where the new poll appears with its status and copy-link action. **SavvyCal: Meeting Poll Setup and Scheduling** (web, 12 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3370) > Happy path for creating a group meeting poll, collecting votes and converting the chosen option into a calendar event. Entry: user lands on the Meeting Polls area (empty state) while signed in on desktop. Exit: a calendar event is created and invitations are sent; the poll shows a scheduled confirmation. Steps: Meeting Polls - Empty State → Poll Editor - Initial → Select Time Slot → Choose Meeting Duration → Select Location Type → Add Attendees and Details → Publish Poll → Copy Link Confirmation → Review Poll Results - Waiting → Review Poll Results - Votes Collected → Schedule Event Modal → Scheduled Confirmation **Reclaim AI: Meeting scheduling** (web, 12 screens, Collaboration Tools, Scheduling Tools) [Refero](https://refero.design/flows/4235) > Happy path for scheduling a one-on-one meeting with a host. Entry: visitor lands on the host's booking landing that lists available meeting types. The visitor selects a specific meeting type card, then selects a date and a time from a calendar/time list, optionally toggles timezone and calendar comparison, and proceeds to a short booking form (name, email, CC, message, question). Exit: booking confirmation screen shows meeting details and confirmation actions. Steps: Choose meeting type → Open calendar selector → Change timezone → Select day → Compare calendars → View granular times → Inspect conflict details → Confirm time availability → Proceed to booking → Fill attendee information → Review filled form → Show confirmation **Calendly: Calendly Account Setup** (web, 29 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/14006) > Happy path begins on the Calendly marketing homepage for a new, logged-out user. The user selects a free-account entry point, enters an email address, chooses the email-based signup route, provides a name and password, completes human verification, and proceeds through a six-step setup wizard. They select a role, review calendar connection options, confirm availability, configure meeting locations and Notetaker preferences, and continue into the scheduling workspace. The flow assumes the user is on a desktop browser and has access to the entered email address and a Google Calendar account. The Google Calendar connection appears to fail because of a domain issue, but the user still reaches the scheduling dashboard with a default 30-minute meeting type. **SavvyCal: New account setup and guided onboarding** (web, 27 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3346) > Summary: This is the new-user acquisition -> onboarding flow. Steps: Marketing landing / entry → Sign up (method selection) → Sign up (form completion) → Welcome — prefilled account settings → Edit scheduling URL (modal) → Edit URL — example placeholder → Edit URL — save new slug → Time zone selection (expanded list) → Time zone search results → Theme selection and time format → Usage type: individual vs team → Team size selection → Role selection (which best describes you) → Role selected (example) → Primary goal / prioritization → Primary goal selected → Connect your calendars → Wire up video conferencing → Set your working hours → Onboarding complete — Guided Setup dashboard **Cofounder: Automated Email & Calendar Assistant** (web, 14 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/7897) > Happy-path: A signed-in user on the home dashboard writes a natural-language instruction to the assistant to email a meeting attendee and then approve scheduling. The assistant runs a multi-step automation: it drafts and sends an email via the connected mail integration, verifies the message appears in Sent, summarizes the result to the user, and (after user approval) creates a calendar hold via the connected calendar integration. Entry: authenticated user on the main dashboard/chat screen. Exit: calendar hold created and assistant prompts to invite the attendee. Steps: Dashboard / New Chat → Compose Instruction → Chat Created / Instruction Sent → Todo List Creation → Timeline Progress (Draft + Send) → Email Draft Preview → Verify Sent Mail → Assistant Summary & Prompt → User Confirms Calendar Hold → Calendar Setup & Slot Search → Create Event Draft → Event Verification → Complete Calendar Hold → Final Prompt / Next Action **Homerun: Candidate Interview Scheduler** (web, 10 screens, Productivity, Communication Tools, Job Search) [Refero](https://refero.design/flows/7186) > Happy-path: A logged-in recruiter opens a candidate profile, uses the Schedule control to create a new interview event, picks a template/date/time/duration and location, adds or confirms attendees, inserts a video-call link and message, previews the email invitation, and sends it. Steps: Candidate profile — Schedule CTA → Schedule actions menu → Event creation modal — initial → Insert template → Set date → Select start time → Set duration → Choose location type → Add video-call link and message → Invitation preview and send **Wrike: In-chat meeting scheduling and GIF send** (web, 9 screens, Business, Productivity, Project Management, Collaboration Tools, Paperwork, Scheduling Tools) [Refero](https://refero.design/flows/1498) > Happy path: A logged-in user on the Inbox (Archive tab) opens the floating Wrike Help/Support chat, requests a sales call, walks through the embedded scheduler (select date → select time → confirm), sees an in-chat confirmation that an email was sent, then uses the GIF picker to insert and send a GIF. Steps: Inbox (Archive) — no archived notifications → Open chat thread → Request sales call → Select date → Select time / timezone info → Confirm meeting slot → Meeting confirmation in chat → Open GIF picker (GIPHY) → Insert and send GIF ##### Task: care plan **Mozi: Share Plans via Messages** (ios, 4 screens, Social Networking, Travel & Booking) [Refero](https://refero.design/flows/4147) > User is on the My Plans screen (mobile, logged-in) and wants to share their upcoming plans with others. From the My Plans list the user opens a "Share my plans" modal, chooses to share, selects Messages from the system share sheet, and lands in the Messages composer with a preview card of their plans inserted. Steps: My Plans list → Share options modal → iOS share sheet → Messages composer with preview **Structured: Structured Planner Download** (ios, 4 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3286) > Happy path: A user on the Structured AI home screen taps a promotional CTA to obtain a printable planner sheet. The CTA opens an external link (Linktree) listing Weekly / Monthly / Annual versions. The user selects a version, confirms the download in the browser prompt, and the PDF opens in the device's PDF viewer where they can share, save, or print it. Steps: Home - Download CTA → Link hub - Choose Version → Download Confirmation → PDF Preview ##### Task: case management **Kickstarter: Reward pledge and checkout** (ios, 23 screens, Fundraising) [Refero](https://refero.design/flows/4631) > Happy-path and edge-case flow for backing a Kickstarter project from the project overview through reward selection, shipping-country selection, customizing pledge amount, entering payment, handling card errors, and seeing confirmation. Steps: Project overview → Select reward (initial) → Open country picker → Country list - browse → Country picker - numeric input (mistype) → Country search - text → Country selected → Customize your reward (open) → Adjust pledge amount (increment) → Manual pledge entry (numeric keyboard) → Pledge amount validation error → Checkout start / loading → Payment modal - choose method → Enter card (invalid data) → Attempt payment — declined → Add new valid card → Processing payment → Pledge success / Thank you → Post-purchase UX — suggestions → Project view updated — backer state ##### Task: code repository pull request **Wabi: Invite Collaborator** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12045) > Happy path for inviting another user to join a shared space from a collaboration screen. The user starts on a member management view showing current participants and invitation options. They can search for a person by username, tap an Invite action next to a matching result, and return to the main view where the member count updates. Steps: Member Overview → Search Active → Search Results → Invite Confirmed → Updated Member Count **Dropbox: Share File Request Link** (web, 3 screens, Collaboration Tools, Business, Productivity) [Refero](https://refero.design/flows/13056) > This is a Dropbox-style file requests area for a signed-in user with access to manage requests. The user starts on the File requests list, opens a sharing dialog for an existing request, copies the request link, and receives confirmation that the link is ready to share. The flow assumes the user is already authenticated and has permission to share the request. **Dropbox: Share File Request** (web, 6 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/13057) > A logged-in Dropbox user opens the File requests area, starts sharing an existing file request, selects a recipient by email, and sends the request. The flow begins on the File requests list and ends with the modal closing and a confirmation toast indicating that the Personal files request was shared. Steps: File Requests List → Share Request Dialog → Search Recipient → Select Recipient → Confirm Request Details → Share Confirmation **Blackbird: Referral Invite** (ios, 2 screens, Food & Drink) [Refero](https://refero.design/flows/9024) > A mobile referral/promotional modal that shows a scannable QR code, an explanatory headline and copy about the reward, the user's referral code, and two primary actions: Copy Code and Share Link. Steps: Referral Promo Display → Copy Confirmation **Ground News: Invite Friends - Referral Code** (ios, 2 screens, News) [Refero](https://refero.design/flows/7580) > Logged-in user lands on the Invite Friends screen (accessed from Profile tab). The screen displays an explanation of the referral promotion, the user's unique invite code in a bordered input, a Copy control, and a large Share Referral Code button. Primary actions: copy the code to clipboard or open sharing. Secondary content: promotion terms and a link to Terms & Conditions. Exit points: user copies the code, shares the code (system share sheet or similar), navigates back via the back arrow, or taps another bottom nav item. Steps: Invite Screen (Initial) → Copy Confirmation **Leonardo.Ai: Copy Generation API Code** (web, 4 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/12888) > Logged-in user on Leonardo.ai’s AI Creation screen generates or reviews an audio asset, opens the item actions menu, chooses the option to view the Generation API Code, and then copies that code from a modal. Steps: AI Creation Workspace → Actions Menu → Generation API Code Modal → Copied Confirmation **Raycast: Question search and select** (ios, 2 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/10193) > User is on a mobile "Questions" screen that shows a searchable list of question titles. The header shows a rounded search input with placeholder text and a numbered list of question rows; each row has a right chevron indicating it leads to details. Happy path: user taps the search field to open the keyboard, types a query, the list filters in real time to matching questions, then the user taps a row to navigate to that question's details (or taps the X control to dismiss). Steps: Questions list (initial) → Search with keyboard and filtered results **Linear: Project selector** (ios, 3 screens, Collaboration Tools, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/6614) > This is a mobile project-picker modal used to reassign an item to an existing project. Entry: user opens the "Project" picker from an item (implicit). The modal lists available projects and exposes a search field at the top. The happy path: open picker → type or scroll to find a project → tap a project to select it → close the picker. Assumptions: user is on a mobile device, keyboard appears when the search field is focused, user has permission to reassign the item, and there are multiple existing projects to choose from. Exit points: selecting a project (selection shown with a checkmark) or tapping the top-right close button to dismiss the modal. Steps: Project picker (initial list) → Search — empty results → Search — single match **Rork: Invite Collaborators** (web, 2 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12169) > A signed-in user opens an invite/collaboration panel for a project, adds collaborators by email or creates a shareable invite link, then gets confirmation when the link is created or copied. The flow starts from a project workspace and ends with invitation access being generated and ready to share. Steps: Invite Collaborators Panel → Link Copied Confirmation **Comet: Edit and rerun assistant query** (ios, 7 screens, AI Tool, Utility) [Refero](https://refero.design/flows/10937) > Happy path: a user opens an Assistant response, uses the thread overflow menu to choose Edit, updates the query in a full-screen edit modal, submits the revised query, watches an ephemeral "Working" state while the assistant searches, and receives an updated answer with reviewed sources and related links. Steps: View Assistant thread → Open thread actions menu → Enter edit mode → Edit the query text → Assistant working/searching → View updated assistant answer → Explore related links and follow-ups **Linear: Assign Project** (ios, 4 screens, Collaboration Tools, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/6613) > Happy path: A user viewing an issue in mobile detail opens the Project picker, searches for a target project (Pulse), selects it, and returns to the issue which now shows the updated project tag. Entry: issue detail screen where project is displayed as a pill. Exit: issue detail updated to show the new project. Assumptions: user is signed in, has permission to change the issue's project, and is on a mobile device (keyboard can appear). Steps: Issue detail (view) → Project selector (list) → Project search result → Issue detail (updated) **Lovable: Git Integrations Management** (web, 4 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12354) > This is a workspace settings area for managing Git integrations. The user starts on a Git integrations landing screen, opens GitHub connection management, and creates or selects an installation to connect. The happy path assumes the user is already logged in with access to the workspace settings and can authorize or install GitHub access. The flow ends after a GitHub connection appears in the connections list. Steps: Git Provider Selection → GitHub Connections → Select Installation → Connected Accounts List **Comet: Assistant command copy** (ios, 2 screens, AI Tool, Utility) [Refero](https://refero.design/flows/10936) > Happy path: user opens the Assistant overlay on a mobile device, reads the help/integration instructions, locates a copyable command (code snippet) and copies it to the clipboard. Steps: Assistant modal — command view → Copy confirmation toast **HYPE: Comment Discussion Panel** (ios, 2 screens, News, Lifestyle) [Refero](https://refero.design/flows/12248) > A mobile comments panel is opened over an underlying content detail screen. The user can view one existing comment, react to it with like/dislike, tap Reply, or type a new comment in the input field at the bottom. Entry appears to be from a content item with a comments trigger; exit is via Close or the back navigation on the underlying screen. Steps: Comments Open → Comment Liked **Podimo: Podcast Comments Thread** (ios, 2 screens, Entertainment) [Refero](https://refero.design/flows/12270) > A mobile comments drawer is open for a podcast episode/post titled “1×01 La partida con SoyUnaPringada y Dar…”. The user is viewing a threaded discussion with top-level comments, nested replies, like counts, and a text field to add a new comment. Assumption: the user is already inside the content and has permission to read and interact with comments. Steps: Comments View → Liked Comment State **Wabi: Recipe Comments** (ios, 4 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12066) > This is a mobile comment view for a shared collection called “Family recipe hub.” The user enters the screen from the comments tab of the collection, views existing comments and engagement counts, then taps the comment field to open the composer and type a response. A send action posts the comment and updates the comment count/list. Steps: Comments View → Comment Composer Open → Comment Drafting → Comment Submitted **Factory: Configure GitHub repository selection** (web, 3 screens, Business, Development Tools, AI Tool) [Refero](https://refero.design/flows/7745) > User opens the "Configure GitHub Integration" modal (Repository Selection tab) from Settings to enable or disable individual GitHub repositories for the product (labelled "Factory"). The interface shows a searchable list of repositories, a count of currently enabled repositories (e.g., Enabled Repositories: 1 / 25), toggles to enable/disable each repo, and a link to manage GitHub permissions. Entry: user launches the Configure GitHub Integration modal. Exit: user finishes selecting repositories (toggles) and closes the modal. Assumptions: user is signed in, the GitHub connection exists, and the user has permission to view/manage repository selections. The UI appears optimized for desktop. Steps: Repository list — default view → Search repositories — no matches → Search repositories — filtered result **Cursor: Generate and review AGENTS.md** (web, 17 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9924) > Happy path: A user opens an agent workspace, types a request to create or update AGENTS.md, selects a model, and runs the agent. The agent inspects the repository, generates the AGENTS.md file, commits and pushes changes, and presents a code diff on a right-side panel where the user can review and create a PR. Steps: Workspace — new prompt → Update prompt and select model → Agent session starts → Activity log — inspection → Activity log — file content discovered → Activity log — git commands → Agent summary and created file → Preview of file content (diff) → Open diff panel → Scroll diff and annotate → Agent run completes / updates → Open full diff editor → Add inline question → Agent generates response draft → Agent elaborates and references repo → Final session state shown → Ready for PR or next steps **fal: Request Code Language Selection** (web, 4 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/13529) > A user opens a request details drawer with the Code section expanded and switches the displayed code sample from JavaScript/Node.js to Java. The flow starts with an existing generated image request and ends with the Java implementation visible. The user is assumed to have access to the request details view; no additional authentication or input is shown. Steps: View Request Code → Open Language Selector → Choose Java → Review Java Sample **Factory: Attach context to Droid session** (web, 4 screens, Business, Development Tools, AI Tool) [Refero](https://refero.design/flows/7757) > Happy-path (attempted): User opens a Droid session (logged in) and uses the inline composer at the bottom to add external context. They invoke the inline insert menu (via @), choose to add a pull request (or other context type), paste a URL, and submit it. The system validates the URL and, in this sequence, rejects the submitted URL with an "Invalid Pull Request URL format" error. Entry: Droid session screen with composer focused. Exit: composer remains open with a validation error preventing the URL from being attached until corrected or canceled. Assumptions: user is logged in (profile shown), workspace is connected and the repository is visible in the session header, user is on a desktop viewport and using the dark theme. Steps: Open Droid session and focus composer → Open inline insert menu → Enter URL for chosen source → URL validation fails and shows error **Cohere: Chat code snippet language selector** (web, 3 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3857) > Happy path: A signed-in user on a desktop visits the Cohere Playground 'Chat' area, reviews a generated chat response on the left, and uses the right-side 'Parameters / Code' panel to view a client code snippet in a preferred language. The user opens the language selector and picks a different language (e.g., JavaScript). The code block updates to the selected language template. Entry: user is viewing a generated chat result. Exit: the code panel displays the selected-language snippet (ready to be copied or used). Assumptions: user is on a desktop browser, is able to view the two-column layout, and has permission to view Playground content (likely signed in). Steps: View generated chat and default code snippet → Open language selector → Select JavaScript and update snippet **Cohere: Inspect and Export Chat Request** (web, 3 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3859) > Happy-path: A user composes or views a chat prompt in the Playground, opens the “Under the hood” panel to inspect the raw API request and response, selects a target code language for the request snippet, and copies the generated code to the clipboard. Entry: user is signed into the Playground and viewing a chat/conversation. Exit: code snippet copied and confirmation shown. Assumptions: user is authenticated, on a desktop or tablet with pointer input, and working in the Playground environment where the ‘Under the hood’ panel and Parameters panel are available. Steps: Open 'Under the hood' to inspect request → Select code snippet language → Copy code snippet to clipboard **Cursor: Repository selection** (web, 3 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9944) > Happy path: From the Agents dashboard the user opens the repository filter control, types to find a repository, and selects a match. Steps: Open Repo filter → Search by numeric ID (no results) → Search by name and view suggestions **Factory: Droid session setup and repo inspection** (web, 17 screens, Business, Development Tools, AI Tool) [Refero](https://refero.design/flows/7748) > Happy path: A logged-in user (Sofia Delgado) starts a Code Droid session, chooses a Droid type, selects a remote workspace (paired via Factory Bridge or chosen from a remote list), and opens a live session. The system connects to the remote machine, runs repository inspections (git status, file reads), and assembles a set of outputs: 'Thinking' analysis, a TODO list, folder view attempts, and a clarification request because the repository is empty. The user can expand sections, view the assistant's thinking, and open the Context panel. Steps: Session start / initial state → Select Droid type → Select workspace (Local pairing) → Select workspace (Remote list) → Confirm remote workspace → Workspace summary / session start → Connecting to machine → Automated inspection begins → Clarification needed prompt → View assistant 'Thinking' details → Todo list expanded → Folder view error → Git status output → Clarification message persisted → Clarification message with reactions → Open Context panel → Two-column session view with context **Rork: Comments Panel** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12170) > A comments side panel is opened over a workspace showing a mobile preview in the background. The user can view existing comments, add a new note, and submit it from the bottom input field. Assumptions: the user is already signed in and has access to the project/review area. The flow starts when the comments panel is opened and ends after a comment is posted and appears in the thread. Steps: Open Comments Panel → Compose Comment → Post Comment **Cohere: API Request Composer** (web, 2 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3860) > A two-step modal that helps users compose a Chat API request and then generates runnable code in multiple languages. Steps: Request Composition → Code Snippet Preview **Lovable: GitHub Connection Management** (web, 5 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12355) > This is a Git/GitHub connections management area within workspace settings. The user opens a specific GitHub connection from the connections list, reviews its details and linked-project status, then initiates deletion. A confirmation modal requires typing the connection name before enabling deletion. After confirmation, the system removes the connection, returns to the connections list, and shows a success toast. Assumption: the user is already signed in and has permission to manage workspace Git connections. Steps: Connections list → Connection details → Delete confirmation → Deletion complete → Post-delete state ##### Task: compliance dashboard audit **OKURI: Daily Task Checklist** (ios, 3 screens, Utility) [Refero](https://refero.design/flows/11866) > Happy-path task management screen in the OKURI experience. The entry point is a daily dashboard for May 13 (Wednesday), with date, sunrise/sunset, zodiac/astrology-style visuals, and a to-do list. The user is assumed to already be on the daily planner screen. The exit state is a completed checklist item, shown as crossed out and disabled/checked, with the list remaining on the same screen. Steps: Daily Dashboard → Task Completion Update → Multiple Task Completion ##### Task: connect bank account open banking **Spoil me: Create wishlist and add first items** (ios, 37 screens, Fundraising) [Refero](https://refero.design/flows/5343) > Happy-path new-user onboarding that creates a Spoil.me wishlist (vanity link), verifies identity via an email magic link, imports product photos from the device, adds item details (price, crowdfunding toggle), and lands the user in their wishlist dashboard. Entry: first-time visitor (not signed in) at the hero screen. Exit: activated wishlist/profile screen showing created wishes. Steps: Hero - Choose vanity link → Typing vanity handle → Vanity handle entered → Email input - prompt → Email validation (invalid) → Email input - valid → Confirmation - check inbox → Resend active state → Open verification email (inbox) → Waiting for verification → Photo picker - select images → Photo selection confirm → Item details - edit → Alternate suggestions - browse images → Confirm item images → Add another wish - title input → In-app notifications prompt → OS-level notification permission → Optional payout connection (Stripe) → Wishlist dashboard / onboarding state **Verse: Connect Streaming Account and Load Playlists** (ios, 7 screens, Social Networking, Design & Media Editing) [Refero](https://refero.design/flows/4000) > Happy path for connecting a music streaming account (Spotify shown) from an in-app "Add a song" modal so the app can export songs to a user playlist. Entry: user opens the Add a song modal inside the mobile app (iOS). If no streaming account is connected the modal shows a Connect Streaming Service CTA. Exit: streaming account is authorized and the user's playlists become available in the Add a song modal so the user can pick a playlist to export songs to. Assumptions: user is inside the host app and wants to link an external streaming account; the streaming service (Spotify) is installed or available on the device; the user is able to authenticate with the streaming service when prompted. Steps: Add a song (no streaming connected) → Select streaming provider → System open-app confirmation → Spotify authorization (intro) → Spotify authorization (consent) → Connection success confirmation → Add a song (playlists loaded) **Wabi: Connect Gmail in Settings** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12042) > Happy path for connecting a Gmail account from the Settings screen to enable a power-up. The user starts inside Settings, triggers the Gmail connection, passes through Google sign-in and consent, and returns to Settings with Gmail enabled. Assumption: the user is already inside the product on a mobile device and has at least one Google account available to choose from. Steps: Settings and Power-ups → Permission Prompt → Account Selection → Google Access Consent → Settings Enabled State ##### Task: contract management **Cohere: Document-based conversation review** (web, 3 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3832) > A desktop chat workspace where a user interacts with an AI assistant to extract or summarize information from uploaded documents. Steps: Conversation view with attachments visible → Activate collapse control for left panel → Expanded conversation (left panel collapsed) **Structured: Task action sheet** (ios, 2 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3260) > Mobile day-timeline view (December 23, 2024) showing a vertical timeline of scheduled items. The user taps a task node (Eat lunch) to open a modal bottom sheet that exposes management actions for that task: Delete, Duplicate, Complete/Incomplete toggle, and Edit Task. Steps: Open actions for a completed task → Open actions for an incomplete task ##### Task: credit score report **Tidal: Artist Credits and Track Liking** (web, 2 screens, Music Streaming) [Refero](https://refero.design/flows/3234) > Desktop, logged-in user (profile visible in left nav) is viewing a credits list for a single artist. The screen shows a header "Credits for [Artist]", a row-based table of tracks with columns (#, Title, Role, Artist, Album, Year, Time), and filter chips across the top (All, Songwriter, Producer, Engineer, Production team). A persistent mini player sits at the bottom showing the currently playing track. Steps: Browse artist credits → Like/favorite a track from list ##### Task: direct debit **Gizmo: Direct message - compose, react, share liked post** (ios, 12 screens, Entertainment, Social Networking, AI Tool) [Refero](https://refero.design/flows/10028) > Private one-to-one conversation where the user composes text, sends messages, receives short replies, reacts, and shares a liked post into the chat. Entry: user opens a conversation with a contact (Danny Go). Exit: shared content card and final text message delivered and marked as seen. Steps: Conversation open (empty) → Start composing message → Message sent → Incoming reply → Compose follow-up question → Confirm total → Recipient short confirmation + typing → Open emoji / reaction picker → Add reaction → Open content menu → Select liked item → Send post with message **Instacart: Payment Methods Management** (web, 3 screens, Food & Drink, Marketplace, Food Delivery) [Refero](https://refero.design/flows/10102) > This is the Edit / Manage Payment Methods area of a logged-in user's account settings (left navigation has Payment methods selected). Happy path shown: user views saved payment methods, chooses to delete a saved Visa card, confirms deletion in a modal, and returns to the add-payment-method list with the saved card removed. Steps: Saved methods and add options → Delete confirmation modal → Post-deletion state ##### Task: doctor video consultation **District: Restaurant Question Assistant** (ios, 5 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12847) > Happy-path mobile assistant experience for asking a restaurant-related question and continuing with a follow-up. Entry is a restaurant detail or assistant overlay with an active query context (“Asking about Hotel Rishi”). The user asks a question by voice, the system transcribes and processes it, shows an answer with suggested next actions, and keeps a persistent follow-up input available. The flow ends when the user has a ready-to-send follow-up request, such as asking for nearby breakfast options. **Voicenotes: Voice query to notes** (web, 5 screens, Collaboration Tools, AI Tool) [Refero](https://refero.design/flows/6977) > User opens the 'Ask' modal to query their notes using a voice recording. The modal shows an Ask/Create toggle, a notes filter (All notes), a prompt area labeled "Ask a question" with a mic icon, and a prominent floating Record control. Happy path: user opens the Ask interface, starts recording, stops the recording (Done), the system transcribes the audio and displays the recorded audio/transcription as a message bubble, and then the assistant posts a textual reply referencing the user's notes. Steps: Open Ask modal → Start recording → Stop recording → Transcription / audio message appears → Assistant reply displayed ##### Task: document editor collaboration **Google Gemini: Export Chat Response to Google Docs** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/3188) > Happy path: A user in the Gemini Introduction chat opens the contextual actions for a specific assistant reply, chooses Export to Docs, waits for the system to create a document, and then opens the newly created Google Doc to review the exported text. Steps: Chat message with action bar → Contextual actions menu → Document creation progress → Creation success confirmation → Google Docs viewer with exported text **Miro: AI-Generated Retrospective Summary Creation** (web, 6 screens, Productivity, Business, Collaboration Tools) [Refero](https://refero.design/flows/143) > This flow demonstrates the user leveraging an AI-powered document generation feature within a collaborative whiteboard environment to create a retrospective summary. The user starts by selecting or highlighting relevant content on the board, then inputs a prompt or selects a document type for AI generation, triggers the generation process, and finally reviews the generated retrospective summary content within the board. The flow assumes the user is authenticated and working within a project board that contains relevant elements such as ideas and solutions. Steps: Access AI document creation panel → Input or select prompt for retrospective summary → Trigger AI document generation → View initial generated content → Observe continued AI document completion → Review completed retrospective summary **Around: Collaborative Meeting Notes** (web, 6 screens, Business, Communication Tools, Utility) [Refero](https://refero.design/flows/2001) > A logged-in participant uses an in-meeting notes editor that opens from the meeting's bottom toolbar. Steps: Open Share Menu → Open Notes Editor → Enter Text → Open Color Picker → Inspect More Options → Add Mentions and Finalize Content **Granola: Edit Shared Meeting Notes** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14165) > The user starts on a shared note titled “Dashboard design and analytics,” opens the note actions menu, chooses to edit the notes, modifies the content in a full-screen editor, and confirms the changes. The updated note is displayed with the revised “Next Steps” content. The flow begins on an existing note detail view and ends on the saved note. The screenshots imply a mobile device and an existing note that the user has permission to edit; authentication and permission states are not shown. Steps: Note Detail → Note Actions → Edit Notes → Review Changes → Saved Note **Instagram: Share Live Video to Feed** (ios, 20 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/3649) > Happy path: User ends an ongoing live video, chooses to share the recorded video, configures post metadata (caption, tags, location, series, advanced settings including profile cover and comment/like toggles), sees a remix consent modal, and publishes the video to their feed. Steps: End Live Confirmation → Post-live Bottom Sheet → Discard Confirmation → New Post - Metadata Overview → Activate Caption Input → Type Caption → Tag People - Entry → Tag Search (empty) → Tag Search Results → Add Tag to Post → Add to Series - Entry → Create New Series → Save Series & Select → Advanced Settings → Edit Profile Cover (Crop) → Save Profile Cover → Remix Consent Modal → Begin Posting → Feed - New Post Preview → Published Video Playback **Raycast: Edit Generated Prompt** (ios, 3 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/10176) > Happy-path: A user reads a generated writing prompt and then selects a portion of the text to edit, invoke helper actions (Proofread / Rewrite), and continue inline editing with the keyboard and formatting controls. Entry: user has a generated prompt open in a mobile text editor (document visible, character count shown). Exit: the user has an insertion point or edited text in the same document and the editor remains active. Assumptions: user is on a mobile device, document is open and editable, keyboard can appear, and basic text actions (select, format, undo/redo) are available. Steps: Prompt Viewer → Select Text & Open Actions → Inline Editing & Formatting **Dropbox: Document Keyword Search** (ios, 5 screens, Business, Collaboration Tools, Productivity) [Refero](https://refero.design/flows/2439) > Happy path: User opens a PDF in a mobile document viewer, invokes the in-document search, types a short query (e.g., "anim"), reviews a list of matching snippets with thumbnails and page numbers, then selects a result to jump to and zoom on the highlighted match. Entry: viewing a PDF (document open). Exit: viewer scrolled/zoomed to the selected match or user closes search. Steps: Document open (viewer) → Open search input → Type query → Review results list → Jump to match and zoom **Raycast: Mobile rich-text edit & link insertion** (ios, 31 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/10169) > Happy-path mobile text-editing session on a rich text editor: user opens a long document, selects text using touch handles, applies inline formatting and block transformations (bulleted list and a link), types new list items, uses AI helpers (Proofread / Rewrite) and voice transcription, and finalizes edits. Steps: Open document (read mode) → Start text selection → Adjust selection → Select paragraph range → Open list type menu → Insert bullet list → Type list items → Move to next section → Voice transcription starts → Continue typing after transcription → Select phrase for hyperlink → Open Add Link modal → Enter URL → Confirm and insert link → Visual confirmation of link → Apply block/code style to a prompt → Use AI helper chips → Make an inline text selection → Edit or re-open Add Link modal → Final review and saved state **Willow: AI Message Drafting** (ios, 6 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/13234) > A messaging conversation screen with an AI writing assistant overlay (Scribe) used to draft or edit a message. The user is in an existing chat with Matt, opens the Scribe composer from the message area, speaks or types a prompt, reviews the generated text, can switch between multiple generated variants, and then either confirm to insert it or cancel to exit. Assumption: the user is already inside the conversation and has permission to use the writing assistant on this device. Steps: Conversation View → Scribe Prompt Editor → Voice Editing State → Generating Draft → Review Generated Draft → Confirmed Draft **Linear: Attach document to draft** (ios, 5 screens, Collaboration Tools, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/6624) > Happy-path: A user is composing a note/task and attaches an existing document from the device Files picker, then returns to the composer with the file displayed as an attachment. Steps: Open composer / prepare to attach → Choose Attach file → Browse recent files → Confirm selection → Finalize or continue editing **Grok: Attach and summarize document** (ios, 9 screens, AI Tool) [Refero](https://refero.design/flows/6088) > Happy-path conversation where a user edits an image then attaches a PDF from iCloud and asks the assistant to summarize it. Steps: Conversation with image edits → Open attachment picker → Pick Files (iCloud) → File selected and attached → Compose instruction with keyboard → Submit request (processing) → Assistant returns summary (first view) → Scroll through full summary → Final state and next actions **Trip Way: Manage Documents** (ios, 3 screens, Travel & Booking) [Refero](https://refero.design/flows/8457) > Happy path: a user opens a modal-style Documents view that displays a scrollable 2-column grid of document thumbnails and filenames. The user can inspect the list, open a contextual action sheet for any tile (Edit or Delete), and use the Add Doc CTA to add more documents. Steps: Documents gallery → Document contextual actions → Gallery updated / modal state **Dropbox: Add date text to scanned form** (ios, 5 screens, Business, Collaboration Tools, Productivity) [Refero](https://refero.design/flows/2444) > Happy-path summary: A user opens a scanned document in an annotation editor and adds a dated text field (or signature) to the form, positions and resizes it, edits the text with the keyboard, changes its color, then saves and/or shares the updated document. Steps: Open annotation editor → Insert text field → Position and resize text → Edit text content → Style and finalize annotation **Linear: Editor — Create & Format Description** (ios, 3 screens, Collaboration Tools, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/6625) > Happy path: A signed-in user opens the 'Refero Design' create modal on an iPhone to enter and format a descriptive paragraph, assign tags/metadata, then create the item. Steps: Select Text in Editor → Apply Inline Formatting → Place Caret and Access Block Controls **Willow: Scribe Message Drafting** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/13233) > This is a mobile messaging conversation where the user opens a writing assistant called Scribe from the message composer, dictates or enters a prompt, reviews generated text, and then confirms it to insert into the message field. Entry appears to be inside an existing chat thread on iPhone with the keyboard open. The flow ends back in the conversation composer with assistant-generated text ready to send or edit. Steps: Chat Composer → Scribe Prompt Entry → Prompt Suggestions → Generating Draft → Review Generated Draft **Givingli: Card Message Editor** (ios, 2 screens, Lifestyle, Scheduling Tools) [Refero](https://refero.design/flows/8159) > This is a mobile (iPhone) in-place text editing experience inside a greeting card template. Steps: Enter message text → Adjust text frame and finish **Verse: Text overlay editor** (ios, 2 screens, Social Networking, Design & Media Editing) [Refero](https://refero.design/flows/3984) > Happy path: A user is editing a multiline text overlay on a dark canvas. The screen presents a live preview of the text in the center, a vertical slider on the left for size/weight adjustment, a horizontal font style selector along the bottom of the preview, and a control row with color, alignment, background toggle, and a Done action. The keyboard is visible for typing. Steps: Edit text content and style → Confirm or clear edits **Substack: Convert paragraph to heading** (ios, 4 screens, News, Social Networking) [Refero](https://refero.design/flows/4477) > A mobile (iOS) rich-text post editor where the user writes a title and body, selects a line of text in the body, opens the text-style menu, and changes that line into a heading. Entry: user is in a 'New post' editor with the onscreen keyboard visible. Exit: the selected paragraph is converted to a heading and the editor returns to the normal editing state; the Next button remains available to continue publishing. Steps: Edit post - typing → Select text → Open style menu → Apply heading style **Linear: Create and edit project document** (web, 27 screens, Business, Productivity, Project Management, Collaboration Tools) [Refero](https://refero.design/flows/6660) > Happy path: A user on a project overview opens the Resources menu to create a new project document, writes and formats content, adds inline comments, inserts images (handling an oversize upload), adds and edits a table, and returns to the project overview where the new document appears under Resources. Steps: Project overview → Add resource menu → Blank document editor → Enter title and content → Open style menu → Apply inline formatting → Start inline comment → Compose comment → Post comment → Image upload attempt → Insert alternative image → Resize image → Insert table block → Populate table → Open column menu → Rearrange columns → Block-level options → Document saved → Return to project view → Project resources updated **Linear: Edit Product Brief Content** (web, 2 screens, Business, Productivity, Project Management, Collaboration Tools) [Refero](https://refero.design/flows/6665) > Happy path: A user with edit permissions opens a project document titled "Real-Time Insight Dashboard", reviews the Overview text, then selects and edits content in the Problem section. The interface is a desktop, Notion-like rich text editor with inline formatting controls and a comment thread visible on the right. Steps: Select Overview text → Edit Problem section **Linear: Inline Document Editing** (web, 2 screens, Business, Productivity, Project Management, Collaboration Tools) [Refero](https://refero.design/flows/6661) > Happy path: A team member opens a project brief titled "Real‑Time Insight Dashboard" inside a project workspace, reads the content sections (Overview, Problem, Solution), then selects a paragraph to edit or format using the inline rich-text toolbar. Steps: Open document → Select paragraph for editing **Microsoft Copilot: Document drafting with Copilot** (web, 12 screens, AI Tool) [Refero](https://refero.design/flows/5529) > Happy-path: User starts at a central co-pilot prompt hub, opens the create menu and creates a new blank document (untitled). They paste or type a long article into the editor, select text and use the inline formatting toolbar to change a heading style. The user then composes a question to the Message Copilot from the right-hand panel; Copilot returns a summarized answer with suggested follow-up prompts and citation links. Finally the user inspects the left navigation which lists saved documents and conversations. Steps: Compose hub → Create menu → New blank document → Insert content → Select text - inline toolbar → Open style menu → Apply header → Document review → Compose assistant query → Assistant answer appears → Sources and follow-ups → Left navigation - documents list **Slite: Create and format discussion document** (web, 3 screens, Paperwork, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/1697) > Happy path: A user composes a discussion document titled "Universe" inside a team workspace. The user types or pastes body text, selects text to reveal inline formatting controls, opens color/highlight options, and prepares the draft for creation. Steps: New document editor → Select text to reveal inline formatting → Open color/highlight palette **Craft: Invite collaborator to document** (web, 6 screens, Productivity, Project Management, Paperwork) [Refero](https://refero.design/flows/7956) > Happy path: A signed-in user opens the document "Materials", clicks the Share control, and invites an external email address to collaborate. The user can choose a permission level (Editor or Viewer) before sending. After sending, the document shows a Shared indicator and the invited collaborator appears in the share panel. Steps: Open share panel → Open invite modal → Select collaborator role → Enter collaborator email → Send invite → Confirm shared state **Skiff: Document Sharing & Invite** (web, 6 screens, Communication Tools, Scheduling Tools, Paperwork) [Refero](https://refero.design/flows/1917) > Happy path: A signed-in user viewing a document opens the Share dialog, adds a collaborator email, assigns a role and scope (only this file), sends the invite, and (optionally) enables a public link. Steps: Open document viewer → Open Share dialog → Enter collaborator email → Set role and invite scope → Invite sent — pending state → Confirmation and toast **Jasper AI: Document Sharing and Link Copy** (web, 2 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/5298) > Happy path: A signed-in workspace member opens the Share modal for the document titled "Login UI Guide" from the editor/preview. The modal lets the user search for and add workspace collaborators, enable a "Make Document public to the Workspace" toggle, copy a shareable invite link, and save or close the dialog. Steps: Open Share dialog → Copy link confirmation **Hello Ivy: Inline comment and highlight** (web, 4 screens, Productivity, Project Management, Collaboration Tools) [Refero](https://refero.design/flows/4514) > Happy path: A signed-in user in the "Office" workspace opens the "Pitch deck / Plan" document, selects a portion of text, opens the inline comment composer, types a comment, and posts it. Steps: Open document → Select text to annotate → Compose comment → Post comment and annotate **Framer: Delete screen with confirmation** (web, 3 screens, Collaboration Tools, Design & Media Editing, Productivity) [Refero](https://refero.design/flows/282) > Happy path: A logged-in editor with edit permissions opens the left "Pages" panel, selects a screen from the list, issues a delete command, confirms deletion in a modal, and the editor updates selection afterwards. The flow ends with the /404 screen selected and an AI assistant panel active (generating or preparing content for that screen). Steps: Open item menu and choose Delete → Confirm deletion in modal → Editor shows new selection and AI assistant active **Manus: Create report from voice** (web, 8 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/10459) > Happy path: A logged-in desktop user creates a new task by speaking into the prompt, transcribes the audio to text, submits it, and the system generates a long-form report which the user can review and scroll through. Steps: Dashboard prompt → Recording in progress → Recording finished / Transcribe option → Transcribed prompt placed in input → Document generation (initial) → Generated report displayed → Report conclusion and suggestions → Return to prompt / Choose new content type **Skiff: Inline text commenting with mention** (web, 5 screens, Communication Tools, Paperwork, Scheduling Tools) [Refero](https://refero.design/flows/1923) > Happy path: A logged-in user in a desktop environment opens a document in a Skiff Drive workspace, selects a block of text, opens the inline comment composer, mentions a collaborator (and optionally sets a permission/role), sends the comment, and sees the comment anchored to the selected text with a thread panel open. Steps: Document view / read mode → Select text and open comment composer → Choose role/permission and start mention → Compose mention and send comment → Comment thread posted and anchored **n8n: Edit workflow description** (web, 4 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9526) > Happy-path: A user working in a workflow editor opens the workflow options and chooses to edit the workflow description. A modal dialog appears with a multiline text field. The user enters or pastes the description and saves it. Steps: Workflow canvas → Open options menu → Description modal (empty) → Description modal (filled) and save **Hello Ivy: Workspace document editing** (web, 9 screens, Productivity, Project Management, Collaboration Tools) [Refero](https://refero.design/flows/4513) > Happy-path: a signed-in user (desktop browser) working inside a shared workspace called "Office #1" opens the Pages section, views an existing meeting note, creates a new blank document, and then creates or edits a second document (Pitch deck / Plan). While editing that document the user opens the inline formatting toolbar to change text color, change background/fill color, and add a hyperlink. Steps: Open Pages — view Meeting notes → Reveal Add new document control → Create blank document → Open an existing draft (Pitch deck / Plan) → Select text to format — show formatting toolbar → Open fill/background color picker → Open link input control → Enter URL for hyperlink → Link applied and styling visible **Bezi: Signup & First-Time Onboarding** (web, 24 screens, Design & Media Editing) [Refero](https://refero.design/flows/178) > Happy path for a new desktop user who discovers the product, signs up with an email address, verifies their account with a code, provides their display name, and completes the in-app onboarding that includes creating a new file, taking an editor tour, and creating a team with invited collaborators. Steps: Marketing hero / Get started → Sign up modal (entry) → Credentials populated → Verification prompt (code error shown) → Verification prompt (correct code) → Verification in progress → Welcome — enter name → Workspace — Drafts with onboarding callout → Editor new file modal prompt → Editor first tour slide — share/export → Editor tour — play mode slide → Back to Drafts — create a team tooltip → Create a team modal (empty) → Create a team modal (ready) → Add collaborators modal → Add collaborators (filled) → Join community prompt → Team workspace (empty) → Onboarding completion toast → Recently opened (final state) **Craft: Inline collaborative whiteboard** (web, 16 screens, Productivity, Project Management, Paperwork) [Refero](https://refero.design/flows/7945) > Happy path: a user working inside a document called “Materials” inserts a Whiteboard block, opens the whiteboard in a full‑canvas editor, uploads and arranges images, adds shapes and text, adjusts font/size/color, uses alignment/measurement guides and element context actions, then closes the whiteboard and returns to the document where the whiteboard is embedded as a block preview. Steps: Document with blank whiteboard block → Open block contextual menu → Launch full whiteboard editor → Open drawing/media submenu → Upload images to canvas → Place and accept/reject images → Add more images and switch tools → Open font family menu → Set text size → Choose a shape/icon → Pick color for annotations → Draw/position shape with guides → Add text inside shape → Open element context menu → Duplicate and arrange items → Close whiteboard and return to document **Jasper AI: Create and share blog draft** (web, 53 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/5288) > Happy-path: A logged-in workspace user starts at the 'My content' list, opens the Library to discover content templates, applies marketing filters (Marketing Function, Content Type, Funnel Stage), doesn't find a matching app, selects the 'Blog Post' template to create a long-form document, adds and inserts an image (via Unsplash, upload, or text-to-image), generates content, then shares the document to the workspace and returns to the My content list where the new draft appears. Steps: Open My content (empty state) → Open Library modal → View template cards → Open Marketing Function filter → Select marketing function options → Confirm marketing filter selection → Open Content Type filter → Select Blog & Long-form → Open Funnel Stage filter → Select Consideration stage → No results for combined filters → Open Blog Post template → Open Add Image modal (Text-to-image active) → Search Unsplash from Add Image modal → Switch to Upload Image tab → Upload image (start) → Upload in progress → Image inserted into document → Open Share modal → Save share settings and return to My content **Slite: New user signup and workspace setup** (web, 16 screens, Paperwork, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/1692) > Happy-path new-user onboarding for Slite. Steps: Marketing landing / CTA → Create account - method selection → Email entry → Verify email - code entry (valid example) → Verify email - invalid code handling → Set password - entry → Set password - strength and enable → Create team - collect metadata → Create team - submit → Invite teammates - options → Invite teammates - confirm → Tell us about yourself → Select starter docs / templates → Team ready confirmation → Workspace: Getting started doc (main UI) → First-run experience complete ##### Task: email inbox **Givingli: Send a Card to a Contact** (ios, 12 screens, Lifestyle, Scheduling Tools) [Refero](https://refero.design/flows/8216) > Happy-path: A logged-in user starts at an empty Inbox and chooses to send a digital card. They select one or more recipients from their contacts, pick a card from categories or popular cards, personalize the card (choose template and message), optionally add a gift, review delivery options (send now or schedule), then send. The flow ends at a send-confirmation screen with sharing options and a return to the Inbox where the sent item appears. Steps: Empty Inbox → Select Recipients (list) → Recipient confirmed → Card Gallery → Card Detail → Choose a Template → Card Editor (Message tab) → Add a Gift → Review & Delivery → Send Confirmation → Inbox Updated → Conversation Detail (recipient) **Fernand: Support tool sign in** (web, 4 screens, Business, Communication Tools) [Refero](https://refero.design/flows/11909) > Happy path for a support agent or founder starting from the public marketing site, moving through sign-in, and landing in the product’s main support workspace. Steps: Marketing homepage → Login screen → Login submission → Support dashboard **PayPal: Email linking for package tracking** (ios, 9 screens, Finance) [Refero](https://refero.design/flows/7253) > Happy path: A logged-in user in the Activity area discovers the empty package-tracking state and taps the 'Get live package tracking' CTA. The app offers email linkage options (Gmail or Outlook). The user chooses Gmail which opens a Google OAuth sign-in/consent flow in a webview. After signing in and accepting permissions, the app shows a success confirmation and returns to the package tracking screen where the system checks the user's inbox for package-related emails. Entry: Activity tab (logged-in user). Exit: Package tracking view with background check in progress and option to enter a tracking number manually. Assumptions: user is already authenticated in the main app; they have a Google account available and can complete OAuth in the embedded browser/webview. Steps: Activity — empty state → Link email provider selection → Google sign-in (email entry) → Google email confirmed → Google password entry → Google consent and permission review → Linking in progress → Link success confirmation → Package tracking — scanning state **HYPE: Password Recovery and Reset** (ios, 7 screens, News, Lifestyle) [Refero](https://refero.design/flows/12243) > Happy path starts on the Hypebeast account login screen after a failed sign-in attempt. The user selects Forgot Password, enters an email or username, receives a reset email, opens the email link, sets a new password, and returns to the login screen with a success confirmation. Steps: Login Error State → Forgot Password Entry → Email Input While Typing → Recovery Sent Confirmation → Reset Email Inbox → New Password Form → Reset Success Return **Meta AI: Gmail Email Audit** (ios, 12 screens, AI Tool) [Refero](https://refero.design/flows/13354) > A user starts in Meta AI and asks it to identify active paid subscriptions and assess which are useful. Meta AI explains that it needs access to an email inbox, presents Gmail and Outlook connection options, and guides the user through Gmail authorization. After Gmail is connected, the user initiates a separate request to summarize unread Gmail messages. Meta AI reads the messages and returns a structured summary with priority items, recommended next actions, and follow-up options. The flow assumes the user is signed in to Meta AI and has access to a Google account, but Gmail is initially unconnected. **Linear: Passwordless email sign-in to Home** (ios, 10 screens, Collaboration Tools, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/6603) > Happy path for a passwordless email sign-in that starts at a branded welcome splash and ends at the authenticated Home screen. The user chooses email login from a webview, enters their email, receives a temporary verification code in their inbox, copies/pastes or types the code back into the webview, submits it, then the app loads and prompts for push notifications. Steps: Welcome / Get started → Choose sign-in method → Enter email address → Check inbox / open email → Return to code entry → Submit verification code → Spinner / loading → Notification permission prompt → Home (initial state with overlay) → Home (ready) **XChat: Sign‑up and Chat Onboarding** (ios, 25 screens, Communication Tools, Social Networking) [Refero](https://refero.design/flows/11642) > Happy-path sign-up and chat onboarding for a mobile X Chat experience. Steps: Branded splash / Sign in CTA → Sign-in options → Birthday picker (initial) → Birthday picked & confirm → Enter email (empty) → Enter email (filled) → Enter verification code (empty) → Email with code → Code entry with suggestion → Code submitted (full) → Choose a password (empty) → Choose a password (filled) → Chat onboarding summary → Create passcode (enter) → Create passcode (partial) → Confirm passcode (empty) → Confirm passcode (partial) → Confirm passcode (invalid) → Notification permission prompt → Empty Chat inbox **Kin: Onboarding to Chat & Reminder Setup** (ios, 34 screens, AI Tool) [Refero](https://refero.design/flows/5431) > Happy‑path new user onboarding: user starts at an introductory hero screen and taps Get started. They answer a referral source question, then provide a display name and (optional) email. The app validates the email inline. After submitting, the user enters the conversational interface with an AI assistant (Kin). The user provides personal details via typed text or voice, the assistant proposes a weekly reminder, the user approves it, and the assistant confirms the reminder and continues the conversation. Steps: Intro hero → Referral source question → Referral selection confirmation → Name and optional email form → Email inline validation (error) → Correcting email input → Form submitted (valid email) → Assistant welcome screen → Typing first user message → Assistant proposes a plan → Offer voice input (microphone) → Microphone permission modal → Recording started → Voice recording in progress → User confirms via text → Reminder card with actions → Reminder approved confirmation → Assistant continues conversation → Conversations list → Home / welcome dashboard **Daze: Account creation & friend onboarding** (ios, 36 screens, Social Networking) [Refero](https://refero.design/flows/9114) > Happy-path new account onboarding for a social messaging product (Daze). Entry: user opens the app and sees a welcome splash. Flow collects birthdate, then asks for phone number or email; the user chooses email, enters it, and confirms via a 6-digit code. After verification, the app shows benefit screens (notifications, contact sync) and requests system permissions for notifications and contacts. The app then offers options to share contacts (select or share all), lets the user add up to 3 close friends (by contact or username), and finishes at the main Messages list. Steps: Welcome splash → Birthday picker (initial) → Birthday picker (select date) → Birthday confirm → Phone number entry → Email entry (keyboard) → Email typed → Confirm email modal → Email verification (enter code) → Verification email (external) → Feature intro: notifications → System notification permission → Contacts sync intro → System contacts permission → Contacts share options → Add 3 best friends (list) → Search username to add friend → Friend(s) added preview → Messages list (initial) → Onboarding complete / home state **komoot: Password recovery via reset email** (ios, 4 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8422) > Happy-path recovery after a failed login attempt on a mobile device. Entry: user lands on the login screen with an email prefilled and an "Incorrect password" inline error. From there the user taps "Forgot password?" to open a Reset password screen (email address prefilled). They submit the email, receive an in-app confirmation modal "Email Sent", then check their inbox and receive an email containing a temporary password and instructions to change it. Exit: user has a temporary password in email to use to sign in and update their password in account settings. Assumptions: user is on a mobile device, not currently authenticated, and has access to the listed email inbox. Steps: Sign-in with inline error → Reset password (enter/confirm email) → In-app confirmation modal → Received reset email with temporary password **Nooka: Password recovery and reset** (ios, 6 screens, AI Tool, Education, Books) [Refero](https://refero.design/flows/8125) > Happy-path recovery: a signed-out user on the Login screen uses the "Forget password?" link to request a password reset. The app sends a reset request to the backend and shows an in-app confirmation. The user opens their email, taps the reset link (a Firebase-hosted URL), is taken to a “Reset your password” form, submits a new password, and sees a final confirmation that the password was changed. Steps: Login (starting point) → Request in progress → Email sent confirmation → Reset email (inbox) → Reset password form → Password changed confirmation **Fresha: Password reset and update** (ios, 9 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4981) > Happy path: A logged-out user who has previously entered their email proceeds to sign in, chooses ‘Forgot your password?’, requests a reset link, receives it in their email, opens the link, sets a new password (meeting inline validation rules), and then returns to the sign-in screen to log in with the updated password. Steps: Sign-in (password entry) → Reset request → Reset email sent confirmation → Inbox — reset email → Change password (initial) → Validation — length error → Validation — numeric character error → Validation pass (accept new password) → Sign-in after update (success confirmation) **Wispr Flow: Password Recovery** (ios, 9 screens, AI Tool, Utility, Productivity) [Refero](https://refero.design/flows/13689) > The user starts on the login screen after an unsuccessful sign-in attempt and dismisses the error message. They choose Forgot password, complete the human-verification step, and request a reset email. After opening the email in Gmail, they follow the secure reset link, enter and confirm a new password, and submit it. The system confirms that the password was reset successfully. The flow assumes the user has an account, knows the associated email address, can access that inbox, and is using a mobile device. **Netflix: New account signup and onboarding** (ios, 41 screens, Entertainment, Video Streaming) [Refero](https://refero.design/flows/10554) > Happy path for a new user creating a Netflix account on mobile (web flow that redirects to app). Entry: marketing hero / promotional screen. Exit: user finishes onboarding and is prompted to open/download the Netflix app to start watching. Steps: Marketing Hero / Call to Action → Account Entry Choice → Email Collection (empty) → Email Input (autofill visible) → Send Sign-up Link (passwordless option) → Check Inbox / Next Options → Create Password Instead → Profile Details (name & DOB) → Language Preferences → Content Preference Grid Intro → Pick Favorites (selecting) → Finish Selections → Processing / Personalization Loading → App Download Prompt → Final Marketing / Confirmation → Email input validation state → Send Link confirmation with checkbox → Password creation validation → Language multi-select confirmation → Completed onboarding redirect **Netflix: Email code sign-in to locked profile** (ios, 12 screens, Entertainment, Video Streaming) [Refero](https://refero.design/flows/10557) > Happy-path: user starts on a promotional welcome screen, chooses to sign in, enters their email, receives a 4‑digit code by email, returns to the app to enter that code, passes a short onboarding about profiles, and then selects a locked profile which requires a PIN. After entering the PIN the user arrives at the profile’s home. Steps: Welcome / Promotional screen → Sign in — Email entry → Email entered / Continue → Code entry — blank state → Email inbox — verification code → Code entry — partial auto-fill / input → Onboarding — profiles intro (slide 1) → Onboarding — profiles slide 2 → Onboarding complete — profile lock info → Choose profile (locked indicated) → PIN entry modal → Profile home — authenticated state **WishWish: Email Sign-In and First Box** (ios, 16 screens, Entertainment, Kids) [Refero](https://refero.design/flows/14234) > The flow begins with a logged-out user on the WishWish welcome screen. The user chooses email sign-in, enters an email address, retrieves a one-time code from their email, and submits the code. After authentication, the product loads the initial experience, introduces daily free boxes, requests notification permission, presents an updates announcement, and guides the user to an available collectible box. The flow ends on the authenticated home screen with the box ready to open. The screenshots show an iPhone-sized device and assume network access, a valid email address, and successful code verification. Steps: Welcome and Sign-In → Email Entry → Email Entered → Send Verification Code → Confirmation Code Prompt → Email Code Message → Partial Code Entry → Code Verification → Account Initialization → Loading Progress → Loading Completion → Free Box Introduction → Notification Permission → Product Updates → Collection Introduction → Available Free Box **HBO Max: Password Reset and Recovery** (ios, 8 screens, Entertainment, Video Streaming) [Refero](https://refero.design/flows/10275) > Happy path for a user who cannot sign in and uses the "Forgot Password" flow to reset their HBO Max password. Steps: Sign In — Error → Reset Your Password — Enter Email → Check Inbox — Reset Requested → Email — Reset Link → Create New Password — Entry → Create New Password — Input → Password Changed — Confirmation → Sign In — Ready to Authenticate **Jace AI: Select and mark email as done** (web, 3 screens, Communication Tools, AI Tool) [Refero](https://refero.design/flows/10215) > Happy path: A signed-in user on a desktop/large-screen inbox selects a single email from the main inbox list, uses the contextual action bar that appears to mark that email as done, and receives an in-app confirmation while the inbox view updates. Steps: Inbox list (default) → Message selected — action bar appears → Confirmation and list refresh **Jace AI: Sign in to inbox** (web, 4 screens, Communication Tools, AI Tool) [Refero](https://refero.design/flows/10211) > Happy path: A new or returning user visits the product marketing site (hero/feature sections) and clicks the primary CTA to start or sign in. The user chooses an identity provider on the product sign-in screen (Google or Outlook). Selecting Google opens the Google account chooser; after selecting an account the user is authenticated and redirected into the product’s inbox interface. Steps: Marketing landing (hero) → Sign-in provider selection → Google account chooser → Authenticated inbox (dark theme) **Cushion: Email Login to Inbox** (web, 6 screens, Communication Tools, Collaboration Tools) [Refero](https://refero.design/flows/12506) > Happy path starts on the Cushion marketing/landing page for a signed-out visitor. The user can either choose the login entry point from the header or hero CTA, then move to a minimalist authentication screen where they enter an email address and request a one-time login link. After submission, the system confirms the link was sent and, once authenticated, takes the user into the workspace inbox. The exit point is the logged-in inbox view with navigation, message lists, and an empty-state/read-state mix visible. Assumptions: user is on desktop, not already logged in, and email-link authentication is available. Steps: Marketing Landing → Log In Start → Email Entered → Sending Login Link → Link Sent Confirmation → Inbox Home **Fernand: Send New Support Email** (web, 7 screens, Business, Communication Tools) [Refero](https://refero.design/flows/11894) > Happy path for sending a new support email from an inbox/inbox-style messaging workspace. The user starts on the Inbox list, opens the compose modal, enters or selects recipients, adds a subject and message, optionally formats selected text, then sends the email. After sending, a confirmation toast appears and the new conversation is added to the inbox list. Assumptions: user is already logged in, has access to the inbox, and can send from the shown support address on desktop. Steps: Inbox Overview → Open Composer → Recipient Entry → Draft Message → Edit Formatting → Send Confirmation → Inbox Update **Jace AI: Jace Trial Onboarding** (web, 35 screens, Communication Tools, AI Tool) [Refero](https://refero.design/flows/10210) > Happy path from the public Jace marketing experience into a new-user onboarding journey. The user starts logged out, selects a free trial, reviews social proof and privacy information, signs up with Google, grants requested Google permissions, answers a short discovery question, optionally invites a teammate, and enters the Jace inbox. The final state shows the connected inbox being prepared and then populated with categorized email. The flow assumes the user has a Google account and is willing to grant access to Gmail and related Google data. Steps: Product Introduction → Social Proof → Privacy Assurance → Account Signup → Google Account Selection → Initial Google Consent → Discovery Question → Google Permission Review → Permission Authorization → Team Invitation → Add Teammate → Send Invitation → Invitation Confirmation → Welcome Preparation → Inbox Synchronization → Initial Inbox → Prepared Workspace → Signup Testimonial Carousel → Authorization Loading → Permission Selection State **Missive: Team sign-up and onboarding** (web, 11 screens, Business, Communication Tools, Collaboration Tools) [Refero](https://refero.design/flows/6245) > Happy-path: A visitor arrives on the product marketing landing hero, chooses to create an account, completes a short sign-up form (SSO or email + password), answers a short onboarding questionnaire (goals, team size, industry, referral), and then lands inside the product (empty team inbox). Steps: Marketing hero / CTA → Create account — auth options → Email entry → Password entry and agreement → Onboarding questionnaire — start → Onboarding — select goals → Onboarding — select team size → Onboarding — select industry → Onboarding — review and continue → Onboarding — ready to start → Authenticated workspace — empty inbox **Jace AI: Snooze an inbox message** (web, 5 screens, Communication Tools, AI Tool) [Refero](https://refero.design/flows/10216) > Happy path: a signed-in user in the Main inbox suspends a single message until a later time. Steps: Main inbox overview → Reveal quick actions → Open time input dialog → Enter and select time → Confirmation and removal from inbox **Jace AI: Find messages by contact** (web, 3 screens, Communication Tools, AI Tool) [Refero](https://refero.design/flows/10223) > Happy path: A signed-in user on desktop is viewing their main inbox list, uses the top search bar to find messages by typing a short query, selects a contact suggestion from the autocomplete dropdown, and is shown a filtered list of emails matching that contact. Steps: Inbox Overview → Search Autocomplete → Search Results (Contact) **Skiff: Inbox search and results** (web, 3 screens, Communication Tools, Scheduling Tools, Paperwork) [Refero](https://refero.design/flows/1901) > Happy path: A signed-in user (desktop) in their mail workspace uses the global search box at the top of the message list to find messages. Steps: Inbox overview → Search with no matches → Search with matches **Missive: Collaborative Inbox Sign-in to Mailbox** (web, 4 screens, Business, Collaboration Tools, Communication Tools) [Refero](https://refero.design/flows/6302) > Happy path: A visitor lands on the product marketing page, reads feature highlights and testimonials, then navigates to sign in. The sign-in screen offers social SSO (Google, Apple), SSO for enterprises, and an email/password form. After entering credentials and submitting, the system authenticates and redirects the user to their collaborative inbox (three-column layout: mailbox list, message list, message view). Steps: Marketing / Feature Overview → Sign-in Options → Credentials Entry (Form Filled) → Authenticated Inbox **Fernand: Fernand Trial Onboarding** (web, 12 screens, Business, Communication Tools) [Refero](https://refero.design/flows/11874) > A prospective customer starts on the Fernand marketing site, reviews the product positioning and feature sections, and begins a 14-day trial. They create an account by entering an email address, password, product name, and product website, then accept the terms and privacy policy. After account creation, they configure a support inbox name and email address, review the generated inbox address, and choose whether to set up email forwarding or skip setup. The user then identifies their current support tool, selects Intercom, and reaches the initial Fernand workspace dashboard. The screenshots assume a desktop browser and a new, logged-out user. Steps: Product Overview → Account Details → Completed Signup Form → Password Visibility Check → Support Inbox Details → Inbox Setup Options → Forwarding Instructions → Forwarding Instructions Continued → Setup Guidance Completion → Migration Source Selection → Intercom Selected → Workspace Welcome Dashboard **Matter: Connect and manage email sources** (web, 4 screens, Magazines & Newspapers, News) [Refero](https://refero.design/flows/9396) > Happy path: A logged-in user with an empty Inbox adds content sources (newsletters or RSS) by opening the Add Sources dialog, choosing to connect Gmail, completing the connection, and then following specific newsletter senders discovered in their Gmail account. Steps: Empty Inbox / Add Sources → Add Sources Modal → Manage Gmail — Connect Prompt → Sources from Gmail List **Square: Password reset and confirmation** (web, 5 screens, Business, Finance) [Refero](https://refero.design/flows/1950) > A password recovery path that starts from a failed sign-in and guides the user to request a password reset by entering their registered email. Steps: Sign-in attempt — incorrect password → Reset request — empty email → Reset request — email entered → Confirmation — check your inbox → Confirmation — additional guidance **Skiff: Select and Move Messages to Trash** (web, 4 screens, Communication Tools, Scheduling Tools, Paperwork) [Refero](https://refero.design/flows/1887) > Happy path: a signed-in user is viewing an alias inbox with two messages, selects both messages using the list checkboxes, and deletes/moves them to Trash via the list toolbar. Steps: Inbox with unread messages → Select messages → Execute delete (move to Trash) → Empty inbox state after action **Lemni: Inbox Conversation Triage** (web, 3 screens, Business, Productivity, AI Tool) [Refero](https://refero.design/flows/4594) > Three-column inbox workspace for triaging and acting on conversations. Steps: Open conversation in inbox → Adjust list order and filters → Run quick actions and view details **Skiff: Manage Email Aliases** (web, 3 screens, Communication Tools, Scheduling Tools, Paperwork) [Refero](https://refero.design/flows/1912) > User navigates to Settings > Aliases while signed in (desktop/web view). The Aliases panel lists configured email aliases with avatars/initials, marks the account default, and shows an Add alias button. Below the list are wallet connection buttons for Web3 identities (MetaMask, Coinbase, Phantom, etc.) and a toggle for sorting alias emails into separate inboxes. From this screen the user can remove an alias via an inline menu which opens a confirmation modal. After confirming deletion the modal closes and the aliases list refreshes to reflect the removed address. Steps: Aliases overview → Confirm delete alias → Updated aliases list **Tidal: Profile email verification** (web, 5 screens, Music Streaming) [Refero](https://refero.design/flows/3278) > Happy path: A logged-in user visits their account profile which shows an unverified email banner. From the profile the user initiates email verification, confirms the target email, receives a verification email, clicks the link in that email (external/inbox step), and lands on a success confirmation. After verification the user returns to the profile and the UI reflects the verified status (banner/indicator removed). Entry: logged-in user on Profile. Exit: profile view with verified email and normal subscription details visible. Steps: Profile overview (unverified email) → Confirm email before sending → Check your inbox (instructions) → Verification success confirmation → Profile overview (verified email) **Missive: Email conversation with inline tasks** (web, 3 screens, Business, Communication Tools, Collaboration Tools) [Refero](https://refero.design/flows/6260) > Happy-path: A signed-in user opens the team inbox, selects an email from the list, reads the message in the right-hand reading pane, interacts with task items embedded in the conversation (opens status controls and updates a task), and then continues messaging or closes the thread. Steps: Open message in three-column inbox → Open task status menu → Apply status and return to thread **HBO Max: Password Reset via Email Link** (web, 8 screens, Entertainment, Video Streaming) [Refero](https://refero.design/flows/10265) > Happy-path: A signed-out user who cannot sign in uses the "Forgot Password?" recovery path. From the Sign In screen they open the Reset your password form, submit their account email, receive an email with a reset link, follow that link to a Create New Password form, save a new password that meets displayed requirements, and return to the Sign In screen. Steps: Sign In (error shown) → Reset your password (empty) → Reset your password (email entered) → Check Your Inbox (confirmation) → Password reset email (inbox) → Create New Password (empty) → Create New Password (filled) → Password Changed (confirmation) ##### Task: eye test vision **MyMind: Visual Reference Search** (web, 3 screens, Productivity, Reference, Inspiration Board) [Refero](https://refero.design/flows/13313) > A user starts on a personal visual search dashboard with content grouped across images, books, web pages, and videos. They enter a search term, first producing an empty state for "00000," then replace or refine the query with "or" to view matching content. The flow assumes the user is already in the search workspace and can access indexed personal or curated references. It ends on a results view showing tags and a masonry-style collection of books, artwork, design references, and video content. Steps: Search Dashboard → Empty Search Results → Browse Matching Results ##### Task: file sharing storage **WhatsApp: Export Chat and Share** (ios, 5 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12669) > Happy path for exporting a WhatsApp chat and sharing the exported ZIP file through iOS Messages. **Voicenotes: Add recording by upload** (web, 5 screens, Collaboration Tools, AI Tool) [Refero](https://refero.design/flows/6996) > Happy path: A signed-in user on a desktop-like interface views their list of voice recordings and uses the persistent bottom control to open quick actions, select Upload, and submit an audio file. The UI shows an upload progress state inside the action sheet. After the upload completes, a new recording item is added to the top of the list with a transcribing status, then becomes available for playback. Steps: Recordings list (idle) → Open quick-action sheet → Start upload (progress) → Upload complete — new recording created → Recording processed and playable **Dropbox: Add Photos to Mobile Uploads** (ios, 8 screens, Business, Collaboration Tools, Productivity) [Refero](https://refero.design/flows/2421) > Happy path: A signed-in user on the Home tab is prompted to add content. The user taps Add content, which opens the system photo picker. They select multiple images (checkmarks appear), preview the selected items, confirm the selection (Done/Add), and start the upload. The app shows a preparing/uploading overlay, then navigates to a folder (Mobile Uploads) that displays upload progress for each file and a warning about backgrounding. When uploads finish the folder list updates with the new filenames and modification timestamps. Entry: Home screen with empty Dropbox. Exit: Mobile Uploads folder showing the uploaded files. Assumptions: user is logged in, has granted photo library access, and is on an iOS device using the native photo picker. Steps: Home - Add content prompt → System Photo Picker - Grid → Multi-select state → Selected Preview - Fullscreen → Confirm Selection → Preparing to upload overlay → Mobile Uploads - Uploading → Mobile Uploads - Completed list **PocketShelf: Data Backup & Import** (ios, 4 screens, Books) [Refero](https://refero.design/flows/8300) > Happy path: a user creates an export backup of their PocketShelf data and saves it to a local folder via the iOS share/save-to-files flow. Steps: Your Data (Overview) → Backup Card — Save Action → iOS Share Sheet → Files Picker — Choose Location **Dropbox: Subscription selection and trial start** (ios, 6 screens, Business, Collaboration Tools, Productivity) [Refero](https://refero.design/flows/2419) > User selects a paid Dropbox plan from a modal-like subscription sheet and begins a 30-day free trial. Entry: user opens the “Do more with Dropbox” subscription sheet while using the mobile product. Happy path: user reviews plan cards (Plus vs Professional), taps the primary CTA 'Try free for 30 days' on the Professional plan, reviews trial details and FAQs, expands subscription details/terms as needed, and then proceeds — after confirmation they land on a short onboarding screen with setup options. Assumptions: user is already signed in; the interface shows there is a card on file (billing will use that card); this is a mobile native web/modal experience with top-level Back/Cancel controls. Steps: Plan comparison / selection → Alternate plan view / highlight → Trial explanation / FAQ entry → FAQ expanded answers → Subscription details expanded → Post‑confirmation onboarding choices **Dropbox: Add file to folder** (ios, 3 screens, Business, Collaboration Tools, Productivity) [Refero](https://refero.design/flows/2429) > Happy path: a signed-in user in the Files tab opens a folder and uploads a document. Entry: user is inside a folder view (folder header visible). Exit: the uploaded file appears in the folder list with size and timestamp and any temporary upload UI is removed. Assumptions: user is authenticated, has permission to add files to this folder, and is on a mobile device. Visible controls include Share, Upload, Folder, Scan buttons, a sort dropdown (Modified), per-file overflow menus, and bottom navigation (Home, Files, Create, Photos, Account). Steps: Saving overlay → Upload progress → Upload complete / File listed **UNHINGED: Share Day Entry** (ios, 5 screens, Health & Wellness) [Refero](https://refero.design/flows/6339) > Happy-path: A user in the Journal view exports the currently selected day's timeline as Markdown and shares it via the iOS share sheet (example chooses Messages). Steps: Journal — Day Selected → Open Share Day Menu → Select Export Format → System Share Sheet → Compose with Exported Markdown **Meta AI: High-Protein Meal Plan Artifact** (ios, 11 screens, AI Tool) [Refero](https://refero.design/flows/13375) > A user starts in the Artifacts library, selects the “Build a high protein meal plan” preset, and submits a prefilled request to Meta AI. The system shows staged generation progress, then presents a completed “Viral meal prep plan” response with a linked HTML artifact. The user opens the artifact, reviews the five-day meal overview, and scrolls through recipe ingredients, storage guidance, preparation steps, timing blocks, nutrition labels, and source notes. The flow assumes the user can access the Artifacts library and is using a mobile device. The final state is the fully generated meal-prep artifact; no sharing or editing action is shown as completed. **Dropbox: Make file available offline** (ios, 6 screens, Business, Collaboration Tools, Productivity) [Refero](https://refero.design/flows/2434) > Happy-path: The user is viewing a folder containing a PDF and wants to make that file available offline. From the folder listing they open the file's action sheet, toggle the "Make available offline" option, observe download progress, and return to the folder where the item shows as downloaded and available offline. Steps: Folder listing → File actions sheet → Download started → Download progress → Download completes → Available offline confirmation **Binsoo: Settings: Clear Cache** (ios, 2 screens, Design & Media Editing) [Refero](https://refero.design/flows/6811) > Summary: This is a two-step happy-path of a mobile settings screen on a modern iPhone-style device. Steps: Settings Overview → Cache Cleared Confirmation **komoot: Save or Share Activity Image** (ios, 3 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8414) > Happy path: a logged-in user is creating a post for a completed activity (Run) and wants to save or share an image of the activity summary. Entry: user is on the 'Create Post' screen showing the activity background, headline (Run), primary stats, and a segmented selector (All / Stats / Map) with Stats selected. Exit: after choosing 'Save image', the app shows a Saved confirmation and the user remains on the Create Post screen. Assumptions: user is on a mobile device, has an activity loaded into the Create Post editor, and the Share control triggers the OS-level share/save sheet. Steps: Create Post (preview) → Share Sheet → Saved Confirmation **Moonly: Meditation Player — Full Moon** (ios, 5 screens, Lifestyle) [Refero](https://refero.design/flows/9903) > Single-content audio player (mobile). Steps: Player opened (idle) → Start playback → Initiate download (in-progress) → Download complete → Bookmark / secondary interactions **Dropbox: Create new folder** (ios, 8 screens, Business, Collaboration Tools, Productivity) [Refero](https://refero.design/flows/2424) > Happy-path: A logged-in user creates a new folder, picks or changes the destination (Dropbox or Mobile Uploads), and confirms creation. Steps: New folder form → Rename folder (keyboard) → Rename confirm (return to form) → Choose a different folder (destination picker) → Browse Mobile Uploads → Confirm destination and create → Creation progress feedback → New folder view (empty) **Moises: Song Library** (ios, 3 screens, AI Tool, Music Streaming) [Refero](https://refero.design/flows/6842) > Happy path: A signed-in mobile user opens the Songs tab (bottom navigation) to view their music library. From the list view they open the filter control, choose the "Shared" subset, and the list refreshes. If there are no shared songs the interface shows an empty-state illustration and instructions to add songs. Steps: Songs list (All) → Filter selection → Shared list (empty state) **Playground: Birthday Card — edit & save** (ios, 61 screens, AI Tool, Design & Media Editing) [Refero](https://refero.design/flows/8991) > Happy-path: user searches/browses birthday card templates, selects one, waits for the editor to load, then customizes the design by changing the background, importing an image, adding a rounded shape and text label, adjusting shape appearance (color, radius, opacity, blend), then saves/exports the final image. Entry: search or template selection. Exit: edited design saved to camera roll and appears in the user's My Designs gallery. Assumptions: user is on a mobile device, has photo-library permission, and has at least one edit credit or required model available. Steps: Template search results → Editor loading (initial) → Editor loading (progress) → Template opened in editor → Model selection dropdown → Edit Background — source picker → Photo library chooser → Selected background preview → Edit canvas and select insert tool → Shape selected — property toolbar → Adjust corner radius → Change stroke / border and opacity → Add text overlay → Style text (underline & alignment) → Layer and final layout adjustments → Share / export → Save to camera roll → Return to My Designs gallery → Alternate background edit quick actions → In-editor object actions summary **Dropbox: Share file from folder** (ios, 3 screens, Business, Collaboration Tools, Productivity) [Refero](https://refero.design/flows/2433) > Happy path: a signed-in user opens a folder, selects a file and taps the Share control. A share sheet (bottom sheet) slides up showing recipient input and link-sharing options. The user types an email address, confirms the default "Can view" permission, and taps Share. The app validates the recipient locally (shows "No Dropbox contacts found"), then performs the share and displays a temporary confirmation banner. Steps: Open share sheet → Enter recipient and confirm permission → Confirm share and show confirmation **WhatsApp: Multi-file Document Sharing** (ios, 8 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12657) > A user in a one-to-one conversation with Samuel opens the attachment menu, chooses the document option, selects multiple recent files from the system file picker, reviews an audio file, and sends the selected files. The conversation then displays the files as outgoing messages with upload indicators that later change to sent checkmarks. The user is assumed to be on a mobile device and already inside the conversation. **MD Vinyl: Music Share Options** (ios, 2 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/12541) > A share screen for a music release opens with a large preview card and quick actions. The user can copy a link, save the image, share to Instagram Stories, or open more options. In the shown happy path, the user taps Copy Link and receives a brief confirmation message. Steps: Share Options → Link Copied Confirmation **District: Movie Share Sheet** (ios, 2 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12820) > A logged-in mobile user opens the share menu for a movie and sees a bottom sheet with movie details and sharing actions. The entry point is the movie playback/detail screen, with the share icon already available on the video. The user can dismiss the sheet, copy the movie link, use the system share action, save the movie to Hotlist, or send it via chat. In the visible interaction, selecting Copy link triggers a confirmation toast while the sheet remains open over the movie screen. Steps: Share Menu → Link Copied Confirmation **ElevenReader: Add document for listening** (ios, 4 screens, AI Tool, Books) [Refero](https://refero.design/flows/8969) > Happy-path: A signed-in user on the app Home screen starts adding a document to convert/listen to by choosing an upload method (tile or central add). The app opens a file picker rooted at Recents, the user selects a file, an uploading overlay appears, and the app returns to Home showing an "Importing content" status banner while processing. Steps: Home – upload entry → Recents file picker → Uploading indicator → Importing status on Home **Dropbox: Create Team Folder** (web, 7 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/13048) > A logged-in user is viewing a team file repository and creates a new shared folder from the Team area. They enter a folder name, review access permissions, optionally open folder settings to adjust notification/automation options, and then create the folder. The flow starts on the Team files view and ends back in the Team folder list with the new folder visible. **Dropbox: Delete Shared File** (web, 4 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/13029) > A logged-in user is viewing a shared team folder in a file-management interface, selects a single asset from a grid, chooses delete, confirms a destructive action that removes the file for everyone with access, and then sees a success toast with undo/close controls. Entry is the team folder file browser; exit is the updated folder view with the file removed and a temporary deletion confirmation message. **Dropbox: Star Folder from Files** (web, 3 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/13050) > A logged-in cloud file manager view showing the user browsing all files. The user can hover over a folder card to reveal quick actions such as Share, then choose a starring/favoriting action that updates the left sidebar’s Starred section and shows a confirmation toast. Entry is the main file library; exit is the folder being added to Starred. **Dropbox: Dropbox Upgrade Prompt** (web, 4 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/13066) > This is a Dropbox file-management experience where a signed-in user starts in the All files view, is shown plan-limitation upgrade messaging, opens a modal to review or buy a plan, then exits back to the file list. The entry point is the main All files dashboard; the exit point is returning to the same dashboard after closing the upgrade prompt. Steps: All Files Dashboard → Plan Limitation Modal → Trial and Plan Comparison → Return to File Browser **Dropbox: Team File Type Filter** (web, 4 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/13037) > A logged-in user is viewing a team file library and uses the Type filter to narrow visible items by content category. The entry point is the Team files area with all items shown; the exit point is the same library view updated to show only matching files. **Dropbox: Create Team Folder** (web, 6 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/13026) > Happy path in a cloud file manager: the user starts in the All files area, opens the create-folder dialog, names a new team folder, chooses who can access it, and confirms creation. The interface then shows the new folder workspace with a success message, where the user can immediately upload files or create more folders. **Dropbox: Add Folder Tag** (web, 3 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/13034) > A logged-in user is viewing a team folder in a cloud file manager with the Info side panel open. The user adds a new tag from the folder’s tag input, then the tag is created and applied to the folder. The flow starts on the folder contents screen and ends with a confirmation message that the folder has been tagged and an option to auto-tag current and future files in that folder. **Shuttle: Access and Workspace Onboarding** (web, 4 screens, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/6382) > Happy path: an unauthenticated desktop user visits the marketing landing, clicks the primary CTA to get started, signs in (email + password), and lands in their workspace where they can browse a "Space" containing folders and media assets. Steps: Marketing landing and CTA → Login — email entry → Login — password entry → Workspace — Space view (files & folders) **Clipchamp: Create and back up a video project** (web, 57 screens, Design & Media Editing) [Refero](https://refero.design/flows/10860) > Happy path: A logged-in user starts at the home dashboard and creates a new video project. They import media from local or cloud sources, wait for uploads to finish, then drag clips onto the timeline. The user selects clips for trimming and re-positioning, plays back the edit, opens the Transitions panel and drags a transition between clips. When the user attempts to share, the editor prompts to enable OneDrive backup; the user enables backup, media uploads to OneDrive and a confirmation toast appears. The user returns to the editor (now with cloud-backed media) and then back to the home dashboard where the new project appears as a thumbnail. Assumptions: user is signed in, has permission to connect OneDrive (or other cloud sources), and has sufficient cloud storage. Steps: Home — Dashboard → Editor Loading → Empty Project Workspace → Import Sources Dropdown → Upload Progress → Media Thumbnails Appeared → Drag Clip to Timeline → Select and Adjust Clip → Play Preview → Scrub and Fine Trim → Open Transitions Panel → Insert Transition → Preview Transition → Full-screen Preview → Share Requires Backup → Enable Backup — Progress → Backup Uploading → Backup Completed Confirmation → Final Playback & Save State → Return to Dashboard — Project Saved **Clipchamp: OneDrive Backup for Project Media** (web, 5 screens, Design & Media Editing) [Refero](https://refero.design/flows/10878) > Happy path: A user in the video editor attempts to share a project. The product requires project media to be backed up to OneDrive before sharing, so the editor shows a modal asking to turn on backup. The user confirms, a backup status popup appears and shows progress, then a success state and toast confirms all media are backed up. Steps: Project Editor (Initial) → OneDrive Backup Prompt → Backup Started — Status Popup → Backup In Progress — Progress Update → Backup Complete — Confirmation **Skiff: File Import and Upload** (web, 4 screens, Communication Tools, Scheduling Tools, Paperwork) [Refero](https://refero.design/flows/1916) > Happy-path: A signed-in user in the "Skiff Drive" workspace opens the General folder and uploads files (single or multiple) using the Upload control. The import menu (Upload Files, Upload Folder, Google Drive) appears as a floating modal and shows a running upload queue with progress for each item. As uploads complete the folder list is populated with file rows showing name, type, size, and upload date. Steps: Open Import Menu → Upload Single File → Upload Multiple Files / Folder → Review Uploaded Files **Dropbox: Copy Uploaded File Link** (web, 4 screens, Collaboration Tools, Business, Productivity) [Refero](https://refero.design/flows/13028) > A signed-in user is viewing a team folder in a cloud storage interface after uploading files. The main content shows selected files in the folder, while an uploads panel on the right lists completed uploads and provides per-file actions. The user opens the link-copy dialog for one uploaded file, copies the shareable link, and receives confirmation that the link has been copied. The flow starts from the file list view and ends with a confirmed copied link, with the file still visible in the uploads panel and sharing controls available. **Shuttle: Upload Items to Space** (web, 4 screens, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/6375) > Happy path: A signed-in user (Miles Everett) opens a named storage area called "My First Space" on a desktop (macOS-style) layout, uses the top-right upload control to choose files or a folder, and uploads multiple items. The system shows an upload progress notification, renders placeholders while uploading, and then shows the completed items in the space. Steps: Open Space Overview → Open Upload Menu → Start Upload and Show Progress → Upload Complete and Items Visible **Cake Equity: Grant existing share offer** (web, 7 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/6143) > Happy-path: A logged-in company user on a specific share plan opens the Offers section and uses the 'Add new or existing offer' action to record/grant shares that were signed outside the platform. The user selects a grant date using a calendar, optionally uploads a signed contract, and confirms the grant. Steps: Share plan — Offers list → Grant existing offer — initial modal → Choose grant date (calendar) → Upload contract prompt → File selected and upload ready → Confirm grant with contract attached → Offers list — updated status **Shuttle: Create and open new space** (web, 4 screens, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/6380) > Happy path: a signed-in user opens an existing file space, invokes the control to add a new space, names the space in a modal, confirms creation, and is taken to the newly created (empty) space. Steps: Open existing space → New Space modal (empty) → Confirm creation → New empty space view **Cohere: Document Review Conversation** (web, 3 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3825) > Happy-path: a user opens a conversation titled "Business Plan for a Marketing Agency," reads the inline document, attaches PDF files in the message composer, optionally opens the right-hand Tools/Connections panel to manage integrations (e.g., Google Drive), and then closes the Tools panel to focus on the conversation. Steps: Open conversation and view document → Open Tools / Connections panel → Hide tools panel to focus **Shuttle: Edit Filename in Space** (web, 4 screens, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/6374) > Happy path: A signed-in user opens a named storage space, inspects the file grid, opens the context menu for an image file, chooses Rename, edits the filename inline, and the system saves the change and refreshes the space. Steps: Open space and view files → Open file context menu → Edit filename inline → Save rename and refresh **Shuttle: New account signup and first-run onboarding** (web, 12 screens, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/6366) > A new user discovers the product from the marketing hero, signs up with name/email/password, completes an email verification code, and lands in an empty dashboard where an in-product onboarding tour (series of tooltips) guides them through branding, storage/upload and sharing features. Steps: Marketing hero → Sign up - blank form → Sign up - filled form → Verify email - empty code → Verify email - partial input → Verify email - error state → Post-signup welcome modal → Dashboard - initial empty state with tooltip → Settings - Branding panel → Dashboard - Upload files tip → Dashboard - Share tip → Dashboard - empty state (tour complete) **Vimeo: Video editing and export** (web, 35 screens, Social Networking, Entertainment, Video Streaming) [Refero](https://refero.design/flows/712) > Happy-path sequence for editing a single video in Vimeo's in-browser editor: start from the video's metadata/edit screen, open the built-in editor, add/replace media, adjust visual properties (color/crop), add and upload a music track (accepting submission terms), arrange clips and audio on the timeline, encounter an upgrade prompt when accessing brand colors, save and render the edited video, and then return to the library/metadata screen once rendering completes. Steps: Video details / Edit entry → Editor loading → Editor canvas and timeline → Open color picker → Open Media panel (Stock view) → Library tab selected → Uploads tab (empty state) → Upload in progress / new thumbnail → Place uploaded clip on timeline → Transform / replace clip → Open Music panel (Featured) → Music uploads tab (empty) → Music submission terms modal → Uploaded track appears → Place music on timeline → Attempt to use Brand Colors (upgrade prompt) → Continue editing and save → Render / export loading → Rendering progress → Return to library / updated video ##### Task: fraud alert suspicious transaction **Ground News: Alerts Inbox Cleanup** (ios, 3 screens, News) [Refero](https://refero.design/flows/7577) > Happy path: A signed-in mobile user opens the Alerts tab, sees one or more alert items, uses the trash (bulk delete) action in the top-right to remove all alerts, confirms the destructive action in a modal, and is returned to an empty-alerts state with a message indicating no new alerts. Entry: Alerts tab is active. Exit: Alerts list is cleared and empty-state message shown. Steps: Alerts list → Delete confirmation → Empty alerts state ##### Task: grant application **Rox: Google Calendar & Gmail Connect** (web, 7 screens, Business, Productivity, AI Tool) [Refero](https://refero.design/flows/10759) > Happy path: A signed-in user navigates to the Integrations area, selects Google Calendar and Gmail from the catalog, reviews and confirms product-level permission dialogs, is redirected to Google's OAuth consent screens to grant account access, and returns to the Integrations list where each integration's card updates to show connecting/connected states. Steps: Integrations Gallery → Calendar Permission Modal → Google OAuth Consent (Calendar) → Integrations — Connecting State → Gmail Permission Modal → Google OAuth Consent (Gmail) → Integrations — Connected & Connecting ##### Task: health onboarding profile **Make: Signup, onboarding, and dashboard setup** (web, 17 screens, Business, Productivity) [Refero](https://refero.design/flows/10956) > Happy path: A new user arrives on the marketing landing page, starts account creation, verifies their email, signs in, completes a multi-step onboarding questionnaire (discovery, experience, skillset, company size, AI intent, tech stack) and is then routed into the initialized organization dashboard. Steps: Marketing landing → Sign up (initial) → Sign up (form filled) → Email verification notice → Sign in (post-verification) → Sign in (credentials entered) → Onboarding - Discovery → Onboarding - (duplicate view) → Onboarding - Experience → Onboarding - Skillset → Onboarding - Company size → Onboarding - AI intent → Onboarding - Tech stack selection → Tech stack - selections visible → Workspace initializing (partial load) → Loading spinner → Organization dashboard (ready) ##### Task: health record medical history **Granola: Search Chat History** (ios, 3 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14161) > The user starts in an open conversation and opens the Chat History panel. They select the Search field, enter a query, and submit it using the keyboard search action. The history list is then replaced with an empty-state message because no chats match the entered value. The flow begins with an existing conversation visible and ends in the Chat History search results state. The user is assumed to be on a mobile device with access to stored chat history; authentication and permissions are not shown. Steps: Open Chat History → Enter Search Query → View Search Outcome ##### Task: insurance claim health **Wabi: Service Access Settings** (ios, 3 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12043) > This mobile settings experience lets the user enable or disable data connections for several services. Steps: Service Toggles → Health Permission → Settings Updated **komoot: Device Connection - Apple Health** (ios, 4 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8347) > Happy path: A signed-in iOS user is prompted to connect a device/service (Apple Health) so Komoot can sync activities and routes. Steps: Device & Services List → Health Access Prompt (initial) → Enable Permissions & Confirm → Connected State & Continue **Brink: Enable Health Integration** (ios, 5 screens, News) [Refero](https://refero.design/flows/10923) > Happy path: while a user is in an active Walking Mode session, they open the Health Integration modal, choose to connect to Apple Health, grant the permission, and return to the running session with a confirmation banner showing the connection is active. Steps: Active Walk — Health Not Connected → Health Integration Overview → Connect Apple Health — Permission → Connecting (Progress State) → Active Walk — Health Connected ##### Task: integrations connect apps **Matter: Onboarding → Newsletter Import → Subscription** (ios, 33 screens, Magazines & Newspapers, News) [Refero](https://refero.design/flows/9331) > Happy-path onboarding, account connection and subscription purchase for new users being migrated from Pocket into Matter. Steps: Pocket migration prompt → Usage reason question → Newsletter volume question (options) → Newsletter volume confirmed → Sync newsletters prompt → System sign-in permission prompt → Google account chooser → Google basic info grant → Gmail access permission → Gmail connect success → Alternate: Enter email for magic link → Email input with keyboard → Check your email confirmation → Open Magic Link email → Subscription promo (post-auth) → Plan selection → In-app purchase / App Store confirmation → Purchase success modal → Discover (home) screen → Background import & processing (implicit) **Apple TV: Open title in HBO Max** (ios, 6 screens, Entertainment, Video Streaming) [Refero](https://refero.design/flows/11088) > Happy-path: From an Apple TV movie detail the user taps the Call-To-Action to open the title in HBO Max. The system shows a branded modal explaining a connection will be made, then switches to the HBO Max app (splash + welcome screens). HBO Max presents onboarding/television-notification prompts. After the user continues, control returns to the Apple TV detail which now shows an inline Play button (indicating the title is playable via HBO Max). Steps: Movie detail (source) → Provider connect consent → HBO Max launch (splash) → HBO Max landing / welcome → Watch on television prompt → Return to movie detail (playable) **Verse: Connect & Add Song** (ios, 12 screens, Social Networking, Design & Media Editing) [Refero](https://refero.design/flows/3999) > Happy path for attaching a streaming song to a wallpaper composition. Entry: user is in the wallpaper editor with a music control visible (not yet connected to a streaming account). Exit: the user's streaming account is connected, a song is added to a chosen streaming playlist, and the wallpaper editor shows the added song (small album indicator). Assumptions: user is on an iPhone, has Spotify installed (or Apple Music available), and is logged into the streaming app; the Verse app is allowed to open other apps and start an OAuth authorization flow. Steps: Editor - Music Control → Add a song - Popular → Add a song - Your Top Tracks (empty) → Connect Streaming Service - Provider Choice → iOS Handoff Prompt → Spotify Authorization Start → Spotify Authorization - Agree → Connection Success → Add a song - Your Playlists → Playlist - Track List → Add Confirmation Toast → Editor - Song Attached **Tinder: Spotify connect for music profile** (ios, 6 screens, Dating & Relationships) [Refero](https://refero.design/flows/3433) > Happy path for connecting a Spotify account to a Tinder profile to surface a user's anthem and top artists. Entry: user viewing Tinder profile music section (not connected). Exit: Tinder shows Spotify-linked content (anthem, top artists) and returned controls for sharing. Assumptions: user is on iOS, logged into Tinder, has a Spotify account (may be logged out), and Tinder uses OAuth via a webview/redirect to Spotify for authentication and consent. Steps: Music preferences (Tinder) → iOS external sign-in prompt → Spotify sign-in (webview) → Spotify home (authorizing) → Spotify authorize consent → Tinder connected (top artists shown) **Longevity Deck: Enable Health integration** (ios, 5 screens, Health & Wellness) [Refero](https://refero.design/flows/9883) > Happy path for a user enabling Apple Health integration from the app's main screen. Steps: Home with Connect prompt → Connect panel (in-app) → Health Access - initial → Health Access - enable categories → Post-permission state **Joi: First‑launch setup** (ios, 25 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/4374) > A first‑launch onboarding sequence for Joi (iOS). Steps: Welcome / Intro → Set wake / wind‑down → Native time picker (open) → Adjust wake time → Times confirmed → Connect calendar (explain) → Calendar permission prompt → Calendar allowed state → Enable notifications (explain) → Notification permission dialog → Pick starter habits → Subscription prompt (in‑flow) → App Store purchase sheet → Purchase success → Add Reminders (explain) → Reminders permission dialog → Connect HealthKit (explain) → HealthKit permission sheet → Onboarding complete → Main daily timeline (first run) **BeReal.: Connect Music Service** (ios, 5 screens, Social Networking) [Refero](https://refero.design/flows/4940) > Happy path for linking a third‑party music service (Spotify shown) from the user's Settings. Steps: Settings — Audio entry → Audio — Service list → iOS domain sign‑in prompt → Spotify OAuth consent → Audio — Connected confirmation **Pinkllama: Apple Health Sync Setup** (ios, 8 screens, Health & Wellness) [Refero](https://refero.design/flows/5660) > Happy path: from the app's Settings list the user opens the Apple Health section, reads a short explanation, and is guided through iOS Health permission screens. After granting access the app returns to its Apple Health Sync settings, enables automatic export and offers a manual "Sync All Data" action. Entry: Settings > Apple Health row. Exit: Apple Health Sync screen with SYNC toggle ON (and confirmation). Assumptions: user is on iOS, has the built-in Health app, and the app can call the HealthKit permission UI. Steps: Settings — Apple Health row → Apple Health Sync Intro (bottom sheet) → Permission Instructions → iOS Health Access Prompt → Permissions Confirmed → Start Sync (manual) → Sync Success Alert → Final Sync State **Instagram: Instagram account signup and onboarding** (ios, 45 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/2833) > Happy-path new user signup and initial onboarding on iOS. Entry: unauthenticated user opens Instagram and taps "Create new account" (from the login screen). The flow collects contact (phone/email) info, full name, password, birthday, and username; asks permission to sync contacts; optionally connects to Facebook/Accounts Center; presents follow suggestions; optionally saves login info to device; then lands the user on the home feed. Assumptions: user is on iOS, not logged in, has a working network connection, and grants or denies optional permissions (contacts/Facebook) which affect later steps. Steps: Login / Entry → Contact / Start Signup (blurred) → What's your name? (empty) → What's your name? (filled) → Create a password → Save login info? (iCloud suggestion) → What's your birthday? → Create a username (suggestion accepted) → Create a username (validation error & suggestions) → Sync contacts intro → iOS Contacts permission modal → Get Facebook suggestions (Accounts Center) → Facebook webview login → Remember login info? (post-Facebook / account) → Discover people (initial suggestions) → Discover people (follow state updates) → Loading home feed (spinner) → Now in Accounts Center (welcome) → Home feed (initial) → Onboarding complete / steady state **GO Club: Go Plus subscription purchase** (ios, 14 screens, Health & Wellness) [Refero](https://refero.design/flows/6353) > Happy-path for upgrading to the app's premium subscription (Go Plus). Steps: Dashboard (entry) → Subscription modal — hero slide → Subscription modal — widgets slide → Subscription modal — golden hour slide → Subscription modal — Apple Watch slide → Subscription modal — activity planner slide → Pricing card — initial Continue → Pricing list expanded → Family Sharing toggle → Select plan → Confirm purchase (App Store sheet) → Authenticate purchase → Purchase success confirmation → Welcome / onboarding **Canva: Manage connected social accounts** (web, 3 screens, Design & Media Editing, Collaboration Tools) [Refero](https://refero.design/flows/8078) > Happy path: A signed-in user navigates to their personal profile settings to manage connected social accounts. From the 'Your profile' screen they locate the connected Pinterest card and choose to disconnect it. A confirmation modal appears; the user confirms the action. The UI returns to the profile screen with the Pinterest connection removed. Steps: Your profile — Connected accounts → Confirm disconnect modal → Your profile — Post-disconnect state **Canva: Bring Google Drive files into design** (web, 7 screens, Design & Media Editing, Collaboration Tools) [Refero](https://refero.design/flows/8063) > Happy-path: A signed-in user in Canva's main workspace discovers and opens the Google Drive integration from the Apps area, chooses to bring Drive files into an existing design, connects their Google Drive account, and then sees Drive folders and assets in the editor's left panel ready to be used. Entry: user on Canva "All apps" Apps marketplace. Exit: Google Drive connected and Drive assets available inside the editor. Assumptions: user is signed in to Canva, has at least one existing design/project, and has Google Drive files accessible in the Google account used during connection. Steps: Apps marketplace – All apps → Google Drive app preview modal → Folder / project picker (collapsed) → Project list selection → Editor — Google Drive connect prompt → Connection in progress → Editor — Google Drive browser **Parallel: Slack notification integration** (web, 4 screens, AI Tool) [Refero](https://refero.design/flows/9077) > Happy path: A logged-in user on the product's Monitor screen chooses to connect Slack so monitor alerts can be delivered to Slack channels. The user clicks the Connect Slack control, reads the integration details, clicks Add to Slack, completes Slack's OAuth consent for a chosen workspace, and receives an in-app Installation Successful confirmation with next steps for adding the bot to a specific Slack channel. Steps: Monitor list / integrations entry → Connect Slack — details and CTA → Slack OAuth consent → Installation Successful / next steps **Manus: GitHub connector installation** (web, 10 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/10427) > Happy-path summary: A logged-in user opens the Connectors modal, searches for and selects the GitHub connector, reviews the connector details, and clicks Connect. The app initiates an OAuth install flow that redirects the user to GitHub. On GitHub the user reviews requested permissions, selects where to install the Manus Connector (personal account or organization), chooses repository access scope, and verifies their identity if prompted. After authorization, GitHub redirects back to the original app and the connector appears as successfully installed. The user can then add repositories, manage the integration, or start using GitHub features inside the product. Entry point: Connectors modal in the product (user is already signed in). Exit point: Connector marked installed and actionable inside the product (user can manage repos or try the integration). Assumptions: User is authenticated in the product and in GitHub; user has sufficient GitHub permissions to install apps on the chosen account/org; browser permits OAuth redirects; network and GitHub availability are normal. Steps: Connector Browser → Connector Overview → Connector Metadata → Initiate OAuth → GitHub Permission Request → Choose Install Target → Repository Access Scope → Confirm Access (verification) → Return to Product — Installed → Post-install Actions **Visitors: Stripe integration setup** (web, 4 screens, Business) [Refero](https://refero.design/flows/9818) > Happy path: A user chooses to connect a payment provider (Stripe) so the product can attribute revenue back to visitors. Steps: Integration selection → Connect Stripe — empty key → Key entered — ready to submit → Validation error shown **Factory: New account setup and onboarding** (web, 19 screens, Business, Development Tools, AI Tool) [Refero](https://refero.design/flows/7723) > Happy-path new user signup and first-time onboarding. Entry: marketing/home hero or direct sign-in link. The flow guides a new user to sign up (or sign in), verify their email with a one-time code, read two short product microcopy pages, create an organization name, choose where they will use the product (Browser/IDE/Terminal), connect repository integrations (org + personal), install/connect the local Bridge (CLI package + connection code + root directory), then finishes to a welcome screen and the initial dashboard. Assumptions: user is on desktop, not previously registered, has an email for verification, and has access to their Git provider if connecting integrations. Steps: Marketing hero / entry → Sign in → Sign up (empty) → Sign up (filled) → Email verification (code entry - empty) → Email verification (invalid code) → Email verification (valid code) → Product interstitial: FACTORY → Product interstitial: DROID → Organization name (placeholder) → Organization name (entered) → Platform selection (Browser / IDE / Terminal) → Platform selection (details / continue) → Integrate your codebase (initial) → Integrations (org GitHub connected) → Integrations (personal GitHub connected) → Integrate to local (Bridge) → Final welcome → Initial dashboard / workspace **Linear: New workspace onboarding** (web, 24 screens, Business, Productivity, Project Management, Collaboration Tools) [Refero](https://refero.design/flows/6631) > Happy-path new-user onboarding that takes a visitor from the marketing landing through account sign-up, workspace creation, a short interactive tutorial and optional integrations/invitations, and finally into the main app. Entry: marketing/landing. Exit: user lands in the app Issues view with onboarding hints. Steps: Marketing landing → Choose auth method → Email entry (empty) → Email entry (filled) → Check email / code option → Enter login code → Code verified / create workspace visible → Workspace form (empty) → Workspace details filled → Region selection → Interactive tutorial: command palette → Tutorial confirmation → Connect GitHub prompt → Invite co-workers (empty) → Invite co-workers (emails added) → Subscribe to updates → Ready to go → App view with onboarding popover → App view (popover dismissed) → Onboarding complete **Cofounder: Sign-up to Workspace Onboarding** (web, 3 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/7906) > Happy path: an anonymous visitor arrives at the Cofounder marketing/home page, reads product messaging and clicks to sign up or log in. The user selects the sign-up method (prominently a Google sign-up) and authenticates, which transitions them into the main authenticated workspace. In the workspace the user sees a personalized greeting, a large chat input for natural-language commands, a templates grid showcasing ready-made automations, a left navigation for Chats/Flows/Memory/Integrations, and status indicators (e.g., memory indexing). Steps: Marketing / Landing → Authentication / Sign-up → Authenticated Workspace **Copperx: Deposit and cross-chain bridge to USDC** (web, 17 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8894) > Happy-path: A logged-in user on the Copperx dashboard uses the Deposit flow to convert and bridge funds from a source chain/token (ETH on Optimism) into USDC on a destination chain (Polygon) and deposit that USDC to their Copperx wallet. Steps: Dashboard – start deposit → Choose USDC network → Deposit fund modal – connect wallet prompt → Select source token → Enter amount & view quote → Quote details expanded → Get gas toggle & Connect CTA → Choose wallet connector → WalletConnect QR / app selection → Wallet connected – refreshed quote → Ready to review → Bridge summary / Start bridging → In-progress – initial steps → In-progress – detailed step list → Waiting for destination chain → Bridge successful modal → Dashboard updated balance **Make: Create and configure MCP toolbox with AI module** (web, 15 screens, Business, Productivity) [Refero](https://refero.design/flows/10976) > Happy path: A user creates a new MCP Toolbox, adds an AI module (Anthropic Claude -> Simple Text Prompt) as a tool, configures the module (name/description/model/prompt), creates and copies a client key, and views the toolbox detail. Steps: Empty toolboxes / start → Create toolbox modal → Choose tool type → App search list → Filter providers → Provider detail & module list → Add module tool modal → Select model → Module add in progress → Tool selected and toolbox named → Create key modal → Select connect target → Copy key confirmation → Toolbox detail overview → Toolbox list updated **Marathon: Account & Workspace Setup** (web, 28 screens, Business, Productivity, Project Management, Collaboration Tools, Scheduling Tools) [Refero](https://refero.design/flows/4840) > Happy-path new-user onboarding: user arrives on the marketing/home hero, clicks the primary CTA to sign up, completes an email + password form (with inline validation), sets up a profile (photo + name), chooses to create a new team, provides a team name and optional logo, then creates at least one project (selecting color, sharing rules and optionally connecting a Figma file). The flow ends at the app dashboard with the new projects listed and an empty board ready for items. Assumptions: user is not signed in, has permission to create a team and projects, and is using a desktop browser. Steps: Marketing hero → Sign up form (empty) → Sign up form (filled) → Password mismatch error → Sign up success (submit) → Profile setup (empty) → Profile setup (photo + name) → Team requirement prompt → Create team - basics → Create team - logo uploaded → Create project modal (color picker open) → Change project color → Project sharing options → Figma integration - add file → Figma file connected (preview) → Project create summary → Add more projects → Edit project settings → Finish onboarding → Dashboard → Empty board / initial dashboard **Visitors: Account and Project Setup** (web, 16 screens, Business) [Refero](https://refero.design/flows/9816) > Happy path: A new user creates an account using email, verifies with a code, provides personal info (name + optional avatar), creates a project by naming it and providing a domain, optionally connects a payment provider, and is taken to the project dashboard which shows an install modal with a tracking script and a globe-based live view waiting for events. Steps: Sign up entry → Email entered → Verification code prompt → Verification code entered → Name capture (empty) → Name entered → Profile picture prompt (default) → Profile picture uploaded → Project name prompt → Project name entered → Domain prompt (empty) → Domain entered → Integrations prompt → Dashboard — install modal → Dashboard — waiting view → Live metrics preview **Make: Gmail to Google Calendar event** (web, 15 screens, Business, Productivity) [Refero](https://refero.design/flows/10964) > User is building an automation scenario that triggers from Gmail and creates events in Google Calendar. Steps: Scenario canvas with Gmail module → Open module picker → Search for calendar integrations → Select Google Calendar → Open Google Calendar action → Connection required prompt → Validation message for missing connection → Create connection dialog → Authorization in progress → Connection added and fields shown → Select calendar ID → Map event name from Gmail output → Insert mapping token → Confirm mapped Event Name → Save module and show link **Cal: New user onboarding and setup** (web, 20 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3480) > Happy‑path new user onboarding for Cal.com: user starts on marketing/home, chooses to create an account, completes account creation (email/password or SSO), verifies email, and proceeds through a 5‑step guided setup: basic profile (username/timezone), connect calendars, connect video apps, set weekly availability, and add profile photo/about. The flow finishes in the authenticated dashboard (Event Types). Steps: Marketing / Home (Entry) → Create account - choose method → Create account - form (empty) → Create account - form (valid) → Email verification (Check your email) → Welcome — basic profile (username/timezone) → Welcome — profile filled → Connect your calendar → Connect video apps → Set your availability - weekly ranges → Copy times helper (open) → Copy times helper (selected) → Availability updated → Profile — add photo and about → Upload avatar — modal open → Upload avatar — cropping → Upload avatar — finalise → Profile — completed and Finish → Confirm timezone update (modal) → Event Types dashboard (Exit) **Make: Gmail to Google Calendar Automation** (web, 36 screens, Business, Productivity) [Refero](https://refero.design/flows/10960) > Happy path: A user creates a new automation that listens for a Gmail reply action and creates a Google Calendar event. Steps: Open Scenario Builder → Open module picker → Select Gmail integration → Choose module action → Create connection prompt → Fill connection details → Authorize and wait → Configure Gmail module fields → Add Google Calendar module → Configure Calendar module → Add secondary Gmail module → Open module context menu → Set up a filter between modules → Open scenario settings → Run with existing data → Run or save scenario → Canvas shows validated modules → Return to Scenarios list → Final verification of scenario → Scenario listed and reusable ##### Task: interview scheduling **Daylish: Insert timer into schedule** (ios, 5 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/8012) > Happy-path: The user views a vertical, chronological timeline of tomorrow's events. They tap an empty gap between two scheduled items, open the "Fill " menu, choose "Timer", review the provisional timer card that is auto-populated with start/end/duration, edit the timer title inline, and save it. Steps: Timeline overview → Open 'Fill Gap' menu → Edit timer title (initial) → Rename timer → Final timeline insertion **Tripsy: Venue Visit — Date & Duration** (ios, 4 screens, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/2342) > A mobile venue-detail panel lets the user schedule a visit by selecting a start date/time and either an explicit end or a preset duration. Steps: Venue card with scheduling controls → Calendar modal — choose start date → Duration picker — choose visit length → Venue card updated with End and Duration **Instagram: Schedule Live Video** (ios, 5 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/3645) > Happy-path: An authenticated mobile user schedules a live video by entering a title, picking a start time, confirming the selection, and then choosing whether to share a post about the scheduled event. Steps: Compose scheduled live → Open start time picker → Confirm start time selection → Submit schedule request → Scheduled confirmation & share **Maze: Manage scheduled interview** (web, 4 screens, Business, Communication Tools) [Refero](https://refero.design/flows/3679) > Happy path: A recruiter views the Recruitment section, opens an interview's detail drawer, chooses to cancel and delete the session, confirms in a modal, and returns to the updated scheduled-interviews list. Steps: Recruitment — Scheduled interviews list → Interview details drawer → Confirm cancel and delete modal → Recruitment list (post-deletion) **Maze: Interview recruitment setup** (web, 18 screens, Business, Communication Tools) [Refero](https://refero.design/flows/3677) > Happy-path: A signed-in team owner starts on the dashboard, creates a new interview study, assigns it to an existing project, and walks through the recruitment setup. The user chooses to set up recruitment, optionally connects a calendar (or skips), and configures recruitment settings across Details, Attendees, Availability, Contact form, and Reminders. The flow ends on the study's Recruitment overview where the recruitment link and controls are visible. Assumptions: user is logged in with permission to create studies, has access to at least one project, and can edit study recruitment settings. Steps: Dashboard - Recent studies → Create a study - choose type → Create interview study - choose project → Create interview study - confirm name → Start point selection → Study Recruit overview (empty) → Connect your calendar prompt → Manual scheduling explanation → Recruitment settings - Details → Recruitment settings - Attendees (initial) → Recruitment settings - Add extra attendee → Recruitment settings - Availability options → Recruitment settings - Add time slot → Recruitment settings - Contact form (initial) → Recruitment settings - Require additional fields → Recruitment settings - Reminders (default) → Recruitment settings - Adjust reminder timing → Recruitment overview - Link ready **Maze: Interview session supervision** (web, 35 screens, Business, Communication Tools) [Refero](https://refero.design/flows/3680) > Happy-path summary: A recruiter selects a scheduled interview from the Recruitment dashboard, grants camera/microphone permissions in a pre-join dialog, and joins a live meeting. During the meeting they open Participants, invite a candidate (copy link or email), admit the candidate from the waiting room, manage recording (pause/resume) and end the meeting. After the meeting ends the platform processes the recording, attempts to create a transcript, and surfaces a playable recording or an error state; the user can then add/save a manual session summary. Steps: Recruitment dashboard → Session selection highlighted → Permissions prompt → Pre-join camera preview → Joining meeting → In-meeting controls visible → Network status shown → Participants panel open → Invite dialog options → Participant in waiting room → Recording controls → Recording paused indicator → Resume recording → End meeting confirmation → Meeting disconnected → Session detail - processing start → Processing recording → Video load error state → Playback available → No transcription / manual summary **Maze: Recruitment link copy** (web, 3 screens, Business, Communication Tools) [Refero](https://refero.design/flows/3678) > Happy path: A logged-in user viewing an interview's Recruitment section copies a public recruitment link to share with participants. Steps: Recruitment overview → Copy recruitment link modal → Copy confirmation toast **Homerun: Candidate Review Submission** (web, 3 screens, Productivity, Communication Tools, Job Search) [Refero](https://refero.design/flows/7184) > Happy path: A recruiter (already signed in) opens a candidate's profile for a specific job and submits an application review. Steps: Candidate detail — review prompt → Enter rating and comment → Save review and view activity **Homerun: Candidate management — profile actions** (web, 3 screens, Productivity, Communication Tools, Job Search) [Refero](https://refero.design/flows/7190) > Happy-path: A recruiter opens a candidate record from a job and inspects the profile, scheduling status, and recent activity. From the candidate header they can send a message, schedule an interview, move the candidate between stages, or access more actions via an overflow menu. In this flow the recruiter opens the overflow menu and selects an action (copy to talent pool), then confirms the activity appears in the right-hand activity feed. Steps: Candidate profile overview → Open overflow actions menu → Action taken and activity logged **Homerun: Candidate message & template email send** (web, 7 screens, Productivity, Communication Tools, Job Search) [Refero](https://refero.design/flows/7185) > Happy path: a recruiter opens a candidate profile, opens the in-product messaging composer, inserts an email template (Interview Invitation), reviews/edits the message, and attempts to send. The system detects that company email verification is required and prompts the user to verify/upgrade before outgoing email can be sent. Steps: Candidate profile → Inline message bubble → Fullscreen message composer → Composer fields and verification warning → Open template menu → Template inserted and editable → Verify company modal **Wrike: Chat-based meeting scheduling** (web, 5 screens, Business, Productivity, Project Management, Collaboration Tools, Paperwork, Scheduling Tools) [Refero](https://refero.design/flows/1523) > Happy-path: A logged-in user, viewing their Wrike Inbox on desktop, uses the right-side chat assistant to book a 30-minute meeting. Entry: user has Wrike open and the chat assistant visible. The user triggers the scheduling flow, selects a date, picks a time slot (timezone shown), confirms the slot, and receives an in-chat confirmation indicating an email confirmation was sent. Exit: chat shows scheduled slot details and confirmation message; user can continue interacting or close the chat. Assumptions: user is authenticated, has the chat assistant available in the interface, and is on a desktop/larger-screen layout (two-column: inbox + chat). Steps: Open chat assistant and invite to schedule → Choose a date → Select a time slot → Confirm and schedule → Receive confirmation in chat **Calendly: Copy Email Availability Suggestions** (web, 11 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/14042) > A logged-in Calendly user starts in Scheduling > Event types and opens the actions for the One-on-One event. They choose to offer time slots in an email, review the generated availability, edit the suggested times for a date, remove one date, and copy the final suggestions to the clipboard. The flow begins on the event list and ends with a confirmation that the customized availability text was copied. The user is assumed to have permission to manage the listed event type and availability. The flow is shown on a desktop-sized interface. **Calendly: Email Availability Sharing** (web, 3 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/14043) > The user starts in the Scheduling area with the “Offer time slots in email” dialog open for a one-on-one event. They review the suggested availability for the next three available days, optionally remove an entire date from the preview, and then copy the remaining time suggestions to the clipboard for use in an email. The flow ends when the availability text and booking link are copied. The user is assumed to be logged in and authorized to manage the selected event. The screenshots show a desktop interface and do not show the destination email client or a confirmation toast after copying. **Airtable: Schedule unscheduled assets onto calendar** (web, 3 screens, Project Management, Collaboration Tools, Scheduling Tools, Paperwork, Productivity) [Refero](https://refero.design/flows/1567) > Happy path: a user in the 'Creative Asset Development' workspace is viewing the 'Creative Assets' table in Calendar (Month) view filtered to show items without dates. The user schedules assets by dragging items from the right-hand 'Without dates' list onto days in the calendar. Each drag creates a dated entry on the calendar and removes or hides the item from the unscheduled list under the current filter. Steps: Calendar view with unscheduled list → Drag an unscheduled asset onto a date → Calendar updates and unscheduled list refreshes **Understory: Create Calendar Event** (web, 11 screens, Business, Scheduling Tools) [Refero](https://refero.design/flows/12719) > Happy path for creating a calendar event from the Calendar section. The user starts on an empty day view with a + Event action, opens an experience picker, selects an existing experience, configures event details in a modal, optionally edits capacity, spoken languages, and resources, then saves. After submission, the modal shows a loading state and closes back to the calendar, where the new event appears on the selected date and time. Steps: Calendar Overview → Select Experience → Event Details → Edit Capacity → Capacity Updated → Select Language → Language Selected → Add Resource → Review Event Settings → Saving → Event Added to Calendar ##### Task: inventory warehouse **Loowner: Inventory search and filter** (ios, 3 screens, Marketplace) [Refero](https://refero.design/flows/6599) > Mobile inventory list with an inline search bar. Entry: user opens "All your stuff" (logged-in, mobile device). User can browse the saved items list or focus the search field to filter. Typing a query shows either filtered results or an empty-state with a "No results" message. Exit: user either taps a listed item control to proceed (detail/navigation), clears/cancels the search to return to the full list, or leaves the screen via the back arrow. Steps: Inventory list → No results (numeric ID search) → Filtered results (text search) **ElevenReader: Search Collections** (ios, 2 screens, AI Tool, Books) [Refero](https://refero.design/flows/8942) > User starts on a 'Recommended collections' list (home/explore area) that includes a prominent search field in the header. From this screen the user can either browse collection tiles or tap the search field to enter a query (numeric or text). In the shown scenario the user taps the search field, types a numeric identifier (likely an ISBN or ID), and runs the search. The flow ends at a 'No results found' state when the query returns no matches. Assumptions: mobile device with soft keyboard, network available, user has access to the collections and search functionality from the bottom navigation Home/Explore area. Steps: Recommended collections list → Search results — no matches **Cushion: Inbox Management** (web, 3 screens, Communication Tools, Collaboration Tools) [Refero](https://refero.design/flows/12510) > Logged-in workspace inbox screen with a left navigation sidebar and an inbox panel in the main area. The user can review active notifications/messages, mark them read, archive individual items, and switch between Active and Archived views. The flow begins with unread or read inbox items visible and ends when the inbox is fully processed, showing an empty-state message such as "Inbox complete," "All caught up," or "Inbox clear." Assumes the user already has access to the workspace and is viewing the inbox on desktop. Steps: Review Inbox Items → Archive Item → Inbox Empty State **Fernand: Knowledge Base Search** (web, 3 screens, Business, Communication Tools) [Refero](https://refero.design/flows/11902) > This is a workspace settings area for managing a knowledge base. The user is assumed to be logged in with access to workspace settings and is viewing the Knowledge base section. The entry point is the Articles tab with a list of collections and root articles, plus search and add controls. The user can search articles by typing in the search field, which filters the visible results in place. The exit state is the filtered article list, showing matching results or an empty-state message if nothing matches. No navigation away from the section occurs in this sequence. Steps: Knowledge Base Overview → Search with No Matches → Search with Matching Result ##### Task: invoice create send **Daze: New conversation creation** (ios, 5 screens, Social Networking) [Refero](https://refero.design/flows/9150) > Happy-path for creating a new one-to-one conversation from the Messages list on a mobile device. Entry: user is on the Messages list (logged-in). Exit: user returns to Messages list where the newly created conversation appears with a preview/timestamp. Steps: Messages list → New chat — contact list → New chat — recipient selected → Conversation composer → Messages list — updated ##### Task: medication reminder **Calendly: Create Meeting Reminder** (web, 10 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/14031) > A logged-in Calendly user starts from the Automations area, chooses an email reminder template, configures the event types it applies to, changes the trigger from 24 hours to 1 day before the event, reviews the email content, and saves the automation. The flow begins on the Automations landing screen and ends with the new automation visible in the user’s automation list. The user is assumed to have permission to create automations and is using a desktop browser. ##### Task: meeting notes transcription **Granola: Search Notes and Meetings** (ios, 4 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14168) > A signed-in user starts on the notes home screen and opens the search tool. The search field is focused automatically, allowing the user to enter a query. The interface displays a no-results state for a numeric query, then updates to show matching previous meetings when the query is changed to “meeting.” The flow begins on the notes home screen and ends with matching search results displayed. It assumes a mobile device, an authenticated user, and access to the visible notes and meetings. Steps: Notes Home → Search Entry → No Results → Matching Results **Wispr Flow: AI Note Summary** (ios, 3 screens, AI Tool, Utility, Productivity) [Refero](https://refero.design/flows/13679) > The user is viewing the Scratchpad notes list and wants to understand the content of a saved note. From the Scratchpad screen, they select the note’s AI action, wait while the system generates a summary, and review the completed summary in a bottom sheet. The user can close the sheet or regenerate the summary. Steps: Scratchpad Notes → Summary Generation → Summary Review **Granola: Voice Note Transcription** (ios, 9 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14151) > The flow starts on the signed-in My notes home screen on a mobile device. The user taps New note, is prompted to enable microphone access, grants the operating-system permission, and is taken to a new untitled note with recording active. The user records for a period of time, taps End, and waits while the audio transcript is uploaded and notes are generated. The flow ends on the completed note containing a generated title, structured summary, action items, and sharing controls. The screenshots assume microphone access has not yet been granted and that the user has permission to create notes and record audio. Steps: Notes Home → Microphone Introduction → Microphone Permission → Recording Started → Recording In Progress → End Recording → Transcript Upload → Note Generation → Generated Meeting Note **Voicenotes: Record and Transcribe Note** (ios, 9 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14264) > A signed-in user starts from the Voicenotes library on a mobile device, begins a recording using the prominent red record control, monitors the live recording and transcript in a bottom sheet, then ends the recording. The system uploads and transcribes the audio before presenting a generated title and formatted note. The flow starts at the Voicenotes home screen and ends on the completed note detail view. The user is assumed to have microphone permission and network access; no permission prompt or authentication step is shown. **Voicenotes: Create Note Summary** (ios, 4 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14261) > The user starts on an existing voice note detail screen containing transcript content, an image attachment, a link, and tags. They open the Create menu, choose Summary, wait while the system generates the output, and review the completed summary. The flow ends on a summary result screen where the user can copy the generated text or delete it. The screenshots show a mobile device and an existing note; authentication, permissions, and network state are not shown. **Wispr Flow: Copy Transcript** (ios, 2 screens, AI Tool, Utility, Productivity) [Refero](https://refero.design/flows/13666) > The user is viewing a chronological transcript on the Home screen and copies the transcript content to the clipboard. The flow starts with an existing transcript list visible and ends with a confirmation toast indicating the copy succeeded. The screenshots do not show the initiating control, so the copy action is inferred from the resulting confirmation. The user appears to be on a mobile device, with the Flow service active and a Pro Trial label visible; login and permission state are not shown. Steps: Transcript History → Copy Confirmation **Voicenotes: Regenerate Voice Note Title** (ios, 4 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14263) > The user starts on a mobile voice-note detail screen showing an automatically generated title and transcript. They open the overflow actions menu, expand the Fix section, and select Regenerate title. The interface shows a temporary processing state, then replaces the original title with a more descriptive AI-generated title. The flow starts on an existing voice note and ends when the regenerated title is displayed. The user is assumed to have access to the note and its available actions; no sign-in or permission state is visible. **Voicenotes: Voicenotes First-Time Setup** (ios, 42 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14245) > Happy path begins with a first-time user opening Voicenotes on an iPhone. The user reviews the product introduction, accepts the Privacy Policy and Terms of Service, continues through the iOS tracking disclosure, selects how they heard about Voicenotes, chooses note-taking interests, selects a preferred language, reviews Apple Watch support, enables the Voicenotes keyboard in iOS Settings, and learns how to switch to it and dictate. The user then records a spoken note, confirms the transcription, and arrives at the main Dictation workspace with the note saved. The flow assumes an iOS device, a first-time or not-yet-configured user, access to system Settings, and no existing authenticated session visible in the screenshots. The shown path ends in the Dictation workspace; account creation or login is not shown. **CocoonWeaver: Share Transcription via Messages** (ios, 4 screens, Productivity) [Refero](https://refero.design/flows/4613) > Happy-path: User opens a single note that contains an audio recording and an automated transcription. From the note detail screen the user opens the note options sheet, chooses to share the transcribed text, picks Messages as the target in the iOS share sheet, and lands in a new message composer with the transcription and metadata prefilled. Steps: Note detail / Playback → Note action sheet → System share sheet → New Message composer (prefilled) **Voicenotes: Copy Note Summary** (ios, 2 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14262) > A user views an existing voice-note summary in a mobile bottom-sheet interface and copies its contents. The flow begins with the Summary sheet already open for a saved note dated 6 October. The user taps Copy, and the interface provides temporary visual confirmation while keeping the sheet open. The flow ends with the note content copied to the device clipboard. The user is assumed to be viewing an existing note on a mobile device; authentication and permissions are not visible. **Abode: Voice Note Transcription and Reaction** (ios, 3 screens, Collaboration Tools, Communication Tools, Social Networking) [Refero](https://refero.design/flows/4314) > User views an in-conversation message thread that combines an activity feed (recent list updates), text messages, an embedded GIF, and a sent voice message. The interface exposes an inline 'Transcribe' CTA next to the voice note and shows a composer area (camera, GIF, text input, voice/emoji). Steps: Conversation view with voice note → Transcribe CTA still visible; reaction appears → Composer and interaction state **Voicenotes: Voice Note Capture** (ios, 7 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14248) > The user starts from a dictation exercise screen, activates the microphone, speaks a note, reviews the live recording state, and confirms the transcription. The completed note is then saved and displayed in the main Dictation workspace. The flow begins on a mobile device with the user already inside the dictation experience and ends on the authenticated or initialized Dictation dashboard. Microphone permission is assumed to be available, as no permission prompt is shown. **CocoonWeaver: Re-transcribe audio note** (ios, 6 screens, Productivity) [Refero](https://refero.design/flows/4614) > Happy path: A user opens a note that contains an audio recording, opens the note's options, chooses to re-transcribe the existing audio, confirms the action, watches a transcription-processing screen with progress, and then sees the updated transcription in the note. Steps: Note detail with audio playback → Note action sheet → Re-transcribe confirmation → Transcription processing — start → Transcription processing — in progress → Note detail — updated transcription **Wabi: Voice Input Transcription** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12057) > This mobile chat-based experience starts in a conversation view where the user can review a proposed concept and use the voice input control to speak a follow-up request. If microphone access is not yet granted, the system shows an OS permission prompt. After the user allows access, the recording state appears at the bottom with an active audio indicator. When the user stops recording, the system switches to transcription mode and attempts to convert speech to text. The flow ends either with the transcribed input ready for use or, in the shown case, an error toast indicating the data could not be read in the correct format. Assumptions: the user is on a mobile device, inside a chat interface, and interacting as an existing user with access to the voice input feature. Steps: Chat Review → Microphone Permission → Recording → Transcribing → Transcription Error **pillowtalk: Voice Reflection Entry** (ios, 6 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13505) > A logged-in user starts from a personalized daily reflection screen and chooses the “transcribe” input method to record a spoken thought. The interface displays an active recording state with animated audio visualization, then processes the recording into text. The completed entry is presented with a generated title, timestamp, transcript, and a “dive deeper” reflection prompt. The user can continue the conversation by typing, using the microphone, or using the “yap” voice input, or finish the entry. Steps: Daily Reflection Home → Start Voice Recording → Continue Recording → Stop Recording → Review Transcribed Entry → Finish or Continue Reflection **Wispr Flow: Voice Note Transcription** (ios, 7 screens, AI Tool, Utility, Productivity) [Refero](https://refero.design/flows/13678) > The user is viewing a note titled “Jodie” with an empty editor. They start voice input using the floating microphone button, speak while the interface displays recording activity, and confirm the recording with the checkmark. The system then processes the audio and inserts the transcribed text into the note. The flow starts in an active note editor and ends with a completed transcription available for editing. The screenshots show an iPhone-style device with the keyboard visible before and after recording. Steps: Empty Note Editor → Recording Started → Active Voice Capture → Ongoing Recording → Confirm Recording → Transcription Processing → Transcribed Note **Wispr Flow: Voice Note Capture** (ios, 13 screens, AI Tool, Utility, Productivity) [Refero](https://refero.design/flows/13676) > A user starts in an empty Scratchpad and creates a new note. They enter a title, record a spoken thought using the microphone control, confirm the recording, and wait while the system transcribes it. The completed note can be reviewed and then returned to the Scratchpad list, where it appears as a saved note with its title, transcript, timestamp, and AI indicator. The flow assumes the user is already inside the Scratchpad section on a mobile device and has microphone access available. Steps: Empty Scratchpad → New Note Editor → Title Entry Dialog → Entered Note Title → Titled Note Editor → Recording Started → Active Voice Recording → Recording Progress → Confirm Recording → Transcription Processing → Transcription Review → Saved Note View → Scratchpad Note List **Programa: Meeting Recording and Summary** (web, 7 screens, Project Management, Business) [Refero](https://refero.design/flows/13432) > A logged-in project user opens the Meetings section, creates a new meeting, starts and ends an in-product recording, and waits while the system processes the transcript and generates a summary. The flow ends on the completed meeting overview, which contains a summary, decisions, and tasks. **Voicenotes: Join and record meeting** (web, 8 screens, Collaboration Tools, AI Tool) [Refero](https://refero.design/flows/6995) > Happy path: A signed-in user on a desktop browser starts on the Voicenotes dashboard, opens the record control, chooses "Join meeting as a bot", pastes a meeting link, starts an automated recording, watches the recording progress, and sees the new recording added to the note list while it is processed and transcribed. Steps: Dashboard / Ready to record → Open recording options → Join meeting modal → Paste meeting link → Recording in progress → Recording saved / processing started → Recording actions available → Transcription added to notes **Voicenotes: Create summary from notes** (web, 4 screens, Collaboration Tools, AI Tool) [Refero](https://refero.design/flows/6978) > Happy path: A signed-in desktop user opens the Create modal, picks a deliverable type (Summary), selects one or more existing notes, and requests generation. The system shows an in-progress state while creating the summary and then displays the generated text with basic actions (Copy, Retry). Steps: Open Create modal & choose deliverable → Select notes and enable Create → Generation in progress → View generated summary **Manus: Meeting recording and auto-summary** (web, 4 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/10451) > Happy path: A user in a project/task view starts an audio recording of a meeting, the system shows a live waveform and incremental recognition text while recording, the user stops the recording, and the system processes the audio to produce meeting minutes/summary. Steps: Ready to record → Recording in progress → Live transcription visible → Processing and summary created **Claude: Meeting Notes Summarization and Action Extraction** (web, 5 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/1816) > Happy-path: A logged-in user opens Claude, composes a specific instruction asking the assistant to review meeting notes and create a concise summary with key takeaways and action items, attaches the meeting-notes file, and receives a structured summary with action-item owners. Steps: Open chat and prepare instruction → Enter detailed instruction and attach file → Upload or display source notes → Assistant returns a structured summary (partial) → Assistant completes summary and offers follow-up **Voicenotes: Generate AI content from a voice note** (web, 4 screens, Collaboration Tools, AI Tool) [Refero](https://refero.design/flows/6982) > Happy-path: A signed-in user on desktop views their list of voice notes, opens the per-note 'Create' menu, selects an AI template (e.g., Summary / Meeting report), waits for the system to generate content, and sees the generated blocks inserted under the note. Steps: Notes list with playback and record controls → Open Create menu → AI generation in progress → Generated content inserted below the note **Voicenotes: Create from notes with custom prompt** (web, 4 screens, Collaboration Tools, AI Tool) [Refero](https://refero.design/flows/6979) > Happy-path: A signed-in user on a desktop opens the Create modal, chooses a creation type (here: Custom prompt), selects one or more existing notes as input, optionally edits and saves a custom instruction, and runs the generator. The system shows a progress state while generating, then presents the generated result with quick actions (Copy, Retry). Entry: user invokes the Create modal (Create tab active). Exit: generated content is displayed and can be copied or regenerated. Assumptions: user is authenticated (typical app header is visible), has existing notes to select, and is on a desktop/laptop viewport. Microphone/record controls are present elsewhere but not required for this flow. Steps: Open Create modal & pick type → Edit and save custom instruction → Generation in progress → View and act on generated output **Descript: Live Room — Screen Share with Chat** (web, 5 screens, Business, Collaboration Tools) [Refero](https://refero.design/flows/2144) > A live meeting room where one participant is sharing their screen. The main area shows the shared screen, the right column contains participant video thumbnails and a collapsible Chat panel. The footer has meeting controls (Record/Start, Audio, Video, Stop, Leave). The user starts in a state where the Chat is open but empty; they type a message, send it, and subsequent messages appear from other participants. Steps: Room open — screen sharing → Compose message → Message sent — local display → Message confirmation and minor UI affordances → Other participant replies **Handshake: Document Upload & Preview** (web, 7 screens, Social Networking, Job Search) [Refero](https://refero.design/flows/8772) > Happy path: A logged-in user navigates to the Documents section, uploads a PDF (resume), selects a document type, optionally marks visibility and adds a description, then saves. The system shows a saved notification and displays a document detail view with an embedded PDF preview. The user is returned to a Documents list that now contains the uploaded file. Steps: Empty Documents - CTA → Upload Form Displayed → Document Type Selection → File Selected & Fields Filled → Document Detail (Preview - Loading) → Embedded PDF Viewer → Documents List Updated **Manus: Meeting recording to summary** (web, 30 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/10450) > Happy path: A user records a meeting audio from the main task entry, stops the recorder, and the system transcribes the audio and generates a concise meeting summary automatically. The generated summary is shown inline and saved as document files (including a markdown file). The user can preview the document, batch-download files, open task options (rename, details), submit a rating/comment, and confirm a new title. Steps: Task entry / Recorder CTA → Recording pending start → Recording in progress (recognizing) → Live transcript appears → Recording ended — generating → Generation progress checklist → Summary delivered inline → Document stored (editor panel) → Task completed / suggestions → Document full-screen preview → Select files for download → Download confirmation → Open task menu → Open task details card → Rate and comment → Rating submitted → Rename task — edit title modal → Confirm new title → Back to task overview → Return to dashboard **Memotron: Create interview with audio and transcription** (web, 8 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/8325) > Happy path for creating a new note titled "Interview" using an audio capture input, linking that note to existing collections or nodes, recording audio, saving it, and viewing the generated transcription. Entry: user opens a new capture/note composer (assumed authenticated). Exit: saved note with attached audio file and a generated transcription shown in a playback/transcription viewer. Assumptions: user is signed in, has permission to create notes and link to collections/nodes, microphone access is available, and the device supports audio recording and playback (desktop or laptop). Steps: Composer - choose input type → Audio composer - ready to record → Search collections - results shown → Search audio node - preview → Collections attached → Recording in progress → Saving / processing → Playback & transcription viewer **Hello Ivy: Create meeting notes from template** (web, 3 screens, Productivity, Project Management, Collaboration Tools) [Refero](https://refero.design/flows/4492) > Happy path: A logged-in user, inside a workspace (Office #1) on a desktop, opens the Pages area, uses the Templates picker and applies a 'Meeting notes' template to the current document. Steps: Pages list and empty canvas → Templates modal → Populated meeting notes document **Around: Live transcript during gallery meeting** (web, 4 screens, Business, Communication Tools, Utility) [Refero](https://refero.design/flows/2004) > Happy path: A participant in a gallery-style meeting opens the right-hand panel, starts live transcription, watches the transcript populate in real time (with speaker labels and continuously appended lines), and then stops the recording, which triggers a confirmation control. Steps: Meeting open — transcript panel idle → Start transcription — initial transcript lines → Transcript continuing — more content appended → Stop recording — confirmation surfaced **Cohere: Document summarization with attachments** (web, 3 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3833) > Happy path: A signed-in user opens a conversation titled "Business Plan Outline for a Marketing Agency in French", uploads one or more documents as references, and asks the assistant to summarize the documents. The conversation pane shows the extracted document text as chat messages, the message composer for user input and attachments, and a right-side reference panel that lists uploaded files. Steps: Conversation with document content → Attach file from composer → File type validation and helper text **Figma: Summarize team sticky notes** (web, 5 screens, Design & Media Editing, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/5020) > Happy path: A user working inside a collaborative infinite canvas selects a framed area containing colored sticky-note cards titled "Team Roles & Responsibilities," uses a built-in summarize/organize command from the floating toolbar/context menu, and inserts the generated summary as a text/document object on the board. Steps: View roles board → Select frame → Open organize menu → Place summary placeholder → Insert generated summary **Gladia: Asynchronous transcription job** (web, 6 screens, AI Tool) [Refero](https://refero.design/flows/10895) > Happy path: A logged-in user creates an asynchronous transcription job by uploading a file or pasting a public audio URL, configures options (language, toggles), starts the job, and observes progress in the Transcriptions list until completion. Steps: Choose file or URL → URL provided → Configure transcription options → Transcribing in progress → Transcriptions list (initial) → Filter and confirm completion ##### Task: music player playlist **YouTube Music: Liked Music collection** (ios, 4 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11347) > Happy path: A signed-in user opens the Liked Music collection (mobile/iPhone). The screen shows a large artwork, playlist title, owner/avatar, timestamp and a short description. Primary actions (download, big play, more) are centered under the description. Below are horizontal, selectable filter chips that narrow the visible tracks. The track list appears as rows (artwork, title, artist, duration, overflow). A persistent mini player sits above the bottom navigation. Steps: Collection header and primary controls → Scroll to track list with chips → Apply / clear filter chips → Interact with a track and playback **ElevenMusic: Music Library Browsing** (ios, 3 screens, AI Tool, Entertainment, Music Streaming) [Refero](https://refero.design/flows/12451) > A mobile music library screen where the user browses saved content through segmented tabs: Liked, Playlists, Albums, and Artists. The flow starts in Library with the Liked tab selected, then the user can switch between content categories to view lists or grids of saved music. The screen includes a persistent mini-player at the bottom and a bottom navigation bar for moving to Explore, Library, Artist, Create, or Search. Steps: Liked Songs → Playlists → Artists **MD Vinyl: Music Player Onboarding** (ios, 3 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/12538) > A 3-step in-product onboarding tour for a music player interface. It starts on the Now Playing screen, then introduces the Library, then the Queue. The user is assumed to be in the music player with no special permissions required. Each step uses a tooltip anchored to the bottom control bar and a Next button, ending with a primary Start to Explore action that exits the tour and lets the user use the player normally. Steps: Now Playing Intro → Library Intro → Queue Intro **ElevenMusic: Add Song to Playlist** (ios, 5 screens, AI Tool, Entertainment, Music Streaming) [Refero](https://refero.design/flows/12448) > Happy path starts on a now-playing screen for the track 'Quiet Horizon' with playback controls and an Add action. The user opens the add-to-playlist sheet, chooses to create a new playlist, enters a name, and confirms creation. After creation, the new playlist appears in the song’s playlist list alongside existing playlists, with the option to remove the track from any listed playlist. Steps: Now Playing → Playlist Picker → New Playlist Naming → Create Playlist → Updated Playlist List **ElevenMusic: Playlist Track Actions** (ios, 4 screens, AI Tool, Entertainment, Music Streaming) [Refero](https://refero.design/flows/12449) > A mobile music playlist screen for the collection titled “Quiet Horizon.” The user is viewing a scrollable list of tracks with artwork thumbnails, track names, and per-item overflow menus. A mini-player is docked at the bottom, and the top bar includes back and share controls. The flow starts on the playlist screen, opens a track’s overflow menu, triggers a download, and ends with a download-complete confirmation banner. Steps: Playlist View → Track Actions Menu → Download In Progress → Download Confirmation **YouTube Music: Add Track to Playlist from Album** (ios, 5 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11358) > Happy path: A signed-in user views an album detail screen with artwork, album title and a list of tracks. From the track list the user opens the track’s contextual actions and chooses to add a song to an existing playlist. The interface shows an inline action reveal (and a green add action when tapped), then opens a modal listing recent and all playlists with a prominent New playlist button. After picking a playlist the UI shows a brief confirmation bar. Entry: album detail screen. Exit: user returns to album screen with confirmation that the song was saved. Assumptions: user is logged in, has playlists, and the device is a vertically-oriented mobile phone. Steps: Album detail / track list → Open track actions (inline) → Trigger add-to-playlist (quick action) → Playlist selection modal → Save confirmation **Carbon: Playlist Save and Confirm** (ios, 2 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/7777) > Happy path: a signed-in mobile user views a playlist detail screen (cover art, title, metadata, action buttons, track list) and taps the Add to Library control to save the playlist. Steps: Playlist Details — Initial → Playlist Details — Saved Confirmation **TIDAL: Add Song to Collection** (ios, 2 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/13126) > A mobile music playback screen showing a currently playing track. The entry state is an active now-playing view for the song "pure colour (feat. Erykah Badu)" by Steve Lacy and Erykah Badu, with playback controls, progress, lyrics access, and collection actions visible. The user taps the plus/add control to save the track; the interface confirms success with a checkmark and a temporary snackbar message, while playback remains on the same screen. Steps: Now Playing → Collection Confirmation **YouTube Music: Playlist track list** (ios, 2 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11375) > User is viewing a playlist (9 tracks) in a library context with a persistent mini-player at the bottom. Steps: Playlist overview → Voting on tracks **Carbon: Queue Reorder** (ios, 4 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/7773) > Happy-path: A user opens the in-player queue (a rounded bottom sheet) to review and reorder upcoming tracks. The sheet exposes global controls (Shuffle, Repeat), section headings (Play Next, Continue Listening), and a vertically stacked list of tracks. The user long-presses a track, drags it to a new position, then releases — the list updates and the sheet remains visible. Entry: music player with queue sheet expanded. Exit: updated queue order saved in the player and visible in the sheet. Assumptions: user is in a playing context (player visible behind the sheet), they can open the queue, and they have permission to modify the queue (remove/reorder). The device is a mobile phone and supports touch gestures. Steps: Open Queue / Inspect List → Start Drag (Lift Item) → Drag Over List (Preview Drop) → Drop & Confirm Reorder **TIDAL: Music Video Playback and Browse** (ios, 2 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/13134) > A logged-in mobile music streaming experience moves from a video library/browse screen into a full-screen now-playing player for a selected track, then back to browsing the artist’s video catalog. The user appears to be on a phone in portrait orientation, with playback already active. The entry point is the video library section; the exit point is either continued playback in the player or returning to the videos list. No account creation, payment, or permissions are shown. Steps: Full-Screen Player → Video Library Browse **YouTube Music: Account confirmation and signed-in home** (ios, 2 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11323) > Happy path: A mobile user is presented with a branded promotional welcome that includes an account confirmation prompt. The user confirms the displayed account (or can pick another), then the app navigates to the signed-in Home screen where personalized content and controls are available. Steps: Welcome + Account confirmation → Signed-in Home (Quick picks) **YouTube Music: Track playback & like feedback** (ios, 3 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11325) > Happy-path: A signed-in user views the now-playing song detail, interacts with playback controls and social actions, and receives immediate visual feedback after liking a track. Steps: Now Playing — Track Detail → Like Interaction → Transient Confirmation Toast **YouTube Music: Library Recent Activity Browsing** (ios, 2 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11366) > Mobile Library screen (iPhone portrait). Steps: Recent activity — List view → Recent activity — Grid/card view **YouTube Music: Creator profile subscription** (ios, 4 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11361) > Happy path: a logged-in user views a creator's profile (header with avatar, name, and playlists). The user taps the Subscribe CTA in the profile header, which toggles into a Subscribed state with a brief animated celebration and an incremented subscriber count. A transient confirmation bar (toast) appears near the bottom saying "Subscribed to [creator]". The user stays on the same profile; playlists and the persistent mini-player remain visible. Assumptions: user is signed in, on an iOS device, and network connectivity is available; this is not the user's own profile. Steps: Profile (initial unsubscribed state) → Subscribe tapped — immediate feedback → Subscribed state with microanimation → Final subscribed confirmation (toast) **MD Vinyl: Music Library Browsing** (ios, 4 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/12548) > A music library screen where the user browses albums and playlists, switches between list and grid presentation, filters/sorts the library, and accesses favorites. Entry is the main library view with stacked album cards and a persistent mini-player. The user can open a menu, change view/sort options, switch between Albums and Playlists, or tap the heart icon for favorites. Exit points include viewing all albums, viewing playlists, or using search and playback controls. Steps: Album Library → Library Options Menu → Favorites Empty State → Filtered Album List **MD Vinyl: Music Library Browse** (ios, 5 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/12549) > This screen sequence shows a music library browser with a persistent mini-player. It starts in an album list-style view, opens a translucent menu to change layout or sort/filter the library, then switches to a grid view of album covers. The user appears to be signed in and viewing their local/library collection on a mobile device; no onboarding, login, or permissions steps are shown. The flow ends with the album grid visible and an item count shown at the bottom, while playback controls remain accessible throughout. Steps: Album List View → View and Sort Menu → Grid Layout Preview → Album Grid Library → Library Count and Player **YouTube Music: Personalized home discovery** (ios, 4 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11355) > A mobile music app home experience where a logged-in user browses mood/category chips (Podcasts, Energize, Relax, Feel good, etc.), scans large carousel cards and ranked 'Quick picks' lists, and starts playback with a persistent mini player. Steps: Home — welcome & recommendations → Podcasts category — featured carousel → Energize category — quick picks and community → Relax category — quick picks & speed dial **YouTube Music: Create and Customize Mix** (ios, 6 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11348) > Happy path: A logged-in listener starts on Home and taps the "Create a mix" card to build a custom mix. They choose artists/genres from a grid, review selected items in a compact strip, open the Adjust panel to tune artist variety, discovery preference (Familiar / Blend / Discover) and mood filters, then confirm. The app generates a mix and navigates to the mix detail screen where the user can play the generated playlist. Steps: Home — Create a mix entry → Pick music — selection grid → Selection review — mini strip & Done → Adjust — variety & discovery → Adjust — filters selection → Mix detail & playback **MD Vinyl: MD Vinyl Setup and Purchase** (ios, 18 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/12536) > Happy-path onboarding for MD Vinyl on iPhone, starting from an introductory album-picking screen and ending at the main player/home state after setup. The user is assumed to be on a mobile device, not yet fully onboarded, and may not have granted music/notification permissions yet. The flow includes optional skipping of intro screens, connecting to Apple Music, handling subscription purchase prompts, and arriving at the main turntable interface. Steps: Album Selection Intro → Album Browsing → Turntable Setup → Tonearm Placement → Playback Intro → Music Library Connect → Apple Music Permission → Apple Music Join Screen → Apple Music Trial Offer → Apple Music Subscription Sheet → Pro Upgrade Offer → Pro Purchase Prompt → App Store Confirmation → Purchase Success → Now Playing Tip → Library Tip → Notification Permission → Main Player Home **YouTube Music: Playlist Track Selection & Playback** (web, 2 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11239) > User opens a playlist (Hip Hop Essentials) in a desktop music player and interacts with the track list and playback controls. Entry: playlist view with artwork, playlist metadata and long, scrollable track list. Primary happy-path: user scans the list, clicks a track to play it (or uses header controls to start playback), then clicks another track in the list to change the currently playing song. Exit: mini/now-playing bar and in-list highlight update to reflect the selected track. Assumptions: user is signed in (account avatar present), on a desktop/larger screen, and has permission to play tracks; network playback is available. Steps: Playlist overview with current song playing → Change selection to another track in the list **YouTube Music: Video playback with queue management** (web, 3 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11241) > Happy path: A signed-in user is watching a music video in the desktop music experience and interacting with the currently playing track and queue. Steps: Video playback with visible up-next queue → Open per-track context menu and view playlist details → Display technical overlay (stats for nerds) **YouTube Music: Download a track from a playlist** (web, 4 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11234) > Happy path: A signed-in desktop user is viewing a playlist in the music player and starts a download for an individual track via the track's overflow (kebab) menu. Steps: Playlist overview with track list → Open track overflow menu → Start download and show progress toast → Download complete notification **Epidemic Sound: Track download with usage confirmation** (web, 4 screens, Music Streaming) [Refero](https://refero.design/flows/5108) > Happy-path: A signed-in desktop user browsing a playlist selects a track and downloads it. Steps: Playlist / Player view → Download music track modal → Usage destination confirmation → Download in progress / Return to playlist **ElevenMusic: Remix Discovery and Playback** (web, 2 screens, Music Streaming, Entertainment, AI Tool) [Refero](https://refero.design/flows/12578) > This is a desktop music discovery and playback experience. The user enters on an Explore view while signed in, with a persistent left navigation sidebar, a central browsing area for featured remixes and playlists, and a right-side track detail panel. The user can browse content, inspect a selected song, and interact with playback controls and remix options. The flow appears to start from the Explore section and ends with the selected track playing and its remix list visible in the details panel. Steps: Browse Explore → Track Detail and Playback **YouTube Music: Add Track to Library from Playlist** (web, 3 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11233) > User is viewing a playlist (Hip Hop Essentials) in a desktop music player and wants to add one of the listed songs to their personal library. Steps: Playlist View — locate track → Open track context menu → Save to library and show confirmation **Epidemic Sound: Play and Save Track from Playlist** (web, 2 screens, Music Streaming) [Refero](https://refero.design/flows/5106) > User views a playlist landing for “Laid-back Lofi” on a desktop music product and plays a track using the persistent bottom player. The bottom player shows playback controls, a waveform scrubber, track metadata, and action buttons (like, add, download). From this screen the user can browse playlist content, use filters (Genres, Moods, Vocals, Duration, BPM), change sort order, and interact with the currently playing track. The happy path shown: user listens to “City Lights” and taps the heart to save it to their liked tracks (confirmation tooltip appears). Steps: Playlist Overview with Active Player → Save Track Confirmation **Tidal: Profile — manage play queue** (web, 4 screens, Music Streaming) [Refero](https://refero.design/flows/3271) > Happy path: A logged-in desktop user is viewing their profile (Liam) and browsing the "My Picks" tab. While playback is active in the mini player, the user opens the play queue (right-side overlay), inspects upcoming tracks, removes individual items or clears the queue, then closes the queue to return to the profile view. Steps: Profile (My Picks) shown with active playback → Open play queue overlay → Manage queue (remove items / clear queue) → Close queue and return to profile **Epidemic Sound: Find Similar Tracks** (web, 2 screens, Music Streaming) [Refero](https://refero.design/flows/5112) > User is on a desktop playlist detail (Laid-back Lofi) with an integrated bottom playback bar. From the playing track (City Lights) the user opens a "Find similar" overlay to discover tracks similar to a highlighted section of that track. Steps: Playlist detail with player → Find similar modal **Spotify: Adding a Radio Song to a Playlist** (web, 5 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/53) > The user navigates through a music streaming web interface to add a currently playing song from a mixed radio playlist to a personal playlist within their library. Beginning from the radio playlist screen, the user opens the options menu for the song, initiates the action to add it to a playlist, selects one or more existing playlists, confirms the addition, and sees a success confirmation. The flow ends with the user viewing the updated playlist containing the added song. It is assumed the user is logged into their account and using a desktop web browser. Steps: View Radio Playlist and Open Song Options → Open Add to Playlist Modal → Select Playlists and Confirm → Receive Confirmation of Addition → View Updated Playlist with Added Song **YouTube Music: Play and Like Track in Playlist** (web, 3 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11232) > Happy path: A signed-in desktop user opens a curated playlist, starts playback, and likes the currently playing track. Entry: user is on the playlist detail view (left: artwork & metadata, center controls, right: tracklist). Exit: playback continues while the liked state is applied and a confirmation message is shown. Assumptions: user is signed in (profile visible), on a desktop/larger-screen layout, and has permission to like tracks. Steps: Playlist Details & Playback Start → Like Track from Tracklist → Ongoing Playback & Visual Confirmation **Epidemic Sound: Select and play stem mix** (web, 4 screens, Music Streaming) [Refero](https://refero.design/flows/5109) > Happy-path: A user visits the 'Laid-back Lofi' playlist page, opens the mix selector in the persistent bottom player, chooses the Melody-only stem, and plays the track. Steps: Playlist overview with player → Open mix selector → Select Melody stem → Play selected Melody mix **SoundCloud: Save Track to New Playlist** (web, 5 screens, Music Streaming, Social Networking) [Refero](https://refero.design/flows/6574) > Happy-path: A logged-in desktop user is viewing a SoundCloud track page and saves the currently playing track into a newly created playlist. Entry: user is on the track detail/player view (track is playing). Exit: the app shows a confirmation toast and the Create playlist modal updates to show the newly created playlist with a link to view it. Assumptions: user is signed in (profile avatar present, Follow/Following visible), on a wide/desktop layout, and has permission to create playlists. Steps: Track Player (First Listen) → Track Actions Menu → Create Playlist Modal (empty) → Create Playlist Modal (title entered) → Confirmation and Playlist Link **[untitled]: Delete Track from Album** (web, 3 screens, Music Streaming) [Refero](https://refero.design/flows/5949) > Happy path: a signed-in user viewing an album/playlist opens a track's options, chooses to delete that track, confirms the deletion in a modal, and sees the playlist and metadata update. Entry: user is on an album/playlist detail screen with a visible tracklist and a persistent bottom audio player. Exit: the selected track is removed from the list, the displayed track count and total duration update, the modal closes, and the UI returns to the album view. Assumptions: desktop/laptop layout (wide viewport), user has permission to edit the album (UI shows Add tracks and per-track menus), and the bottom audio player is persistent across the flow. Steps: Album / Playlist detail → Delete confirmation modal → Updated album view after deletion **Deezer: Personalised Flow selection** (web, 3 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/10682) > Desktop Home landing for a logged-in user showing the personalised "Flow: play how you feel" surface. Entry: user opens Home from the left sidebar or lands on Home after signing in. The screen surfaces a horizontal set of selectable chips (default: Moods), a "Continue streaming" artist carousel, several recommendation rows (Mixes, Discover, Playlists you’ll love, Top genres, New releases, personalized artist rows) and the persistent bottom player. Exit points: select a mood/genre chip to filter the Flow content, open an artist/playlist to start playback or navigate to a detail view, or use the left sidebar to change sections. Assumptions: user is signed in (personalised content, playlists and avatar visible) and on a desktop/wide viewport. Steps: Home landing - default Flow (Moods) → Open filter dropdown → Switch to Genres and update chips **Deezer: Personalized track exclusion** (web, 5 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/10683) > Happy-path: A logged-in desktop user visits the personalized "Flow" hub to browse music and listens using the persistent mini player. From the Flow landing they open a contextual feedback menu (labelled “Not satisfied with your Flow?”) and choose to exclude the currently playing track from future recommendations. The system updates personalization, advances/changes playback if needed, and shows an inline toast confirming the exclusion. Steps: Flow hub — browsing and playback → Open contextual feedback → Select exclusion option → Apply exclusion and confirm → Persistent confirmation and updated content **YouTube Music: Video to Playlist Navigation** (web, 2 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11240) > Happy path: a signed-in desktop user is watching a music video and inspects the upcoming track queue, then navigates into a recommended playlist to browse tracks. Steps: Video Player with Queue → Playlist Detail Screen **Deezer: Playlist Quiz Playthrough** (web, 12 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/10697) > Happy path: A signed-in desktop user browsing a playlist opens the playlist's action menu and chooses "Launch a music quiz." The system shows a mode-selection modal (Solo / Multiplayer / Party), then a short "How does it work?" overlay. The quiz begins as a series of timed multiple-choice questions (visible timer, per-question progress, and instant feedback with color-coded answers). At the end the user sees a results screen with total points, tracks played, and a leaderboard view; they can challenge a friend or play again. Steps: Playlist detail / actions → Select 'Launch a music quiz' → Pick your mode → Instructions overlay → Timed multiple-choice question → Correct answer feedback → Incorrect / mixed feedback → Ongoing questions & score accumulation → End-of-quiz summary → Leaderboard & rank → Music quizzes hub → Alternate entry & replay **Deezer: Home Flow listening and favourite** (web, 3 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/10684) > Happy path: A signed-in user arrives at the personalized Home/Flow landing, listens to auto-selected music via the persistent bottom player, browses discovery carousels (Flow genres, artists, playlists, new releases), and marks the currently playing track as a favorite. Steps: Home (Flow) landing with playback → Initiate favourite action → Confirmation and continued browsing **Epidemic Sound: Sign in and browse music** (web, 5 screens, Music Streaming) [Refero](https://refero.design/flows/5064) > Happy path: An unauthenticated desktop user arrives at the marketing landing page, clicks the Log in link, completes credentials (or chooses a third-party sign-in) on the split-screen login view, and is redirected to the logged-in Discover / Music library where they can search, preview, and interact with tracks and playlists. Steps: Marketing landing → Login — choose method → Login — enter credentials → Login — submit → Discover — music library ##### Task: notifications inbox **Gizmo: Manage conversation from inbox** (ios, 3 screens, Entertainment, Social Networking, AI Tool) [Refero](https://refero.design/flows/10036) > User lands on a dark-themed chat inbox showing recent conversations (entry point: chat list). From the list the user opens a contextual actions menu for a specific conversation and chooses Delete. The UI immediately removes that conversation from the list (exit point: updated chat list). Assumptions: user is authenticated and on a mobile device; chat list is the primary inbox view and actions are available per conversation. Steps: Chat inbox → Conversation contextual actions → Updated chat inbox **Linear: Inbox notification delete** (ios, 3 screens, Collaboration Tools, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/6626) > Happy path: a signed-in user opens the Inbox (notifications/reminders) and removes a single notification using a left-swipe gesture. Entry: Inbox screen with a vertical list of notifications, each row showing avatar/icon, title, subtitle, and optional timer/badge. Exit: Inbox returns to list state with the deleted item removed and a temporary confirmation (snackbar) offering Undo. Assumptions: user is authenticated and notifications are loaded; device is a mobile portrait layout; gesture support (left-swipe) is available. Steps: Inbox list → Reveal Delete action → Confirm delete and show Undo **Meta AI: Delete Notification** (ios, 3 screens, AI Tool) [Refero](https://refero.design/flows/13389) > A logged-in user manages an existing notification. The flow starts on the Notifications screen with one feedback-related notification. The user swipes the notification left to reveal a trash action, then activates it. The notification is removed and the screen displays an empty-state message. No confirmation dialog or undo action is shown. **Airbnb: Notification dismissal** (ios, 3 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6437) > Happy path: A signed-in mobile user opens the Notifications view, reviews listed notification rows, swipes a row left to reveal a contextual Remove action, and removes that notification. Steps: Notifications list → Reveal Remove action → Remove notification **Substack: Appearance & Visual Preferences** (ios, 3 screens, News, Social Networking) [Refero](https://refero.design/flows/4506) > Happy path: a logged-in user opens the Settings list, chooses 'Appearance', adjusts visual and layout preferences (auto-play, theme, inbox density, default tab), and observes immediate UI feedback. Steps: Settings list → Appearance (light mode) → Appearance (dark mode applied) **Pinterest: Private message and media actions** (ios, 17 screens, Reference, Inspiration Board, Social Networking, Lifestyle) [Refero](https://refero.design/flows/2542) > Happy-path: user opens the Messages tab, composes a new message to a contact, views a conversation with media, interacts with a media action menu (pin/share/WhatsApp), reacts to a message (heart), types and sends messages, then opens message settings to mute notifications. Steps: Messages list → Recipient picker → Search input active → Open empty conversation → Empty thread interface → Messages appear → Open media actions → Pin action highlight → Share to WhatsApp → Pinned media state → React to message → Type and send message → Continue typing → Open message settings → Adjust notifications → Confirm mute → Return to messages list **Playground: Email Sign-in & First Design Onboarding** (ios, 17 screens, AI Tool, Design & Media Editing) [Refero](https://refero.design/flows/8989) > Happy-path onboarding where a new user signs in with email, verifies via a 6-digit code or magic link, and completes a short guided onboarding that creates a first design. Steps: Welcome & Sign-in Options → Enter Email → Email Filled → Request Sent Loading → Check Email Instruction → Email With Code → Enter Verification Code → Verification Loading → Welcome Carousel / Get Started → Choose Type of Design → Pick a Template → Edit Prompt Input → Generating / Preview → Export & First Design CTA → Usage Preference Question → Enable Push Notifications → Main Design Selection (Onboarded) **Savee: Email & Social Login Onboarding** (ios, 13 screens, Social Networking, Inspiration Board, Reference) [Refero](https://refero.design/flows/5267) > Happy path sign-in and first-run onboarding for a new user on an iOS device. Entry: user launches the app (not authenticated). The user can choose social sign-in (Apple/Google) or email sign-in. If they choose email, they enter their email/username, receive a magic-login link by email, open that link which completes authentication, then the app shows a post-login welcome/onboarding carousel and an optional push-permission prompt. Exit: user finishes the carousel and lands on the main content feed. Assumptions: user is on a mobile device, network connectivity is available, user has access to the email account they provide, and the device can open links back into the app. Steps: Welcome / Sign-in choices → Email / Username entry → Validation error shown → Email corrected / Ready to continue → Email sent confirmation → Login email contents → Opening link / Loading → Authentication success + Notifications prompt → Onboarding carousel — welcome → Onboarding carousel — saving interaction → Onboarding carousel — following people → Onboarding carousel — premium prompt → Main content feed **HYPE: Hype Account Onboarding** (ios, 19 screens, News, Lifestyle) [Refero](https://refero.design/flows/12241) > A first-time mobile user opens the Hype app and encounters a consent banner, then an introductory welcome screen. The observed sequence demonstrates account creation through the Hype account service, profile completion, newsletter selection, and a return to the welcome screen for account login. The user then reaches the content feed, where the system requests notification permission and displays a device-language information dialog. The flow assumes a mobile iOS device and a user who is initially unauthenticated. The observed path includes both sign-up and login states rather than one continuous linear action sequence. Steps: Consent Banner → Welcome Screen → Sign-In Authorization → Account Registration Form → Completed Registration Form → Profile Form → Profile Form Ready → Completed Profile Details → Profile Submission → Gender Selection Menu → Selected Profile Gender → Newsletter Preferences → Return to Welcome → Login Authorization → Login Form → Submitted Login → Notification Permission → Language Information → Hype Content Feed **Artsy: Account Signup and Onboarding** (ios, 33 screens, Marketplace) [Refero](https://refero.design/flows/7591) > Happy-path: An unauthenticated iOS user opens the app, taps to sign up with email, completes a three-step account creation (email → password → name & consents), then completes a short personalization step (choose a content category). The app shows short loading/splash screens, surfaces a curated collection (Curators' Picks), asks for notification permission, and lands the user in the logged-in Home feed. Entry: app launch on iOS with no account. Exit: user is logged in and viewing the For You / Home feed. Assumptions: mobile (iPhone) viewport, network available, user creating a new account (not social sign-in). Steps: Landing / Sign up prompt → Email input (keyboard visible) → Email submit / Loading → Password entry (rules shown) → Password typing and validation → Full name + consents (initial) → Full name validation error → Full name entry and consents checked → Welcome splash (account creation) → Onboarding preference selection → Preference submit / Loading → Curators' Picks preview → Explore More / Curators' Picks expanded → Post-onboarding splash → Home feed initial load → In-app notification prompt (soft) → System notification permission (iOS) → Home feed with notifications handled → Finalized personalized feed → Optional: Reopen permission or onboarding flows **Linear: Filter inbox by issue status** (web, 4 screens, Business, Productivity, Project Management, Collaboration Tools) [Refero](https://refero.design/flows/6681) > Summary: The user is in an inbox-style notification center with an issue detail pane open. They open the notification filter menu, drill into the "Issue status type" category, choose the "Completed" status, and the notification list updates to show only matching items while indicating notifications hidden by filters. Steps: Inbox with issue open → Open filter menu → Drill into Issue status type → Apply Completed filter **Linktree: Notifications Center** (web, 3 screens, Business, Social Networking) [Refero](https://refero.design/flows/12409) > A logged-in user opens a notifications modal from the dashboard and reviews recent activity alerts. The modal overlays the current page and includes category tabs for All, Updates, Opportunities, and Insights, plus a menu action to mark all notifications as read. The user can inspect individual notifications and then close the modal to return to the underlying dashboard. Steps: Open Notifications → Mark All Read Menu → Notifications List **Wittl: Sign-in to Inbox** (web, 6 screens, Job Search) [Refero](https://refero.design/flows/5360) > Happy-path: A visitor starts on the public marketing landing (hero) and uses the page controls (Log in / Get started) to open the sign-in form. The user enters work credentials, submits the form, sees immediate submit feedback (button animation) and then a full-page loading state while the product loads. The flow finishes on the authenticated workspace Inbox/dashboard where recent candidate applications are listed. Entry: public marketing landing. Exit: authenticated Inbox (user workspace). Assumptions: user is on a desktop/laptop viewport, has or can create an account, and has valid credentials. The flow demonstrates a standard sign-in and transition into the app's primary inbox area. Steps: Marketing landing (hero) → Sign-in form (empty) → Credentials entered → Submit feedback (button state) → Workspace loading → Inbox / Welcome dashboard **Linear: Inbox issue browsing and notification action** (web, 3 screens, Business, Productivity, Project Management, Collaboration Tools) [Refero](https://refero.design/flows/6682) > Happy-path: A signed-in user views an inbox-style issue list (middle column), opens the list sorting/filter menu, navigates between individual issues, and performs a notification deletion which shows a transient toast. Entry: user is signed in and on the project workspace with the global left navigation visible. Exit: a different issue may be selected and a confirmation toast appears for the deleted notification. Assumptions: user is authenticated, has permission to view the workspace and issues, and the interface shows a persistent left navigation, a scrollable middle list (inbox), and a right-side issue detail panel. Steps: Open inbox and view ordering options → Select an item and delete notification → Return to original issue and confirm state **Wise: Referral Source Survey** (web, 4 screens, Finance) [Refero](https://refero.design/flows/352) > Happy path: A logged-in desktop user receives a notification in their Inbox prompting them to take a short marketing survey. The user opens the notification, is routed to a single-question survey (“How did you hear about us?”), selects an option. Selecting certain options (Social Media) reveals a conditional sub-question listing platforms. After completing required selections the user taps Complete, sees a thank-you confirmation with a Back to home CTA, and returns to the main/home area. Assumptions: user is authenticated, on a desktop layout with a persistent left navigation and top-right account controls. Steps: Inbox notification → Primary survey question → Conditional sub-question & submit → Submission confirmation **X: Conversation leave confirmation** (web, 2 screens, Social Networking) [Refero](https://refero.design/flows/993) > Happy path: A signed-in user in the desktop Messages area views a selected direct message conversation's info panel, chooses the 'Leave conversation' option, and confirms via a modal to remove the conversation from their inbox. Steps: Conversation info → Leave confirmation modal **Cushion: Inbox Management** (web, 5 screens, Communication Tools, Collaboration Tools) [Refero](https://refero.design/flows/12509) > Logged-in user views the Inbox in a workspace messaging product. The flow starts with unread messages visible in the Active inbox and a top-right overflow menu for bulk actions. The user can open the menu, choose grouping or bulk actions, and then see the inbox update. The flow ends with the inbox cleared of unread items and an empty-state message indicating everything has been sorted. Steps: Inbox View → More Menu Open → Bulk Action Selection → Inbox Cleared → Completed State **Skiff: Secure Account Setup** (web, 19 screens, Communication Tools, Scheduling Tools, Paperwork) [Refero](https://refero.design/flows/1886) > Happy-path onboarding: A visitor arrives on the marketing landing page, clicks the sign-up CTA and proceeds through an account creation flow. The user chooses an @skiff.com address, sets a password, is shown a recovery phrase (and prompted to download it), optionally provides a recovery email, is offered mobile/desktop downloads and browser notifications, then creates up to three email aliases (with inline validation). The user is then offered to set up a custom domain and is taken to the new mailbox, which shows an onboarding checklist modal. Assumptions: user starts logged out on a desktop browser, has network access, and is completing initial account setup for a new workspace. Steps: Marketing landing / Sign-up CTA → Create your account — pick local part → Username entered confirmation → Create a password — entry → Create a password — validated → Show password recovery key (seed phrase) → Add password recovery email — entry → Recovery email — confirmed → Offer mobile and desktop downloads → Browser notifications opt-in → Create aliases — empty fields → Create aliases — sample entries → Alias validation — character rule errors → Alias correction and retry → Alias uniqueness error (taken) → Custom domain setup prompt → Landing in mailbox — onboarding checklist modal → Inbox with aliases visible in sidebar → Mailbox main view — default inbox **Skiff: Archive message from inbox** (web, 2 screens, Communication Tools, Scheduling Tools, Paperwork) [Refero](https://refero.design/flows/1893) > Happy path: A logged-in user on a desktop mailbox workspace views their Inbox and archives a thread. Steps: Inbox list (pre-archive) → Archive confirmation and list update **Fernand: Support Inbox Review** (web, 2 screens, Business, Communication Tools) [Refero](https://refero.design/flows/11893) > This is a support inbox screen in a web-based messaging tool. The user is viewing the conversation list with tabs for Inbox, Waiting for answer, and All, plus toolbar actions in the top-right. The flow appears to start with a populated inbox list showing a conversation card and status badges, then changes to a filtered state where no conversations match the selected filters. Assumptions: the user is already logged in, has access to the inbox, and is using a desktop device. The visible toast indicates a system update about snoozed conversations. Steps: Inbox List → Empty Filter State **Synthesia.io: Clear Workspace Notifications** (web, 3 screens, AI Tool) [Refero](https://refero.design/flows/13759) > A logged-in workspace user opens the notification center from the dashboard to review recent activity. The panel initially shows three unread notifications, including a low-credit warning and completed video task updates. After the notifications are reviewed or marked as read, the unread count changes to zero while the notification list remains visible. The flow begins on the workspace home dashboard and ends with no unread notifications. Steps: Open Notification Center → Review Notifications → Clear Unread Count **Nextdoor: Remove Chat** (web, 4 screens, Communication Tools) [Refero](https://refero.design/flows/11447) > A logged-in user removes an existing conversation from the Chats area. The user opens the conversation’s overflow menu, selects the removal action, confirms the decision in a modal, and returns to the chat list. The removed conversation disappears from the list, the conversation pane switches to a general empty state, and a temporary confirmation banner offers an Undo action. Steps: Open Chat Actions → Choose Remove Chat → Confirm Removal → View Removal Confirmation **Cushion: Draft Message Save** (web, 4 screens, Communication Tools, Collaboration Tools) [Refero](https://refero.design/flows/12529) > This flow starts in a workspace messaging inbox with a selected direct conversation. The user opens a post composer from the conversation, edits the message content in a modal, and then attempts to close or leave the editor. The system prompts to save the unfinished content as a draft or discard it. After choosing Save, the draft is stored and the user returns to the conversation view where a draft confirmation appears and the Drafts item is added in the left navigation. Assumptions: the user is already logged in, has access to the workspace and messaging area, and is using a desktop browser. Steps: Conversation View → Compose Post → Save Draft Confirmation → Draft Created **Missive: Schedule Snooze for Email** (web, 4 screens, Business, Communication Tools, Collaboration Tools) [Refero](https://refero.design/flows/6281) > Happy path: A user selects an email in the inbox, opens the message actions menu, chooses Snooze, picks a preset or custom snooze time, optionally toggles related options, confirms the snooze, and sees a success toast. Steps: Open message actions → Open snooze presets → Adjust snooze options → Snooze confirmation and feedback **Preply: Archive conversation** (web, 3 screens, Education) [Refero](https://refero.design/flows/9224) > Happy path: A logged-in user opens the Messages section (desktop view), locates an inactive/system conversation in the left-hand thread list, opens the thread actions menu (ellipsis) and archives that conversation. Steps: Messages list — thread selected → Open thread actions menu → Thread archived — list updated **ManyChat: Delete an Automation** (web, 5 screens, Business, Productivity, Communication Tools) [Refero](https://refero.design/flows/10332) > Happy path: A signed-in user opens the Automation area, locates an existing automation, opens its context menu, and deletes it. The system shows a confirmation modal explaining side effects (triggers moved to Trash), performs the deletion, displays a success toast, and then shows the empty-state/first-time prompt inviting the user to create a new automation. Steps: Automations list (single item) → Context menu opened → Delete confirmation modal → List updated — empty state → Empty-state onboarding CTA **Missive: Team Meeting Draft Composition and Commenting** (web, 3 screens, Business, Communication Tools, Collaboration Tools) [Refero](https://refero.design/flows/6256) > Happy path: A signed-in user opens the inbox, selects (or continues) a draft titled "Team Meeting," edits recipients/subject/body, attaches files, and posts a private comment on the draft before sending. Steps: Open (or continue) draft in split view → Edit recipients, subject and body → Post a private comment on the draft ##### Task: research collaboration **Meta AI: Mythical Biology Poster** (ios, 3 screens, AI Tool) [Refero](https://refero.design/flows/13353) > A user is viewing an existing creature-biology conversation and selects a suggested request to combine five creature diagrams into one mythical biology poster. The system processes the request, shows a planning state, and then returns a completed master poster containing all five diagrams. The flow assumes an existing conversation and generated infographic assets are available. It is shown on a mobile device; no login or permissions are visible. Steps: Select Poster Request → Poster Layout Planning → Review Completed Poster **MyMind: mymind Onboarding** (web, 17 screens, Productivity, Reference, Inspiration Board) [Refero](https://refero.design/flows/13289) > A prospective user discovers mymind through a long-form marketing landing experience, reviews its privacy and content-management capabilities, and starts account creation. After choosing Google or Apple sign-up, the user moves through a guided onboarding presentation explaining private storage, one-click saving, automatic organization, supported content types, and search. The user then reaches a subscription screen with an option to continue as a guest and enters the mymind workspace. The final state is a populated dashboard with introductory content and setup reminders. The screenshots suggest the user is initially logged out and using a desktop browser. No completed payment or detailed account form is shown. Steps: Product Discovery → Account Creation Choice → Private Oasis Welcome → Privacy Promise → Saved Content Overview → One-Click Saving → Automatic Organization → Book Recognition → Article Recognition → Website Recognition → Search Notes → Search Visual References → Search Articles → Search Books → Plan Selection → Workspace With Setup Guide → Initial Workspace ##### Task: search everything **Substack: Global search with empty-note results** (ios, 3 screens, News, Social Networking) [Refero](https://refero.design/flows/4486) > Happy-path: a logged-in user on the Home feed opens the global search, types a query (numeric in this example), and inspects filtered results. Steps: Home feed — open search → Search input — type query → Results with filter chips — empty state **TIDAL: Music Search and Browse** (ios, 9 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/13150) > A logged-in mobile music app search and discovery experience. The user starts on an Explore screen with genre, mood/activity, decade, and category shortcuts, then taps Search to open a focused search screen. After entering a query, the interface shows categorized results with tabs such as Top results, Profiles, Albums, Tracks, Uploads, Playlists, and Videos. The user can switch between tabs to narrow the result type, or exit search using the close icon. **XChat: Conversation Filtering and Mark-Read** (ios, 4 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/11672) > Happy-path: A signed-in user opens the Chat list, uses the top-right filter button to narrow conversations (Direct / Groups / Unread / All) or mark messages as read. The filter menu appears as a floating rounded card. Choosing a filter updates the header subtitle and the conversation list; choosing "Mark All Read" clears unread messages and shows an empty unread state. The bottom persistent search/compose controls remain visible throughout. Entry point: Chat list screen (user is logged in). Exit point: Filtered conversation list or an empty-state screen indicating there are no matching conversations (e.g., "No Unread Messages"). Assumptions: User is authenticated, on a mobile device (portrait), and has at least one conversation. The app supports conversation-level read/unread state and categorization into direct vs group conversations. Steps: Conversation List (All) → Open Filter Menu → Unread Empty State → Groups Filter Result **SSENSE: Search & Results Discovery** (ios, 6 screens, Marketplace) [Refero](https://refero.design/flows/7104) > Mobile shopping -> Search interaction (happy path and alternative). Entry: user is on the Shop/Home feed (bottom nav present) and taps the top-left SEARCH affordance. A full-screen search overlay appears with category tabs (WOMENSWEAR / MENSWEAR / EVERYTHING ELSE), an active text input and keyboard, suggested searches and recently viewed thumbnails. The user can type free text or numbers, tap a suggested designer, or switch categories. Exit points: product results listing (when matches found) or empty-results state (when no matches). Assumptions: mobile iOS viewport, user reached this flow from the Shop/Home feed, basic navigation (bottom nav, bag count) is available on every screen, keyboard is controlled by the device. Steps: Shop feed / Search entry → Search overlay — input focused → Numeric query entry → Empty results state → Text query & suggestions → Search results listing **CocoonWeaver: Cocoon list and sorting** (ios, 4 screens, Productivity) [Refero](https://refero.design/flows/4624) > Single-screen list of "Cocoons" (collections/categories) with quick sorting/filter chips across the top, a search affordance, large rounded list items that represent each Cocoon, a per-item favorite button (heart), and a floating add action (+ Cocoon). Entry: user opens the All Cocoons view. Happy path: user browses the list, changes the sort via filter chips, favorites/unfavorites items, and uses the + Cocoon control to start creating a new Cocoon. Exit points: tapping an individual Cocoon (expected to open that Cocoon), tapping search (expected to open search), tapping + Cocoon (expected to open creation flow). Assumptions: user is on a mobile device and already in the app (logged in state not shown). The bottom navigation remains visible while interacting with this list. Steps: Cocoon list — overview → Filter: Fresh selected → Filter: Random selected → Filter: Alphabetical selected **Pinterest: Board edit and collaborator management** (ios, 16 screens, Reference, Inspiration Board, Social Networking, Lifestyle) [Refero](https://refero.design/flows/2489) > Happy-path: User opens a board, opens the board options, chooses Edit board, updates the board cover (selecting a pin, cropping and saving), edits the board description, manages collaborators (invite people and set collaborator permissions), then returns to the board with a saved cover, updated description and pending collaborator invites. Steps: Board details → Board options sheet → Edit board (initial) → Edit board (scrolled) → Set board cover list → Select pin as cover → Crop cover (initial) → Crop cover (adjust) → Saving cover → Edit board (cover saved) → Edit description → Collaborators list → Collaborator permissions → Change collaborator permission → Invite members (picker) → Finish and saved board **Spotify: Filter library to my playlists** (ios, 3 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/2190) > Happy-path: a signed-in user opens the Library section on a mobile device (bottom tab bar shows 'Your Library' active). The user switches from the default Artists view to Playlists, then applies the 'By you' filter to see only playlists they've created. Persistent UI elements: search icon, create (+) button, sort control (Recents), layout toggle (grid icon) and the mini player at the bottom. Entry: user in 'Your Library' tab. Exit: filtered list of user-created playlists displayed. Steps: Library — Artists (initial state) → Library — Playlists (unfiltered) → Library — Playlists filtered 'By you' **BeReal.: Account creation and onboarding** (ios, 29 screens, Social Networking) [Refero](https://refero.design/flows/4896) > Happy path onboarding for a new iOS user: app launch → phone number entry (with country code selector) → privacy/consent modals → OS tracking prompt → verification flow → create username → contacts permission and optional contact sharing/invites → notification permission → final home screen with CTA to post first BeReal. Entry: fresh install / first-time open. Exit: user lands on main friends feed (onboarding complete). Steps: Splash → Loading → Phone number entry → Privacy / trackers modal → OS tracking permission (ATT) → Country selector (list) → Country search → Phone formatted input → Send verification → Enter verification code → Create username (unavailable) → Create username (available) → Contacts permission prompt (OS) → Contacts sharing options → Add friends (select contacts) → Invite via SMS composer → Notifications rationale → OS notification permission → Finalizing onboarding (loading) → Home feed (onboarding complete) **Days: Create visual event card** (ios, 20 screens, Scheduling Tools, Lifestyle) [Refero](https://refero.design/flows/7790) > Happy-path for creating a visual event card inside a calendar-style app. Entry: user starts on the Events overview. The primary path is: Create new event → enter title → pick or import date (optionally set repeat) → add photo (library/Unsplash) → crop/position image → style text (font/color) → choose category → finish. Steps: Events overview → Add title (initial) → Calendar connect prompt (card) → System calendar permission → Title suggestions & Continue → Title filled → Pick date (wheel) → All Day toggle → Repeat toggle (disabled) → Set repeat frequency → Add photo options → Photo library grid → Crop / position image → Event card preview (first pass) → Edit Font dialog → Color applied → Edit Category → Final preview with actions → Event detail / full-screen view → Finalized card on overview **Volv: Search from profile** (ios, 3 screens, AI Tool, News) [Refero](https://refero.design/flows/2164) > Happy path: A logged-in user opens the profile/drawer, taps the prominent "search." control, types a query and submits it. The UI transitions from the drawer to a dedicated search screen, focuses the text input (keyboard appears), then displays search results or an empty-state message. Steps: Open profile / invoke search → Input focused — type query → Submit search and show result **Poppy: Memories Search** (ios, 4 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12185) > This is a mobile Memories search screen. The user opens the Memories section from the main experience, can search across memories and saved memory categories, and sees either no results or a list of matching entries. Steps: Memories Search Home → No Results State → Results List → Results List Expanded **District: Search Discovery** (ios, 6 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12842) > A user enters the global search screen, types a query, and reviews results across category tabs such as All, Dining, Movies, Events, Stores, and Activities. The entry point is the search bar from the main discovery area; the exit point is either selecting a result, refining the query, switching tabs, or clearing the search. **HYPE: Discover Search** (ios, 6 screens, Lifestyle, News) [Refero](https://refero.design/flows/12249) > This is a mobile discovery and search experience inside a fashion/streetwear content app. The user starts on the Discover tab, opens Search, enters a query, and views either autocomplete suggestions or a results screen. The happy path shown assumes the user is already signed in and using a phone. The user can return with the back arrow, clear the query with the clear icon, or navigate using the bottom tab bar. Steps: Discover Browsing → Search Entry → Query Input → No Results State → Suggested Search Terms → Search Results **Tripsy: Trip Itinerary Search** (ios, 3 screens, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/2364) > Happy path: User views a trip overview on a map (named "Quest Voyage"), opens the trip-specific search, types a query and reviews matching saved activities and documents. Steps: Trip Map Overview → Search Saved Activities → Filtered Results (Documents & Emails) **Dropbox: PDF in-document search** (ios, 3 screens, Business, Collaboration Tools, Productivity) [Refero](https://refero.design/flows/2438) > Happy path: User opens a PDF in a mobile document viewer, invokes the in-document search, types a query, and reviews results. Steps: Document viewer (PDF open) → Search input (empty) → Search results — no matches **Volv: Search from account menu** (ios, 3 screens, AI Tool, News) [Refero](https://refero.design/flows/2165) > Happy path: A logged-in user opens the app's slide-over menu to access quick actions or search. From the menu they tap the promin _… 100760 more characters in the zip._ ### research/legacy/BLUEPRINTS-FEATURES-JOURNEYS.md `research/legacy/BLUEPRINTS-FEATURES-JOURNEYS.md` · sha256 `f56ba0c3c49a2f7985d4bc5f9eab2dc4c15725471abd848f4359db7ce8f93858` #### Old estate blueprints, features and journeys: what they record and what is still to specify Research note, written 31 October 2026, for the new estate (JudeAI | EDGE ID: EDGE Finance, EDGE Clinic, EDGE Public Consortium, Judiverse). **Status of this note.** Research, not authority. Everything quoted from the old estate is historical. Every number from the old estate is given as "recorded in on (historical)", next to my own recount where I made one. Nothing in this note is approved, built or verified. Old identities (P-codes, rule numbers, contract versions, unit IDs) are named only so that a reader can find the source; they must not be carried into the new estate. **Paths.** Corpus paths are relative to the research corpus root and start with `legacy/` or `old-repo/`. New-estate paths are relative to `new-estate/`. --- ##### 0. How I read the corpus, and how much I covered **Scope.** Catalogue rows with `generated: false` whose categories include `blueprints-specs` or `features-journeys-screens`. | Step | Result | |---|---| | Rows in scope | 6,961 (legacy 5,401; old-repo 1,560) | | Distinct blob oids | 6,838 | | Groups after de-duplication | 6,603. I stripped branch-version hash suffixes (for example `register.490f9012.json` → `register.json`) and folded mirror trees. `legacy/local-source-snapshots/` mirrors `legacy/x-judeai-edge-tok-kol-ndc-ml-x/`; `legacy/github-metadata-20261030/local-working-changes/...` mirrors parts of both; `legacy/governed-intelligence-judeai-edgeid/` mirrors parts of `old-repo/`. | | Run evidence and score records | 3,748 groups. About 2,220 are outcome-register evidence packets dated 19 to 22 October 2026 (`.../programme/outcomes-register/evidence/...`), about 1,155 are benchmark score records (`old-repo/handover/benchmark-standard/scores/...`), and the rest are proof folders. These record runs, not features. I triaged them by name and type and sampled them; I did not read them one by one. | | Specifications, registers and documents | 2,855 groups: 1,652 JSON/JSONL and 1,203 prose, CSV, HTML or PDF files. | | Registers analysed by script | Every JSON/JSONL group: item counts, fields, fill rates, distinct values and samples. The large registers named in sections 1 and 2 were analysed in depth. | | Documents read in substance | About 60 governing and blueprint documents, read in full or in large part (named in section 1). Of the 81 product specifications and about 80 journey contracts under `legacy/edge-judiverse-tech/docs/`, I read P29 Payments in full; for the rest I extracted headings, journey titles and function tables by script, because they share one template. The 40 regulatory-pack documents (`legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/`) were read by title only, because they are assurance documents, not feature specifications. | **Gaps in my coverage.** - I read the 38 MB build specification (`old-repo/handover/tok-20261030/specification/build-spec-20261031/SPEC.md`) by script: its units, outcomes and counts. I read one unit in full and sampled about 40 more. - The authored unit files behind it (`.../authoring-20261031/out/*.jsonl`) are not in this scope, so I did not open them. --- ##### 1. Inventory: blueprints, specifications, ontologies, schemas and registers Blob oid first, then the corpus path. Where a file exists in several mirrors, one representative is listed. ###### 1.1 Feature, capability, journey and screen registers (counted in section 2) | oid | Path | What it is | |---|---|---| | `ec14f56acfd1293443652616b095452643db5685` | `old-repo/handover/journeys-prospective/COUNT.json` | Count and counting method for the prospective journey register | | `29164c03fc55ced371155a5255f9a0e0c09ca0da` | `old-repo/handover/journeys-prospective/JOURNEYS.jsonl` | 2,139 journeys, each with steps, screens, controls, operations, failure paths and variants | | `930d14c1387934625c28ae29447887ba1343c932` | `old-repo/handover/journeys-prospective/FEATURES.jsonl` | 2,897 features (the same 2,897 rows also appear in `CAPABILITIES.jsonl`, oid `93e0e05aa78d159ef6a36caee3d3197f21617a70`, with a disposition added) | | `714c42d64cc3c7879e629dc8490238d442e1eb30` | `old-repo/handover/journeys-prospective/SCREENS.jsonl` | 5,363 generated screens | | `51e591eb75f05e74c5110e6503fe72f1af646231` | `old-repo/handover/journeys-prospective/SCREEN-CONTRACTS.jsonl` | 5,363 screen contracts: layout template, states, roles and action binding | | `a231217f7e8fb997fd2af621dfdd965b196e8a6d` | `old-repo/handover/journeys-prospective/CONTROLS.jsonl` | 7,725 controls | | `0dc5694ef4b815268ce3f31bd9d29a96e285d198` | `old-repo/handover/journeys-prospective/SERVICES.jsonl` | 6,121 backend operations, with authority and lawful basis | | `aa720f13b1408e12319cb01b5606d9c68e2e4dd9` | `old-repo/handover/journeys-prospective/PROOFS.jsonl` | 1,611 operation proofs | | `46af1206050fa16b2411d58abebd47e81aecb7b5` | `old-repo/handover/journeys-prospective/CERTAINTY.json` | Two-enumerator capture–recapture comparison | | `e433b946b1adb644a0ba7274adb8bc2edbafd799` | `old-repo/handover/journeys-prospective/README.md` | How the register was generated | | `39231f12533298405d7a192fac3fec3752100ffb` | `old-repo/handover/scope-freeze/package/REQUIREMENTS.jsonl` | **780 Owner-intended capability rows.** The best single list of what the organisation itself intended. | | `32d8e91acb1b08b879b28662f06a0eace7d1e101` | `old-repo/handover/scope-freeze/package/JOURNEYS.jsonl` | 5,047 journeys taken from an evidence matrix of 23 October 2026 | | `3b0d3594ee724d5d9657af6b43ba59b66876c416` | `old-repo/handover/scope-freeze/package/SCREENS.jsonl` | 21,048 generated screen URLs | | `a1a0e781cae7e91b78edeade85a526a74c8ae3a5` | `old-repo/handover/scope-freeze/TARGET-JOURNEYS.json` | A multiplicative target count (72,550 "journey definitions") | | `4474313bd40f2453c40866422efb3b8c181c69bb` | `old-repo/engine2/data/FEATURES.json` | Feature catalogue: 91 capabilities (P01 to P91) and 5,402 entries | | `c52191076ec52195eff871ab524dfdbb8c04a958` | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/teams-full-feature-scope-20261022/compiled/requirements.jsonl` | 3,225 Microsoft Teams parity scope rows | | `e44b83d7779c1927396e8c6582125bb8984a0587` | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/packs/feature-workflows/feature-workflows.json` | 1,038 generated feature workflows | | `c4338830d64b0b2c78a481ada5e0e9126d78c75a` | `legacy/judeai-edge-control/tests/bank_workflows/register.json` | 357 bank role workflows | | `77935fa4b3916afef9ff4ac6779b0d549e0c4fe6` | `legacy/judeai-edge-control/tests/clinic_workflows/register.json` | 332 clinic role workflows | | `1651ab8f8ff3c380af20d5169dbff315e67888c8` | `legacy/judeai-edge-control/tests/consortium_workflows/register.json` | 348 consortium and academic role workflows | | `59166b9500e1d2856c1cfcd082bb3b31da8cb496` | `legacy/judeai-edge-control/tests/government_workflows/register.json` | 479 government role workflows | | `bc933865d2fc9f17b1df134924c972b631f492a7` | `legacy/edge-judiverse-tech/docs/takeover/PLAN/REQUIREMENT-LEDGER.json` | 830 function contracts plus 144 Kubernetes-of-Life requirements | | `0568ee1b8655c34ddc750119667987e855547e3c` | `legacy/edge-judiverse-tech/docs/takeover/PLAN/SCREEN-STATE-MATRIX.csv` | 6,188 screen × required-state rows | | `20b47da1a4f3a1cdb54cc5077a01733a651e3220` | `legacy/edge-judiverse-tech/docs/takeover/SCREEN-STATUS.csv` | 448 registered screens and their status | | `e28763710da5a258294de07da6fc315bc8394bf0` | `legacy/edge-judiverse-tech/docs/finance/bank/03-FEATURE-CATALOGUE.md` | 225 bank features (JB-F-001 to JB-F-225) | | `54c8742c31bf7f5edf534590c5245461a82dacca` | `legacy/jude-mvp/jude-live/file-capabilities.json` | 1,000 file-handling capabilities for the assistant | | `36489c832dc1942115e31347118f24bd5bcd1e29` | `legacy/edge-judiverse-tech/registry/capability-register.json` | 102 agent behaviour capabilities | | `886cd57d90b1cd1ffef92909ec7909023d4640fc` | `legacy/edge-judiverse-tech/registry/kubernetes-of-life-requirements.json` | 144 platform requirements | | `7afa59759b59d657be0f109c8e38d2242639604c` | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/map/LEVEL-REGISTER-20261020.json` | 4,579 mixed nodes (workflows, outcomes, claims, tickets, routes and more) | | `ae146a0a0261312dbe33787b818cc6d7a7b4a021` | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/packs/build-list/compiled/unit-routes.csv` | 3,292 build units across 12 packs (the same set is in `.../handoffs/estate-graph-20261022/original/units.source-fd3653c0.csv`, oid `4cfd743a4a0bc0f8ed2175e85fc55ad55a81891f`) | | `3dcb2c6f6e487c5621e68809d31f9b0c4c4afe79` | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/parallel-proof-20261021/spec/scope/units-01.json` (and `units-02` to `units-12`) | 4,621 scope units: 81 apps and the rest claims | | `204286660dca8314ad0f2c23cf2aae384c06f0ba` | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/outcomes-register/register.json` | 838 outcome records. These are work items, not product features. | | `eaead569d30bd1e4ace79cccc44894a388d78270` | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__parity__DEFINITIONS-REGISTER.jsonl` | 1,391 defined terms | | `da32316488f9e6cf8e1e6f385fdc15133a4a9e65` | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__parity__PARITY-DESIGN.jsonl` | 173 design-tool parity features | | `f0662ac1b99ba74f9047894a5877cd8087d66576` | `legacy/judeai-edgeid-cloud/handover/whole-estate-acceptance-20261026-01a16a2c/REQUIREMENT-TRACE.json` | 195 acceptance requirements | | `b5b144277336aa1fee03dbf5b73dcc57ede4043f` | `old-repo/registers/parity/PARITY-20261029T232500Z.json` | 321 competitor flows (285 of them banking) | | `8d1de824d1566c6f40e7fcb96c34d68c4c46fb82` | `old-repo/registers/parity/GAP-DEFINITIONS-20261029T232500Z.jsonl` | 95 parity gaps | | `7ef8741f2efc65ef1604566bad9622e8201d4061` | `old-repo/engine2/scenarios/research/source/bank-capabilities-20261029.json` | 637 bank operations | | `4c9f1a445976cd21f086f3ed93bf587f39bb4233` | `old-repo/benchmarks/frontier/CAPABILITIES-P92PLUS.json` | 30 further capabilities (P92 onwards) | | `b64a942f1ec83a01a97ce2745f0aff3005992aa8` | `old-repo/handover/benchmark-standard/BENCHMARK-REGISTER.jsonl` | 1,704 benchmark definitions | ###### 1.2 Build specifications and their coverage records | oid | Path | What it is | |---|---|---| | `b046718d7381944f7d6dfb2d02a123af187634fc` | `old-repo/handover/tok-20261030/specification/build-spec-20261031/SPEC.md` | Full rendered build specification: units with workflows, screens, controls, data model, backend contract and acceptance tests | | `95f2e7b05149075edf8f2dc8ef14e38d1504e201` | `old-repo/handover/tok-20261030/specification/build-spec-20261031/COVERAGE.json` | Counts for the build specification | | `6677160f4f5cfd685391e211ad88607e98b04dbd` | `old-repo/handover/tok-20261030/specification/consolidated/SPECIFICATION.md` | Consolidation of 80,818 source units into 7,394 capabilities. Its counts are in `consolidated/COVERAGE.json`, oid `09b6ffaa5835e89b083dbfba1ff16ad1403bf422`. | | `fa9dbe91954c2621a783173a07cbd11609764728` | `old-repo/handover/tok-20261030/specification/named-spec-20261030/README.md` | 1,819 named capabilities and 39,130 named features | | `8f8a6dcf2fcca4ba8994840d516018c64a1c93ac` | `old-repo/handover/tok-20261030/specification/NEXT-AGENT-HANDOVER-20261031.md` | Last handover. It states that the different counting grains are not interchangeable. | | `5a3deb721a5ed7a683fc9da886be714b21d52b3b` | `old-repo/handover/tok-20261030/specification/authoring-20261031/BRIEF.md` | Unit authoring brief: the old generator rules | | `cac6e3cbc01cd04be9fa1a5c3f7c3a0ba02bf5b7` | `old-repo/handover/tok-20261030/specification/authoring-20261031/BRIEF-ADDENDUM-5-AGENT-NATIVE.md` | Agents as native users: human path, agent path and harmony | | `3060f325b19f887f08660bdd796486441bfbb516` | `old-repo/handover/tok-20261030/specification/authoring-20261031/inputs/A1.jsonl` (and A2 to A5, B1 to B5) | Authoring inputs: about 1,527 A-units and about 290 B-units | | `0e5534fc7affe05a294c5c40dba43361e7bcadb1` | `old-repo/handover/tok-20261030/specification/authoring-20261031/TOK-VOCABULARY.json` | 186 ontology kinds | | `4f5406ea8aac2788483c63481679fbeef71ecf41` | `legacy/jude-mvp/docs/SPEC.md` | The earlier "legacy estate as a feature specification" | | `ba18cd2cd99296f3a5e86828a33229e4ddfc06bc` | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/MASTER-PLAN.md` | Compiled master plan of 15 October 2026. It copies the documents dated 12 to 15 October and adds no new features. | ###### 1.3 Product specifications and journey contracts (one template, 13 October 2026) | oid | Path | What it is | |---|---|---| | `c6b77e1aa4a1d462c333f032cd453fcf76c7dbf6` | `legacy/edge-judiverse-tech/docs/products/README.md` | Index of 81 product specifications (P01 to P81) | | `ae3d6a172ebf4e00e154dd569f1812033b20fe78` | `legacy/edge-judiverse-tech/docs/products/P29-payments.md` | Worked example of a product specification | | `683ef37711803cbd6e11801f6fef5b93e08a1fc8` | `legacy/edge-judiverse-tech/docs/journeys/finance/P29-payments.md` | Worked example of a journey contract | | `c73def02e35e40c24f083cd9f3d4c271977940ce` | `legacy/edge-judiverse-tech/docs/journeys/README.md` | Engineering and journey handbook | Also in this template family: - `legacy/edge-judiverse-tech/docs/estate/ESTATE-WHOLE-OF-LIFE-JOURNEYS.md` (journeys EST-J-001 onwards) - `legacy/edge-judiverse-tech/docs/finance/bank/04-CUSTOMER-JOURNEYS.md` and `05-STAFF-REGULATOR-ENGINEER-JOURNEYS.md` - `legacy/edge-judiverse-tech/docs/products/P81-JUDEBOX-JOURNEYS-AND-REQUIREMENTS.md` - `legacy/edge-judiverse-tech/docs/outreach/03-FEATURE-CATALOGUE.md` ###### 1.4 Ontologies, schemas and governing specifications | oid | Path | What it is | |---|---|---| | `fe174571f8dd59724c29dd46b5e05a0775dc8b59` | `old-repo/handover/ontology-20261030/README.md` | Ontology stage 2: kinds and their counted population. The kind catalogue itself is in `stage1-v2/kinds.txt` and `stage1-v2/TYPES.md`. | | `611d7f707e900d51985c16c0a32e231cda6a3a88` | `old-repo/handover/audit-20261031/NEW-REPO-PROPOSAL.md` | Proposal for an ontology-first repository | | `cbaa730945183f7ab7969b37666e989f13839ae7` | `legacy/governed-intelligence-judeai-edgeid-pre-contract/spec/GES-1.1.md` | Governed Estate Specification 1.1 | | `75440f3f75ea6ab2724b50a763817cbe7bcd7558` | `legacy/governed-intelligence-judeai-edgeid-pre-contract/spec/ANNEX-E-ACTION-ENVELOPE.md` | Action envelope rules | | `2708893c8e7c3eb7933deffca2bec6d54b2d709d` | `legacy/governed-intelligence-judeai-edgeid-pre-contract/LEGACY.md` | Which repositories are superseded | | `bc9968ac4066fee79cb04a360d91b76577adc1d1` | `legacy/superseded-contracts-20261030/README.md` | Which contracts are superseded | **Schemas** (fields only, no instances): - `legacy/edge-judiverse-tech/schemas/capability-contract.schema.json` - `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/parallel-proof-20261021/spec/schemas/` (action, unit, receipt, certificate, status, review-request and asset schemas) - `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/outcomes-register/schema.json` - `legacy/jude-mvp/work/jude-lab-training-sprint-20261001/TOK_AXIOM_BLUEPRINT.schema.json` **Screen and design blueprints:** - `legacy/local-source-snapshots/foundations/canvas/templates/TEMPLATES.md` - `legacy/edge-judiverse-tech/docs/design/DESIGN-TEMPLATE.md` - `legacy/edge-judiverse-tech/docs/design/HOUSE-STYLE.md` - `old-repo/handover/tok-20261030/specification/design-system-20261031/DESIGN-SYSTEM.json` - the Figma job lists at `old-repo/handover/tok-20261030/specification/authoring-20261031/figma/JOBS-1.jsonl` to `JOBS-4.jsonl` **Domain workbench registers:** - `legacy/x-judeai-edge-tok-kol-ndc-ml-x/workbenches/materials/MATERIALS-CAPABILITY-REGISTER.md` - `legacy/x-judeai-edge-tok-kol-ndc-ml-x/workbenches/urban/URBAN-CAPABILITY-REGISTER.md` - `legacy/judeai-edge-control/docs/research/life-control-plane/life_control_plane_specification.md` --- ##### 2. Feature coverage: how many features were actually recorded ###### 2.1 The short answer on "near 100,000 features" The registers do **not** show anything near 100,000 distinct features. Figures of that size exist, but each one is one of three things: - a raw count of competitor text fragments; - a multiplication of variants; - a prediction. | Figure | What it actually is | Source | |---|---|---| | 80,818 "source units" | Inputs, not features. Recorded in `old-repo/handover/tok-20261030/specification/consolidated/COVERAGE.json` on 30 October 2026 (historical). Made up of 57,014 competitor requirement clauses, 13,126 competitor catalogue items, 6,426 competitor functional roll-ups, 2,897 estate features, 780 Owner capabilities and 575 parity obligations. I re-added these parts and they do sum to 80,818. | consolidated | | 89,607 "placements" | The same 80,818 source units, counted once per family when they are shared | `.../named-spec-20261030/README.md`, 30 October 2026 (historical) | | 72,550 "journey definitions" and 181,375 "verification runs" | Multiplications: 7,255 feature units × 5 outcome variants × 2 actor classes, and × 5 surfaces. The file marks itself PARTIAL. | `old-repo/handover/scope-freeze/TARGET-JOURNEYS.json`, 30 October 2026 (historical) | | 18,000 capabilities / 30,000 features / 40,000 screens / 400,000 control instances | Marked "PREDICTION … not a quota; not compared with specified counts" | `.../consolidated/COVERAGE.json`, 30 October 2026 (historical) | Feature registers that are reduced to distinct estate outcomes come to the **low thousands**. The best old count is the prospective journey register: 2,139 journeys from 2,897 feature rows. My recount gives 2,088 distinct outcomes (section 2.3). ###### 2.2 Recorded counts against my recount How to read this table: - "Recorded" means recorded in that file on that date (historical). - "Recount" means my own reading of the same file. - "Not semantic" means the register's items are not distinct outcomes, so I did not de-duplicate them further. | Register | Recorded | Recount and de-duplication | |---|---|---| | Owner-intended capabilities, `old-repo/handover/scope-freeze/package/REQUIREMENTS.jsonl` (rows reconciled 30 October 2026) | 780 | 780 rows. 769 distinct capability texts after lower-casing and trimming. Recorded status: 384 SPECIFIED, 315 MISSING, 73 SIMULATED, 8 EXECUTED. Basis: 604 EXPLICIT, 104 NECESSARY, 51 INFERRED, 21 OPTIONAL. | | Prospective journeys, `old-repo/handover/journeys-prospective/COUNT.json`, observed 30 October 2026 19:58 UTC | 2,139 journeys; 2,897 features; 5,363 screens; 7,725 controls; 6,121 operations; 1,611 proofs; 25,736 variants inside journeys | 2,139 unique IDs. 2,088 distinct (action, object) pairs once family is ignored and objects are normalised (lower case, determiners removed, singular). 1,516 journeys are not plain "view" journeys. Of the 2,139, 773 come from Owner rows and 1,365 only from competitor rows (one has both). | | Same register, features | 2,897 (1,666 with journeys; 1,231 "feature areas" with no stated outcome) | 1,630 distinct (application, normalised object) pairs among the features that have a stated outcome | | Same register, independent second count, `CERTAINTY.json` | Enumerator A 2,139; enumerator B 1,436; matched 1,243; Chapman estimate 2,471; **0 of 10 families certain** | Reproduced from the file. The two enumerators agreed on only 1,243 journeys, so the true number of distinct journeys is uncertain within roughly 1,400 to 2,500. | | Scope-freeze journeys, `old-repo/handover/scope-freeze/package/JOURNEYS.jsonl` (evidence matrix, 23 October 2026) | 5,047 journeys; 21,048 screens | 21,048 screen IDs, but every one is a generic typed URL (viewer 4,801, receipt 4,635, form 3,803, decision 3,702, intent 2,917, and others). Only 6 distinct design frame references are used across them. These are not distinct designed screens. | | Engine feature catalogue, `old-repo/engine2/data/FEATURES.json` (undated; its sources name 27 October 2026) | 91 capabilities; 5,402 entries | 4,567 real entries; the other 835 are bare operation stubs with no definition. 3,939 distinct (application, intention) pairs. Kinds: 1,810 workflow, 1,356 feature, 670 role-workflow, 344 requirement, 339 untyped, 29 screen, 19 capability. Only **86 distinct outcome predicates** across 4,567 entries, so predicates were templated. | | Build specification, `.../build-spec-20261031/COVERAGE.json`, 31 October 2026 01:48 UTC | 29,582 capability records; 1,828 authored units; 3,426 screens; 15,392 controls; 17,845 operations; 17,584 tests; 196 units pass the gate; 97 capabilities complete | The 29,582 are mostly competitor documentation sections (15,535), catalogue items (6,843) and roll-ups (6,426); only 777 are Owner capabilities. The rendered `SPEC.md` (snapshot 01:28 UTC the same night) says 1,347 units and "0 of 1347 units pass". My parse finds 1,346 unit headings with 2,670 workflows, 2,407 screens, 6,594 operation rows and 7,867 test lines. The handover file of the same date says "fully passing … 2 / 1,828". **These three records disagree with each other.** | | Consolidated specification (30 October 2026) | 7,394 capabilities; "Fully specified capabilities: 0" | Not recounted; the stage table is in section 3 | | Named specification (30 October 2026) | 1,819 capabilities; 39,130 named features; "0 chains are complete" | Not recounted. Most named features are competitor clauses (for example 10,762 for the old lab family and 8,759 for the old cyber family). | | Role workflows, `legacy/judeai-edge-control/tests/*_workflows/register.json` (source evidence dated 20 October 2026) | 357 bank, 332 clinic, 348 consortium, 479 government; total 1,516 | 1,513 distinct titles. The "Today" field records 178 as "live" (64 bank, 49 clinic, 64 consortium, 1 government; self-reported, historical). Most others are "not started". **The best-written workflows in the corpus:** each has trigger, steps, screens, records, permissions and expiry, the regulation or standard, and a "done when" test. | | Feature workflows, `.../packs/feature-workflows/feature-workflows.json` | 1,038 | 1,032 distinct names: 460 per-screen, 334 per-function, 82 help, 81 record and 72 per-app workflows. Generated: every one has the same five failure paths. | | Teams parity scope (22 October 2026) | 3,225 | 2,729 distinct titles. Only 167 are workflows. 1,495 are official documentation topics, 893 source-row obligations, 618 linked references and 52 frontier controls. This is competitor scope, not estate features. | | Function contracts in `REQUIREMENT-LEDGER.json` (October 2026, undated) | 830 functions | 527 journey functions and 303 viewer controls. Only 6 distinct precondition sets, 6 recovery texts and 5 error contracts across all 830. | | Product specifications P01 to P81 (13 October 2026) | 81 apps; 442 screens (SCREEN-STATUS lists 448 rows) | 425 functions and 303 controls in the specifications. 324 numbered journeys across the journey contracts. 283 distinct screen labels. 438 screens recorded "not started / implementation unverified". | | Bank feature catalogue (13 October 2026) | JB-F-001 to JB-F-225 | 225 distinct features. Well formed: each has a definition, journey, rule, data and measured outcome. | | File capabilities (undated) | 1,000 | 560 tools, 220 platform controls and 220 skills. All are "proposed_not_verified". | | Level register (20 October 2026) | 4,579 nodes | A mixture: 1,516 workflows (the same set as the role workflows above), 838 outcomes, 691 claims, 574 desk tickets, 454 routes, 311 CRL items, 81 apps, 60 risks and others. Only 1 node is recorded at level 9. | | Build units (22 October 2026) | 3,292 | 3,278 distinct. This is the union of the 1,038 feature workflows, the 1,516 role workflows and the frontier packs (447 controls and others). | ###### 2.3 Per application, mapped to the four new applications **Mapping, by my judgement.** The old families were not drawn on the same lines as the new applications, so this mapping is approximate. | Old family | New application | |---|---| | bank / "Jude Bank FI" / "EDGE Bank" | EDGE Finance | | health / "Jude Health CI" / "EDGE Clinic" | EDGE Clinic | | consortium (education and research, plus some office work), cyber, public | EDGE Public Consortium | | office, judeai, lab, JudeBox | Judiverse | | shared foundations, cross-family, organisation onboarding | Cross-cutting (EDGE ID and shared foundations) | | Measure (historical source) | Finance | Clinic | Public Consortium | Judiverse | Cross-cutting | |---|---|---|---|---|---| | Owner capabilities (`REQUIREMENTS.jsonl`) | 134 | 79 | 130 | 243 | 194 | | Prospective journeys | 320 | 223 | 880 | 527 | 189 | | Of which Owner-sourced | 134 | 79 | 131 | 241 | 189 | | Prospective features (with stated outcome) | 248 (243) | 219 (162) | 1,048 (624) | 1,196 (451) | 186 (186) | | Prospective screens | 835 | 546 | 2,086 | 1,502 | 394 | | Engine catalogue entries (distinct intentions) | 1,328 (1,176) | 477 (417) | 811 (619) | 1,951 (1,727) | — | | Role workflows | 357 | 332 | 827 (348 consortium + 479 government) | — (the office, mail and channel workflows sit inside the consortium register) | — | | New-estate units today | 26 | 8 | 8 | 8 | — | **Counting method I used.** 1. I loaded each register in full. 2. I counted unique IDs. 3. I counted distinct normalised outcome texts: lower case, punctuation and determiners removed, plurals made singular, and paired with the action verb where one exists. 4. I mapped families to the new applications as above. 5. I did not merge registers into each other. Cross-register merging would need a reading of each item. The prospective register already merged the 780 Owner rows with 13,126 competitor rows. Adding the role workflows and the bank catalogue would add some hundreds to perhaps a few thousand outcomes; I cannot give a reliable figure without that reading. **Honest summary.** The old estate recorded: - roughly 2,000 to 4,000 distinct estate-side outcomes; - about 770 Owner-intended capabilities; - about 1,500 well-specified role workflows; - 50,000 to 80,000 competitor text fragments. --- ##### 3. Fields the old registers used, and which were consistently missing ###### 3.1 Fields used **Prospective journeys (`JOURNEYS.jsonl`).** - Identity and placement: `id`, `title`, `type`, `application`, `family`, `owner`, `feature_ids`. - Outcome: `outcome` (action, object), `start_condition`, `result_state`, `destination`. - Actors: `actor_classes`, `roles`. - Steps: each step has a screen, screen type, controls, an instruction and a backend operation (with receiving system and state). - Failure and recovery: `failure_paths`, `recovery`. - Variants: actor classes, permission outcomes (granted, denied, step-up), failures, environments, roles and competitor floors. - Other: `handoffs`, `cross_family`, `intended_basis`, `build_state_in_sources`, `source`, `acceptance`, `verification`. - A block called `p68`: audit trail, carbon standard, inspection surfaces, intention and authority. **Features and capabilities.** `id`, `object`, `outcome_stated`, `journeys`, `source_rows`, `acceptance`, `recovery`, `verification`, `disposition`. **Screens and screen contracts.** `screen_type` (list, detail, form, confirm, result, workspace, handoff), `layout_template`, `controls`, `inputs`, `states` (empty, loading, ready, error, denied), `roles`, `action_binding`, `receiving_system`, `reference`, `truth_status`, `acceptance`, `recovery`, `evidence`. **Operations (`SERVICES.jsonl`).** `authority` (grant, identity, regulated statement), `identity`, `lawful_basis`, `receiving_system`, `parity_reference`. **Function contracts (`REQUIREMENT-LEDGER.json`).** `code`, `label`, `kind`, `path`, `permission`, `input`, `preconditions`, `output`, `error_contract`, `idempotency`, `recovery`, `status`. **Role workflows.** `Workflow`, `Trigger`, `Steps`, `Screens`, `Records and receipts`, `Permissions and expiry`, `Regulation or standard`, `Today`, `Done when`, role `group`. **Bank catalogue.** ID, feature, definition, journey, law or rule, data needed, measured outcome, EDGE basis, status. **Build-specification units.** - Outcome and comparison: intended outcome, "beyond competitors", capabilities served, a competitor comparison table. - Workflows: steps, outcome, human path, agent path (with MCP tool), harmony, and TRL/CRL blocks. - Screens: Mobbin and Figma references; controls with kind, effect, operation and a dead-control test. - Data model: entities, fields and types. - Backend contract: method, route, request, response, errors, authority and MCP tool. - Acceptance tests: Given / When / Then, marked person or agent, plus negative tests. **Screen-state matrix.** `required_state` (twelve states: empty, ready, loading, partial, offline, denied, expired, conflict, failed, awaiting-review, running, completed), `themes`, `source_modes`, `privacy_states`, `platforms`. ###### 3.2 Consistently missing or only templated | Missing or templated | Evidence (historical) | |---|---| | **Data model and request/response schemas** | "Data model is a GAP everywhere": 0 of 39,130 named features (`named-spec-20261030/README.md`, 30 October). Consolidated record: data model DEFINED for only 616 of 7,394. | | **Specific acceptance tests** | DEFINED for only 8 of 7,394. 7,303 were "templated only" (consolidated, 30 October). Prospective features use 2 distinct acceptance texts across 2,897 rows; screens use 1. | | **Specific recovery** | 1 distinct recovery text across all 2,897 prospective features and all 5,363 screens. Every journey has the same two failure paths (permission denied, dependency unavailable). | | **Specific outcome predicate** | 86 distinct predicates across 4,567 engine catalogue entries | | **Lawful basis for operations** | OPEN for all 6,121 operations (`SERVICES.jsonl`) | | **Approved design** | Figma frame pending for all 5,363 prospective screens. The build specification records 2 Figma receipts. Mobbin competitor screen unresolved for 1,264 units (COVERAGE) or 2,407 screens (`SPEC.md` snapshot). | | **Controls on screens** | Present on only 198 of 5,363 prospective screens | | **Screen copy, help text and accessibility notes** | Absent from all prospective screen records. Present only as general rules (GES-UX). | | **Workflow and screen for competitor-derived capabilities** | Workflow GAP for 5,812 of 7,394; screen GAP for 6,504 of 7,394 (consolidated) | | **Accountable human role** | Roles often malformed. For example, "Not advice unless permissioned." appears as a role in the bank journeys. There is no consistent answerable owner per capability. | | **Measured outcome** (time, money, lives, planet, resources, value) | Only the bank catalogue gives a "measured outcome" for each row | | **Unknown outcome, receiving confirmation, batch semantics** | Only in prose (the P29 journeys and bank catalogue rule 8). Absent from every register. | | **Agent path equal to the human path** | Required only from 31 October (authoring addendum 5). Present in the build-specification units; absent from every earlier register. | --- ##### 4. Behaviour and generator rules defined before, and where they were incomplete This compares each part of the new behaviour model (`spec/UNIT-FORMAT.md`) with what the old estate defined. | New-estate part | Where the old estate defined it | How complete it was | |---|---|---| | **Actor** | Journeys: `actor_classes` (person, customer, patient, organisation, regulator, assessor, running agent). Authoring brief: each workflow names person, agent, engineer or regulator. Feature workflows: four fixed actors. GES-ID-03: an agent acts under its own identity, never a borrowed person's. | Actor classes were generic. Agent mandates were named but not modelled per operation. | | **Permissions** | Functions: `permission` strings such as `payments.draft`. Role workflows: "Permissions and expiry" (the best source). Operations: `authority.grant`. Product specifications: role sentences such as "Maker and independent authoriser". | Strings with no scope model. Expiry appears only in the role workflows. | | **Guards** | Functions: four repeated preconditions (registered-device session or published read projection; purpose-bound policy allows; correct source scope; expected version matches). GES-EX-04: decide "act, hold, escalate or refuse" before any effect. GES-SA-02: every hold has scope, reason, deadline, alternative and appeal. | Generic. No named guard with a source and a refusal message per operation. | | **Operations** | `SERVICES.jsonl` (6,121 names such as `app:bank.13-week-forecast.get`, with no schema). Build-specification backend contracts (method, route, request, response, errors, authority, MCP tool). GES-CT-02: one operation kind per definition, each with a state machine, terminal states, receiver and compensation. P29 journeys: command states queued, authorising, dispatched, unknown/reconciling, completed, rejected, cancelled; "a 202 acceptance is not completion". | Schemas exist only in the build-specification units. The operation-kind catalogue (`spec/OPERATIONS-CATALOGUE.json` in the pre-contract tree) holds 27 benchmark entries, not state machines. | | **Rehearsal** | Owner capability "Rehearse before live: every workflow plays out first". GES-AU-01 and AU-02: dry-run for every agent tool. | Principle only. No operation-level rehearsal field. | | **Confirmation and approval** | Authoring brief, "glass box": every consequential action previews its effect, needs explicit authorisation, writes an audit record and is reversible or has a recovery. Bank catalogue JB-F-195 "ask-first rule". GES-EX-09: clinical action, contract signature, credential issue and agent payments above the limit are reserved to a person. 876 prospective screens are of type "confirm". P29: "proposer cannot self-approve". | Good principles. Not bound per operation. | | **Receiving confirmation** | Journeys: a step's receiving system. P29: "settled only from provider evidence". GES-AU-04: "original receiving-party outcomes". | Named. No per-operation contract saying what counts as confirmation. | | **Success** | Role workflows "Done when" (specific and testable). Journeys `result_state` (templated). Functions `output` (654 distinct texts). | Strong in the role workflows only | | **Failure** | Functions: five error codes (permission-denied, validation-failed, version-conflict, dependency-unavailable, deadline-exceeded). P29: `AUTHORITY_DENIED`, `VERSION_CONFLICT`, `VALIDATION_FAILED`, `DEPENDENCY_UNAVAILABLE`, `OUTCOME_UNKNOWN` with a trace ID. Feature workflows: five fixed failure paths. | No separate person message and agent structured message. Generic everywhere except the P29-style journey contracts. | | **Unknown** | Bank catalogue rule 8: "'Unknown' is a first-class outcome … A feature that cannot say 'unknown' is not specified". JB-F-202. P29-J05: "Unknown remains visible until external evidence resolves it; no fabricated success and no automatic duplicate dispatch". | Excellent rule, applied only in finance prose | | **Recovery** | Functions: "retain draft and reconcile outcome before retry; compensation must be separately authorised". Journey contracts: a recovery line per journey. GES-EX-08: only a scoped actor or guardian may reverse a hold. Annex E: E-04 stops a repeated failure signature; E-05 requires exactly one terminal decision. Owner capability "Reversibility Engine". | Mostly one template. No link from an operation to the operation that undoes or compensates it. | | **Evidence** | GES-EV-01: every call, success or refusal leaves a receipt. GES-EV-02: before and after hashes. Journeys `p68.audit_trail`. P29: evidence = actor, organisation, device or session, source mode, policy and object versions, timestamps, decision and receipt, never credentials. | Consistent and usable | | **Idempotency** | Functions: "key bound to actor, organisation, mode, operation and request digest". GES-EX-02: a replay returns the stored result. | Consistent | | **Batch** | P29-J03 only: "One rejected row does not hide partial effects; mutation invalidates prior approvals". The engine catalogue has bulk-refund workflows. | **No general batch rule.** Absent from every register. | | **Output verification** | GES-PR-01: "a parent's pass shall not prove a child; … an HTTP 200 shall not prove an outcome; a receipt's existence shall not prove an outcome". GES-UX-05: toast, DOM change, navigation and HTTP 200 alone fail. Annex E E-05: PASS needs recorded evidence. Journeys: verification by a non-builder replay. | Strong as principle. Not attached to each operation's output. | | **Monitoring** | GES-PR-04: re-measure hourly. Role workflows such as "Monitor job runs and leases". | Estate level only | | **Screen blueprints** | See the list below the table. | No copy keys, help markdown or agent equivalent per screen; competitor floors unresolved; designs pending. | | **Journeys by actor kind** | Build-specification units give workflows for person, agent, engineer and regulator. Addendum 5: at least two live end-user outcome tests per workflow, one by a person and one by an agent under its own identity, reaching the same outcome. Harmony: lease and fence, no silent overwrite. | No OPERATOR journey kind. Regulator journeys appear only in finance and the regulator portal. | Old screen blueprint sources (for the "Screen blueprints" row): - Screen contracts: layout template, five states, roles, action binding. - Screen-state matrix: twelve states × three themes × four source modes × privacy on/off. - The product-specification state table: empty, loading or partial, offline, denied or expired, conflict, failed, awaiting review, completed, revoked, recovery. - GES-UX-03: each screen has its own ID, intention, role, guidance, agent tool route, expected and denied results, benchmarks, hold and recovery, and contest path. - GES-UX-06: enumerated states. - `TEMPLATES.md`: measured shell, ribbon, pill and pane dimensions. - `DESIGN-TEMPLATE.md`: brand rules. - The authoring brief: control kinds (button, field, select, toggle, table, table-action, link, menu, upload, search); every non-navigation control names its operation; a dead-control test. **Generator rules that produced the large registers.** These explain why many old items are templated. - **Prospective register** (`COUNT.json`): - a fixed lexicon maps verbs to canonical actions (get, list, search, show → view); - objects are singularised and synonym-mapped, and stop at the first preposition; - rows with the same family and signature merge; - an Owner row with no verb becomes a "use" journey (350 journeys); - marketing phrases are NOT_AN_OUTCOME; - one step template per canonical action. Its own stated limit: "Signatures come from a lexicon, not a reading by a domain expert". - **Target count** (`TARGET-JOURNEYS.json`): units × 5 variants × 2 actor classes (× 5 surfaces for runs). - **Feature workflows:** one workflow per app, screen, function, help page and record, all with five fixed failure paths. - **Authoring brief** (31 October): "Specific, never generic", "No quotas", typical sizes (1–4 workflows, 1–4 screens, 3–12 controls, 1–5 entities, 2–10 operations, 3–8 tests), British English, at least one negative test per unit. --- ##### 5. Old-estate capabilities and features not yet covered by the new-estate units **What I compared.** The 50 units in `spec/units/` (26 finance, 8 clinic, 8 public-consortium, 8 judiverse). For each unit I compared by meaning: its `capability.requirement` and its operation list. **Old sources used:** - Owner capabilities: `old-repo/handover/scope-freeze/package/REQUIREMENTS.jsonl`, cited by capability name. - Role workflows: `legacy/judeai-edge-control/tests/*_workflows/register.json`, cited by role group. - Bank catalogue: `legacy/edge-judiverse-tech/docs/finance/bank/03-FEATURE-CATALOGUE.md`, cited by JB-F number. - Product inventory: `legacy/edge-judiverse-tech/docs/products/`, cited by old P-code. - Build-specification units: `old-repo/handover/tok-20261030/specification/build-spec-20261031/SPEC.md`. **How to read the entries.** - "Partial" means a new unit touches the area but does not specify the capability named. - Each entry is a candidate for a new unit or for an extension of a unit. Each must be re-researched and re-grounded; none is a requirement because it appears here. - The old SPEC.md units for several of these areas already contain draft workflows, screens, operations and tests. Treat those as research input. ###### 5.1 EDGE Finance The 26 existing units cover: - payments: send money, scheduled payments, bulk payments, request money and split bills, Direct Debits; - accounts and cards: cards, savings pots, open personal account, open business account, close account, connect external accounts; - lending: apply for a loan, lend, solution bank funding; - risk and crime: disputes, transaction monitoring, sanctions and PEP screening, fraud-sharing network, confidence report; - identity and security: security and devices, proximity verification; - agents: agent onboarding, mandate and retirement; - other: crypto wallet, regulator inspection. Not yet covered: 1. **Joint accounts and account types.** Joint accounts, separation and continuation after a death; under-18 accounts with guardian controls; basic and student accounts, with graduation step-down. Sources: JB-F-008 to JB-F-014; bank role group "Joint accounts, attorneys and bereavement". 2. **Attorneys, deputies and third-party mandates.** Register and act under a lasting power of attorney; court-appointed deputy; third-party mandate. Sources: JB-F-186, JB-F-187; same role group. 3. **Bereavement and estate settlement.** Notify a death, verify the executor, date-of-death balance letter, release estate funds. Sources: JB-F-184, JB-F-185; that role group. (`apply-for-loan` mentions bereavement only as a support context.) 4. **Vulnerable customers and support needs.** Support-needs flag, trusted contact, gambling block, spending friction, plain-language mode, human handover, mental capacity concerns. Sources: JB-F-015 to JB-F-021; role group "Vulnerable customers"; Owner capability "Vulnerability, support needs and accessibility". 5. **Complaints and Ombudsman.** Acknowledge, investigate, final response, redress, referral to the Financial Ombudsman, root-cause analysis. Sources: JB-F-179 to JB-F-183; role group "Complaints"; Owner capability "Complaints and DISP". (`disputes` covers transaction disputes, not complaints.) 6. **Statements and documents.** Statements, exports, statement letters, statement import and reconciliation. Sources: JB-F-004, JB-F-149; old product P28; journeys P28-J01 to P28-J04. 7. **Savings products beyond pots.** Easy-access, fixed-term and regular savers, cash ISA and ISA transfers, rate changes and notices, maturity choices, savings interest statements. Sources: JB-F-022 to JB-F-029; role group "Products and pricing". 8. **Overdrafts.** Arranged overdraft, alerts, repeat-use support. Sources: JB-F-076 to JB-F-078; workflow "Request an arranged overdraft". 9. **Arrears and collections.** Detect missed repayments, income and expenditure, forbearance, payment holiday, default notice, breathing-space notices, write-off, debt sale. Source: role group "Arrears and collections" (12 workflows). Partial: `apply-for-loan` has support plans and `lend` has forbearance and default. 10. **Other credit.** Mortgages (decision in principle, application, offer, in-life), car finance, buy now pay later, credit builder, credit file view. Sources: JB-F-081, JB-F-083 to JB-F-086, JB-F-088, JB-F-089. 11. **International payments and foreign exchange.** International send and receive, multi-currency balances, FX transparency, CHAPS, remittance. Sources: JB-F-038, JB-F-056 to JB-F-059; Owner capability "Multi-currency accounts, FX and remittance". (`send-money` is limited to UK payees.) 12. **Cheques and cash.** Cheque imaging and stops, counter cash deposits and withdrawals, cash double-beep receipts. Sources: JB-F-060, JB-F-061, JB-F-108, JB-F-109; role group "Branch and colleague servicing"; Owner capabilities "Cash in and out" and "Cash double-beep receipts". 13. **Open Banking as account provider.** EDGE Finance serving third parties: account information, payment initiation, variable recurring payments, funds confirmation, event notifications, consent dashboard, 90-day refresh, TPP directory check. Sources: JB-F-145 to JB-F-148, JB-F-049 to JB-F-052; Owner capability "Open banking: own AIS, PIS, VRP and consents"; several build-specification units for app:bank. (`connect-external-accounts` covers only the customer's side.) 14. **Merchant services.** Acquiring, payment links and QR codes, tap to pay, subscriptions, refunds, payouts, merchant disputes. Sources: JB-F-055, JB-F-106, JB-F-107; Owner capability "Merchant acquiring, payment links and subscriptions"; build-specification merchant units. 15. **Business finance tools.** Multi-user roles and approval policies beyond batches, invoicing and invoice chasing, expenses and receipts, VAT and Making Tax Digital, accountant access, accounting export and import. Sources: JB-F-093 to JB-F-105; old products P33 and P34; Owner capabilities "Jude Financial Accounts: MTD VAT, CT600, CIS, MTD IT", "Payroll with HMRC RTI, holidays and timesheets" and "Import from Xero, Sage and prior providers". (`bulk-payments` only pays a payroll file.) 16. **Treasury, ledger and regulatory reporting.** Intraday liquidity, liquidity coverage, exposure limits, general ledger, end of day and period close, regulatory returns, FSCS single customer view, expected credit losses. Sources: role groups "Treasury and liquidity" and "Finance and regulatory reporting"; old product P35; Owner capabilities "General ledger, end of day and period close", "Regulatory reporting" and "Safeguarding or FSCS single customer view". 17. **Governance, risk and conduct.** Obligations register, Consumer Duty outcome reviews, senior manager responsibilities, breach notification, operational resilience and important business services, product governance and fair value, financial promotions approval. Sources: role groups "Risk, compliance and governance" and "Products and pricing"; Owner capabilities "Operational resilience front and centre" and "Product governance and pricing". 18. **AML case work beyond monitoring.** Enhanced due diligence, source of wealth and funds, perpetual KYC, trigger-event reviews, production orders, the MLRO annual report. Sources: role group "Anti-money laundering"; JB-F-175; Owner capabilities "Enhanced due diligence and source of wealth" and "Perpetual KYC". Partial: `transaction-monitoring` has decide-report, and the sanctions unit exists. 19. **Fraud operations and anti-scam protections.** Mule-account flagging, funds tracing and recovery, receiving-bank APP notices, coercion lock with safe and danger words, a trusted "Local Stability" network, scam-call detection at the point of payment, proving the caller is the bank, SIM-swap defence. Sources: role group "Fraud"; JB-F-164 to JB-F-174; Owner capabilities "Mule network and funds tracing", "Coercion lock with daily safe and danger words" and "Prove the caller is the bank". Partial: `disputes` covers scam reimbursement claims and `transaction-monitoring` covers warnings. 20. **Insurance.** Quote, compare, underwrite, issue, renew, cancel, claims, and reporting to the regulator. Sources: old product P59; JB-F-124 to JB-F-130; role group "Insurance" (12 workflows); Owner capability "Insurance quote, buy, cancel, claim". 21. **Wealth.** Investment accounts, stocks and shares ISA, pensions and pension finder, appropriateness warnings. Sources: JB-F-111 to JB-F-117. 22. **Money insights and forecasting.** Categorisation, budgets, subscription detection and cancellation, price-rise alerts, net worth, carbon footprint, personal and 13-week business cash forecasts, shortfall warnings, bill negotiation. Sources: JB-F-131 to JB-F-143, JB-F-205 to JB-F-211; old products P74 and P33; Owner capability "13-week cash forecast and JudeFI forecasting". 23. **Account switching.** Switch in and out with the Current Account Switch Service, switch preview. Sources: JB-F-212 to JB-F-214; workflow "Switch account to EDGE Bank". 24. **Customer data rights.** Subject access, data export and Smart Data. Sources: JB-F-151 to JB-F-153; workflow "Request my personal data". Partial: `close-account` exports data at closure. 25. **Staff servicing and case management.** Verify a caller, open and move service cases, temporary freezes, branch appointments, the Banking Protocol, relationship management (covenants, reviews, pricing). Sources: role groups "Branch and colleague servicing" and "Relationship management"; old product P38. 26. **Decision passports and recovery authorisation.** Portable evidence of a decision, lawful-scope recovery with dual authorisation. Sources: old products P39 and P40; JB-F-200, JB-F-201. 27. **Partner fraud intelligence as a service.** Onboard a partner, run checks for it, bill it, end the partnership. Source: role group "Partner fraud intelligence". Partial: `fraud-sharing-network` covers member markers and queries. 28. **Agent banking companion and staff copilots.** Jude as personal banker; case, AML, fraud and treasury copilots. Sources: JB-F-216 to JB-F-222; Owner capabilities "Jude as a personal banker" and "Bank agent manager". Partial: the agent units cover mandates, not assistant behaviour. ###### 5.2 EDGE Clinic The 8 existing units cover: front-door check-in, symptom check and triage, online urgent triage, wellbeing questionnaires, social prescribing referral, medication safety check, wearable signals, and retinal screening. Not yet covered: 1. **Person-held health record.** Claim the record, view results, connect to GP, pharmacy and hospital systems, one verified patient identity, request a copy, correct a fact, emergency access, a whole-of-life record. Sources: old product P41 and journeys P41.J01 to P41.J04; role group "Patient"; Owner capabilities "Person-held lifelong health record", "Connection to GPs, pharmacies and hospitals on UK standards" and "Emergency access to the record". 2. **Carers, proxies and parental consent.** Accept a proxy, register a child, book and consent on someone's behalf, young person's own access. Sources: role group "Carer, parent or guardian with consent"; Owner capability "Proxy, parental and capacity-aware consent". 3. **Appointments and waiting lists.** Book, rebook and cancel; clinic templates; waiting-list management; cancellation slots; honest queue position; interpreters and transport. Sources: role groups "Patient" and "Receptionist and appointment scheduler"; Owner capabilities "Smart GP, clinic and hospital scheduling", "Priority-safe rebooking after cancellation" and "Known queue position and honest wait estimate". Partial: front-door check-in routes to care but does not book. 4. **Repeat medicines and pharmacy work.** Order repeats, change pharmacy, clinical check, dispensing, controlled drugs, adverse reaction reporting, patient group directions. Source: role group "Pharmacist and prescriber". Partial: `medication-safety-check` covers interactions only. 5. **Referrals and cancer pathways.** Suspected-cancer referral, advice and guidance, referral tracking, the 62-day and faster-diagnosis clocks, multidisciplinary team meetings, decision to treat, cancer registration. Sources: role groups "GP and referrer", "Consultant and clinician (cancer detection pathway)" and "Multidisciplinary team coordinator"; old product P45; Owner capabilities "Direct-to-test at first presentation from primary care" and "Cancer wait time measurement and forecast". 6. **Imaging and radiology beyond retinal screening.** DICOM store and viewer, vetting and reporting, nodule measurement, critical findings, chest X-ray, endoscopy and CT triage. Sources: old products P42 and P43; role group "Radiologist and radiographer"; Owner capabilities "Medical image store, DICOM pixels and viewer", "Chest X-ray triage" and "CT nodule measurement and growth". 7. **Laboratory and pathology.** Sample booking and custody, quality control, result authorisation, phoning critical results, histology, digital pathology, home sample collection ("Health Riders"). Sources: old product P46; role group "Pathologist and laboratory scientist"; Owner capabilities "Sample custody from collection to result" and "Rider sample collection for home diagnostics". 8. **Genomics and sequencing.** Ordering, run quality control, pipelines, variant classification and reclassification, genomic consent, pharmacogenomics, the urine-DNA bladder pathway. Sources: old products P47 and P48; role group "Genomic scientist and bioinformatician"; Owner capabilities "Variant interpretation with cited evidence", "Pharmacogenomics at prescribing" and "Bladder cancer urine DNA pathway". 9. **Mental health service and crisis care.** Practitioner caseloads, care plans, remote sessions, crisis alerts, risk formulation, suicide prevention, safeguarding. Sources: role groups "Mental health practitioner" and "Crisis, safeguarding and suicide prevention lead"; Owner capabilities "Suicide prevention pathway", "Crisis safety controls" and "Trusted contact alert". Partial: `wellbeing-questionnaires` escalates risk. 10. **Clinical trials and research use.** Trial search and eligibility, recruitment, consent, adverse events, publishing results. Sources: old product P44; role group "Researcher and trial coordinator"; Owner capability "UK clinical trial search and eligibility". 11. **Clinical safety and medical-device governance.** Hazard log, safety case, release review, device classification, post-market surveillance, field safety notices. Source: role group "Clinical safety officer"; Owner capabilities "Medical device decision and post-market surveillance" and "Model registry, cards and drift monitoring". 12. **Information governance.** Data Security and Protection Toolkit, national data opt-out, dataset submissions, access audit, records retention. Source: role group "Data manager and information governance". 13. **Clinic regulators' evidence access.** CQC, MHRA, NHS England, ICO and HRA. Sources: role group "Regulators and auditors"; Owner capability "Regulator and clinical safety portal". Note: only finance has a regulator-inspection unit. 14. **Clinic operations.** Capacity, rota, incidents, inspection preparation, nurse and healthcare-assistant work (observations, samples, vaccinations). Sources: role groups "Clinic manager and operations" and "Nurse and healthcare assistant". 15. **Health insurer link with a clinical firewall.** Pre-authorisation, claims, prevention rewards, a firewall between clinical data and insurance decisions. Sources: role group "Health insurer and payer"; Owner capabilities "Health insurer link for claims and pre-authorisation" and "Firewall between clinical data and insurance decisions". 16. **Forecasting and twins.** Personal health forecast, patient digital twin, retinal progression forecast, a ledger of disease detected and reversed. Source: Owner capabilities of those names. ###### 5.3 EDGE Public Consortium The 8 existing units cover: tender and award, contract management, supplier vetting, procurement intelligence, logistics tracking, cyber posture, security incident response, and research consortium. Not yet covered: 1. **Local government decision-making.** Councillor and cabinet decisions, interests registers, scrutiny and call-in, audit committee, monitoring officer, democratic services, the Section 151 finance role (reserves, savings ledger, Section 114). Source: `legacy/judeai-edge-control/tests/government_workflows/register.json`, role groups for councillors, leaders, chief executives, Section 151 officers, monitoring officers, internal audit, audit committees, scrutiny and democratic services (about 160 workflows). Owner capability: "Council deficit and income recovery". 2. **Council resident services.** Housing and homelessness, benefits, council tax and business rates, planning and building control, licensing and environmental health, registration of births, deaths and marriages, electoral services, freedom of information and complaints, social care assessments and commissioning. Source: the government register's service role groups and its "Resident" group (27 workflows). Owner capability: "Social care and supported living". 3. **Central government and Parliament.** Ministerial submissions, business cases, grant schemes, spend controls, parliamentary questions, MP casework, party and campaign finance returns. Source: the central government, MP and political-party role groups. 4. **Emergency planning and resilience forums.** Source: the government register's emergency-planning role group. 5. **Defence and secure operations.** Single-source contract pricing and reports, security aspects letters, facility security clearance, classified information handling, export licences, defence security incidents, defence recruitment and supply. Sources: role group "Defence official …" (10 workflows); Owner capabilities "Palantir replacement for MoD" and "Defence logistics, recruitment and supply". Partial: `logistics-tracking` covers consignment custody. 6. **Personnel vetting and clearances.** Vet a person or supplier staff member, grant a clearance with expiry, director passkey approvals. Sources: consortium role group "Security officer, CISO and vetting officer"; Owner capabilities "Capita replacement: recruitment and vetting" and "Conveyor-belt vetting with SMF approvals". Partial: `supplier-vetting` vets organisations, not individuals. 7. **Procurement steps not yet specified.** Pipeline notices, below-threshold direct awards, framework and dynamic-market awards, exclusion-ground declarations, payment-compliance publication, security-classified procurement, contract termination notices. Sources: consortium role groups "Procurement officer and public buyer" and "Supplier and bidder"; government group "Procurement and commercial". 8. **Education.** University membership and enrolment, modules, teaching, marking and moderation, transcripts, verifiable credentials and their revocation, evergreen learner record, safeguarding young learners, academic integrity. Sources: consortium role groups for learners, registrars and teaching leads; Owner capabilities "Global online university and academic source of truth", "Verifiable credentials and certificates", "Evergreen learner record replacing exams" and "Safeguarding for young learners". 9. **Peer review and verified knowledge.** Reviewer credentials, review assignments, editor decisions, appeals, retractions, reviewer payment. Sources: role groups "Verified academic reviewer" and "Review editor and coordinator"; Owner capability "Peer review and reviewer credentials". 10. **Research life cycle beyond the consortium unit.** Data management plans, ethics approval, preregistration, dataset deposit with persistent identifiers, ORCID, REF output tracking, supervision. Sources: role groups for doctoral researchers and supervisors; old products P51 and P07; Owner capability "ORCID and scholarly records". Partial: `research-consortium` covers proposals, partners, funding, milestones and outputs. 11. **Grants and funding radar.** Find calls, cost an application, set up the award, spend control, funder reports. Sources: role group "Finance, grants and funding officer"; Owner capabilities "Grants and competitions directory" and "Smart grant portals". 12. **Intellectual property and licensing.** Invention disclosure, ownership, open licences, licences to third parties, takedown requests, royalties. Sources: role group "Intellectual property and licensing officer"; old product P07. 13. **Federated data stewardship.** Dataset registration, purpose views, federated queries with disclosure control, data-sharing agreements, provenance. Sources: role group "Data steward and data holder"; old products P52 and P53; Owner capability "Federated analysis across hospitals". 14. **Libraries and the knowledge graph.** Cataloguing, schemas, merging entities, controlled vocabularies. Sources: role group "Librarian and curator"; old product P03 (Tree of Knowledge); Owner capability "Research Commons". 15. **Employment within the platform.** Vacancies, offers, contracts, right-to-work checks, pay, leaving. Sources: role group "Employer and employee"; Owner capabilities "Living EDGE ID record replacing the CV" and "Search Desk job market". 16. **Cyber governance capabilities beyond posture and incidents.** Decoy "honey" graph, adversarial testing, prompt-injection immunity, SIEM/SOAR-grade correlation, automatic subject-access ledger, lawful disclosure pipeline, software supply-chain assurance, backup restore drills, a security officer console, an EU AI Act conformity file. Sources: Owner capabilities under "Cyber Governance" (48 rows); old capabilities P107 and others in `old-repo/benchmarks/frontier/CAPABILITIES-P92PLUS.json`. 17. **Public-sector identity and directories.** GOV.UK One Login, a directory of councils, departments and decision makers, unclaimed organisation workspaces, a company knowledge graph that finds fake companies. Source: Owner capabilities under "Public EDGE". 18. **Community and civic features.** Public alerts, reports and disputes; community watch; coercion lock; smart-city simulation; waste-materials graph; simulated regulators. Source: Owner capabilities under "Public EDGE". Several of these are speculative and need fresh research before any specification. 19. **Regulator portals for this application.** Source: Owner capability "Regulator portal with mark scheme". Note: only finance has a regulator-inspection unit. 20. **Carbon, cost and compute stewardship.** Energy and carbon per job and per outcome, budgets, low-carbon scheduling. Sources: role groups "Carbon and cost steward" and "Compute operator and scheduler"; old products P58 and P61. This could equally belong to Judiverse. ###### 5.4 Judiverse The 8 existing units cover: Jude assistant, agent builder, procedure compilation, tasks and workflows, documents and forms, calendar and scheduling, mail and agent mailboxes, and JudeBox music. Not yet covered: 1. **Messages and chat.** One-to-one and group conversations, attachments, corrections, blocking, export. Sources: old product P22 and journeys P22.J01 to P22.J04; Owner capability "Disposable group messaging and pitch sharing". 2. **Calls and meetings.** Calls and video meetings, lobby and guests, screen sharing, recording with consent, transcription and notes, live captions and translation, reconnection. Sources: old products P23 and P79; the Teams parity scope (`.../teams-full-feature-scope-20261022/`, 167 workflows); Owner capabilities "Calls with Teams and WhatsApp parity", "Notetaking better than Fireflies" and "Live captions and translation in calls". 3. **Agent Channel and team channels.** Teams, channels, moderation, adding an agent under mandate, retention and export. Sources: old product P14; consortium role group "Channel moderator and agent governor"; Owner capability "Agent Channel as executive group chat". 4. **Baton hand-over.** Reproducible hand-over of work between people and agents. Sources: old product P15; Owner capability "Baton handover from any moment". Partial: `tasks-and-workflows` has hand-off notes. 5. **Office editors beyond documents.** Sheets (formulas, imports, forecasts), Slides (decks, presenting without leaking notes), Markdowns (a Notion-like knowledge base with database views), real-time co-editing with presence. Sources: old products P17, P19 and P20; Owner capabilities "Documents, Sheets and Decks at Microsoft parity" and "Markdowns as a sovereign Notion"; `PARITY-DESIGN.jsonl` (Markdowns, 36 features). 6. **Drive and file storage.** Storage, sharing, sensitivity labels, versions, sensitive-file triage, migration from SharePoint. Sources: old product P80 (353 operations in the engine catalogue); Owner capabilities "Drive at SharePoint parity" and "Sensitive-file risk triage"; `legacy/edge-judiverse-tech/docs/drive/`. 7. **E-signature and contracts.** Templates, negotiation, signing with EDGE ID, NDAs, signing logs. Sources: Owner capability "Secure Contracts and e-signature"; consortium role group "Legal and contracts officer". Partial: `documents-and-forms` has approvals. 8. **Mail features beyond the existing unit.** External delivery and protocol compatibility (IMAP, SMTP, CalDAV, CardDAV), migration from Microsoft, eDiscovery and legal hold, shared mailboxes, scheduled send and recall, outbound hold, aliases. Source: Owner capabilities under "Judiverse Office". Partial: `mail-and-agent-mailboxes` has undo send and delegation. Check each item against that unit. 9. **Contacts, people directory and search.** Sources: old products P75, P76 and P77; Owner capabilities "Governed contact directory" and "Intent-based file search". 10. **Notifications, account, organisation administration, help and learning.** Sources: old products P65, P66, P67 and P69; Owner capabilities "Consumer account, deletion and export", "Permissions as plain toggles" and "Interactive tutorial and guidance". 11. **Personal record, vault and secrets.** Sources: old products P05 and P06; Owner capability "Credential use without exposure". 12. **Connections, integrations, developer portal and MCP foundry.** Connect accounts, bounded capability grants, rotate secrets, public API and SDK, custom MCP servers and plugins, sharing an MCP server between organisations. Sources: old products P13, P70 and P71; Owner capabilities "Public developer API and SDK", "Custom MCP and plugins", "MCP foundry" and "Every capability a standalone MCP, shareable between organisations". 13. **Agent lab engineering.** Git hosting, change requests and review, pipelines and isolated runners, issues and releases, package registry, secret scanning, deploy and rollback, the Terminal. Sources: Owner capabilities under "intenJude Lab" (57 rows); old products P09 and P25; consortium role group "Open-source maintainer and external contributor". Partial: `agent-builder` and `procedure-compilation` cover agents and procedures, not code hosting. 14. **Model training lab.** Datasets, bounded training recipes, evaluation, promotion and rollback, regression evaluation per model, benchmark harness runner. Sources: old products P11 and P12; Owner capabilities "Model training lab" and "Regression evaluation per model and release"; P92 in `CAPABILITIES-P92PLUS.json`. 15. **Compute scheduling.** Placing workloads, pre-emption, quotas, energy per job. Sources: old products P10 and P61; Owner capability "Intelligent compute scheduler". 16. **Digital twins, online testing and simulation sandboxes.** Sources: old products P72, P73 and P78; Owner capabilities "Digital twins as benchmarked flagships" and "Sandbox training ground". 17. **Assistant capabilities beyond ask-and-act.** Deep research, voice (read aloud, voice notes, voice chat), image and video generation, browser and device control, persistent memory and temporary chat, scheduled agent tasks, a live task panel with pause and steer, file handling across types. Sources: Owner capabilities under "JudeAI" (76 rows); `legacy/jude-mvp/jude-live/file-capabilities.json` (1,000 file capabilities). 18. **Plans, credits and billing for the assistant and the estate.** Sources: Owner capabilities "Guest allowance, credits and plans", "Stripe and in-app billing" and "Licence, usage and subscription billing". 19. **Age assurance and child safety.** Source: Owner capability "Age assurance and child safety". 20. **Creative studio and wider JudeBox.** Image studio, video editing, audio workstation, podcasts and audiobooks, offline downloads, recommendations, social layer. Sources: old product P60; Owner capabilities under "JudeBox" (29 rows). Partial: `judebox-music` covers upload, lyrics, playback, royalties and takedown. 21. **Canvas and design engine.** Template generation, deterministic rendering, specialist viewers. Sources: old product P02; `PARITY-DESIGN.jsonl` (JudeCanvas 48, Design engine 29, Template generator 12). 22. **Science workbenches.** Materials, chip design, carbon and biochar, water. Sources: old products P58, P62 and P88 to P91; `legacy/x-judeai-edge-tok-kol-ndc-ml-x/workbenches/`; Owner capabilities "Materials and chip design" and "Jude Water". These are research areas; they need fresh research before any specification. ###### 5.5 Cross-cutting (EDGE ID and shared foundations), not owned by any unit yet 1. **Organisation onboarding and tenancy.** - Every UK organisation pre-represented and claimable. - Organisation checks (KYB) for every organisation type, including charities, trusts and public bodies; `open-business-account` covers companies only. - Domain control, group structures and switching between them. - Tenant isolation, organisation exit and export. - Invite and import all staff, enterprise single sign-on and SCIM provisioning. Source: the 73 Owner capabilities under "Organisation onboarding". 2. **Sign-in, recovery and the "one door" for every application.** Sources: `legacy/edge-judiverse-tech/docs/estate/ESTATE-WHOLE-OF-LIFE-JOURNEYS.md` EST-J-001 onwards, including under-18 sign-up with a guardian (EST-J-006) and vulnerable people with a named helper (EST-J-007). `finance/security-and-devices` covers the finance side only. 3. **Leaver switch across every application.** Source: Owner capability of that name. 4. **Regulator evidence portal for every application, not finance alone.** Sources: Owner capability "TRL 9 and CRL 9 guarantee portal for regulators"; GES-AU-05. 5. **Estate-wide principles that the 121 "intended-estate" Owner rows describe as capabilities:** complete mediation, causal trace, circuit breaker, intent lineage, kernel-level action capture. These may belong in shared foundations or tooling rather than in application units. --- ##### 6. What is superseded 1. **Whole old repositories.** From 27 October 2026, `legacy/governed-intelligence-judeai-edgeid-pre-contract/LEGACY.md` declared seven repositories superseded. Their content in this corpus is under `legacy/x-judeai-edge-tok-kol-ndc-ml-x/`, `legacy/judeai-edgeid-cloud/`, `legacy/edge-judiverse-tech/`, `legacy/judeai-edge-control/`, `legacy/peak-judeai-behavioural-harness-baseline-full-application/` and mirrors. Everything in them (P01 to P81 specifications, journey contracts, role-workflow registers, the 15 October master plan) is historical research. 2. **Old contracts.** Versions 2.3.1 to 2.9.11 are listed as superseded in `legacy/superseded-contracts-20261030/README.md` (30 October 2026). Later versions referenced in the old tree, up to 2.9.26, belong to the old estate too, and the new estate does not load any of them. 3. **GES-1.0** is superseded by GES-1.1 (stated in GES-1.1). Both are old-estate documents. Their principles are summarised in section 4 as research. 4. **Counts superseded within the old estate.** - The multiplicative target (72,550) and the evidence-matrix count (5,047 journeys of 23 October, marked "HISTORICAL" in `TARGET-JOURNEYS.json`) were followed by the prospective register, which says "No earlier journey or screen count is inherited". - The consolidated (7,394), named (39,130) and build-grain (29,582) counts are different grains. The last handover says they "are not interchangeable". - The 01:28 UTC `SPEC.md` snapshot (1,347 units) was overtaken by the 01:48 UTC `COVERAGE.json` (1,828 units). 5. **Branch and file versions.** - `journeys-prospective/VERIFICATION-v8-bdee7fe-stale.json` is self-labelled stale. - `ontology-20261030/stage1/` is kept as lineage behind `stage1-v2/`. - `.../packs/build-list/archive/BUILD-STEPS-SUPERSEDED-20261022.md` is self-labelled superseded. - Hash-suffixed files (for example `README.596a9276.md`, `register.490f9012.json`) are branch versions of the same file. 6. **Old application structure.** The eight old families (bank, health, consortium, cyber, public, office, judeai, lab), the station names and the P-code capability numbering are replaced by the four new applications and the unit-per-capability layout in `spec/units/`. 7. **Old specification formats.** - The authoring-brief chain (source → capability → outcome → feature → workflow → screen → controls → data model → backend contract → acceptance test, with human and agent paths) is the closest predecessor of the new unit format. - The new format (`spec/UNIT-FORMAT.md`) adds per-operation guards, rehearsal, approval, receiving confirmation, unknown, batch, output verification and monitoring, plus nine screen states with copy, accessibility, help and agent equivalents. - The old generic templates (one recovery text, two failure paths, two acceptance texts) should not be reused. 8. **The ontology-first proposal** (`old-repo/handover/audit-20261031/NEW-REPO-PROPOSAL.md`, 31 October 2026) proposed `tools/admit.py`, `amend.py`, `build.py`, `gate.py` and a generated output folder. The new estate's `tools/` and `generated/` folders follow the same shape. That is an observation of resemblance, not a claim that the proposal was adopted as authority. --- ##### 7. Notes for whoever specifies next - **Strongest reusable old material**, for research only: - the four role-workflow registers (specific triggers, permissions with expiry, regulations and "done when" tests); - the bank feature catalogue (JB-F rows with law, data and measured outcome); - the P-series journey contracts (actions, exceptions, recovery and evidence per journey); - the 780 Owner capability rows (outcome, authority needed, uncertainty); - the build-specification units for detailed draft operations and Given/When/Then tests. - **Weakest material.** Anything whose fields are templated (section 3.2). Their counts measure generation, not specification. - **Old status words are not evidence.** "live", "EXECUTED" and "SPECIFIED" in old registers are self-reported and historical. Re-establish every claim from new evidence. - **Before reusing anything, re-check personal data.** Several old records name real people's work identities and contain machine paths. None is copied here. ### research/legacy/COMPETITORS-BENCHMARKS-READINESS.md `research/legacy/COMPETITORS-BENCHMARKS-READINESS.md` · sha256 `971658fccf85aa2e7322ef3b88cfb1d233deed9c7c88049ce0779be058d56be8` #### Legacy competitors, benchmarks and readiness (TRL, CRL): research reading for JudeAI | EDGE ID **What this is.** A reading of the organisation's earlier work (the old estate) for three catalogue groups: documents tagged `competitors-parity`, `benchmarks-baselines` or `readiness-trl-crl`. It records which competitors were researched, which benchmarks and baselines were defined, scored or run, and how readiness (TRL, CRL and the old estate's own scales) was defined and to be proven, so that none of that work is wasted or repeated for the new estate's four applications: EDGE Finance, EDGE Clinic, EDGE Public Consortium and Judiverse. **What this is not.** It is not authority. Nothing here is a rule, requirement, approval, readiness level or count for the new estate. Every figure is historical and is quoted with the corpus-relative file it was recorded in and that file's date. Old-estate rule numbers, agent identities, contract fingerprints and approvals are deliberately not carried over. Anything worth keeping should be admitted into the new estate as research, with new identities. **Markers used** (as in `new-estate/research/README.md`): figures that an old document says it read from a regulator, publisher or leaderboard are **OBSERVATION** as reported by that document (not re-checked by this reading); figures the old estate produced about itself are **CLAIM**; targets are **PREDICTION** or plan; anything looked for and not found is **UNKNOWN**. Where a table mixes these, the column "what was recorded" says which. **Paths and oids.** Every path is relative to the corpus root: `old-repo/...` is the old canonical repository; `legacy/...` is the set of earlier repositories and snapshots. Blob oids are given in full. **Sensitive data.** Several source files contain host names, server addresses, personal or organisational email addresses, remote-shell commands and absolute machine paths. Where a finding came from such a file, the finding is summarised and the detail is marked "contains sensitive data; not reproduced". Vendor web addresses are not copied either; sources are named by publisher. ##### Acronyms used | Acronym | Meaning | |---|---| | TRL | Technology Readiness Level, a 1 to 9 scale of technical maturity (UKRI, NASA and EU Horizon 2020 wordings were used) | | CRL | Commercial Readiness Level, a 1 to 9 scale of commercial maturity (BINDT/RCNDE wording was used; other CRL scales exist) | | AEC | The old estate's own combined level: "AEC n = TRL n + CRL n, done by agents" (what the letters stand for is not stated in any in-scope file) | | DECFAW 9 | The old estate's internal acceptance target; the expansion of the acronym is not stated in any in-scope file | | UKRI, STFC | UK Research and Innovation; Science and Technology Facilities Council | | NASA NPR | NASA Procedural Requirements | | BINDT, RCNDE | British Institute of Non-Destructive Testing; UK Research Centre in Non-Destructive Evaluation | | ARENA CRI | Australian Renewable Energy Agency Commercial Readiness Index (a 1 to 6 scale) | | WCAG | Web Content Accessibility Guidelines | | OWASP ASVS | Open Worldwide Application Security Project, Application Security Verification Standard | | CAF | NCSC (National Cyber Security Centre) Cyber Assessment Framework | | DCB0129 / DCB0160 | NHS clinical risk management standards for manufacturers and for deployers of health IT | | DTAC, DSPT | NHS Digital Technology Assessment Criteria; Data Security and Protection Toolkit | | FHIR | Fast Healthcare Interoperability Resources (HL7 standard) | | FCA, PRA, PSR, FOS, CMA | Financial Conduct Authority; Prudential Regulation Authority; Payment Systems Regulator; Financial Ombudsman Service; Competition and Markets Authority | | APP | Authorised push payment (fraud) | | AML, KYC, KYB | Anti-money laundering; know your customer; know your business | | MCP | Model Context Protocol (agent tool interface) | | SLO, RPO, RTO | Service level objective; recovery point objective; recovery time objective | | DORA | Either the DevOps Research and Assessment delivery metrics or the EU Digital Operational Resilience Act; each use says which | | pass^k | Share of tasks passed on all k repeated trials | | AUROC, AUPRC | Area under the ROC curve; area under the precision-recall curve | | UK AISI | UK AI Security Institute | | P26, P41 and similar | Old-estate capability codes (for example P26 bank, P41 health); quoted only to locate sources | --- ##### 0. Method and coverage (stated honestly) **Source.** `catalogue.jsonl` in the corpus, rows with `"generated": false` whose categories include at least one of the three in-scope tags. | Measure | Count | |---|---:| | In-scope catalogue rows | 3,369 (1,916 under `legacy/`, 1,453 under `old-repo/`) | | Distinct blob oids | 3,268 | | Distinct documents after de-duplication | 3,182 | | of which structured registers (JSON, JSONL, CSV, XLSX) | 2,597 | | of which prose, HTML and other text | 585 | | Distinct oids tagged `benchmarks-baselines` / `readiness-trl-crl` / `competitors-parity` (a document can carry several tags) | 1,975 / 910 / 437 | **De-duplication.** Rows with the same blob oid were counted once (101 duplicate rows). Files whose names differ only by an eight-character branch-variant suffix (for example `STANDARD.bafb4072.md` beside `STANDARD.md`) were treated as one family and diffed; where the variants differ in substance this report says which one it follows. The 70 in-scope files under `legacy/governed-intelligence-judeai-edgeid/` and the 21 under `legacy/governed-intelligence-judeai-edgeid-pre-contract/` are byte-identical to `old-repo/` files except for three. The 25 copies of `old-repo/handover/ESTATE-LEVEL-READINESS*.md` differ only in their plan-version and fingerprint line. **Depth of reading.** - **Read in full: 55 distinct documents.** The benchmark standard (both variants), score-record schema, portal specification, register counts (all variants) and memory note; the frontier run result, completion state and claims boundary; the benchmark-result files; the TRL 9 / CRL 9 execution gate; Annex P (both variants); the route-to-9 order; the DECFAW adoption and proof architecture; the TRL 8 gate incident; every readiness-statement variant; the AEC standard; the predeclared TRL 9 / CRL 9 receipts; the 15 October 2026 CRL 9 count; both parity state handovers; the catalogue monitoring note; the Palantir/Incogni floors lane and its parity count; the horizon parity summary; the agent-governance, compute, executable and frontier benchmark research documents; the UK readiness and benchmark protocol; the Palantir alternative and Palantir procurement register notes; the benchmark, KYC, typology, card-fraud and model register files; the TRL/CRL criteria and the four computed runs; the hard-task, preregistration, readiness-uplift, capability-gap, measurement-contract, research-evaluation and buyer-benchmark notes; and five owner-source files. - **Read in part (every heading, plus every section or table that bears on competitors, benchmarks or readiness): about 70 documents.** These include the long registers (financial intelligence 161 KB, intenJude 129 KB, risk and resilience 75 KB, app generator and office 60 KB, bank competitor benchmarks 60 KB, bank architecture 53 KB), the eight horizon parity family files, the scope-to-TRL-9 file (113 KB), the competitor catalogue README, the screen-flow and vendor-documentation parity matrices, the Office parity file, the four research notes of 13 October 2026, the platform inventories, the JudePay, ChatGPT, music, mail, Teams and GitLab parity files, the mail CRL 9 plan and proof, the operational readiness decree, the value map and the clinical adoption research. - **Analysed programmatically: about 2,530 structured files**, each parsed or counted by script: 1,245 published-benchmark score records; three benchmark registers (1,704, 1,704 and 1,579 rows); the benchmark graph (1,704 rows) and parity obligations (575 rows); 788 clinic TRL 9 / CRL 9 proof contracts; 14 pack benchmark catalogues; 268 outcome-evidence files (counted by folder and file type, with sample evidence matrices and accessibility summaries opened); 9 outcomes registers; 19 programme-stream registers; two level registers (4,038 rows each); the frontier gap file and five discovery files; the TRL/CRL criteria and computed runs; 23 competitor-catalogue JSON files; the parity JSON and JSONL; eight horizon parity JSONL files; and 13 registry JSON files. - **Skimmed or skipped, with reasons (about 520 documents):** 287 files of `legacy/peak-judeai-behavioural-harness-baseline-full-application/` (in scope mainly because the repository name contains "baseline"; a keyword scan found 261 of them contain no competitor, benchmark or readiness term; the 26 that do were skimmed and the value map read in its relevant parts); 100 `fiction_excerpt.rst` files (a test corpus for a measurement kit, not research); 60 GitHub pull-request and workflow-run metadata files (repository metadata only); 16 Teams coverage HTML source captures and 14 Mobbin query logs (counted only); captured competitor web pages (Monzo and Starling HTML; titles only); app and web interface pages tagged by keyword; and the team CV appendices (personal data; not reproduced). --- ##### 1. Inventory by topic Each entry gives the corpus-relative path and the full blob oid. Groups of many similar files give the count and one representative. ###### 1.1 Benchmark standard, registers and scores (old repository, 26 to 30 October 2026) - `old-repo/handover/benchmark-standard/STANDARD.md` — oid `b00ad18e4d2a3b5b9eb98a0aecf4340fabd24b65` - `old-repo/handover/benchmark-standard/STANDARD.bafb4072.md` — oid `bafb40723fc8b724faaac267deb49d025f05bfce` - `old-repo/handover/benchmark-standard/SCORE-RECORD-SCHEMA.md` — oid `b6514f69e9b3f13b96e93c27d430e91bd75f1288` - `old-repo/handover/benchmark-standard/PORTAL-SPEC.md` — oid `52d7c8ac37d92cd578f49356e3fe006a17ddec16` - `old-repo/handover/benchmark-standard/COUNT.json` — oid `a470b2b9e8615829ddd3ecf211d9ac1a24629073` - `old-repo/handover/benchmark-standard/COUNT.265e021f.json` — oid `265e021feb5abc919204b8125ad5751d4e294bea` - `old-repo/handover/benchmark-standard/BENCHMARK-REGISTER.jsonl` — oid `b64a942f1ec83a01a97ce2745f0aff3005992aa8` - `old-repo/handover/benchmark-standard/BENCHMARK-REGISTER.9bf50795.jsonl` — oid `9bf50795c42bff28126c5825d9226bc299fcee5c` - `old-repo/handover/benchmark-standard/BENCHMARK-REGISTER.855f0f80.jsonl` — oid `855f0f80d752e5d622bc8171e8edeeab1ab534b1` - `old-repo/handover/benchmark-standard/UNITS.jsonl` — oid `b616a57447c7a0b8c8f12a736535e6c14ad19843` - `old-repo/handover/benchmark-standard/sources-BENCHMARKS-20261026.json` — oid `fdb11ff73dfaa572014a9b3c74955aac0ca7b42d` - `old-repo/handover/benchmark-standard/portal/board.html` — oid `9aad1bfc830bc4dbc2efb3069b73aed9178324bf` - 1,245 score records, `old-repo/handover/benchmark-standard/scores/*.json`; representative `old-repo/handover/benchmark-standard/scores/about-deliveroo-plus-diamond.json` — oid `99ef375c379933aca9707972f69fb3bb22a72121` - `old-repo/handover/HANDOVER-COORDINATOR-20261030/memory/benchmark-standard-20261030.md` — oid `abbe23ea802176f15f49c59dd2de9be2de50fc8e` - `old-repo/handover/HANDOVER-COORDINATOR-20261030/inputs/baseline-family-coverage.json` — oid `5bfd2036bcb396984b8c1703a191d039304bdfa2` - `old-repo/handover/journeys-prospective/BENCHMARK-GRAPH.jsonl` — oid `5741377d4e51d65d3ea4d725dd1861797c4b5702` - `old-repo/handover/journeys-prospective/PARITY-OBLIGATIONS.jsonl` — oid `a5da9e44f36dc80b823b4c353c4e39abe1d133db` - `old-repo/plan/sources/owner-benchmark-standard-20261030.json` — oid `f204d2783f060a2189134164fe89497d013226ad` - `old-repo/plan/sources/owner-benchmark-taxonomy-20261030.json` — oid `1ec7bed8a9464362957e331172305bdddeb5b8ee` - `old-repo/plan/sources/owner-competitor-grade-programme-20261030.json` — oid `d3a681b8de4e49ba54afdee0fa77b40273a7be2c` ###### 1.2 Frontier benchmark discovery and runs (old repository, 29 to 30 October 2026) - `old-repo/benchmarks/frontier/GAP.json` — oid `bf0582ee7f4e40e08fb446e6d1dbe92738bc895a` - `old-repo/benchmarks/frontier/discover/autonomous-ai.json` — oid `deeaa2242f4b69d27bb1ffff35af7b8df221496b` - `old-repo/benchmarks/frontier/discover/engineering.json` — oid `f94acfcf38552f27a1f726533ce10ab6726bfff0` - `old-repo/benchmarks/frontier/discover/fintech.json` — oid `23a5fcb746778600fe275e453d2d4a53f2b43127` - `old-repo/benchmarks/frontier/discover/healthtech.json` — oid `eb7366f064e3033f512351ba25eb57a4a060743e` - `old-repo/benchmarks/frontier/discover/regtech.json` — oid `942adbffe65a5bf87fc151ae526525ccceb93bfa` - `old-repo/benchmarks/frontier/CAPABILITIES-P92PLUS.json` — oid `4c9f1a445976cd21f086f3ed93bf587f39bb4233` - `old-repo/benchmarks/frontier/harness/CONTROLS-1.1.0.json` — oid `7c4b4e39dbd43623a0d849c338c20846da401b75` - `old-repo/benchmarks/frontier/data/frontier-definitions-v1.jsonl` — oid `c2c6390b05ec53b8923a4102145ca8cd5874e265` - `old-repo/benchmarks/frontier/BUILD-LOG.jsonl` — oid `785d3e0fee074040a3a7b069e11b43e130cf11de` - `old-repo/benchmarks/frontier/inputs/relaxai-measurement-20261029.json` — oid `5ec97cfbe7e776faf266d9c29f3b47aaf2741140` - `old-repo/benchmarks/frontier/codex-20261030/RESULT.md` — oid `605d79548103accbfcddcd90a543f5ac4daababe` - `old-repo/benchmarks/frontier/codex-20261030/COMPLETION-STATE.json` — oid `f50debde9d311ec21506bf36b91608734e30bcd3` - `old-repo/benchmarks/frontier/codex-20261030/frontier/GAP.json` — oid `a2a153f2324049b4f50c705456910967370174f6` - 17 further run-receipt and manifest files under `old-repo/benchmarks/frontier/codex-20261030/` (counted; the recovery read-back contains process and machine detail; contains sensitive data; not reproduced) - `old-repo/benchmarks/BENCHMARKS-RESULT.json` — oid `539bc3032d4eb15565685cdbf3189f7aefa174d3` - `old-repo/registers/benchmarks.json` — oid `5c317e6973280245e85e7d7660608d183e26feb8` ###### 1.3 Earlier benchmark registers and research (legacy repositories, mainly 13 October 2026) - `legacy/edge-judiverse-tech/docs/assurance/BENCHMARK-REGISTER.md` — oid `2ad820349f562b95e1901aa9cb94ad72ffcfa6fa` - `legacy/edge-judiverse-tech/docs/assurance/COMPUTE-AND-AGENT-BENCHMARKS.md` — oid `e8417d2652c8aeec31c6e217f11b9314b11ba1d8` - `legacy/edge-judiverse-tech/docs/assurance/EXECUTABLE-BENCHMARKS.md` — oid `6fcbca9251e42f861602d68849ecfde83072b54f` - `legacy/edge-judiverse-tech/docs/assurance/FRONTIER-BENCHMARK-RESEARCH-2026-10-13.md` — oid `64f119e3438afd6a94ad3b0d031d182dbec166ff` - `legacy/edge-judiverse-tech/docs/assurance/clinical-evaluation-register.md` — oid `3702b915f7ad9507669cec37d1b9622e82cd8634` - `legacy/edge-judiverse-tech/docs/engineering/INTENJUDE-BENCHMARK-REGISTER.md` — oid `a252ef6865c16bdd6dd9c474f3a8e62c34a52750` - `legacy/edge-judiverse-tech/docs/benchmarks/RISK-AND-RESILIENCE-BENCHMARKS.md` — oid `fe1bb472b4377b157236564dc6d99bc99afa0116` - `legacy/edge-judiverse-tech/docs/benchmarks/APP-GENERATOR-AND-OFFICE-BENCHMARKS.md` — oid `6f212ebaf90afdfa6f846da9beea79eba6a24bdf` - `legacy/edge-judiverse-tech/docs/regulation/FINANCIAL-INTELLIGENCE-BENCHMARKS.md` — oid `ec1706661a4c6a61964712330f08d23a49e7fc23` - `legacy/edge-judiverse-tech/docs/research/2026-10-13/01-agentic-coding-benchmarks.md` — oid `b390199e63ca4651caaad004c21f2c98bcd79f18` - `legacy/edge-judiverse-tech/docs/research/2026-10-13/04-reliability-safety-ops-ux-uk.md` — oid `873246b6bc6501f141a20ae18f41c38e43eb1d67` - `legacy/edge-judiverse-tech/docs/research/2026-10-13/08-financial-intelligence-benchmarks.md` — oid `1b620b47cbddfdc4b5b0f807f99c52fedbb363e6` - `legacy/edge-judiverse-tech/docs/research/risk-resilience/R4-RISK-AND-RESILIENCE-BENCHMARKS.md` — oid `4efe64649151b89104f84ca5aea57a75fd76d2b6` - `legacy/edge-judiverse-tech/docs/dossiers/frontier-ai/11-methodology-benchmarks.md` — oid `25697487173eecab137504a47c6278fae8022157` - `legacy/edge-judiverse-tech/docs/bank/BANK-ARCHITECTURE-API-AND-BENCHMARKS.md` — oid `fab0f8932860a3e4787bec6bab377a8ba7a8d630` - `legacy/edge-judiverse-tech/docs/packs/UK-PACK-07.md` — oid `a0a78c35f5c7492c441aaece7f45af697aa4010c` - `legacy/edge-judiverse-tech/registry/baselines/hosted-b.json` — oid `aa121915a06f4a6f3565ae9bfa20c06ee72766d3` (with `hosted-a.json` and `hosted-c.json` beside it) - 13 machine-readable registry copies under `legacy/edge-judiverse-tech/registry/` (benchmark, financial-intelligence, frontier, risk-resilience, JudeBox criteria, mail parity and others; counted) - `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/packs/bank-frontier/benchmarks.json` — oid `e6541ac684bef4cd842665202b1d38c35a0b92a1` - `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/packs/clinic-frontier/benchmarks.json` — oid `2ea26c8d74d0f6d8a04aba5d6ece4c7ad181e1a8` - `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/packs/government-frontier/benchmarks.json` — oid `0a1de5bff7f03f63aa1573b39707bc04027b4b79` - `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/packs/consortium-frontier/benchmarks.json` — oid `84533748d77a84e4385c24beb088d734728eb32b` - `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/packs/agent-estate-frontier/benchmarks.json` — oid `c3c34a74abb5eee476be187c80c3ee76ee050baa` - `legacy/local-source-snapshots/programme/outcomes-register/BENCHMARK-PROTOCOL.md` — oid `e5984a528fd8843da8a465816cfb23955d73b78e` - `legacy/local-source-snapshots/programme/jps/transfers/t1-register-20261019/vc6-v15-source-review-1745/t10-vc6/BUYER-BENCHMARK-SECTION.md` — oid `d425d8662b6a7b4d8afbe028543aa5ce4c8ce679` - `legacy/local-source-snapshots/research/nvidia/NVIDIA-BUILD-ADOPTION.md` — oid `ff7834e31b943b2b5308c35dfbeec0917697805e` - `legacy/jude-mvp/judeai-hosted/benchmarks/hard-tasks-v1/README.md` — oid `a324e5332bb81f007392f8b864e7d4a4d8c67eb2` - `legacy/jude-mvp/work/jude-27b-comparison-20261002/PREREGISTRATION.md` — oid `55210bd7a0de208a2b3642ffbca198db92ae1292` - `legacy/judiverse-next/docs/research/evals.md` — oid `a354b553c11862a6d5033c4bf85324f7517b1407` - `legacy/judiverse-next/docs/research/records/2026-09-25-transfer-baseline.md` — oid `6c3b0b8a03349af4ced4f143c02220fa849235de` ###### 1.4 Domain measurements (finance and clinic) - `old-repo/registers/measurements/bank-kyc-20261029.json` — oid `311edfa3b7a9ef67f30411fdc7f87a1001d6d410` - `old-repo/fc/results/TYPOLOGY-BENCHMARK-dev-seed7.json` — oid `224e33e52b3793392586a43d41f509c9534fa3c1` - `legacy/github-metadata-20261030/local-working-changes/canonical-local/files/fc/results/CARD-FRAUD-BENCHMARK-20261028T020325Z.json` — oid `1cd6d695aa6f81f2dddaf654247b14992ade3f4d` - `old-repo/adoption/research/clinical.json` — oid `cccb3c52758ab92e34a3d990d99fc99d2f1a1d76` ###### 1.5 Competitor research and parity - `old-repo/registers/parity/competitor-catalogue-20261030/README.md` — oid `73c04f54ec6cfccba57d46170159b44cf0af6544` - `old-repo/registers/parity/competitor-catalogue-20261030/summary.json` — oid `d65a34bdd18a966a02118c5ac4df8e2aaa5ca104` - `old-repo/registers/parity/competitor-catalogue-20261030/MONITORING.md` — oid `1f06d9fde9615bc66160db8a216f683e1695e782` - `old-repo/registers/parity/competitor-catalogue-20261030/lanes/competitor-floors-p50/README.md` — oid `5616bc092403af5532ea47eea280ffd2edbc0a5c` - `old-repo/registers/parity/competitor-catalogue-20261030/lanes/competitor-floors-p50/parity-count/PARITY-COUNT.md` — oid `90d0ad05e0fabac7cd6dd110bb11eb14fcc02cd8` - `old-repo/registers/parity/STATE-HANDOVER.md` — oid `9e4274a24810d0b52b790dbf54b292f5b54af64f` (contains sensitive data; not reproduced) - `old-repo/registers/parity/PARITY-20261029T232500Z.md` — oid `9145c55b228cf7def657b660d7098998c4966ae3` - `old-repo/registers/parity/PARITY-20261029T230000Z.md` — oid `50162298f35ff97cccfd4ba9a181f6a6f0f2edd6` - `old-repo/registers/parity/PARITY-DOCS-20261029T231223Z.md` — oid `ccd22da344b6e653901379da5f5afc13f5c641c8` - `old-repo/registers/parity/GAP-DEFINITIONS-20261029T232500Z.jsonl` — oid `8d1de824d1566c6f40e7fcb96c34d68c4c46fb82` - `old-repo/registers/parity/mobbin_pages.jsonl` — oid `855d8caf729cf436de0d0f0394b84b3b44193fb0` - `old-repo/office/PARITY-OFFICE-20261029T231015Z.md` — oid `37f2e2fd8b4ef683c21e5c9131761a262b24320e` - `old-repo/engine2/proof/parity/STATE-HANDOVER.md` — oid `3b83b14ae8e3ad775b56725d69b4483e7bcd8b73` - `old-repo/spec/PL-1-SOVEREIGN-ALTERNATIVE.md` — oid `d868cbb4dbc87d38260828eb75fabe53089c5327` - `old-repo/handover/uk-palantir-register-20261030/README.md` — oid `1472f416c5124a529206b4bd273698fb3581e7b9` - `old-repo/handover/uk-palantir-register-20261030/evidence/fts_keyword_search_results.txt` — oid `2de9da5d99edae57dd6fb8aed5d2d981c1b419b5` - `old-repo/handover/HANDOVER-COORDINATOR-20261030/memory/glass-box-palantir-incogni-20261030.md` — oid `f33495898b5bd1fb6e6482e1b507c2baa8fb7401` - `old-repo/handover/scope-baseline/SCOPE-BASELINE.md` — oid `26de13930cb77c006fa14d057d94bf8bde988758` - `legacy/governed-intelligence-judeai-edgeid-pre-contract/registers/competitors.json` — oid `6a4713c199b4e71cd2a72b46ece43729fd682422` - `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__PARITY-SUMMARY.md` — oid `e1772953d44aab3cf5c500127494e8e4f0a81196` - `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__parity__PARITY-BANK.md` — oid `52a4a1b518b5b5031aab5e88ae9e81e343c6ba53` - `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__parity__PARITY-CLINIC.md` — oid `9d4f001241bb241de7ac1442adc3890c0ac43677` - `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__parity__PARITY-GOV.md` — oid `f9d7b19376acf0cb49080d1566d4ebaf09f1ea39` - `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__parity__PARITY-JUDEAI.md` — oid `a8a34ea6f0c5abb78d2ed9aff62eaa64fee1e9f2` - `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__parity__PARITY-COLLAB.md` — oid `b81895c815108da1428ba969bae5b63dbd5de50d` - `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__parity__PARITY-IDENTITY.md` — oid `89cf2e900f15f6590fa3fb477197a7ff3c05246a` - `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__parity__PARITY-PLATFORM.md` — oid `fcbcf040d2097c27ad8af915e556441fe29eebef` - `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__parity__PARITY-DESIGN.md` — oid `04faa124a2d175d5a3780701f2a21a60d6824344` - 8 machine copies `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__parity__PARITY-*.jsonl`; representative BANK — oid `477fb50e3735b7dc4d04032d06ed82303284a0fa` - `legacy/edge-judiverse-tech/docs/finance/bank/01-COMPETITOR-BENCHMARKS.md` — oid `07327f1fef2e267c6c7b775d0e2caea73897fdad` - `legacy/edge-judiverse-tech/docs/finance/bank/01-COMPETITOR-BENCHMARKS.04bae573.md` — oid `04bae5733696f87e16e4ce5ce6507b304f7d4c2c` - `legacy/edge-judiverse-tech/docs/outreach/01-COMPETITOR-BENCHMARKS.md` — oid `37802a690e1e544dcd353137a3572e5b248f6f19` - `legacy/edge-judiverse-tech/docs/mail/PARITY-MATRIX.md` — oid `7c6de03ac83cddcfc6bb983b8ddc66ccdf57394f` - `legacy/edge-judiverse-tech/docs/mail/PARITY-PLAN.md` — oid `7c1de72ec371ad30ffc307672b0fdcc5c2fc4913` - `legacy/jude-mvp/docs/judepay-vs-sage.md` — oid `c7b7470c4ff2c69cb4c4297afa2090e849f9219b` - `legacy/jude/docs/judepay/parity.md` — oid `4f26736f4253f7d6a7a09c46d0872a700a7a9d20` - `legacy/jude-mvp/docs/plan/PARITY-CHATGPT.md` — oid `d546fd8898ea85287d255cc035a448f04b849a30` - `legacy/jude-mvp/docs/jpen-brief.md` — oid `014d39e4124c7d99d707d4559568af3e7012ae6e` - `legacy/jude-mvp/jmusic/RIVALS.md` — oid `0269534b2927b25f7501ce360c74b492f1147d2e` - `legacy/jude-mvp/jude-live/jmusic/DAW-PARITY.md` — oid `5267d23c92ac50537918dd4d37c83ed3fc3122f2` - `legacy/jude-mvp/work/unified-finance-20261001/reference/starling.txt` — oid `d71276b316844495e267f8168bcde5034cc661db` - `legacy/judeai-edge-control/docs/admin-benchmark.md` — oid `fef8e7660c0e70f1eab20a26cc5fa9d38e226980` - `legacy/judeai-edge-control/docs/jude-agent-toolset-benchmark.md` — oid `6d4fc75d1266d45282d56a6ba7e86123c8558fe8` - `legacy/judeai-edge-control/docs/secure-group-call-benchmark.md` — oid `d3bab1709456200150c96279d7b5890d338928ca` (contains sensitive data; not reproduced) - `legacy/judeai-edge-control/docs/research/platform-inventories/ehr.md` — oid `7ab9496dcb9575bbdf94fd9ca1c6d166e8eb4570` - `legacy/judeai-edge-control/docs/research/platform-inventories/excel.md` — oid `410f9db8e459ee8a1194680a7727f9024e5fdffe` - `legacy/judeai-edge-control/docs/research/platform-inventories/teams.md` — oid `af6682d6bc015f9e81a284b85d7fdff9bb536a86` - `legacy/judiverse-next/docs/parity/judeai.md` — oid `511fda83f887ded02384802ae1627d483865bcde` - `legacy/judiverse-next/docs/parity/mail.md` — oid `cabaa7ab4e7dc297d9caf8539887c4799b969bc7` - `legacy/judiverse-next/docs/parity/sources/chatgpt-research.md` — oid `ba188246373030186d2c8f203784b57c8e882b1c` - `legacy/judiverse-next/docs/parity/sources/outlook-research.md` — oid `5ee96e5e496964192a07a6d0a72e9a724dfd11b4` - `legacy/local-source-snapshots/workbenches/intenjude/GITLAB-PARITY.md` — oid `8084be96a1bb7bcf2587a6bb4ef07ec8f8643c44` - `legacy/local-source-snapshots/workbenches/communications/TEAMS-OUTLOOK-PARITY.md` — oid `218c198acd63bea61b1739fa463d67031584caac` - `legacy/local-source-snapshots/programme/proof/partners/STARLING-BANK-PROPOSED.md` — oid `836f7a8ef458e182f503f0337dbdbd484e73aaac` - `legacy/local-source-snapshots/programme/proof/procurement/FCA-JMLSG.md` — oid `c746939de0b33885417a42e690c137aa49525939` - 16 Teams coverage source captures under `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/proof/teams-coverage-independent-20261022/`; representative `.../sources/accessibility.html` — oid `2c98ecaaeb9f40f7b60c645979e8f9694f95e5ac` - 14 Mobbin query logs `old-repo/handover/tok-20261030/specification/authoring-20261031/mobbin/QUERIES-*.jsonl`; representative QUERIES-1 — oid `42b6422cadeae8372bec4c518442e4abc110cace` - 206 bank-journey evidence files modelled on Starling flows (onboarding, KYC checks, fraud desk, maker-checker) under `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/outcomes-register/evidence/2026-10-2*/claude-raguel-starling/` (counted; sample evidence matrices and accessibility summaries read; they contain host names; not reproduced) ###### 1.6 Readiness definitions, gates and recorded levels - `old-repo/spec/TRL9-CRL9-GATE.md` — oid `993790a9d40cb076ff4ef0b21432cc992b6f639d` - `old-repo/spec/ANNEX-P-PATHWAYS.md` — oid `6f77ecf04e7ec6832d7710e09eeb4964899d9956` - `old-repo/spec/ANNEX-P-PATHWAYS.16869b95.md` — oid `16869b957e98ead410e6ed53adef61d68dba4ef7` - `old-repo/engine2/trlcrl/criteria.json` — oid `a2d3c49de486aa73f0dd3b0cca9e4aab88405c79` - `old-repo/engine2/trlcrl/criteria.2ef35630.json` — oid `2ef3563067282a090ee19ec005812cb0a8cbd027` - `old-repo/engine2/trlcrl/sources/ukri-stfc-trl.html` — oid `af57f87fdbadb9c8da63e8585cc0e4172e13d36c` - `old-repo/engine2/trlcrl/sources/nasa-npr-7123-1d-appendix-e.html` — oid `600f641d7cdeab90ad28e8d5c84f79b46946bb1b` - `old-repo/engine2/trlcrl/sources/nasa-trl.html` — oid `2c60f7a42a450b865c3e94dfe332b272332d4890` - `old-repo/engine2/trlcrl/sources/bindt-crl.html` — oid `d09874f498c9c5186671fecf05d08520e6c830c9` - `old-repo/engine2/trlcrl/sources/rcnde-measuring-tech-progress.html` — oid `88d712895cac21ce4f02e82687f8056759bb22f2` - `old-repo/engine2/proof/trlcrl/run-2026-10-29T204735/summary.json` — oid `6673f80f6da5514943ef3a5b41928c4a743a3c6b` - `old-repo/engine2/proof/trlcrl/run-2026-10-29T220716/summary.json` — oid `27a1e9a82d811d9df2be508a16ba8b7e910ec9db` - `old-repo/engine2/proof/trlcrl/run-r4-20261029T2305Z/summary.json` — oid `b5b6c4bef476b037b948ecfd378b6f73a0d14f50` - `old-repo/engine2/proof/trlcrl/run-2026-10-29T231613/summary.json` — oid `4d4288debaecd7f460047bb5e2c2e7dd2d38ec7c` - `old-repo/engine2/proof/submission/TRL-CRL-EDGE-BANK-r4.json` — oid `2c0802d9d820512c0ac2ba17b6d2ff1a9d4f9efa` - `old-repo/scope/SCOPE-T9C9-20261029T230350Z.md` — oid `566789e8a7279dec47ff3be8ff1ef702a38b0cdd` (contains personal email addresses; not reproduced) - `old-repo/scope/README.md` — oid `d439ba2c500f70e05985db68b0c48b25e36ee1b5` - `old-repo/handover/source-context/ROUTE-TO-9-20261029T2335Z.md` — oid `acf1d3247d0e87c9ba382d0419cbbdff9425435f` - `old-repo/docs/ACTUAL-TRL9-CRL9-JOURNEY-CONTRACTS.md` — oid `ac1f5144def9353f31ab2815b2ab95efc505c251` - `old-repo/docs/DECFAW-ADOPTION-AND-PROOF.md` — oid `d404fb850bd1891c243c55ae647e7e5b10f585a2` - `old-repo/docs/ORACLES.md` — oid `79dac78828a9e5384ccd3b84638cc2d4e2faea82` - `old-repo/docs/incidents/INC-EST-gate-run-isolation-2034.md` — oid `a03fe5b49610565227939fd87a8e3c8f9d8bc944` - `old-repo/handover/ESTATE-LEVEL-READINESS.md` — oid `e03f2e2cd0382aed5d1ae1c6ee647e9fa821f16e` (and 24 near-identical variants) - `old-repo/handover/tok-20261030/specification/authoring-20261031/BRIEF-ADDENDUM-3-NEW-TRL9.md` — oid `bc9332aaab9fb5282c28e2f5aea5d00d4c058887` - `legacy/governed-intelligence-judeai-edgeid-pre-contract/decisions/2026-10-29-judeai-edge-id-product-direction.md` — oid `a8bf234a439fdde2adea778ebfb91eab825c35dc` - `legacy/x-judeai-edge-tok-kol-ndc-ml-x/handoffs/AEC-STANDARD.md` — oid `940e55847b21daa3cead1ec73cdb42f378b97ce2` - `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/ESTATE-LEVEL-READINESS.md` — oid `9a3b37052446cdccf248dc479758de4277dd5174` - `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/aec9/TRL9-CRL9-PREDECLARED-RECEIPTS.md` — oid `2b18fcca257a66bb8dcb9f7a2eb8af7f0d4b9a01` - 788 clinic proof contracts (394 items, each with a TRL 9 and a CRL 9 contract) under `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/packs/clinic-autonomous-proofs/compiled/proofs/`; representative `CL-AGEN-01-CRL9.json` — oid `21dd0e6f55618c68de5718c8dcafa843b2452e73` - `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/map/LEVEL-REGISTER-20261020.csv` — oid `81e3a52755394b33e8b45e0fec0b7c8734e8680a` - `legacy/local-source-snapshots/programme/map/LEVEL-REGISTER-20261020.csv` — oid `0296ea174e995c83c584d25b131ca9f6d95c729d` - `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/outcomes-register/register.json` — oid `204286660dca8314ad0f2c23cf2aae384c06f0ba` (and 8 earlier or mirrored versions) - `legacy/local-source-snapshots/programme/OPERATIONAL-READINESS-DECREE-20261020.md` — oid `a06193d3a5edad64b40c1ce8bce1ad9be459f9e4` - `legacy/local-source-snapshots/map/crl9/JudeAI-EDGE-ID-CRL9-20261015.md` — oid `ba1652cc7c4550923e56ae23e6d7fbe9f061a65c` (with its PDF, HTML and XLSX renderings) - `legacy/x-judeai-edge-tok-kol-ndc-ml-x/stations/mail/MAIL-CRL9-PLAN.md` — oid `9a6b579dd616e8af8b861e2861a308ed30ea485b` (contains sensitive data; not reproduced) - `legacy/x-judeai-edge-tok-kol-ndc-ml-x/stations/mail/MAIL-CRL9-PROOF.md` — oid `74876327492ac8181f779ed3006a9b935e97e79c` - `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/proof/agent-estate-frontier-independent-20261022/baseline/README.md` — oid `aead72ea741e178696d3e2959d16904bedb3da8c` - `legacy/peak-judeai-behavioural-harness-baseline-full-application/docs/VALUE-MAP.md` — oid `5c7975865f123603126613eafd6a5eba3173f0e8` - `legacy/jude-mvp/docs/readiness-uplift.md` — oid `ab298f43b57e3a51902fa2aafc21136d5f9d2728` - `legacy/edge-judiverse-tech/docs/assurance/uk-readiness-and-benchmarks.md` — oid `80f5637087bd2f0548c289679307861702c98ff6` - `legacy/edge-judiverse-tech/docs/assurance/DECFAW9-CONTROL-CLOSURE.md` — oid `2d790e19f0f9db941dd0b0f3dfab7512ccf315f3` - `legacy/judeai-edgeid-cloud/handover/whole-estate-acceptance-20261026-01a16a2c/sources/SECTION-16-OBSERVED.md` — oid `546306ab2be0f12375a613e808ed8d5b4e64ccb8` - `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__BASELINE-20261024.md` — oid `273067f07f0584a5160f013d24fe24940514e081` (contains host names; not reproduced) - `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__applications__IUK-2527__SCORE-CHECK.md` — oid `43282a335c605d0e8792a04bb49ea7f6972ee2ff` - `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__applications__UKDI-SECURITY-C2__SCORE-CHECK.md` — oid `e9965bb9241ad4c2516b974acedf320612586cf1` --- ##### 2. Competitors already researched, by application area The old estate's own rule was that a competitor row is "a source to measure against, not a claim about us". Almost every parity statement in the corpus was either a reading of public documentation or a keyword match against the old estate's own feature definitions, not a test of the competitor's working product. Competitor figures below are OBSERVATIONS as reported by the old document; the old estate's parity percentages are CLAIMS. ###### 2.1 Findings that apply to every area - **The largest competitor catalogue** — recorded in `old-repo/registers/parity/competitor-catalogue-20261030/README.md` on 30 October 2026 (historical): 407 product or reference records and 13,126 item records (4,538 features, of which 2,139 documented API operations and 2,399 documented feature claims; 6,257 assurance controls; 671 interaction controls; 348 workflows; 1,312 baselines), from 1,492 distinct source addresses, across seven lanes (ai-lab 116 products, bank 79, clinical 38, cyber 45, office 54, patterns 16, public-learning 59). Its own qualification, recorded in `old-repo/registers/parity/competitor-catalogue-20261030/summary.json` on 30 October 2026 (historical): enumeration was complete only for the frozen inputs, "worldwide competitor discovery remains open", 946 gap records and 91 open discovery records, zero estate nodes admitted and zero parity tests executed. - **Institutional and privacy "floors"** — recorded in `old-repo/registers/parity/competitor-catalogue-20261030/lanes/competitor-floors-p50/README.md` on 30 October 2026 (historical): 57,014 de-duplicated requirements from 65,871 mentions and 5,518 sources for Palantir (Foundry 48,432; Apollo 3,101; AIP 2,486; Gotham 1,242) and Incogni (1,571; Ironwall 182). The deterministic parity count, recorded in `old-repo/registers/parity/competitor-catalogue-20261030/lanes/competitor-floors-p50/parity-count/PARITY-COUNT.md` on 30 October 2026 (historical): 10,163 documented requirement records; 6,426 distinct functional capabilities; 32 excluded (Defence targeting and fires, order of battle, intelligence and mission planning, Gotham Target Workbench and some Incogni procedures); 6,394 admitted parity obligations; 0 verified. An independent audit of a 60-record sample found 7 errors (11.7 per cent, 95 per cent Wilson interval 5.8 to 22.2 per cent), all over-counts; the estimated true number of distinct capabilities was about 5,366 (range 4,486 to 5,897). Gotham user documentation is not public. - **Screen-flow parity by keyword match** — recorded in `old-repo/registers/parity/PARITY-20261029T232500Z.md` on 29 October 2026 (historical): 321 competitor flows and 1,996 screens from the Mobbin design library; "feature parity" 93.5 per cent, "flow parity" 93.1, "control parity" 67.9, "agent-executable" 91.9, "audit superiority" 76.9; 95 gap definitions. The rule was textual ("feature intention contains any flow term"), so these percentages measure vocabulary overlap with the old estate's 5,402 feature definitions, not working parity. The earlier run, recorded in `old-repo/registers/parity/PARITY-20261029T230000Z.md` on the same date (historical), gave 86.1 per cent on 108 flows and is superseded by the later file. - **Vendor-documentation parity** — recorded in `old-repo/registers/parity/PARITY-DOCS-20261029T231223Z.md` on 29 October 2026 (historical): 68 vendor documentation addresses attempted, 51 answered, 17 blocked or missing; 55 flows; control parity 12.7 per cent; GOV.UK and Babylon/eMed not reached. - **Parity obligations per floor** — recorded in `old-repo/handover/journeys-prospective/PARITY-OBLIGATIONS.jsonl` on 30 October 2026 (historical): 575 obligations, 25 per competitor floor per family across "depth" (207), "maturity" (184) and "reliability" (184); every one `NOT_YET_DEMONSTRATED`. - **Competitor-parity benchmark rows** — recorded in `old-repo/handover/benchmark-standard/BENCHMARK-REGISTER.jsonl` on 30 October 2026 (historical): 119 rows over 7 scorecard dimensions (autonomous effectiveness, correctness, efficiency per verified outcome, experience, functional coverage, recovery, security) for 17 floors: Monzo, Revolut, Starling, Barclays, Tide, Spotify (for JudeBox), GitHub, GitLab, every Microsoft product, Microsoft Office, Microsoft Outlook, Incogni, a leading clinical EHR, Oracle Fusion, Palantir, Salesforce and ServiceNow. Required result: "at least equal to the competitor's measured result on the same task; exceeds requires strictly better with the difference outside the declared tolerance"; a planned feature never scores. - **Horizon parity across eight families** — recorded in `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__PARITY-SUMMARY.md` on 25 October 2026 (estate statuses read 24 October 2026, historical): 3,231 parity features, 1,430 live (44 per cent), 876 partial, 925 missing; without the Government family's 839 contract lines, 2,392 curated features of which 468 were live (20 per cent). Fourteen live defects were recorded (sign-in, missing status, terms and accessibility pages and others; the route details contain host names; not reproduced). ###### 2.2 EDGE Finance (banking, payments, accounting, financial crime, insurance) | Competitor or reference | What was recorded | Recorded in, on (historical) | |---|---|---| | Starling, Monzo, Revolut, Chase UK, Wise, Zopa, Tide, Allica, Barclays, Lloyds, NatWest, HSBC UK and first direct, Nationwide, Santander, TSB, Kroo | OBSERVATIONS from regulators and surveys: licence status; customers and deposits (Monzo 15.2 million customers and £25.7 billion deposits; Starling 6.2 million accounts and £12.7 billion; Wise 18.9 million active customers worldwide); CMA Ipsos service quality (Monzo 82 per cent GB overall, Starling 77); Which? scores (Starling 86 per cent); Financial Ombudsman cases and uphold rates (average 26 per cent; Wise 19, Starling 23); FCA complaints per 1,000 accounts (Chase 0.97 lowest, NatWest 2.90); PSR APP reimbursement (88 per cent of in-scope value, 7 October 2024 to 31 March 2026); 446 major incidents reported by retail banks in 2025; accessibility surveys. PREDICTIONS: a 40-metric scorecard setting each target as "beat the best by a stated margin, to be measured". No bank published availability, support response time, false-hold rate or agent task success | `legacy/edge-judiverse-tech/docs/finance/bank/01-COMPETITOR-BENCHMARKS.md`, 13 October 2026. A branch variant (`legacy/edge-judiverse-tech/docs/finance/bank/01-COMPETITOR-BENCHMARKS.04bae573.md`, same date) has a 36-metric scorecard; this report follows the 40-metric file | | Starling, Monzo, Revolut, Barclays, Nationwide, HSBC, Tide, Wise; Xero, QuickBooks, FreeAgent, Sage; Stripe, GoCardless; Feedzai, Featurespace (Visa), Napier AI, ComplyAdvantage | The bar per bank product (for example Starling's £1,000,000 Faster Payments limit; Wise from 0.24 per cent; vendor CLAIMS of 62 per cent more fraud caught by Feedzai and 75 per cent fewer false positives by Featurespace). Old-estate CLAIM: bank family 295 features, 68 live, 74 partial, 153 missing | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__parity__PARITY-BANK.md`, 24 October 2026 | | Monzo, Starling, Revolut Business, Xero, Tide, Barclays Business, British Business Bank, Oracle Fusion/NetSuite | Screen-flow and documentation parity: Monzo 111 flows and 757 screens; Starling 86 and 507; Revolut 85 and 471; Tide 13 documented flows with 15 gaps; Tide CHAPS limit £1 million with a £15 fee | `old-repo/registers/parity/PARITY-20261029T232500Z.md` and `old-repo/registers/parity/PARITY-DOCS-20261029T231223Z.md`, 29 October 2026 | | Sumsub, Entrust (Onfido), ComplyAdvantage | KYC/KYB/AML feature comparison from vendor documentation. Old-estate verdict: "not ahead yet on KYC itself"; identity and business verification, screening and monitoring were only specified; difference claimed only on receipts, reproducibility and hold-by-default | `old-repo/registers/measurements/bank-kyc-20261029.json`, 29 October 2026 | | Temenos, Mambu, Thought Machine, Sumsub, Entrust, ComplyAdvantage | Must-match lists and benchmark sources for core banking and verification | `legacy/governed-intelligence-judeai-edgeid-pre-contract/registers/competitors.json`, 29 October 2026 | | 79 bank-lane products: 13 direct competitors (Barclays, HSBC, Lloyds, Metro, Mettle, Monzo Business and current accounts, NatWest, Revolut Business, Santander, Starling business and current, Tide), platforms (Adyen, Backbase, ClearBank, Finastra, Mambu, nCino, Temenos, Thought Machine, Plaid, TrueLayer, Yapily, Stripe, Wise Platform and others), substitutes (Xero, NetSuite, Pleo, Moss and others) and the Open Banking UK 4.0.1 specifications | Item counts per product (Revolut Business 117, Revolut Merchant API 113, Payment Initiation API 4.0.1 48) | `old-repo/registers/parity/competitor-catalogue-20261030/README.md`, 30 October 2026 | | Sage 50 Accounts, Sage 50 Payroll, Sage HR | Old-estate CLAIM from a code audit: about 46, 52 and 22 per cent parity; a later check gave 32, 46 and 27 per cent (40 per cent for Sage Payroll), rising to 46 to 71 per cent including unmerged work | `legacy/jude-mvp/docs/judepay-vs-sage.md`, 6 September 2026; `legacy/jude/docs/judepay/parity.md`, 25 September 2026 | | Starling Bank (regulatory record) | OBSERVATION: FCA Final Notice dated 27 September 2024 imposing a financial penalty of £28,959,426 on Starling over financial crime controls, after a 30 per cent settlement discount | `legacy/jude-mvp/work/unified-finance-20261001/reference/starling.txt` (copy of the notice, filed 1 October 2026) | | HubSpot, Salesforce Marketing Cloud, Mailchimp, Brevo, Klaviyo, ActiveCampaign, Pipedrive, Apollo, lemlist, Instantly, Customer.io, Attio | Outreach and CRM scorecard of 34 metrics (mailbox providers require spam complaints below 0.30 per cent; Mailchimp all-industry unsubscribe 0.22 per cent, open 35.63 per cent, click 2.62 per cent) | `legacy/edge-judiverse-tech/docs/outreach/01-COMPETITOR-BENCHMARKS.md`, 13 October 2026 | | Starling (as a proposed partner, not a competitor) | Proposal pack for a second regulated environment; the old estate recorded that no partner agreement was signed | `legacy/local-source-snapshots/programme/proof/partners/STARLING-BANK-PROPOSED.md`, 20 October 2026 | ###### 2.3 EDGE Clinic (social prescribing, early detection, records and care pathways) | Competitor or reference | What was recorded | Recorded in, on (historical) | |---|---|---| | NHS App, Epic MyChart, Patients Know Best, Kry (Livi), Push Doctor; Epic, EMIS (Enlivio Health), TPP SystmOne, Oracle Health (Cerner), Accurx; Doctolib; Huma; Cera; the NHS national services (NHS login, PDS, GP Connect, e-RS, EPS, Spine, MESH) | Why each is the bar: NHS App 42,751,011 registrations and 7,781,808 repeat prescriptions in September 2026; MyChart 56 named features; Patients Know Best over 8 million users; SystmOne in over 2,900 GP practices; Accurx in 98 per cent of GP practices (undated); Doctolib no-show rate near 2 per cent (secondary source); Huma CLAIMS over 30 per cent readmission reduction (method not stated); Cera hospitalisations halved (partner page); NHS cancer Faster Diagnosis Standard target 80 per cent by March 2026. Old-estate CLAIM: clinic family 322 features, 33 live, 230 partial, 59 missing | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__parity__PARITY-CLINIC.md`, 24 October 2026 | | 38 clinical-lane products: direct (Accurx, Altera Sunrise, Dedalus, Docplanner, Doctolib, EMIS Web, Epic, MEDITECH Expanse, Oracle Health, SystmOne, Vision), substitutes (Abridge, Ada, Babylon legacy, Buoy, Doccla, DrDoctor, eMed, Heidi, Kry, Livi, Luscii, MyChart, Nabla, Patient Access, Patients Know Best, Push Doctor, Suki, Teladoc, Zocdoc), platforms (Amwell, Bahmni, Eka Care, OpenEMR, OpenMRS, ThinkAndor) | Item counts (Epic 622, OpenEMR 148, OpenMRS 53, EMIS Web 19) | `old-repo/registers/parity/competitor-catalogue-20261030/README.md`, 30 October 2026 | | Epic, Oracle Health, NHS Spine and NHS App, openEHR | 80 capabilities across 10 areas, with interface conventions (Epic keyboard accelerators, dense tabbed workspace) | `legacy/judeai-edge-control/docs/research/platform-inventories/ehr.md`, 10 October 2026 | | Zocdoc, NHS App, Patient Access, Babylon/eMed | Screen-flow and documentation parity (Zocdoc 3 flows; Babylon/eMed not reached) | `old-repo/registers/parity/PARITY-20261029T232500Z.md` and `old-repo/registers/parity/PARITY-DOCS-20261029T231223Z.md`, 29 October 2026 | | Epic, Oracle Health (Cerner) | Must-match lists (patient record, scheduling, orders, results, portal, FHIR R4 interoperability) | `legacy/governed-intelligence-judeai-edgeid-pre-contract/registers/competitors.json`, 29 October 2026 | | Moorfields Eye Hospital (research partner candidate, not a competitor) | A retinal early-detection evaluation contract: freeze the intended purpose; DTAC form 2.0 (February 2026); DCB0129 and DCB0160 versions; CQC registry. No partnership, data access or approval was held, and zero messages were sent. The contact route is an email address; contains sensitive data; not reproduced | `old-repo/adoption/research/clinical.json`, 29 October 2026 | UNKNOWN: social prescribing as such was not researched as a competitor set in any in-scope file. The nearest material is the cancer-pathway, check-in and remote-monitoring bars above. ###### 2.4 EDGE Public Consortium (cyber governance, procurement, vetting, defence and government logistics, education and research) | Competitor or reference | What was recorded | Recorded in, on (historical) | |---|---|---| | Palantir Foundry, AIP, Apollo, Gotham | MOD three-year enterprise agreement of about £240.6 million from 1 April 2026, directly awarded under a defence and security exemption; NHS England Federated Data Platform (£330 million, 2023; 45 of 215 trusts adopted at the end of 2024); capability map against the old estate; what engineering alone cannot win (Cyber Essentials Plus, ISO/IEC 27001, Defence Cyber Certification, List X facility, SC and DV clearances, DSPT, DCB0129/0160, framework access, track record) | `old-repo/spec/PL-1-SOVEREIGN-ALTERNATIVE.md`, 27 October 2026 | | Palantir (UK public contracts) | OBSERVATIONS from official notices: 37 register rows (30 contracts counted, 6 framework or market admissions, 1 award at £0); for example MOD enterprise agreement £240,600,000 (notice December 2025); NHS England FDP £182,242,760 (November 2023); FDP transition and exit £24,925,000 (June 2023); NESO interim capability £16,799,683 (August 2026); Police Digital Service firearms licensing £7,500,000 (May 2026); FCA enterprise search proof of concept £375,001 (January 2026); 19 notices found by keyword search | `old-repo/handover/uk-palantir-register-20261030/README.md` and `old-repo/handover/uk-palantir-register-20261030/evidence/fts_keyword_search_results.txt`, 30 October 2026 | | Canvas LMS, Moodle 4.5, Anthology Blackboard, Coursera for Campus; Palantir; Microsoft Dynamics 365 for government; Salesforce Public Sector; Capita; Serco; Jaggaer; Coupa; Proactis; Find a Tender and Contracts Finder; Salesforce Sales Cloud, HubSpot, Clay, Apollo; Companies House; Creditsafe; Dun and Bradstreet | Why each is the bar (Canvas publishes 99.9 per cent uptime; Moodle 16 activity types; Find a Tender thresholds £12,000 and about £139,688). Old-estate CLAIM: Government family 1,147 features, 962 "live" (largely contract lines), 97 partial, 88 missing | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__parity__PARITY-GOV.md`, 24 October 2026 | | Cyber lane (45 platforms): AuditBoard, Auth0, Cloudflare Zero Trust, Credo AI, CrowdStrike, CyberArk, Darktrace, Datadog, Drata, Elastic, Grafana, Hyperproof, IBM watsonx.governance, Keycloak, LogicGate, MetricStream, Microsoft Defender, Microsoft Entra ID, Okta, OneTrust, Ping Identity, Prisma Cloud, Qualys, SailPoint, Saviynt, Secureframe, SentinelOne, ServiceNow GRC, Snyk, SonarQube, Splunk, Sprinto, Tenable, Vanta, Wiz, Workiva, Zscaler and others | Item counts (Keycloak 171, Tenable 17, SonarQube 16, Credo AI 13) | `old-repo/registers/parity/competitor-catalogue-20261030/README.md`, 30 October 2026 | | Vanta, Drata, Darktrace (cyber); Vanta, Drata, Credo AI, IBM watsonx.governance, Palantir AIP/Foundry, Salesforce Agentforce, ServiceNow (assurance and auditing); GOV.UK One Login, Companies House, GOV.UK (public sector) | Named incumbents per domain; education and engineering incumbents explicitly "NOT_MEASURED: no register row exists" | `old-repo/scope/SCOPE-T9C9-20261029T230350Z.md`, 29 October 2026 | | Vanta, Drata, Okta, Twingate, 15Five, Customer.io, GOV.UK One Login, Companies House, Innovate UK funding service | Screen-flow and documentation parity (Vanta 9 flows and 5 gaps; One Login 5 flows and 11 gaps) | `old-repo/registers/parity/PARITY-20261029T232500Z.md` and `old-repo/registers/parity/PARITY-DOCS-20261029T231223Z.md`, 29 October 2026 | | Public-learning lane (59): Accela, Blackboard, Brightspace, Canvas, Consensus, Coursera, Dimensions, edX, Elicit, FutureLearn, GoVocal, Granicus, LinkedIn Learning, OpenGov, SciSpace, Scite, Scopus, Tyler, Udemy Business, Web of Science; substitutes (Anki, Duolingo, Khan Academy, Mendeley, Overleaf, Perplexity, Quizlet, ResearchRabbit, Zotero); references (GOV.UK Forms, Notify, One Login, Pay; Crossref; OpenAlex; PubMed; Wikidata) | Item counts (Canvas 2,288; Brightspace 126; Moodle 30) | `old-repo/registers/parity/competitor-catalogue-20261030/README.md`, 30 October 2026 | | Salesforce, ServiceNow (consortium); Palantir, Credo AI, IBM, Snowflake, Databricks (lab); Reapit, Goodlord, Propertymark (public) | Must-match lists and benchmark sources | `legacy/governed-intelligence-judeai-edgeid-pre-contract/registers/competitors.json`, 29 October 2026 | | Funding competitions as a scored bar | Independent score checks of two draft applications: Innovate UK 2527 (questions scored 7 to 8 of 10 on second read) and UKDI Security Open Call 2026 cycle 2 stage 1 (14 of 20; weakest points the young company, absent end user and unnamed team) | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__applications__IUK-2527__SCORE-CHECK.md` and `...__UKDI-SECURITY-C2__SCORE-CHECK.md`, 25 October 2026 | ###### 2.5 Judiverse (office, AI assistants, agent lab) | Competitor or reference | What was recorded | Recorded in, on (historical) | |---|---|---| | ChatGPT, Claude, Gemini, Microsoft Copilot, Perplexity, Grok, Mistral Le Chat; Claude Code, Codex CLI; Apple Spotlight and Apple Intelligence; Google Drive search; Raycast; Neo4j Bloom; the MCP specification | Plans compared and what each sets as the bar (projects, memory, scheduled tasks, context size, connectors, agent harness features); OpenAI help pages refused automated reads, so ChatGPT facts came from search-index readings. Old-estate CLAIM: JudeAI family 246 features, 9 live, 156 partial, 81 missing | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__parity__PARITY-JUDEAI.md`, 24 October 2026 | | ChatGPT | 48-row matrix: of 34 must-have rows, 7 working, 7 partial, 4 unknown and 16 not working in the then-hosted product | `legacy/jude-mvp/docs/plan/PARITY-CHATGPT.md`, 24 September 2026; feature inventory in `legacy/judiverse-next/docs/parity/sources/chatgpt-research.md`, 25 September 2026 | | Outlook on Exchange Online, Gmail, Proton, Apple Mail, Superhuman; Teams, Google Meet, Zoom, Slack; Bookings, Google Calendar, Notion Calendar; OneDrive, Google Drive, Proton Drive, iCloud; Microsoft Office, Google Docs, Notion | Published limits used as bars (Teams meetings up to 300 or 1,000 by plan; Exchange mailbox 100 GB on E3; Gmail undo send up to 30 s; Slack free plan 90 days of history). Old-estate CLAIM: collaboration family 468 features, 325 live, 83 partial, 60 missing | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__parity__PARITY-COLLAB.md`, 24 October 2026 | | Microsoft 365, Google Workspace, fireflies.ai | Exercised parity with accepted receipts: 26 covered, 43 partial, 49 missing (Microsoft 18, 14, 12; Google 3, 5, 9; fireflies.ai 5, 24, 28) | `old-repo/office/PARITY-OFFICE-20261029T231015Z.md`, 29 October 2026 | | 54 office-lane products, 46 direct (Airtable, Asana, Basecamp, ClickUp, Confluence, Gmail and the Google suite, Jira, Linear, the Microsoft suite, Miro, monday.com, Notion, Otter, Read AI, Slack, Trello, Webex, Zoom and others) | Item counts (Figma Design 251, Microsoft Teams 180, Outlook 149, OneDrive 121, Slack 121) | `old-repo/registers/parity/competitor-catalogue-20261030/README.md`, 30 October 2026 | | Base44, Lovable, Bolt, Replit Agent, v0, Softr, Bubble, Glide, Retool; Notion, Teams, Excel, PowerPoint, Google Sheets and Slides, Airtable, Coda | Pricing, data location, SOC 2 statements, audit logs and export per app generator (Base44 from $16 a month; Retool £8 per builder plus £4 per user); no independent head-to-head benchmark of the nine app generators was found | `legacy/edge-judiverse-tech/docs/benchmarks/APP-GENERATOR-AND-OFFICE-BENCHMARKS.md`, 13 October 2026 | | Microsoft 365 admin centre, Entra ID, Exchange admin, Intune, Purview | Admin parity and minimum security benchmark for a regulated buyer, with a 40-item gap list | `legacy/judeai-edge-control/docs/admin-benchmark.md`, 7 October 2026 | | Microsoft Teams (meetings and calling); Excel | 67 Teams capabilities across 15 areas; 73 Excel capabilities across 9 areas | `legacy/judeai-edge-control/docs/research/platform-inventories/teams.md` and `legacy/judeai-edge-control/docs/research/platform-inventories/excel.md`, 10 October 2026 | | Outlook | Mail parity matrix (80 rows, all "specified; implementation and parity unverified") | `legacy/edge-judiverse-tech/docs/mail/PARITY-MATRIX.md`, 13 October 2026 | | GitLab, GitHub, Linear | GitLab parity matrix (nothing claimed met); GitHub and Linear recorded as having no estate application | `legacy/local-source-snapshots/workbenches/intenjude/GITLAB-PARITY.md`, 15 October 2026; `old-repo/registers/parity/PARITY-20261029T232500Z.md`, 29 October 2026 | | 116 ai-lab platforms (agent frameworks, model APIs, inference and observability, for example Claude Code 161 items, Mistral AI API 52, Kestra 36, CrewAI 25, LangGraph, Microsoft Agent Framework, Pydantic AI, n8n, Palantir AIP) | Item counts; many platforms have no retained items because their sources were unreadable | `old-repo/registers/parity/competitor-catalogue-20261030/README.md`, 30 October 2026 | | Figma, UX Pilot, Penpot, Untitled UI, Storybook, Tokens Studio, Framer, Canva, Notion, Obsidian, GitHub, Vercel, Netlify, Linear, Jira Product Discovery | Design-family bars. Old-estate CLAIM: 173 features, 14 live, 53 partial, 106 missing | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__parity__PARITY-DESIGN.md`, 24 October 2026 | | Datadog, Grafana Cloud, New Relic, PagerDuty, Opsgenie, incident.io, Statuspage, ServiceNow, Jira Service Management, Zendesk, Electricity Maps, Climatiq, Feedly, Google News, Apple News, LinkedIn, TikTok, YouTube | Platform-family bars. Old-estate CLAIM: 337 features, 2 live, 85 partial, 250 missing | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__parity__PARITY-PLATFORM.md`, 24 October 2026 | | Spotify (JudeBox); Serum 2, Ozone 12, Neutron 5, Moises; Logic Pro, Ableton Live; Microsoft Word | Music and document-editor parity tables (built, partial, gap) | `legacy/jude-mvp/jmusic/RIVALS.md` and `legacy/jude-mvp/jude-live/jmusic/DAW-PARITY.md` (September 2026); `legacy/jude-mvp/docs/jpen-brief.md`, 5 September 2026 | ###### 2.6 EDGE ID (identity, cross-cutting) Recorded in `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__parity__PARITY-IDENTITY.md` on 24 October 2026 (historical): Apple Account, Google Account, Microsoft Entra ID, Okta, Auth0, GOV.UK One Login; Yoti, Onfido (Entrust), Apple Wallet IDs, GOV.UK Wallet, Mastercard ID; Microsoft Authenticator, Okta Verify, Apple Passwords, Google Password Manager, 1Password, Bitwarden, Dashlane; Signal. Old-estate CLAIM: identity family 243 features, 17 live, 98 partial, 128 missing. Incogni (Standard, Unlimited, Family, Protect, Concierge) and Ironwall were set as the personal-privacy floor; Incogni's broker table held 420 rows (419 distinct names) and Incogni publishes no public API, recorded in `old-repo/registers/parity/competitor-catalogue-20261030/lanes/competitor-floors-p50/README.md` on 30 October 2026 (historical). --- ##### 3. Benchmarks, test suites and baselines ###### 3.1 How the old estate defined a benchmark claim (method worth reusing) Recorded in `old-repo/handover/benchmark-standard/STANDARD.md` and its variant `old-repo/handover/benchmark-standard/STANDARD.bafb4072.md` on 30 October 2026 (historical), status "proposed standard": - **Register classes:** published benchmark; regulatory or standard; competitor parity; estate gate; depth, maturity and reliability; resource-to-outcome (money, tokens, kWh, gCO2e, litres and seconds per verified outcome); owner superiority question; plus taxonomy and standards-foundation rows. - **Claim classes that never substitute for one another:** compliance, parity, superiority and certification (certification only from an accredited assessor, never self-awarded); the variant adds "measurement (no claim)". - **Mandatory run record:** benchmark id and version, harness digest, hardware, model routing and settings, seeds, task-set digest, per-task results, raw outputs and logs, resource ledger, declared limitations, builder and reproducer identities, tolerance and truth status. - **Measurement rules:** every task attempted (a skipped task scores zero); fixed seeds and pinned versions; the builder never scores its own run and a non-builder reproduces it within a declared tolerance; truth status is one of verified, observed, simulation, prediction or unknown; competitor comparisons run on the competitor's working product with its version and date. - **Gaming voids a result:** favourable subset selection, undeclared retries, tuning on the test set, unreported failures, self-scoring, or changing the acceptance test after the run. - **Score records** (`old-repo/handover/benchmark-standard/SCORE-RECORD-SCHEMA.md`, 30 October 2026): a published best is "observed" only when value, system, date and conditions were all read from the fetched page; otherwise "unknown" with a reason; nothing is filled from memory. - A public **Benchmark Portal** was specified (`old-repo/handover/benchmark-standard/PORTAL-SPEC.md`, 30 October 2026) in which the estate is an ordinary registrant with no scoring privilege, the builder may not be the reproducer and certification needs an accredited assessor's signed record. It was not shown to be operating. The earlier repositories recorded the same discipline in different words on 13 October 2026: report denominators, failures, uncertainty and baselines; never publish a smoke run as official; separate task success from policy compliance; a forbidden action that succeeds fails the security gate (`legacy/edge-judiverse-tech/docs/assurance/FRONTIER-BENCHMARK-RESEARCH-2026-10-13.md`), and "there is no single public score proving the estate exceeds every fraud or scheduling product" (`legacy/edge-judiverse-tech/docs/assurance/uk-readiness-and-benchmarks.md`). ###### 3.2 The benchmark register in numbers Recorded in `old-repo/handover/benchmark-standard/COUNT.json` on 30 October 2026 at 19:49 UTC (historical): 1,704 rows — published benchmark 1,245; regulatory or standard 151; competitor parity 119; taxonomy 110; depth, maturity and reliability 25; owner superiority question 21; standards foundation 15; estate gate 11; resource-to-outcome 7. Required scores known 392, unknown 1,312. In the later count (`old-repo/handover/benchmark-standard/COUNT.265e021f.json`, 30 October 2026 at 20:11 UTC, historical): of the 1,245 published-benchmark score records, 43 were observed at source and 1,202 were unknown with a reason (1,173 because the row is a specification with no leaderboard). Status "partial". By lane (from parsing all 1,245 score records): cyber 892 (NIST SP 800-53 baselines, OWASP ASVS levels and MITRE ATT&CK techniques), office 200 (mostly Microsoft 365 service limits), public-learning 51 (GDS Service Standard and Technology Code of Practice points, 19 BEIR retrieval sets, KILT, FEVER, Open Badges, xAPI), ai-lab 49, clinical 38, bank 13, patterns 2. The 11 estate gates (proposed, not met), recorded in `old-repo/handover/benchmark-standard/BENCHMARK-REGISTER.jsonl` on 30 October 2026 (historical): 100 per cent of applicable parity obligations passed; zero dead controls; 99.99 per cent monthly end-to-end availability; every declared safety-critical case passing; no acknowledged durable transaction lost in tested failures; RPO 5 minutes and RTO 15 minutes (provisional, tier-specific); zero unresolved release-blocking critical vulnerabilities with applicable OWASP ASVS 5.0 requirements verified; published per-journey latency, throughput and concurrency budgets; complete journeys on web, desktop, iOS, Android and APIs; reproduction by an agent other than the builder; at least 90 days of production observation before claiming sustained maturity. The 15 standards-foundation rows were ISO/IEC 25010:2023, 25023, 25040, 42001 and 23894; NIST AI RMF with the GenAI profile; OWASP ASVS 5.0; WCAG 2.2; ISO/IEC 21031:2024 (Software Carbon Intensity); FCA operational resilience; NHS DCB0129 and DCB0160; UK Core FHIR; GHG Protocol; ISO 14064-2. Every one of the 1,704 benchmark-graph rows was `NOT_YET_DEMONSTRATED` (1,689) or `UNKNOWN` (15), recorded in `old-repo/handover/journeys-prospective/BENCHMARK-GRAPH.jsonl` on 30 October 2026 (historical). ###### 3.3 Published best scores observed at source Scored by the benchmark's publisher; read at source by old-estate agent readers. Every row was recorded in `old-repo/handover/benchmark-standard/scores/.json` on 30 October 2026 (historical). "System" is the system the publisher reported, not the old estate. | Benchmark (what it measures) | Version or board | Published best, as read | System and publisher's date | |---|---|---|---| | AgentDojo (prompt-injection security of tool-using agents) | results table, no defence | 88.66 per cent utility without attack; 77.27 per cent under attack | claude-3-7-sonnet-20250219, 24 February 2025 | | BFCL (function calling) | v4 | 77.47 per cent overall accuracy | Claude-Opus-4-5-20251101 (FC), 12 April 2026 | | FinanceBench (financial question answering) | paper abstract, 150-case sample | 81 per cent of questions answered incorrectly or refused (headline failure rate) | GPT-4-Turbo with retrieval, 20 November 2023 | | GAIA (general assistant tasks) | test split | 94.02 per cent accuracy | Shadow Harness V8.01, 12 October 2026 | | MCP-Atlas (MCP tool use) | April 2026 update | 88.1 per cent pass rate | Muse Spark 1.1, 8 April 2026 | | MCPMark Verified | July 2026 board | 96.06 per cent pass@1 | kimi-k3-max, 20 July 2026 | | MedHELM clinical suite | Nature Medicine paper | 66 per cent win rate (model jury) | DeepSeek R1 and o3-mini, 1 March 2026 | | MedHELM public | v5.0.0 | 0.652 mean win rate | Gemini 3.1 Pro (Preview), 14 May 2026 | | Medmarks suite and 28 member benchmarks | leaderboard snapshot 30 October 2026 | suite win rate 0.645; for example MedQA 0.9539, HealthBench 0.6039, MedCalc-Bench 0.8409, MedCaseReasoning 0.526 | Gemini 3 Pro Preview, GPT-5.1, GPT-5.2 and others | | MedXpertQA | main board | 49.89 per cent average (text and multimodal) | o1, December 2024 | | PubMedQA | reasoning-required setting | 82.0 per cent accuracy | GPT-4 (Medprompt), 28 November 2023 | | SWE-rebench (fresh software-engineering tasks) | window 15 May to 1 July 2026 | 64.5 per cent resolved (±1.41) | Fable 5 [high] | | TAT-QA (table and text financial QA) | test board | 81.4 exact match (F1 88.4) | TAT-LLM (70B), 20 January 2024 | | Terminal-Bench | home leaderboard | 64.85 per cent accuracy | Claude Code with Opus 5.5, 22 September 2026 | | TheAgentCompany (simulated office work) | board embedded in site | 52.4 per cent score | TTE-MatrixAgent + DeepSeek-V3.2, 10 November 2025 | | Vals Finance Agent | v2.1, test 449 | 67.85 per cent accuracy | Claude Opus 5.5, 29 October 2026 | Further frontier figures were recorded on 29 October 2026 (historical) in `old-repo/benchmarks/frontier/discover/autonomous-ai.json`, `old-repo/benchmarks/frontier/discover/engineering.json`, `old-repo/benchmarks/frontier/discover/fintech.json`, `old-repo/benchmarks/frontier/discover/healthtech.json` and `old-repo/benchmarks/frontier/discover/regtech.json`, for example: tau2-bench text pass^1 87.9 per cent (Qwen3.5-397B-A17B) and tau3 banking 55.2 per cent (Qwen 3.8 Max); SWE-bench Verified 97.00 per cent on an independent board (Claude Opus 5, 1 September 2026); SWE-Bench Pro public 61.50 per cent (Muse Spark 1.1); Terminal-Bench 2.0 73.20 per cent (archived by its evaluator as saturated); HealthBench Professional 66.0 per cent length-adjusted (GPT-6 Astra, independent evaluator, 19 October 2026); MedQA 96.52 per cent (o1, independent evaluator sample, 16 April 2026); LegalBench 88.56 per cent (Claude Fable 5, page updated 13 October 2026); AIR-Bench 2024 refusal rate 0.9315 (Claude 4.5 Haiku); TPC-C 2,055,076,649 tpmC (Alibaba Cloud PolarDB Limitless, 27 January 2025). Many entries in those files are labelled secondary, aggregator or unverified. An earlier reading on 13 October 2026 (historical), recorded in `legacy/edge-judiverse-tech/docs/research/2026-10-13/01-agentic-coding-benchmarks.md`, set "frontier thresholds" (for example SWE-bench Verified official Bash Only at least 75 per cent; tau2 at least 85 per cent; MCP-Atlas at least 85 per cent; METR 50 per cent time horizon at least 12 hours) and what a sovereign UK platform could credibly commit to (for example tau2 at least 80 per cent pass^1; Terminal-Bench 4.0 at least 40 per cent). Research note 8 of the same date, recorded in `legacy/edge-judiverse-tech/docs/research/2026-10-13/08-financial-intelligence-benchmarks.md` (historical), found that no public financial-intelligence benchmark is solved: best verified analyst-work agent score 65.40 per cent, policy-bound banking service 55.2 per cent, and the strongest published laundering detector reached minority-class F1 68.16 per cent on the easiest AMLworld set and 36.07 per cent on a hard one. ###### 3.4 What the old estate itself measured (all CLAIMS; none independently accepted) | Run | Result as recorded | Who ran it | Recorded in, on (historical) | |---|---|---|---| | Frontier two-arm runs (the same model called directly, and through the old estate's controls), six suites | AgentDojo full set (1,014 tasks): attack success 1.775 per cent baseline and 1.874 per cent with controls; utility fell from 82.150 to 75.740 per cent. MedQA (1,273): 92.773 and 92.302 per cent. tau2 banking (97 tasks per arm): 27.957 and 34.043 per cent. AgentHarm benign mean score 0.6532 and 0.4533. Token-priced cost estimate £61.5383 for the six, £77.0235 for all runs. All published comparisons "NOT_COMPARABLE"; no ranking or maturity claim | An old-estate agent completing inherited runs (not an independent party) | `old-repo/benchmarks/frontier/codex-20261030/RESULT.md` and `old-repo/benchmarks/frontier/codex-20261030/COMPLETION-STATE.json`, 30 October 2026 | | Three hosted model endpoints, regression set (20 tasks, three trials) | Tasks passed on all three trials: 3, 15 and 6 of 20 (15, 75 and 30 per cent); capability probes 6, 13 and 9 of 14; the verifier-integrity probe failed on all three; the estate's own model endpoint had no baseline | The old estate's harness, run by its builders | `legacy/edge-judiverse-tech/docs/engineering/INTENJUDE-BENCHMARK-REGISTER.md` and `legacy/edge-judiverse-tech/registry/baselines/hosted-b.json` (with a and c), 13 October 2026 | | Live service sample | 100 of 100 HTTPS requests returned success across 5 paths; p50 296 to 444 ms and p95 517 to 816 ms; 13 releases on 13 October 2026; 18 of 56 risk and resilience benchmarks had a reading | Builders, one client in one place | `legacy/edge-judiverse-tech/docs/benchmarks/RISK-AND-RESILIENCE-BENCHMARKS.md`, 13 October 2026 | | Financial-intelligence unit tests and regression tasks | Scam interception 12 of 12 scenarios gave the expected decision; regression task R-010 (structuring request) passed 1, 3 and 0 of 3 trials on three endpoints; 18 entries with results, none passing the six required gates | Builders | `legacy/edge-judiverse-tech/docs/regulation/FINANCIAL-INTELLIGENCE-BENCHMARKS.md`, 13 October 2026 | | Laundering typology detector on self-generated synthetic data (20,000 accounts, 145,234 transactions, seed 7) | Any laundering: precision 0.93, recall 0.809, false-positive rate 0.01412; F1 1.0 on cycle and scatter-gather, 0.0 on stack and 0.008 on gather-scatter | Builders | `old-repo/fc/results/TYPOLOGY-BENCHMARK-dev-seed7.json` (date not stated in the file) | | Card fraud on IBM TabFormer (Apache-2.0 data) | Published baseline from the NVIDIA blueprint notebook: test AUPRC 0.1424 and AUROC 0.9914. The old estate's two arms: test AUPRC 0.0747 and 0.0934, AUROC 0.9582 and 0.9592, i.e. below the published baseline; two GPU arms not run | Builders, on a laptop CPU | `legacy/github-metadata-20261030/local-working-changes/canonical-local/files/fc/results/CARD-FRAUD-BENCHMARK-20261028T020325Z.json`, 28 October 2026 | | Hard tasks v1 (60 real requests the assistant had failed, judged by a model) | Strict pass 5 of 60 (8 per cent) without a "teacher note", 4 of 60 (7 per cent) with it | Builders; model judge | `legacy/jude-mvp/judeai-hosted/benchmarks/hard-tasks-v1/README.md`, 8 October 2026 | | Hidden research benchmark (30 PhD-level items) | Only a null solver (0 of 30) and a reference solver (30 of 30) were recorded; no model run | Builders | `legacy/judiverse-next/docs/research/evals.md`, 25 September 2026 | | Ontology and knowledge-plane checks (14 properties) | 14 of 14 passed, including two identical ontology builds and 4,455 of 4,455 shards verified; of 3,523 labels, 479 resolved | Builders | `old-repo/benchmarks/BENCHMARKS-RESULT.json` (source probe dated 29 October 2026) | | Mail workflows (39 planned) | 12 workflow groups marked "pass" with screenshots and API logs | One agent acting as a test mailbox | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/stations/mail/MAIL-CRL9-PROOF.md`, 18 October 2026 | ###### 3.5 Internal benchmark catalogues (targets only, unless stated) - intenJude register: 34 benchmarks with minimum-to-ship and best-in-class targets (for example tau2 at least 80 per cent pass^1, AgentDojo attack success at most 5 per cent at one attempt, pass^3 at least 98 per cent), recorded in `legacy/edge-judiverse-tech/docs/engineering/INTENJUDE-BENCHMARK-REGISTER.md` on 13 October 2026 (historical). - Financial-intelligence register: 58 benchmarks in 10 groups (19 external, 39 built by the estate), recorded in `legacy/edge-judiverse-tech/docs/regulation/FINANCIAL-INTELLIGENCE-BENCHMARKS.md` on 13 October 2026 (historical). - Risk and resilience: 56 benchmarks (governance, delivery, availability, recovery, security response, agent safety and reversibility, third party, people), recorded in `legacy/edge-judiverse-tech/docs/benchmarks/RISK-AND-RESILIENCE-BENCHMARKS.md` on 13 October 2026 (historical). - Agent governance: 12 benchmarks (audit completeness, kill-switch latency, once-only approval, payload binding, mandate forgery, data denial, chain tamper, two reserved replay and reversal entries, AgentDojo, AgentHarm, tau2), recorded in `legacy/edge-judiverse-tech/docs/assurance/BENCHMARK-REGISTER.md` on 13 October 2026 (historical). - UK protocols BENCH-UK-01 to 08 (fraud triage, entity and network evidence, time to evidence, recovery authorisation, operational resilience, procurement interoperability, accessible native and web journeys, agent execution), recorded in `legacy/edge-judiverse-tech/docs/assurance/uk-readiness-and-benchmarks.md` on 13 October 2026 (historical). - Clinical evaluation suite CE-01 to CE-14 (image fidelity, cancer detection, screening benefit, calibration, human-AI teams, trial eligibility, genomics, evidence retrieval, record reconciliation, clinical load, accessibility, adaptive upcycling, sequential monitoring, public benefit), recorded in `legacy/edge-judiverse-tech/docs/assurance/clinical-evaluation-register.md` on 13 October 2026 (historical). It warns against arbitrary universal targets such as "99 per cent accuracy" or "zero missed cancers". - Standards catalogues per domain (bank 32, agent estate 28, clinic 23, government 22, consortium 21, mail 21, Teams 14 entries), recorded in `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/packs/*/benchmarks.json` on 22 October 2026 (historical). - Office-task evaluations cited as published references, recorded in `legacy/edge-judiverse-tech/docs/benchmarks/APP-GENERATOR-AND-OFFICE-BENCHMARKS.md` on 13 October 2026 (historical): SpreadsheetBench (912 questions); OfficeBench (best 47.00 per cent pass rate); PPTC (75.1 per cent turn accuracy against 6 per cent session accuracy); WebDev Arena. - A measurement protocol for "time to verified outcome", recorded in `legacy/local-source-snapshots/programme/outcomes-register/BENCHMARK-PROTOCOL.md` (date not stated): report median, p90 and p95 among verified cases with the unresolved cohort alongside. - Not runnable, as recorded in `old-repo/benchmarks/frontier/GAP.json` on 29 October 2026 (historical): Vals private sets; HealthBench Professional; MedArena; OSWorld (no vision model available); METR; AILuminate official sets; official TPC-C (audit required); PaySim (rejected because "balance columns leak the label"); and standards such as ISO/IEC 42001 and DCB0129 ("standards and assessments, not scored benchmarks"). ###### 3.6 Which benchmarks apply to the four applications | Application | Benchmarks and baselines the old estate mapped or researched | |---|---| | **EDGE Finance** | tau2/tau3 banking; Vals Finance Agent v2.1; FinanceBench; FinQA; ConvFinQA; TAT-QA; FinSearchComp; FinBen credit tasks; IBM AMLworld; AMLSim; SAML-D; Elliptic; IBM TabFormer card fraud; ISO 20022 message validation (no recognised benchmark exists); BFCL for bank APIs; AgentDojo for payments; FAPI 1.0 and 2.0; Open Banking UK 4.0.1; PSR APP, FOS and FCA complaints data; FCA operational resilience and EU DORA incident clocks; PCI DSS; the 40-metric bank scorecard; the 58-entry financial-intelligence register | | **EDGE Clinic** | HealthBench (and Professional); MedHELM; Medmarks and its members (MedQA, MedMCQA, PubMedQA, MedXpertQA, MedCalc-Bench, AgentClinic, LongHealth and others); MedAgentBench (FHIR agent); Inferno ONC (g)(10); HL7 FHIR UK Core validation; Synthea synthetic patients; DCB0129 and DCB0160; DTAC; DSPT; NICE evidence standards; clinical suite CE-01 to CE-14; NHS Faster Diagnosis Standard. UNKNOWN: no benchmark specific to social prescribing was found | | **EDGE Public Consortium** | LegalBench; LEXam; ObliQA; PRBench Legal; AIR-Bench 2024; HELM Safety; AILuminate; AgentHarm and StrongREJECT (UK AISI Inspect); b3; CyberGym; Cybench; OWASP ASVS 5.0; CIS Benchmarks; OpenSSF Scorecard; NIST SP 800-53 baselines; MITRE ATT&CK; NCSC CAF 4.0; GovAssure; JSP 440; Def Stan 05-138; Cyber Essentials Plus; GDS Service Standard and Technology Code of Practice; Procurement Act 2023; BEIR, KILT and FEVER for research retrieval; Open Badges and xAPI for learning records | | **Judiverse** | SWE-bench Verified and Pro; SWE-rebench; Terminal-Bench 2.x and 4.0; GAIA; WebArena-Verified; OSWorld; TheAgentCompany; MCP-Atlas; MCPMark; BFCL v4; MultiAgentBench; METR time horizons; the HAL methodology; SpreadsheetBench; OfficeBench; PPTC; Microsoft 365 service limits; Core Web Vitals; WCAG 2.2; DORA delivery metrics; MLPerf Inference and Endpoints | --- ##### 4. Readiness definitions and how readiness was to be proven ###### 4.1 Scales recorded - **TRL (technical).** UKRI STFC wording (page last updated 12 May 2025), NASA NPR 7123.1D Appendix E and the EU Horizon 2020 Annex G, all retrieved on 29 October 2026 and stored with hashes, recorded in `old-repo/engine2/trlcrl/criteria.json` on 29 October 2026 (historical). TRL 9 is "actual technology qualified through successful mission operations"; TRL 8 "actual technology completed and qualified through test and demonstration". - **CRL (commercial).** BINDT/RCNDE 1 to 9 wording in the same file: CRL 1 "basic hypothesis that a new technology would be commercially viable", CRL 5 "financial model for sales, costs and margins established", CRL 7 "financial model/commercial viability validated", CRL 8 "first commercial system built; route to market established", CRL 9 "technology has been introduced to market". Other CRL scales recorded: the old estate's nine-gate **CRL-ESTATE-1.0** (`legacy/edge-judiverse-tech/docs/assurance/uk-readiness-and-benchmarks.md`, 13 October 2026), which also notes that CRL here means Commercial, not Clinical, Readiness Level; the NITI Aayog nine-level scale and six-level frameworks (`legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/aec9/TRL9-CRL9-PREDECLARED-RECEIPTS.md`, date not stated); and a warning that the Australian ARENA CRI tops out at level 6, so "quoting CRL 9 against ARENA uses the wrong scale" (`legacy/peak-judeai-behavioural-harness-baseline-full-application/docs/VALUE-MAP.md`, 20 October 2026). - **AEC.** "AEC n = TRL n (UKRI) + CRL n (UK CRL scale, BINDT/RCNDE), done by agents"; AEC 9 adds "exhaustively evidenced with independently verifiable and reproducible proof" (`legacy/x-judeai-edge-tok-kol-ndc-ml-x/handoffs/AEC-STANDARD.md`, 21 October 2026). A later statement kept TRL, CRL, AEC and DECFAW 9 as separate records with no automatic conversion between them (`legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/ESTATE-LEVEL-READINESS.md`, date not stated). - **DECFAW 9.** An internal acceptance target (zero dead controls; independent proof). Its exact frozen definition was recorded as pending, and DECFAW 10 as undefined (`old-repo/handover/source-context/ROUTE-TO-9-20261029T2335Z.md`, 29 October 2026). A runnable "zero dead control" gate was specified, release status "blocked until independently demonstrated otherwise" (`legacy/edge-judiverse-tech/docs/assurance/DECFAW9-CONTROL-CLOSURE.md`, date not stated). - **Operational readiness OR-1 to OR-9** (workflows, detection, resilience, response, authority and receipts, sign-in "doors", records, complaints, evidence), declared by the Owner on 20 October 2026 together with the statement that "a convincing simulation alone can establish operational readiness" (`legacy/local-source-snapshots/programme/OPERATIONAL-READINESS-DECREE-20261020.md`). This was later reversed (see 4.4). ###### 4.2 Evidence required per level (the most reusable artefact) Recorded in `old-repo/engine2/trlcrl/criteria.json` on 29 October 2026 (historical); award status "computed assessment, not an award"; an award needs a competent assessor. | Level | Evidence required (summarised) | |---|---| | TRL 1 to 2 | A source-bound definition; outcome, ordered steps, actor and receiver named | | TRL 3 | Code coverage bound to a module hash; unit tests run and all passed | | TRL 4 | At least one case through prepare, validate, commit and receive with accepted receipts; relevant environment defined | | TRL 5 | End to end through the signed agent door, with a receiver different from the preparer and committer; receipt chain verified; operating rate predicted | | TRL 6 | Population-scale run with dataset licences recorded; race-track trials in five variant classes (normal, hard, multi-step, random, human error); an unauthorised call refused; a recovery executed and verified | | TRL 7 | Three or more distinct cases on the operational platform bound to a release; sustained frontier load for at least 300 s with every SLO met; one-sided 95 per cent Clopper-Pearson upper bound on the failure rate below a declared target | | TRL 8 | All evidence bound to the same release digest; independent reproduction by a different actor (re-check or isolated replay, never a duplicated effect); fault recovery; user, training and maintenance documentation; reconciliation | | TRL 9 | Three or more genuine operations by an organisation other than the developer, in production, on the TRL 8 release, with authoritative receiving receipts; sustaining support in force | | CRL 2 to 4 | Market awareness with dated sources; real application identified; supply chain, certification and regulatory requirements listed | | CRL 5 | Price from an accepted quote (no invented price) and measured unit cost | | CRL 6 | Exploitation route; signed external partnership; certification process acknowledged by the competent body | | CRL 7 | Evaluation by a real organisation under an evaluation licence, with unit cost at or below price | | CRL 8 | Published artefacts with digests, evaluation licence, billing through a real payment processor (fixtures are not payment), route to market | | CRL 9 | An external organisation's authorised assent and a payment receipt or signed contract, with entitlement read back | Validity rule in the same file: a record counts only if it passed, is not a fixture, has not expired and, for adoption, mission and commercial kinds, is not synthetic. The scope file applied this per application and listed who could produce each record: agents alone, the organisation as its own first customer, or an external counterparty (`old-repo/scope/SCOPE-T9C9-20261029T230350Z.md`, 29 October 2026). ###### 4.3 Other gate designs recorded - **Execution gate for TRL 8, TRL 9 and CRL 9**, recorded in `old-repo/spec/TRL9-CRL9-GATE.md` on 27 October 2026 (historical): TRL 8 on offline replicas of the frozen release, run by a prover who is not the builder; TRL 9 only from receipts of the organisation's own real work in production, observed by a separate receiver, with no open incident; CRL 9 requires three independent agent families reproducing the TRL 8 run with identical verdicts plus signed Owner acceptance. Its inventory: 28,648 screens, 262,666 bound controls, 115,280 executors, 71,685 journeys and 372,208 steps. Status: 25,840 screens passing on replicas; 33,107 executor pairs with receipts, all from agent identities; **0 operational receipts from the organisation's own identities**; 0 journeys evaluated. Worst-case cost of a three-family run £45.80. - **Annex P readiness pathways**, recorded in `old-repo/spec/ANNEX-P-PATHWAYS.md` (date not stated; its variant `old-repo/spec/ANNEX-P-PATHWAYS.16869b95.md` has a shorter screen check): receipts fixed in advance per item kind (screen, control, executor, journey step, journey, capability) and seven ordered gates (identity and scope, intended behaviour, safety and recovery, TRL 8 qualification, TRL 9 operation, CRL 9 commercial proof, final acceptance). "Evidence never substitutes across levels"; levels only rise, and a regression returns an item to the last level whose receipts still verify. - **Journey contracts and six release gates C1 to C6** (complete enumeration, genuine TRL 9, genuine CRL 9, scoped law and standards, independent original-source verifier, complete signed evidence), recorded in `old-repo/docs/ACTUAL-TRL9-CRL9-JOURNEY-CONTRACTS.md` on 27 October 2026 (historical). It records a conflicting journey count (73,872) and says the conflict must be settled by source hashes, not by choosing a convenient figure. - **Node oracles**, recorded in `old-repo/docs/ORACLES.md` on 27 October 2026 (historical): four states per node that are never merged; "an HTTP 200 does not prove an outcome; a signed receipt does not prove an outcome; a self-stamp does not prove an outcome". - **Predeclared receipts**: TRL 7, TRL 8, TRL 9 and CRL 9 stages per node; even when all are evidenced the result is only "ready for independent issuer" (`legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/aec9/TRL9-CRL9-PREDECLARED-RECEIPTS.md`). 788 such contracts for 394 clinic items were compiled with every actor binding "not proven by contract" (`legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/packs/clinic-autonomous-proofs/compiled/proofs/`, 22 October 2026). - **What CRL 9 requires of every app**, recorded in `legacy/local-source-snapshots/map/crl9/JudeAI-EDGE-ID-CRL9-20261015.md` on 15 October 2026 (historical): validated market, successful pilots, applicable authorisations, secure production operations, customer acceptance, repeatable contracted deployments, sustainable economics, renewals and sustained independently reviewed performance, plus twelve proofs (complete product, independent technical verification, scientific validity, security, resilience, data protection, AI governance and error control, environmental performance, accessibility, operational deployment, independent customer acceptance, commercial readiness). It also named who must accept per station: FCA and PRA for the bank; MHRA and NHS England for clinical; the MOD Cyber Security Model, NCSC and the Export Control Joint Unit for defence; the Procurement Act 2023 for procurement; ICO and the Online Safety Act for the assistant. - **The CRL-ESTATE-1.0 gates**, recorded in `legacy/edge-judiverse-tech/docs/assurance/uk-readiness-and-benchmarks.md` on 13 October 2026 (historical): from a named user problem (1) to sustained commercial operations with independently reviewed outcomes and renewals (9), mapped to an eleven-pack submission evidence index (UK-PACK-01 to 11) with a named independent reviewer per pack. - **Lessons from a failed TRL 8 gate run**, recorded in `old-repo/docs/incidents/INC-EST-gate-run-isolation-2034.md` on 27 October 2026 (historical): shared state written without a lock; reused workspaces without a run namespace; processes stopped by name pattern; evidence not bound to its run; production dependencies inside offline replicas. Each became a preflight refusal. - **Readiness uplift for applicants**, recorded in `legacy/jude-mvp/docs/readiness-uplift.md` (date not stated): a tool to move applicants from TRL 4 towards TRL 8 before submission ("government buys the ability to develop and test more viable solutions"). ###### 4.4 Readiness actually recorded (all historical CLAIMS) - Computed levels over 5,402 to 5,588 workflows on 29 October 2026: a mix of TRL 1 to 3 at 20:47 UTC; all at TRL 3 at 22:07; 2,650 to 2,792 at TRL 3 and 2,796 at TRL 5 at 23:05 and 23:16; CRL 1 for every workflow throughout. The main blockers were market awareness (CRL 2), lifecycle evidence (TRL 4), race tracks and recovery (TRL 6). Recorded in `old-repo/engine2/proof/trlcrl/run-2026-10-29T204735/summary.json`, `old-repo/engine2/proof/trlcrl/run-2026-10-29T220716/summary.json`, `old-repo/engine2/proof/trlcrl/run-r4-20261029T2305Z/summary.json` and `old-repo/engine2/proof/trlcrl/run-2026-10-29T231613/summary.json` on 29 October 2026. The EDGE Bank submission extract (`old-repo/engine2/proof/submission/TRL-CRL-EDGE-BANK-r4.json`, same date) carries the same per-workflow levels (TRL 3 or 5, CRL 1). - Levels reachable, recorded in `old-repo/scope/SCOPE-T9C9-20261029T230350Z.md` on 29 October 2026 (historical): TRL 8 and CRL 4 by agents alone on a synthetic population; TRL 9 and CRL 5 to 6 only with the organisation as the real customer; standard-route TRL 9 and CRL 6, 7 and 9 only with external counterparties. The same file found a defect: the decision tool defaulted to an older definitions file, under which 1,174 completed rows fell to TRL 1. - Level register of 20 October 2026: 4,038 items, none at TRL 9 or CRL 9, every item marked "funding risk RED"; ten Bank apps at AEC 5 (`legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/map/LEVEL-REGISTER-20261020.csv`; the copy in `legacy/local-source-snapshots/programme/map/LEVEL-REGISTER-20261020.csv` differs only in one workflow row). - Outcomes register: 838 outcomes, 797 at AEC 1, verification pending on all but one (`legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/outcomes-register/register.json`, entries captured from 18 October 2026). - Self-assessed bands on 20 October 2026: authority, passports and receipts TRL 6 to 7; cross-institution protocol TRL 4 to 5; scale TRL 3 to 4; agent delivery TRL 5 to 6; commercial CRL 1 to 2; "TRL 9 and CRL 9 are not met anywhere" (`legacy/peak-judeai-behavioural-harness-baseline-full-application/docs/VALUE-MAP.md`). - CRL 9 count on 15 October 2026: 81 apps, 8 live, none at CRL 9 (`legacy/local-source-snapshots/map/crl9/JudeAI-EDGE-ID-CRL9-20261015.md`). - Horizon baseline on 24 October 2026: 5,348 workflow records in the proof gate, 5,331 recorded as "admitted and frozen at DECFAW 9", 17 below 9 (`legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__BASELINE-20261024.md`). This internal level is not TRL or CRL; the 27 October gate above found zero operational receipts from real identities. - **A conflict and how it resolved.** The 20 October decree said a simulation alone could establish operational readiness. The 29 October criteria reversed this: virtual staff and customers of synthetic organisations "are fixtures for adoption purposes and never count", and earlier readiness claims "are not inherited" (`old-repo/engine2/trlcrl/criteria.json`). Owner statements of 31 October 2026 then treated TRL 9 and CRL 9 as build requirements for a new build, with earlier work as lineage only (`old-repo/plan/sources/owner-build-requirements-live-agents-20261031.json` and `old-repo/handover/tok-20261030/specification/authoring-20261031/BRIEF-ADDENDUM-3-NEW-TRL9.md`, 31 October 2026), and stated that agents are native users who must be able to do everything humans can. --- ##### 5. Reuse in the new estate ###### 5.1 The most valuable findings to carry forward (as research, with new identities) 1. **The claim discipline of the benchmark standard** (section 3.1): non-substitutable claim classes, a mandatory run record, non-builder reproduction within a declared tolerance, "unknown until read at source", and the gaming conditions that void a result. This is method, not data, and does not go stale. 2. **The level-by-level TRL and CRL evidence criteria** (section 4.2), including the rules that fixtures and synthetic populations never count as adoption, that prices come only from accepted quotes, and the Clopper-Pearson bound for failure rates. They fit a design where agents are native customers and staff: agents can produce TRL 4 to 8 evidence themselves, but TRL 9 and CRL 6 to 9 need real receiving organisations. The ARENA and Clinical-versus-Commercial cautions should travel with them. 3. **The 43 published-best scores read at source on 30 October 2026**, especially for EDGE Finance (Vals Finance Agent, TAT-QA, FinanceBench), EDGE Clinic (MedHELM, Medmarks members, PubMedQA, MedXpertQA) and Judiverse agents (GAIA, MCP-Atlas, MCPMark, BFCL, Terminal-Bench, SWE-rebench, TheAgentCompany, AgentDojo). Leaderboards move; each must be re-read at source before use. 4. **The UK bank competitor baseline of 13 October 2026** (section 2.2): regulator-published FOS, FCA complaints, PSR APP, outage and accessibility figures per bank, and the 40-metric scorecard with its finding that no bank publishes availability, support response time, false-hold rate or agent task success. It is the strongest competitor research for EDGE Finance and almost all of it is re-checkable at the regulator. 5. **The Palantir public-procurement register and capability map** (27 and 30 October 2026) and the deterministic Palantir/Incogni count with its measured 11.7 per cent over-count. Directly relevant to EDGE Public Consortium (defence, government and health data platforms) and to EDGE ID privacy services. 6. **The clinic bars** (NHS App statistics, Epic, EMIS, SystmOne, Accurx, Doctolib, Huma, Cera, NHS national services) and the clinical evaluation suite CE-01 to CE-14 with its warnings about arbitrary targets, data leakage and prevalence. 7. **Honest negative results as baselines**: the controls arm cut AgentDojo utility by about 6 points without lowering attack success; the card-fraud arms fell below the published baseline; the hosted endpoints failed verifier integrity; zero operational receipts came from real identities. These are what a new build should expect to beat, not assume beaten. 8. **The runnability map** (what can be run, what needs a licence, a GPU, an external evaluator or membership) and the external recognition the old estate said it needed: independent reproduction outside the organisation, submissions to official boards, UK AISI review of Inspect results, evaluators such as Vals, Artificial Analysis, Stanford CRFM and MLCommons, ISO/IEC 42001 certification by an accredited body, Clinical Safety Officer sign-off, and FCA, PRA, PSR and ICO engagement (`old-repo/benchmarks/frontier/GAP.json`, 29 October 2026). 9. **Operational lessons from the failed gate run** (run namespaces, locks, evidence bound to its run, no production dependency inside replicas). ###### 5.2 Superseded or unsafe to reuse - Every **parity percentage produced by keyword matching** (the 86 to 94 per cent screen-flow and documentation figures of 29 October 2026, and the COVERED/PARTIAL text-rule coverage in the scope file). They measure vocabulary overlap with old definitions, not working parity. The 23:00 run is superseded by the 23:25 run; both are superseded as a method. - The **horizon "live" counts** of 24 October 2026 where "live" counted contract lines (for example 962 of 1,147 Government features); the summary itself recomputes 20 per cent live without them. - The **operational readiness decree** of 20 October 2026 (simulation alone suffices): reversed by the 29 October criteria. - The **AEC combined scale**, the 4,038-row level register, the outcomes register levels and the internal "DECFAW 9" counts: tied to old identities and rules; later files say earlier readiness claims are not inherited. - **Old inventory counts** of screens, controls, executors and journeys (28,648; 262,666; 115,280; 71,685, with a conflicting 73,872): an inventory of the old build, not of the new estate. - **13 October 2026 frontier figures** where a later reading exists (Terminal-Bench 2.0 recorded as saturated and archived by its evaluator; tau2 moved to a 2.0 line with a banking grading change). - **The owner superiority questions** in the register ("the best in the world", "the single source of truth for all human domains"): recorded only as questions to be answered by measurement against pinned comparators; never as claims. - Music, DAW and JudeBox-against-Spotify parity, and app-generator pricing: low relevance to the four applications and dated September to October 2026. - Anything naming hosts, server addresses, mailboxes, remote-shell commands or local machine paths: contains sensitive data; not reproduced, and not to be reused. ###### 5.3 Gaps the new estate must research afresh Social prescribing competitors and outcome benchmarks (UNKNOWN in the corpus); education and engineering incumbents (recorded as not measured); signed-in competitor product behaviour (most readings were of public documentation); measured availability for any UK bank (none publishes it); Gotham user functions (not publicly documented); and any benchmark run by a party independent of the old estate (none existed). ### research/legacy/STANDARDS-PAPERS-GUIDES.md `research/legacy/STANDARDS-PAPERS-GUIDES.md` · sha256 `9e98dcbe938cf7f74a2ff5e49d84b1b1114094e0cb246c5e70992f2aa349576c` #### Legacy research: standards, regulations, papers, manuals and guides **What this is.** Research notes for the new estate, JudeAI | EDGE ID (EDGE Finance, EDGE Clinic, EDGE Public Consortium and Judiverse). They summarise what the organisation's old estate wrote about standards and regulation, the papers and white papers it produced, and the manuals, handbooks, guides and playbooks it kept. **What this is not.** It is not authority. Nothing here is a rule, plan, approval, readiness level or identity for the new estate. Two labels are used throughout: - **CLAIM**: something the old estate said about itself, its products, its tests or its intentions. A CLAIM was never independently verified unless the text says so, and most were not. - **AGREEMENT**: an external law, regulation, regulator rule or published standard, with the date the old estate recorded for it. These dates were read by the old estate between 13 and 30 October 2026. Many could not be opened at source at the time (the old estate tagged them "unverified" or "secondary"). Re-check every AGREEMENT against the primary source before the new estate relies on it. Date of this report: 31 October 2026. Paths are relative to the corpus root (they start `legacy/` or `old-repo/`). ##### Scope, method and coverage **Scope.** Every catalogue row with `generated: false` whose categories include `standards-regulation`, `papers-whitepapers` or `manuals-guides`. That is 2,181 rows (887 from `legacy/`, 1,294 from `old-repo/`), 2,156 distinct blob identifiers, about 52 MB. **Triage (scripted).** 1. Rows were sorted into substantive documents and record families by path rules. 1,736 rows fall into 15 record families that are not documents in the ordinary sense: 1,245 benchmark score records, 100 synthetic exam papers, 90 publication receipts and evidence records, 74 in-app guide data files, 68 registers and data files, 56 rendered PDF editions of the regulatory pack, 20 test outputs, 19 third-party licence texts, 15 alert logs, 13 proof records, 12 statistics-handbook provenance stubs, 11 security-patrol working files, 5 copies of third-party web pages, 4 continuation-kit files and 4 rendered-answer captures. 2. The remaining 445 rows were deduplicated. Rows were treated as one document when they differ only by branch (a file name carrying an eight-character content suffix such as `.4654f34c`), or are mirror copies between repositories (`legacy/local-source-snapshots/` mirrors `legacy/x-judeai-edge-tok-kol-ndc-ml-x/`; `old-repo/` mirrors `legacy/governed-intelligence-judeai-edgeid-pre-contract/` and `legacy/governed-intelligence-judeai-edgeid/`). This leaves **357 distinct documents**: 50 papers or paper-related documents, 155 standards, regulatory or policy documents, and 152 manuals, handbooks or guides. Each variant keeps its own row and oid in the inventory. 3. Where two documents share a subject but differ in content (for example `docs/papers/risk-aware-never-risk-averse.md` and `docs/risk/RISK-01-WHITE-PAPER-...md`), they are listed separately and related in section 2. **Reading.** Reading depth varied, and is stated here so that nobody mistakes an opening-level read for a full one. - *Read in depth* (whole text, or every section that carries obligations, status and plans): the five public papers and their examinations; the research and governance white papers in section 2 (abstract, structure, findings and status sections); PART-1, PART-2 and PART-3 of the FCA and PRA framework (commitments, perimeter, build-first tables, gap tables, standards map, dates and fixes); the bank regulatory requirements (dates, perimeter, corrections, open decisions); the procurement, ISO and licence checklist; the regulatory-pack index, the 15 October 2026 review and EDGE-REG-04, 05, 07, 08, 10, 12, 21 and 29; the procurement packs (DSPT, DTAC, ISO/IEC 27001, FCA and JMLSG, NHS frameworks); the organisational-requirements audit rows; the 24 October 2026 law and standards register sets A to X and funding routes; the benchmark standard, its count file and the regulatory rows of the benchmark register; the standards and frameworks registers. - *Read at heading and opening level* (title block, status line, headings and first part of the text): every other distinct document, including the remaining EDGE-REG documents (DPIAs 01, 02, 13 to 20 and 40, hazard logs 22 to 28, the trial, processing, service-level, incident, continuity, vulnerability-disclosure, exit and acceptable-use documents 32 to 39, the online-safety documents 30 and 31), the FCA and PRA framework head document, the insurance product standard, the financial-intelligence benchmark register, the bank staff and regulator journeys, the intenJude frontier standard, the verifiable-work and render-integrity standards, the UK pack, the Challenge 1 use-case, contract-schedule and evidence-index documents, the legal web pages, the trust page, the policy drafts, the research notes R1, R2, 03, 05, 06 and 07, and every manual in section 4. - *Template families*: the 79 product manuals and 17 engineering journey manuals share one template; their status lines and purpose statements were read for all of them, and the full text of P26 and P57. - *Record families* were read by script (field counts, statuses, lanes, results) and a few files by hand. The 74 in-app guide data files were identified by name and size only. **Honest limits.** - One file, `legacy/jude-mvp/docs/reference/XX_London_Regulated_Search_700.xlsx`, has an empty text extraction. Its contents were not read. - Three HTML files have no text extraction (`old-repo/office/it-onboarding/index.html`, `old-repo/office/it-onboarding/office-shell-source.html`, `old-repo/handover/benchmark-standard/portal/board.html`). The first two have mirror copies in `legacy/governed-intelligence-judeai-edgeid/office/it-onboarding/` which were read; the board page was not. - The 56 PDFs are rendered editions of Markdown sources in the same folders. The Markdown sources were read; the PDF text was compared only by title and size. - Dates come from the document text where one is stated, else from a date in the path, else from the earliest date in the names of the branches that hold the file (marked "branch"). 12 distinct documents have no date at all. - Some catalogue categories are wrong. For example 1,245 benchmark score records are tagged `standards-regulation`, and copies of third-party web pages are tagged `papers-whitepapers`. They are inventoried as record families, not as the organisation's own papers. - Personal names, contact details, host names, network addresses, credential locations and access commands that appear in some documents are deliberately not reproduced. ##### Acronyms used | Short form | Meaning | |---|---| | 4R | The old estate's own proposed standard: retraceable, recoverable, reversible, reproducible | | AEC | "Agentic exhaustive completeness", the old estate's own nine-level readiness scale for agent-played journeys | | AML | Anti-money laundering | | APP fraud | Authorised push payment fraud | | BCBS 239 | Basel Committee principles for risk data aggregation and reporting | | CAF | NCSC Cyber Assessment Framework | | CASS | FCA Client Assets sourcebook (CASS 15 is the new safeguarding regime) | | CE / CE+ | Cyber Essentials and Cyber Essentials Plus (UK government-backed certification) | | CQC | Care Quality Commission | | CSO | Clinical Safety Officer (named clinician required by DCB0129) | | DCB0129 / DCB0160 | NHS clinical risk management standards for manufacturers and for deploying organisations | | DISP | FCA Dispute Resolution: Complaints sourcebook | | DPIA | Data protection impact assessment | | DPO | Data protection officer | | DSPT | NHS Data Security and Protection Toolkit | | DTAC | NHS Digital Technology Assessment Criteria | | DUAA | Data (Use and Access) Act 2025 | | EIOPA | European Insurance and Occupational Pensions Authority | | EMI | Electronic money institution | | FCA / PRA | Financial Conduct Authority / Prudential Regulation Authority | | FOS / FSCS | Financial Ombudsman Service / Financial Services Compensation Scheme | | GES | "Governed intelligence" specification, the old estate's own (GES-1.0, 27 October 2026) | | IAIS | International Association of Insurance Supervisors | | ICB | Integrated care board (NHS) | | ICO | Information Commissioner's Office | | ICOBS / PRIN / SYSC / PROD / SUP / COCON | FCA Handbook sourcebooks: insurance conduct; principles (PRIN 2A is the Consumer Duty); senior management systems and controls; product governance; supervision; code of conduct | | ISMS | Information security management system (ISO/IEC 27001) | | MCP | Model Context Protocol (agent tool interface) | | MHRA / UK MDR | Medicines and Healthcare products Regulatory Agency / Medical Devices Regulations 2002 | | MLR / MLRO | Money Laundering Regulations 2017 / money laundering reporting officer | | NCSC | National Cyber Security Centre | | NIST AI RMF | US National Institute of Standards and Technology AI Risk Management Framework | | OFSI | Office of Financial Sanctions Implementation | | OSA | Online Safety Act 2023 | | PECR | Privacy and Electronic Communications Regulations 2003 | | POCA | Proceeds of Crime Act 2002 | | PPN | Procurement Policy Note (Cabinet Office) | | PSR | Payment Systems Regulator | | ROPA | Record of processing activities (UK GDPR Article 30) | | RTO / RPO | Recovery time objective / recovery point objective | | SM&CR / SMF | Senior Managers and Certification Regime / senior management function | | SS1/21, SS2/21, SS1/23 | PRA supervisory statements on operational resilience, outsourcing and third parties, and model risk | | TOK | "Tree of Knowledge", the old estate's knowledge graph | | TRL / CRL | Technology readiness level / commercial readiness level | | WCAG | Web Content Accessibility Guidelines | ##### 1. Inventory Section 1.1 to 1.3 list every distinct document, with its primary path (the copy held on the most branches), its blob oid, its date and every near-identical variant with its own oid. Section 1.4 summarises the record families, and section 1.5 lists the two families that carry regulatory or benchmark substance row by row. "Date (how dated)": *text* = stated in the document; *path* = a date in the file path; *branch* = the earliest date in the names of branches that hold the file; *undated* = none found. ###### 1.1 Papers, white papers, research papers and paper reviews (50 distinct documents, 63 catalogue rows) | # | Title | Corpus-relative path | Blob oid | Date (how dated) | Near-identical variants (path, oid) | |---|---|---|---|---|---| | P1 | Trail for research file 03 (internal, not for publication) | `legacy/edge-judiverse-tech/docs/dossiers/frontier-ai/03-medicines-discovery-and-multi-omics.trail.md` | `cdc76797f203826d4192e6e57cbd422b53d0d825` | 13 October 2026 (text) | - | | P2 | Trail for research file 04 (internal, not for publication) | `legacy/edge-judiverse-tech/docs/dossiers/frontier-ai/04-advanced-materials-with-ai.trail.md` | `e4079319046a0e883019033d7e7ffb81c2f2ebec` | 13 October 2026 (text) | - | | P3 | Trail for research file 05 (internal, not for publication) | `legacy/edge-judiverse-tech/docs/dossiers/frontier-ai/05-simulation-research-and-the-compute-stack.trail.md` | `1f892d09d702418c6dff4eb920bef78a0a8180f0` | 13 October 2026 (text) | - | | P4 | Trail for 06-capability-checklist-and-funding-rules (internal, not for publication) | `legacy/edge-judiverse-tech/docs/dossiers/frontier-ai/06-capability-checklist-and-funding-rules.trail.md` | `51fbe97f8a369147bfcce7e010a4826cc849d6e8` | 13 October 2026 (text) | - | | P5 | Trail for research file 07 (internal, not for publication) | `legacy/edge-judiverse-tech/docs/dossiers/frontier-ai/07-estate-additions-and-whole-of-life-journeys.trail.md` | `58cfaea27681dfd5907430df8946af6c8139558d` | 13 October 2026 (text) | - | | P6 | file": "docs/papers/the-failures-worth-finding.md | `legacy/edge-judiverse-tech/docs/evidence/documents/2026-10-14/public-papers-examination-2.json` | `bfa94acd91e5b1dcacc7a5633628adf7778cf755` | 2026-10-14 (path) | - | | P7 | Second examination of the five public papers, 14 October 2026 | `legacy/edge-judiverse-tech/docs/evidence/documents/2026-10-14/public-papers-examination-2.md` | `f4e6db9acc145561c5e21a48c6bd69d68a76cd04` | 14 October 2026 (text) | - | | P8 | file": "docs/papers/the-failures-worth-finding.md | `legacy/edge-judiverse-tech/docs/evidence/documents/2026-10-14/public-papers-examination-3.json` | `a0ce9f963030001aacaf16ddef4446f5e4d197cc` | 2026-10-14 (path) | - | | P9 | file": "docs/papers/the-failures-worth-finding.md | `legacy/edge-judiverse-tech/docs/evidence/documents/2026-10-14/public-papers-examination.json` | `cdbae34f263445ff304d7c77855a277b5e11529c` | 1 August 2012 (text) | - | | P10 | Public papers examination, 14 October 2026 | `legacy/edge-judiverse-tech/docs/evidence/documents/2026-10-14/public-papers-examination.md` | `0c5acba595a6fa46d12236e2c3aa26dc7aa7315e` | 14 October 2026 (text) | - | | P11 | Handover: intenJude frontier standard, FCA and PRA framework, 4R white paper | `legacy/edge-judiverse-tech/docs/handover/2026-10-13-frontier-standard-and-4r-handover.12dce119.md` | `12dce119e8f9323237f724ef6d63a0af54318d22` | 13 October 2026 (text) | `legacy/edge-judiverse-tech/docs/handover/2026-10-13-frontier-standard-and-4r-handover.md` `6c6fdc5b5d0d9493a17d7ecb8f2166eb77e6aa13` | | P12 | MEET BUILD PLAN (judged and synthesised, 13 Oct 2026) | `legacy/edge-judiverse-tech/docs/meet/takeover/BUILD-PLAN.md` | `83de4f3837c98e9f43ada79950df8a1ba3238b81` | 2026-10-13 (branch) | - | | P13 | Outreach: agentic email marketing and CRM with a hold button | `legacy/edge-judiverse-tech/docs/outreach/06-WHITE-PAPER.md` | `f4f8f23b3d704001557d12fe12e1bbaf349e1077` | 2026-10-13 (text) | - | | P14 | The consent debt problem | `legacy/edge-judiverse-tech/docs/papers/outreach-with-a-hold-button.md` | `1564992fc8b0fdd50dfbe3fe4979a81d5534e707` | 13 October 2026 (text) | `legacy/edge-judiverse-tech/docs/papers/outreach-with-a-hold-button.4654f34c.md` `4654f34c1f5d99d054eab01180bdd7a79c142c63` | | P15 | The claim | `legacy/edge-judiverse-tech/docs/papers/resilience-is-the-product.md` | `cec189f3bc30508045538eb1d3e0df078c6f6546` | 13 October 2026 (text) | - | | P16 | The doctrine | `legacy/edge-judiverse-tech/docs/papers/risk-aware-never-risk-averse.md` | `a9d0ca003f17b080bacaf2d19b9dfc64bed67e56` | 13 October 2026 (text) | - | | P17 | The problem | `legacy/edge-judiverse-tech/docs/papers/the-4r-standard.md` | `b729c807dc606ca87a665ba9f288aaf1cc908edc` | 13 October 2026 (text) | - | | P18 | The claim | `legacy/edge-judiverse-tech/docs/papers/the-failures-worth-finding.md` | `950c66d700b4c89b0f3b08cc713769bd25d0dd14` | 13 October 2026 (text) | `legacy/edge-judiverse-tech/docs/papers/the-failures-worth-finding.4cba8d78.md` `4cba8d78e49cdce13362d04fb2c51041efa1be14` | | P19 | White paper programme and documentation architecture | `legacy/edge-judiverse-tech/docs/research/04-WHITE-PAPER-PROGRAMME.6fc300ce.md` | `6fc300ce86f236c9b1bd965055f6de701867b64e` | 2026-10-13 (text) | `legacy/edge-judiverse-tech/docs/research/04-WHITE-PAPER-PROGRAMME.md` `1700344758c28e4db397c690e0d358a986928bae` | | P20 | Reversible by default: governing AI agents with holds, authority envelopes and receipts | `legacy/edge-judiverse-tech/docs/research/05-WHITE-PAPER-REVERSIBLE-AGENTS.05fead68.md` | `05fead682abc6226a227757f2e64dc480482fed7` | 2026-10-13 (text) | `legacy/edge-judiverse-tech/docs/research/05-WHITE-PAPER-REVERSIBLE-AGENTS.md` `7fc635ee47a03d7e77368623a62e79da1843d693` | | P21 | Resilience is the product: built to fail well, recover fast and prove it | `legacy/edge-judiverse-tech/docs/resilience/OPRES-01-WHITE-PAPER-RESILIENCE-IS-THE-PRODUCT.md` | `18b5cb903065c24fec513aeef57142929e3dcffe` | 2026-10-13 (text) | - | | P22 | Risk aware, never risk averse: how X-X.TECH takes risk on purpose and stays standing | `legacy/edge-judiverse-tech/docs/risk/RISK-01-WHITE-PAPER-RISK-AWARE-NEVER-RISK-AVERSE.md` | `8b7c414276b0aee06802ad83794f107674b3b993` | 2026-10-13 (text) | - | | P23 | Research question | `legacy/edge-judiverse-tech/docs/whitepaper/ADAPTIVE-ADVERSARIAL-SAFETY-TESTING.md` | `703d197fea3c3574b87fd03ca3ce1828bafc2a07` | 13 October 2026 (text) | `legacy/edge-judiverse-tech/docs/whitepaper/ADAPTIVE-ADVERSARIAL-SAFETY-TESTING.4821a278.md` `4821a2784980c372b08d30263af53732920f5a49` | | P24 | The gap in materials discovery | `legacy/edge-judiverse-tech/docs/whitepaper/ADVANCED-MATERIALS-WITH-AI.md` | `52652f913abac564cc2052212a1896ccbf3a72ce` | 13 October 2026 (text) | `legacy/edge-judiverse-tech/docs/whitepaper/ADVANCED-MATERIALS-WITH-AI.344ca6d8.md` `344ca6d8e602f601f61b9d40f4e2f94a857cf200` | | P25 | The problem | `legacy/edge-judiverse-tech/docs/whitepaper/MEDICINES-DISCOVERY-AND-MULTI-OMICS.md` | `65f112fb8c0fd0a4064dddd09efa5a8113f47f89` | 13 October 2026 (text) | `legacy/edge-judiverse-tech/docs/whitepaper/MEDICINES-DISCOVERY-AND-MULTI-OMICS.89da3680.md` `89da3680298a2967d8eaa5877afe90c021c94a4d` | | P26 | The question | `legacy/edge-judiverse-tech/docs/whitepaper/RESILIENT-EDGE-INTELLIGENCE.md` | `b9dd9ed0e30ee696e5390e4bb3606a0a65f4516e` | 13 October 2026 (text) | `legacy/edge-judiverse-tech/docs/whitepaper/RESILIENT-EDGE-INTELLIGENCE.4d33b534.md` `4d33b534f4716eaeb9da4b9352a570f6a0ca1454` | | P27 | The problem with a single success | `legacy/edge-judiverse-tech/docs/whitepaper/SIMULATION-RESEARCH.md` | `a05cdcc75f583c4b60bfbd67079319a574caf917` | 13 October 2026 (text) | `legacy/edge-judiverse-tech/docs/whitepaper/SIMULATION-RESEARCH.94d42f2d.md` `94d42f2dd5ab36938d3a2b78644a4b8f4d41ac52` | | P28 | The 4R standard: retraceable, recoverable, reversible, reproducible | `legacy/edge-judiverse-tech/docs/whitepaper/THE-4R-STANDARD.md` | `e52de5c2d4d3d40f652eaf3c8cc504f1e41130e2` | 2026-10-13 (text) | - | | P29 | document": "docs/whitepaper/SIMULATION-RESEARCH.md | `legacy/edge-judiverse-tech/registry/authorship/simulation-research.json` | `a88d0ce17e58d0259dffb756eec153816f814ea9` | 2026-10-13 (branch) | - | | P30 | document": "docs/papers/the-4r-standard.md | `legacy/edge-judiverse-tech/registry/authorship/the-4r-standard.json` | `4a7e9c7bee5a14ae293ce3e5a76fdff8c3a4e2a4` | 2026-10-13 (branch) | - | | P31 | Governed intelligence you can check | `legacy/governed-intelligence-judeai-edgeid-pre-contract/docs/WHITEPAPER.md` | `7d34b197c020914985839ab3c2341c001df15f16` | 27 October 2026 (text) | `old-repo/docs/WHITEPAPER.md` `7d34b197c020914985839ab3c2341c001df15f16` | | P32 | JudeKey privacy disclosure draft; human review required before publication. | `legacy/jude-mvp/apps/JudeKey/Store/PRIVACY.md` | `33a29d6f06cf1f2ba009406da47a124f4f1c4516` | 2026-09-13 (branch) | `legacy/jude-mvp/apps/JudeKey/Store/PRIVACY.95648eb8.md` `95648eb836027c01e1df4b0ca3059f7f51bf3854` | | P33 | Sovereign Authority over Machine Action | `legacy/jude-mvp/docs/research-paper.md` | `cbd405a4f97d3bf4d27d80138ebad056e1cd4dc7` | undated | - | | P34 | EDGE: authority-bounded agentic AI with precursor-signal learning | `legacy/judeai-edge-control/docs/frontier-ai-whitepaper.md` | `8b0e7066a5c36a52030f2ba72e87d66a8d708276` | 12 October 2026 (text) | `legacy/local-source-snapshots/services/edge/docs/frontier-ai-whitepaper.md` `3cf9309f062f68b8dbc0cf04a29a217f380a87ea` | | P35 | Architecture white paper — design and observed gaps | `legacy/judeai-edgeid-cloud/handover/whole-estate-acceptance-20261026-01a16a2c/docs/03-ARCHITECTURE-WHITE-PAPER.md` | `daba10924f489b32d80354be014e2e0e86fe744e` | 2026-10-26 (path) | - | | P36 | A measurable standard for useful interfaces | `legacy/local-source-snapshots/foundations/canvas/reference/canvas/canvas-review/ui-excellence/whitepaper/document.md` | `547edd91a1dbf33ff91fee92b76eb5c74ca543fe` | 13 October 2026 (text) | - | | P37 | (no title line) | `legacy/local-source-snapshots/foundations/canvas/reference/canvas/canvas-review/ui-excellence/whitepaper/index.html` | `6ab7e2979f381309366868ec397c19142ce631f5` | 13 October 2026 (text) | - | | P38 | United Kingdom Sovereign Capabilities Architecture (UKSCA) white paper: research brief | `legacy/local-source-snapshots/programme/proof/UKSCA-RESEARCH-BRIEF-20261020.md` | `3fb75989c91918808614147f4be19237dfa2d5de` | 20 October 2026 (text) | - | | P39 | Evidence matrix: consortium-022 Write a paper in Markdowns with co-authors | `legacy/local-source-snapshots/programme/proof/journeys/consortium-022-EVIDENCE-MATRIX.md` | `c312519d80e555ef09dfd9b9180854fd03c0552f` | 2026-10-20 (text) | - | | P40 | Write a paper in Markdowns with co-authors | `legacy/local-source-snapshots/programme/proof/journeys/consortium-022.json` | `bf769395f893f53e2126c2e547309294990cec46` | 2026-10-18 (branch) | - | | P41 | Evidence matrix: consortium-023 Submit a paper to Consortium review | `legacy/local-source-snapshots/programme/proof/journeys/consortium-023-EVIDENCE-MATRIX.md` | `feadaecf6ba236ba639819388830f9c19565eb09` | 2026-10-20 (text) | - | | P42 | Submit a paper to Consortium review | `legacy/local-source-snapshots/programme/proof/journeys/consortium-023.json` | `95b202a8bd38125e81a2e3c51d28d3532ef2839a` | 2026-10-18 (branch) | - | | P43 | Intellectual property and publication terms | `legacy/local-source-snapshots/programme/proof/partners/agreements/IP-AND-PUBLICATION-TERMS.md` | `4fedc9987677d5d074e48f906f406e406c6ae0cd` | 20 October 2026 (text) | - | | P44 | Programme white paper: error-control methods and limitations | `legacy/local-source-snapshots/programme/white-paper/ERROR-CONTROL-METHODS-AND-LIMITATIONS.md` | `54b9c4681e1399d2e1916124ec778f2e42615a39` | 2026-10-18 (branch) | - | | P45 | White paper: | `legacy/local-source-snapshots/workbenches/capabilities/WHITE-PAPER-TEMPLATE.md` | `b892b1534e2cf3e97e351aeb6f2dbfdd9f228a22` | 2026-10-16 (branch) | - | | P46 | What this addendum is | `legacy/local-source-snapshots/workbenches/materials/WHITE-PAPER-ADDENDUM-AI-DEFINED-MATERIALS.md` | `7ff0f318f8886702b5a353908249cf45f86fa6bd` | 15 October 2026 (text) | - | | P47 | The question | `legacy/local-source-snapshots/workbenches/urban/WHITE-PAPER-GOVERNED-CITY-AGENTS.md` | `07303cd7f9b223ebaf6a1c0ecb8bd5d0e191077c` | 15 October 2026 (text) | - | | P48 | TOK ROOT. Living evidence | `legacy/peak-judeai-behavioural-harness-baseline-full-application/tok/root/whitepaper/TOK_White_Paper.md` | `1f47328831861f42db923c6ab4c7183e5a1c948a` | 2026-10-05 (text) | - | | P49 | Judiverse live publication — recorded releases, independent acceptance pending. | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/handoffs/estate-graph-20261022/live-publication/README.md` | `b06610294e62cf7102cce44c410fc065896f3084` | 2026-10-22 (path) | - | | P50 | Judiverse estate UI publication packet — prepared, not deployed. | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/handoffs/estate-graph-20261022/native-estate-ui/README.md` | `e18e8f943ee2a212da6ab6c46a92be6ed92916f0` | 2026-10-22 (path) | - | ###### 1.2 Standards, regulatory registers, regulatory packs and policy documents (155 distinct documents, 202 catalogue rows) | # | Title | Corpus-relative path | Blob oid | Date (how dated) | Near-identical variants (path, oid) | |---|---|---|---|---|---| | S1 | JudeAI / EDGE ID Estate — Bank Regulatory and Product Requirements | `legacy/edge-judiverse-tech/docs/bank/BANK-REGULATORY-AND-PRODUCT-REQUIREMENTS.md` | `117424297044f7155c8a3187b97e1ec4852d255d` | 13 October 2026 (text) | - | | S2 | Retraceable, recoverable, reversible, reproducible: the standard AI that acts must meet | `legacy/edge-judiverse-tech/docs/blog/2026-10-13-the-4r-standard.md` | `4d309c4943cfc729529ae2b73797a1ad4811a21f` | 2026-10-13 (text) | - | | S3 | Research file: editorial and drafting standard | `legacy/edge-judiverse-tech/docs/dossiers/editorial-standard.md` | `2bdf98fe945482956b6f6b3fdb0eebefe532c1b0` | 2026-10-13 (branch) | - | | S4 | intenJude frontier standard: what best in class means, how we prove it | `legacy/edge-judiverse-tech/docs/engineering/INTENJUDE-FRONTIER-STANDARD.md` | `832b0ac1a21524a4a753f5d1c7e52fcdb09102d8` | 13 October 2026 (text) | - | | S5 | 02 Laws and regulations register for Jude Bank | `legacy/edge-judiverse-tech/docs/finance/bank/02-LAWS-AND-REGULATIONS.md` | `723640a846be7867b60d394d4600be0fd91edcca` | 13 October 2026 (text) | - | | S6 | 05 Jude Bank: staff, regulator and engineer journeys | `legacy/edge-judiverse-tech/docs/finance/bank/05-STAFF-REGULATOR-ENGINEER-JOURNEYS.md` | `c251bb4fd0ff1cf98f2baefc3936750d48183def` | 13 October 2026 (text) | - | | S7 | P82 Outreach: laws, regulations and standards register | `legacy/edge-judiverse-tech/docs/outreach/02-LAWS-AND-REGULATIONS.md` | `fd8acab3cb04ff238201e454bfa451b8060241ab` | 13 October 2026 (text) | - | | S8 | UK-PACK-02: Architecture, schemas, routes, dependencies, software bill of materials and li | `legacy/edge-judiverse-tech/docs/packs/UK-PACK-02.md` | `7fefad55c77dcd7f23754d5f47f9301c39d8f321` | 13 October 2026 (text) | - | | S9 | FCA, PRA and insurance intelligence framework | `legacy/edge-judiverse-tech/docs/regulation/FCA-PRA-INSURANCE-INTELLIGENCE-FRAMEWORK.md` | `4b7c3f80280523007a34de67de83fcdbfb65607c` | 13 October 2026 (text) | - | | S10 | Financial intelligence benchmarks: the register | `legacy/edge-judiverse-tech/docs/regulation/FINANCIAL-INTELLIGENCE-BENCHMARKS.md` | `ec1706661a4c6a61964712330f08d23a49e7fc23` | 13 October 2026 (text) | - | | S11 | Insurance intelligence product standard | `legacy/edge-judiverse-tech/docs/regulation/INSURANCE-INTELLIGENCE-PRODUCT-STANDARD.md` | `1827af85c3bd01676bc7de97c2d20ba7924a8b85` | 13 October 2026 (text) | - | | S12 | Part 1. The FCA regime: obligation to control to evidence to benchmark | `legacy/edge-judiverse-tech/docs/regulation/PART-1-FCA.md` | `14d58b94a0af69ba9bfa2c6641688956db0d2583` | 13 October 2026 (text) | - | | S13 | Part 2: the PRA and the Bank of England | `legacy/edge-judiverse-tech/docs/regulation/PART-2-PRA-AND-BANK-OF-ENGLAND.md` | `33c4f1463c5c156d98cd2db72bfcb9ae8a7e526b` | 13 October 2026 (text) | - | | S14 | Part 3: data, AI and international regimes, and the standards map | `legacy/edge-judiverse-tech/docs/regulation/PART-3-DATA-AI-AND-INTERNATIONAL.md` | `8812fd6c44276f739a34639298765d888c5c61a5` | 13 October 2026 (text) | - | | S15 | UK legal and regulatory register: agent governance | `legacy/edge-judiverse-tech/docs/regulation/UK-LEGAL-AND-REGULATORY-REGISTER.md` | `ec4adf72d88de46ebefdacc48b2bf4cb7e9e1b43` | 13 October 2026 (text) | - | | S16 | JudeAI / EDGE ID — Regulator review frontend | `legacy/edge-judiverse-tech/docs/regulators/PORTAL-FRONTEND.md` | `77a45767fd3c9fa46b6fdc10da177e06d1d465a4` | 2026-10-13 (branch) | - | | S17 | Research 3: MCP, agent identity, security, supply chain, governance and observability stan | `legacy/edge-judiverse-tech/docs/research/2026-10-13/03-mcp-identity-security-governance.md` | `ec75a2105f105cdf9d17fa56e4715d4084b990a0` | 13 October 2026 (text) | - | | S18 | Research 5: the FCA regime for insurance and AI (13 October 2026) | `legacy/edge-judiverse-tech/docs/research/2026-10-13/05-fca-insurance-and-ai.md` | `98ae1b632f539918bc581e0ed2dd0462ae493350` | 13 October 2026 (text) | - | | S19 | Research 6: the PRA and Bank of England regime (13 October 2026) | `legacy/edge-judiverse-tech/docs/research/2026-10-13/06-pra-bank-of-england.md` | `2e7e9dae209294193df955a14007c5654b4d48a8` | 13 October 2026 (text) | - | | S20 | Research 7: standards landscape for an insurance and financial-intelligence platform (13 O | `legacy/edge-judiverse-tech/docs/research/2026-10-13/07-iso-international-standards.md` | `5dc8f2dbfce73f3e549bc1a5d2bb26a4586830b6` | 13 October 2026 (text) | - | | S21 | R1. Risk doctrine and standards | `legacy/edge-judiverse-tech/docs/research/risk-resilience/R1-RISK-DOCTRINE-AND-STANDARDS.md` | `7fac9a648ebf80973e8f6a8f7215b70e047e8f6d` | 13 October 2026 (text) | - | | S22 | R2. UK regulators on risk and growth: research note for Master principle 21 | `legacy/edge-judiverse-tech/docs/research/risk-resilience/R2-REGULATORS-ON-RISK-AND-GROWTH.md` | `7a6e66ed38e793a8922abd6135bddb2f9fafbd79` | 13 October 2026 (text) | - | | S23 | 4R Specification | `legacy/edge-judiverse-tech/docs/standard/4R-SPECIFICATION.md` | `08b3feb58b7a53d643219db68230a1baa03a9d71` | 13 October 2026 (text) | - | | S24 | Purpose | `legacy/edge-judiverse-tech/docs/standard/RENDER-INTEGRITY-AND-PROGRESSIVE-REVELATION.md` | `987a4da8b41894045909b62e46219a1f2035fe80` | 14 October 2026 (text) | - | | S25 | Purpose | `legacy/edge-judiverse-tech/docs/standard/VERIFIABLE-WORK-STANDARD.md` | `43640f90694e56a2592ea2d793737f513a64d774` | 13 October 2026 (text) | `legacy/edge-judiverse-tech/docs/standard/VERIFIABLE-WORK-STANDARD.2a379d25.md` `2a379d25b139d8a17bc6074b5f5f0facf674a297` | | S26 | Drafting standard for every submission document | `legacy/edge-judiverse-tech/docs/submission/LEGAL-STYLE.md` | `6be5438a526e773a5dd9b25f26630275e9eddfff` | 13 October 2026 (text) | - | | S27 | What must be in place: procurement audits, ISO standards, FCA licences, and Sovereign AI c | `legacy/edge-judiverse-tech/docs/submission/REQUIREMENTS-AUDITS-ISO-FCA.md` | `c235ec16b5c7a6a4f0662e695206199f5757462d` | 13 October 2026 (text) | - | | S28 | Sovereign AI R&D Procurement Scheme, Challenge 1: AI for a more productive NHS. Full Appli | `legacy/edge-judiverse-tech/docs/submission/challenge-1/C1-01-FULL-APPLICATION.md` | `413d877c7f3837fbf141ae37a47ac4fcdcb95c4a` | 13 October 2026 (text) | `legacy/edge-judiverse-tech/docs/submission/challenge-1/C1-01-FULL-APPLICATION.deb271df.md` `deb271df2e75222d7b5c61e729a15cedf74759ff` | | S29 | Sovereign AI R&D Procurement Scheme, Challenge 1: AI for a more productive NHS. Use Cases | `legacy/edge-judiverse-tech/docs/submission/challenge-1/C1-02-USE-CASES.md` | `6a113eabefda515ec6c94ada5aaa417364716b2a` | 13 October 2026 (text) | `legacy/edge-judiverse-tech/docs/submission/challenge-1/C1-02-USE-CASES.e3ea7760.md` `e3ea77605b7758eb22865432e0d7c66a2732758b` | | S30 | Sovereign AI R&D Procurement Scheme, Challenge 1: AI for a more productive NHS. Contract S | `legacy/edge-judiverse-tech/docs/submission/challenge-1/C1-03-CONTRACT-SCHEDULE.md` | `7610705972d67d10069169c44e7c306d9b27f581` | 13 October 2026 (text) | `legacy/edge-judiverse-tech/docs/submission/challenge-1/C1-03-CONTRACT-SCHEDULE.bcdb7c59.md` `bcdb7c59841e0a60779b94dad51d85b9a2271f44` | | S31 | Sovereign AI R&D Procurement Scheme, Challenge 1: AI for a more productive NHS. Evidence I | `legacy/edge-judiverse-tech/docs/submission/challenge-1/C1-04-EVIDENCE-INDEX.md` | `e96d4fcab3a197e78e9541aefe4fa2d273b51b38` | 13 October 2026 (text) | `legacy/edge-judiverse-tech/docs/submission/challenge-1/C1-04-EVIDENCE-INDEX.90d05d60.md` `90d05d6073544dfa11287974f0d87223e63c7851` | | S32 | document": "docs/design/HOUSE-STYLE.md | `legacy/edge-judiverse-tech/registry/authorship/editorial-standard.json` | `4f7b0ada1b46c3500425280e3c635c5e5ef084d3` | 2026-10-13 (branch) | - | | S33 | document": "docs/standard/VERIFIABLE-WORK-STANDARD.md | `legacy/edge-judiverse-tech/registry/authorship/verifiable-work-standard.json` | `5bd962e2f4d9c7bae60f3d4c70ab74c142c69c1f` | 2026-10-13 (branch) | - | | S34 | JudeAI / EDGE ID Estate — Bank Regulatory and Product Requirements | `legacy/edge-judiverse-tech/registry/bank-regulatory-requirements.json` | `3a1d7490580d74f6821f549973d27b6f8030e7af` | 2026-10-13 (text) | - | | S35 | UK GDPR Article 30: records of processing activities | `legacy/edge-judiverse-tech/registry/uk-legal-register.json` | `bac39695db025ef44d361edf9ea5393d3e426fee` | 13 October 2026 (text) | - | | S36 | Estate policy locator | `legacy/github-metadata-20261030/local-working-changes/canonical-local/files/AGENTS.md` | `66787295381215ab30eadb5f1aa83e3d27a27981` | 28 October 2026 (text) | - | | S37 | spec": "registers.standards | `legacy/governed-intelligence-judeai-edgeid-pre-contract/registers/standards.json` | `9293ae8eea7b76b1b985b479fc13c7efb94e09de` | 2026-10-29 (branch) | `old-repo/registers/standards.json` `9293ae8eea7b76b1b985b479fc13c7efb94e09de` | | S38 | REG-1 — Regulator verification portal (GES-1.0 annex R) | `legacy/governed-intelligence-judeai-edgeid-pre-contract/spec/REG-1-REGULATOR-PORTAL.md` | `f0acfb32348d5e01cb1cf1b6a61d3b3286802ce3` | 2026-10-27 (branch) | `old-repo/spec/REG-1-REGULATOR-PORTAL.md` `f0acfb32348d5e01cb1cf1b6a61d3b3286802ce3` | | S39 | Regulator submission | `legacy/governed-intelligence-judeai-edgeid/publish/submission/submission.html` | `86a19a93bd7dac2c25aa404963abdcaa7e4a68af` | 2026-10-30 (branch) | - | | S40 | Challenge 1: one live care state | `legacy/jude-mvp/docs/challenge-1-nhs.md` | `6294057bd6143c7a175c50fa5470f493a39cabc5` | undated | - | | S41 | What the wall has to pass | `legacy/jude-mvp/docs/compliance.md` | `115fb90723742066b1fc536207e4562e4d691ee6` | undated | - | | S42 | Jude and the FCA's review of frontier AI and cyber resilience | `legacy/jude-mvp/docs/fca-frontier-ai.md` | `0c8b737a6c8eb9e61dbec9aba6ddbcb91a297770` | 2 September 2026 (text) | - | | S43 | (no title line) | `legacy/jude-mvp/docs/reference/XX_London_Regulated_Search_700.xlsx` | `0f7df8ad291d941f9942ac36c0d42729def8638f` | 2026-09-13 (branch) | - | | S44 | Next implementation — real practice, Agentic Judgement first | `legacy/jude-mvp/handover/08-training-plan.md` | `3382054f91b2dd53b406e54b680f0cbc469f3ad4` | undated | - | | S45 | What counts as evidence | `legacy/jude-mvp/handover/09-proof-standard.md` | `65b605f0eb6b07cb709a60ce7404a1c2b6bd3f4a` | undated | - | | S46 | Existing runtime integration — inspected source, not a running practice harness | `legacy/jude-mvp/handover/14-runtime-integration.md` | `f09d5da7072db7b10a80969a22e7862bfbb0bd6e` | undated | - | | S47 | The Agent Egress Manifest (AEM), version 0.1 | `legacy/jude-mvp/jbox/Standards/aem.md` | `d017b621f86fc9281edbbffbbde1892e5d0c1cdd` | 2026-09-03 (text) | - | | S48 | Agent evidence to a courtroom standard | `legacy/jude-mvp/jude-live/docs/evidence-standard.md` | `2f0dc821cb7e0708b9b72857d9ee0e1ac7b80593` | 2026-09-13 (branch) | - | | S49 | Company Policy | `legacy/jude-mvp/jude-live/policies/company-policy.md` | `5cd19b02657409df82a24daa1d92ce83f3875ace` | 3 September 2026 (text) | - | | S50 | Privacy Policy | `legacy/jude-mvp/jude-live/policies/privacy-policy.md` | `8c972a57629878b99feb615a525bd3a01bd95d30` | 3 September 2026 (text) | `legacy/jude/jude-live/policies/privacy-policy.md` `da46a2d206f948a0e2718e454208ef1cea9902c1` | | S51 | Trust & legal · JudeAI | `legacy/jude-mvp/judeai-hosted/web/trust.html` | `95ed86fcc135a5cc76909a97b28532602cc95c24` | 2026-10-01 (branch) | `legacy/jude-mvp/judeai-hosted/web/trust.663086fa.html` `663086fa8e7b885fc832a77b7687c504e4c5af3a`; `legacy/peak-judeai-behavioural-harness-baseline-full-application/web/trust.html` `7b464a9cb00793dca8eef6774d5ce502daf32a09`; `legacy/jude-mvp/judeai-hosted/web/trust.43fcd933.html` `43fcd9332c04a547c1db0f6c2bb130233af26d4f`; `legacy/jude-mvp/judeai-hosted/web/trust.a16a7d74.html` `a16a7d74d6d8631cc815700bfd0239533ab109b2` | | S52 | Leakage policy and partitions | `legacy/jude-mvp/judeos-live/training/frontier-gauntlet/holdouts/POLICY.md` | `cacc67b784bbff98068533977a5bb5564f7f759f` | 2026-09-25 (branch) | - | | S53 | X-X.TECH — Cyber Governance, Resilience, Risk & Compliance | `legacy/jude-mvp/judiverse-live/app/home.html` | `e7efbc9d803f4a790faccd866c11ef945cb258ad` | 2026-09-25 (branch) | `legacy/judiverse/app/home.html` `46681b895864c3f2b6304b6047d778bf4a1286b4` | | S54 | Electronic health record systems: Epic, Oracle Health (Cerner Millennium), NHS Spine and N | `legacy/judeai-edge-control/docs/research/platform-inventories/ehr.md` | `7ab9496dcb9575bbdf94fd9ca1c6d166e8eb4570` | 10 October 2026 (text) | - | | S55 | Secure Group Call link: benchmark against Microsoft Teams and UK regulated-sector rules | `legacy/judeai-edge-control/docs/secure-group-call-benchmark.md` | `d3bab1709456200150c96279d7b5890d338928ca` | 7 October 2026 (text) | - | | S56 | Jude Bank Business · X-X.TECH Ltd · Judiverse | `legacy/judeai-edge-control/web/estate/bank-10-business.21a9fcaa.html` | `21a9fcaacbc91ba040ef1394165a621699f48301` | undated | - | | S57 | Climate, carbon efficiency and the Charcoal standard: measured in every workflow | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__research__RESEARCH-CLIMATE-CARBON-CHARCOAL.md` | `a0dcc214bc7dd355cb8aed281764e20fde9f0901` | 24 October 2026 (text) | - | | S58 | Government, MOD and FCA engagement, applications, funding and the real-world use that prov | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__research__RESEARCH-GOVERNMENT-MOD-FCA-FUNDING.md` | `d63542711672b11f6e1ceacab43281347a74e202` | 24 October 2026 (text) | - | | S59 | Research: the law and standards register · every clause that binds a frozen-scope lane, li | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__research__RESEARCH-LAW-AND-STANDARDS-REGISTER.md` | `5f862b0e715b9fa92e367c357d7a7c1dd36845f2` | 24 October 2026 (text) | - | | S60 | Copyright 2026 X-X.TECH. All rights reserved. | `legacy/local-source-snapshots/LICENSES/LicenseRef-X-X-TECH-Proprietary.txt` | `3450e1914d9817999e870d79121f47c476f47e6e` | 2026-10-18 (branch) | - | | S61 | Security policy | `legacy/local-source-snapshots/SECURITY.md` | `996f198171322e2429e1e19ed746fed99b58d653` | 2026-10-18 (branch) | - | | S62 | EDGE-REG-00 Regulatory and submission documents: contents | `legacy/local-source-snapshots/assurance/regulatory/REGISTER.tok.json` | `06b86fd3fba2273f5d1eac0d2e147672b9f2dd7b` | 15 October 2026 (text) | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/REGISTER.tok.json` `56d0fc7998c8d3778737e6a039253a23d599b6ba` | | S63 | Regulator pack review, 15 October 2026 | `legacy/local-source-snapshots/assurance/regulatory/REVIEW.md` | `4c5bf0421407060cda67f40e214559e60f2225ab` | 15 October 2026 (text) | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/REVIEW.md` `4e355bec46cbfb0ce2a1afedf3d0b2c2e6b2c5f4` | | S64 | EDGE-REG-00 Regulatory and submission documents: contents | `legacy/local-source-snapshots/assurance/regulatory/source/00-index.md` | `8d52ba842a88fa57cd3c3148a16c57c756dbb890` | 15 October 2026 (text) | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/00-index.md` `81b3f0cb220c1ff3c2102324ca7961bf468239a8` | | S65 | EDGE-REG-01 Data protection impact assessment | `legacy/local-source-snapshots/assurance/regulatory/source/01-dpia.md` | `ad43bde72a7ef858c1d14228bc890ed8998f7908` | 2026-10-16 (branch) | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/01-dpia.md` `e75ad5b9fca7cfec3872fa5727875408ab07cfd9` | | S66 | EDGE-REG-02 Record of processing activities | `legacy/local-source-snapshots/assurance/regulatory/source/02-ropa.md` | `8819d46ae5ca1c132bb20ca3222c92827abcb18b` | 2026-10-16 (branch) | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/02-ropa.md` `164b7728e6707c0af8b5869cab9830207684bc21` | | S67 | EDGE-REG-03 Privacy notice | `legacy/local-source-snapshots/assurance/regulatory/source/03-privacy-notice.md` | `38c2b53a95961c9de58fe65b09fc622cc96aafc6` | 2026-10-16 (branch) | - | | S68 | EDGE-REG-04 Information security statement | `legacy/local-source-snapshots/assurance/regulatory/source/04-security.md` | `683939838b1eb934b979fc21f3510909c15bf1e0` | 2026-10-16 (branch) | - | | S69 | EDGE-REG-05 AI governance statement | `legacy/local-source-snapshots/assurance/regulatory/source/05-ai-governance.md` | `6db733a41fba8961e654a553ba469b79015903c5` | 2026-10-16 (branch) | - | | S70 | EDGE-REG-06 Accessibility statement | `legacy/local-source-snapshots/assurance/regulatory/source/06-accessibility.md` | `f9b45d79c5c5e502c434323718e3572052c6735c` | 15 October 2026 (text) | - | | S71 | EDGE-REG-07 Clinical safety statement | `legacy/local-source-snapshots/assurance/regulatory/source/07-clinical-safety.md` | `3c82ad4e5adda31744b9fb04a549004d2432f1d7` | 2026-10-16 (branch) | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/07-clinical-safety.md` `768b8ebae2e34fb6d0fdbc22ecbc6f15e0875d05` | | S72 | EDGE-REG-08 Financial services regulatory position | `legacy/local-source-snapshots/assurance/regulatory/source/08-fs-perimeter.md` | `65094154ffe116a0e0ddd95899d14b32fc22bf7a` | 2026-10-16 (branch) | - | | S73 | EDGE-REG-09 Financial crime statement | `legacy/local-source-snapshots/assurance/regulatory/source/09-fin-crime.md` | `3e3a41ed0d3d53a59e51f8d8bc2339896112e556` | 2026-10-16 (branch) | - | | S74 | EDGE-REG-10 Export control and national security statement | `legacy/local-source-snapshots/assurance/regulatory/source/10-export-nsi.md` | `5e0e67d99e335fe4852a460e9646218a81f77ec6` | 2026-10-16 (branch) | - | | S75 | EDGE-REG-11 Environmental and compute statement | `legacy/local-source-snapshots/assurance/regulatory/source/11-environment.md` | `8950eac88140617005b9f9a2eded86647f7ad2e4` | 2026-10-16 (branch) | - | | S76 | EDGE-REG-12 Supplier information | `legacy/local-source-snapshots/assurance/regulatory/source/12-procurement.md` | `d0fb02408eed23fe31f42b3bc6184524c0b6a039` | 14 October 2026 (text) | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/12-procurement.md` `fd61d642ddb82abb9302269b781fd34e709c79be` | | S77 | (no title line) | `legacy/local-source-snapshots/foundations/canvas/reference/canvas/regulator-review/index.html` | `a5083da3f2d3fd8c4a9791c9608ac52204c67688` | 2026-10-15 (branch) | - | | S78 | Sign-in door policy (binding) | `legacy/local-source-snapshots/handoffs/DOOR-POLICY.md` | `9e4fc00f7fc9a81f27a091d31c6ecfd14588c6b1` | 19 October 2026 (text) | - | | S79 | Privacy · EDGE ID | `legacy/local-source-snapshots/handoffs/ui-closeout-20261014/source/edge/web/estate/legal-privacy.html` | `308efb5019dd718b44c492a26a9e14c651d085e4` | 14 October 2026 (text) | `legacy/local-source-snapshots/services/edge/web/estate/legal-privacy.html` `6b80e4258267178c68ed1cce7be6600d5076ef5d`; `legacy/judeai-edge-control/web/estate/legal-privacy.html` `0f5f71c5d6a1c30ec0aa3cc3baafe1a496e1bc6f`; `legacy/judeai-edge-control/web/estate/legal-privacy.4473bdab.html` `4473bdab3ec6f71ecb34bc35b140bcb5bee93881`; `legacy/judeai-edge-control/web/estate/legal-privacy.3b152233.html` `3b15223361488908aa4ff1c301c586e5c527109f` | | S80 | Engineering standard: no single point of failure, joint liability, magic to the naked eye | `legacy/local-source-snapshots/programme/ENGINEERING-STANDARD-20261020.md` | `85a7f4834165613f9f2fca72dd5decba2f2fdd1a` | 20 October 2026 (text) | - | | S81 | Error-control standard | `legacy/local-source-snapshots/programme/ERROR-CONTROL-STANDARD.md` | `660cd7b2276041b292f8417a67c5f6277a89a759` | 19 October 2026 (text) | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/parallel-proof-20261021/spec-source/ERROR-CONTROL-STANDARD.md` `2e09ae81afd0b9862f5fe5a022f3621c62b5be32` | | S82 | Regulated continuity and all-agent deployment/recovery | `legacy/local-source-snapshots/programme/jps/enforcement/regulated-continuity-20261020/CONTINUITY-STANDARD.md` | `fcd2c51d2b38fdb34df6b111e4e72bf0ecadcd7a` | 2026-10-20 (path) | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/jps/enforcement/regulated-continuity-20261020/CONTINUITY-STANDARD.md` `f67a71a4ae340a76609975d1dcbb3a885f4d43a4` | | S83 | End-User Licence Agreement | `legacy/local-source-snapshots/programme/product/LICENCE-EULA-v1.0.md` | `6f213b968ceb9f8557823a8701a50ca24359f7f1` | 20 October 2026 (text) | - | | S84 | Lewisham and Greenwich NHS Trust, University Hospital Lewisham (proposed, later stage): a | `legacy/local-source-snapshots/programme/proof/partners/LEWISHAM-AND-GREENWICH-PROPOSED.md` | `e4d9751cdb2af5b6af0e831181c78ad7355dd77d` | 20 October 2026 (text) | - | | S85 | Starling Bank (proposed): second regulated environment for fraud, KYC and sanctions (Starl | `legacy/local-source-snapshots/programme/proof/partners/STARLING-BANK-PROPOSED.md` | `836f7a8ef458e182f503f0337dbdbd484e73aaac` | 20 October 2026 (text) | - | | S86 | TransPlus (56 Dean Street; Chelsea and Westminster Hospital NHS Foundation Trust): propose | `legacy/local-source-snapshots/programme/proof/partners/TRANSPLUS.md` | `77aaef3f81fd6423180133ba516741e5d48ab9f2` | 20 October 2026 (text) | - | | S87 | Data processing terms (UK GDPR article 28) and data-sharing schedule | `legacy/local-source-snapshots/programme/proof/partners/agreements/DATA-PROCESSING-TERMS.md` | `8f871f352d6df83896a244f6692f4be5f60c7e17` | 20 October 2026 (text) | - | | S88 | Data Security and Protection Toolkit (bid page 55: NHS England, year 2a; Challenge 1 M2) | `legacy/local-source-snapshots/programme/proof/procurement/DSPT.md` | `767b8cbc45cd4d9474e256ab27342f0e6e90cbc0` | 20 October 2026 (text) | - | | S89 | Digital Technology Assessment Criteria pack (bid page 55: NHS England, year 2a; Challenge | `legacy/local-source-snapshots/programme/proof/procurement/DTAC.md` | `bfe4c003b75b8cf80b73fbe995932fc1d3fa41be` | 20 October 2026 (text) | - | | S90 | FCA and JMLSG financial-crime controls (bid page 55: before partner data; bid page 51: Sta | `legacy/local-source-snapshots/programme/proof/procurement/FCA-JMLSG.md` | `c746939de0b33885417a42e690c137aa49525939` | 20 October 2026 (text) | - | | S91 | ISO/IEC 27001 (Challenge 4 M4: audit at month 7; EDGE-REG-04 clause 3.1: no commitment mad | `legacy/local-source-snapshots/programme/proof/procurement/ISO-27001.md` | `7b8c5ff2d18b4f54ef016dac68e8aea909bf5cd8` | 20 October 2026 (text) | - | | S92 | NHS frameworks (bid page 62: procurement listing 2028; bid page 54: three or more NHS serv | `legacy/local-source-snapshots/programme/proof/procurement/NHS-FRAMEWORKS.md` | `138456b572352a8ca0b4ddc3f597767e20e3235f` | 20 October 2026 (text) | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/proof/procurement/NHS-FRAMEWORKS.md` `e91b3f5b38784f8e0d833118434b28027c61e939` | | S93 | Data Security and Protection Toolkit: assessment evidence by standard | `legacy/local-source-snapshots/programme/proof/procurement/documents/DSPT-ASSESSMENT.md` | `6642411934854021f70b1cc96b6ab88a6f43c486` | 20 October 2026 (text) | - | | S94 | Charcoal Standard and the consortium proposal: Owner's brief of 15 October 2026 (verbatim) | `legacy/local-source-snapshots/research/charcoal-standard/CHARCOAL-STANDARD-CONSORTIUM-ALIGNMENT-20261015.md` | `41dc0a47ce7878acb35369ae6909427a36f4c263` | 15 October 2026 (text) | - | | S95 | The charcoal standard (JudeDeck, 10 lines per slide, 15 October 2026) | `legacy/local-source-snapshots/research/charcoal-standard/CHARCOAL-STANDARD-JudeDeck.md` | `997adda4946145ac0c696203dbc44f9bf8cc6e8d` | 15 October 2026 (text) | - | | S96 | The Charcoal Standard: Owner's brief of 15 October 2026 (verbatim) | `legacy/local-source-snapshots/research/charcoal-standard/CHARCOAL-STANDARD-OWNER-BRIEF-20261015.md` | `a729eb070d9fcd640152cc7a438d08cc77ac0c36` | 15 October 2026 (text) | - | | S97 | The charcoal standard: money backed by carbon taken out of the air | `legacy/local-source-snapshots/research/charcoal-standard/CHARCOAL-STANDARD.md` | `78b9b2b55c608de33b63710f491f65108f819cd6` | 15 October 2026 (text) | - | | S98 | Agentic exhaustive completeness (AEC): binding standard | `legacy/local-source-snapshots/services/edge/Master/AEC-STANDARD.md` | `8643f73ced91fbbb052426753ed3841655b52439` | 19 October 2026 (text) | - | | S99 | Accessibility · EDGE ID | `legacy/local-source-snapshots/services/edge/web/estate/legal-accessibility.html` | `f11ee1106118e28432dcf1b9ebb097fec6e64086` | 2026-10-19 (branch) | `legacy/judeai-edge-control/web/estate/legal-accessibility.html` `2ef6cf77108ac4a887ed67196e24baed31973e85`; `legacy/judeai-edge-control/web/estate/legal-accessibility.a87238ad.html` `a87238ad9211d980fedd8193c8ce6e979c5caa45`; `legacy/judeai-edge-control/web/estate/legal-accessibility.05caa91d.html` `05caa91d4601fed085ea9e6ec630299c0f960a9e` | | S100 | Support · EDGE ID | `legacy/local-source-snapshots/services/edge/web/estate/legal-support.html` | `2c7c8f834c7520b67e3159d140be3927a69c5acc` | 2026-10-19 (branch) | `legacy/judeai-edge-control/web/estate/legal-support.html` `4923ed3fa8c27282ee932dd454f431a7a4c1892c`; `legacy/judeai-edge-control/web/estate/legal-support.acffeb80.html` `acffeb80b19a86d04d1c5ecbd91da647e61e5716`; `legacy/judeai-edge-control/web/estate/legal-support.5b0cfe48.html` `5b0cfe487eadb3bc6f9e13ad9eb78250f66d9db9` | | S101 | Terms of use · EDGE ID | `legacy/local-source-snapshots/services/edge/web/estate/legal-terms.html` | `2b77228117fb0b32d254306fd5d8c141ef1f1fc8` | 2026-10-19 (branch) | `legacy/judeai-edge-control/web/estate/legal-terms.html` `a51339e32b58658e599bf65a89cee350a72f460e`; `legacy/judeai-edge-control/web/estate/legal-terms.7393885e.html` `7393885e4390f224bc1ca2d65642f763542daace`; `legacy/judeai-edge-control/web/estate/legal-terms.4c742c7b.html` `4c742c7bc78fa3c66a30da8935e418ca24d3ba23` | | S102 | Estate agents | `legacy/mac-retired-20261030/policy-history/2.5.0-39f35e28-before-agent-mailbox/AGENTS.md` | `8215fd4b0268851ab5cc31025b9c66707b07626f` | 28 October 2026 (text) | - | | S103 | JudeAI / EDGE ID — canonical plan 2.5.0 | `legacy/mac-retired-20261030/policy-history/2.5.0-39f35e28-before-agent-mailbox/handover/ESTATE-LEVEL-READINESS.md` | `65585f9917af105e1b7ea605e66b4c3dce0153b5` | 2026-10-30 (path) | - | | S104 | JudeAI / EDGE ID — canonical plan 2.5.0 | `legacy/mac-retired-20261030/policy-history/2.5.0-39f35e28-before-agent-mailbox/handover/RULES-CLAUSES.md` | `ed1fb2177b3ea2ed118091331d495be3f9e2a68c` | 2026-10-30 (path) | - | | S105 | JudeAI / EDGE ID — canonical plan 2.5.0 | `legacy/mac-retired-20261030/policy-history/2.5.0-39f35e28-before-agent-mailbox/handover/RULES-CONFLICTS.md` | `ed682ae8fb8be33173989090adbba56ffeeaca7e` | 2026-10-30 (path) | - | | S106 | JudeAI / EDGE ID — canonical plan 2.5.0 | `legacy/mac-retired-20261030/policy-history/2.5.0-39f35e28-before-agent-mailbox/handover/RULES.md` | `a88cd66f617afb0cd704001fbdd32f29ac35ee29` | 2026-10-30 (path) | - | | S107 | JudeAI / EDGE ID — canonical plan 2.5.0 | `legacy/mac-retired-20261030/policy-history/2.5.0-39f35e28-before-agent-mailbox/plan/PLAN.md` | `be1e22db8acfe4d91667e91b099eb810f414f545` | 2026-10-30 (path) | - | | S108 | Production service standard: continuous service for JudeAI | `legacy/peak-judeai-behavioural-harness-baseline-full-application/docs/Production service standard, continuous service for JudeAI - claude-estate-policy-20261020T2110Z.md` | `631f31edfd329b4064d64102637c519e16551c01` | 20 October 2026 (text) | - | | S109 | Regulator portal and executable estate gate — one-pass integration contract | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/REGULATOR-PORTAL-ONE-PASS.md` | `48577c4a8a1eb84e16b855475663d7b25d0fb7ad` | 2026-10-27 (branch) | - | | S110 | EDGE-REG-13 Data protection impact assessment and record of processing: EDGE ID identity a | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/13-dpia-identity.md` | `d0221ed84beed1bbff9b22b002d6de9f128c08bc` | 19 October 2026 (text) | - | | S111 | EDGE-REG-14 Data protection impact assessment and record of processing: personal records, | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/14-dpia-records.md` | `86fd0aadc1f6075c3112587cd2745560b7af2a84` | 19 October 2026 (text) | - | | S112 | EDGE-REG-15 Data protection impact assessment and record of processing: JudeAI chat | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/15-dpia-chat.md` | `c0db8fb33844a38e02b036ce99afb7c02157c15d` | 19 October 2026 (text) | - | | S113 | EDGE-REG-16 Data protection impact assessment and record of processing: Judiverse Mail | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/16-dpia-mail.md` | `c6521bc3157f03b726f36bc896de0936f0d493fa` | 19 October 2026 (text) | - | | S114 | EDGE-REG-17 Data protection impact assessment and record of processing: Judiverse Calls an | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/17-dpia-calls.md` | `e27aa0c73499d46fc4cb5c3dbb33b1ac19a7fde1` | 19 October 2026 (text) | - | | S115 | EDGE-REG-18 Data protection impact assessment and record of processing: EDGE Bank, Open Ba | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/18-dpia-bank.md` | `ef088386a59b003535d56c98b040568ca99e79f5` | 19 October 2026 (text) | - | | S116 | EDGE-REG-19 Data protection impact assessment and record of processing: Clinic | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/19-dpia-clinic.md` | `ac59327a4f9f29c5f0c72ec70014d0b088b52f63` | 19 October 2026 (text) | - | | S117 | EDGE-REG-20 Data protection impact assessment: urine-test-first pathway pilot for suspecte | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/20-dpia-clinical-pilot.md` | `24a62334d12aa356e0a399e094cab2b8f5d6670e` | 19 October 2026 (text) | - | | S118 | EDGE-REG-21 Clinical risk management file (DCB0129) | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/21-clinical-risk-management-file.md` | `f4ca0adff3bdbd00d7c94c459e6ef68911bc4c58` | 19 October 2026 (text) | - | | S119 | EDGE-REG-22 Hazard log: Health Record | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/22-hazard-log-health-record.md` | `4a2f966259bf0058664affaaa71e480e1fac13ab` | 19 October 2026 (text) | - | | S120 | EDGE-REG-23 Hazard log: Imaging | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/23-hazard-log-imaging.md` | `06927dce172488660272bf656f1f2179379dafde` | 19 October 2026 (text) | - | | S121 | EDGE-REG-24 Hazard log: Radiology | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/24-hazard-log-radiology.md` | `1c4bd53cfb84fd10ce666ba2094275a70089ef5e` | 19 October 2026 (text) | - | | S122 | EDGE-REG-25 Hazard log: Care Pathways | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/25-hazard-log-care-pathways.md` | `d475a7cf1f942f0a06f277535a07164f5e3d410f` | 19 October 2026 (text) | - | | S123 | EDGE-REG-26 Hazard log: Lab and Pathology | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/26-hazard-log-lab-and-pathology.md` | `d237f2a9f53098da570de96f75829fa3e76fa073` | 19 October 2026 (text) | - | | S124 | EDGE-REG-27 Hazard log: Genomics | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/27-hazard-log-genomics.md` | `b6e2d2de80887339911df9e4fb4593004efa42c4` | 19 October 2026 (text) | - | | S125 | EDGE-REG-28 Hazard log: Sequencing | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/28-hazard-log-sequencing.md` | `e060b0030ca965bbcf3bbd1f77aa72f56438a240` | 19 October 2026 (text) | - | | S126 | EDGE-REG-29 Digital Technology Assessment Criteria: form version 2.0, completed | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/29-dtac.md` | `6f3393858e2cbcb8d6006e3b8de7471dd860b642` | 19 October 2026 (text) | - | | S127 | EDGE-REG-30 Online Safety Act 2023: illegal content risk assessment for JudeAI, the Agent | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/30-osa-illegal-content.md` | `40827b08194e34a3c1a72d96a4eb0a91f072d248` | 19 October 2026 (text) | - | | S128 | EDGE-REG-31 Online Safety Act 2023: children's access assessment for JudeAI, the Agent Cha | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/31-osa-children.md` | `20aff96595d9e8a9638a3f575a199d014e912b00` | 19 October 2026 (text) | - | | S129 | EDGE-REG-32 Trial agreement: Consortium sandbox trial for a university | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/32-trial-agreement.md` | `39303c6a72e1332bb1ad777bcb89a1ec849ced9c` | 19 October 2026 (text) | - | | S130 | EDGE-REG-33 Data processing agreement | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/33-data-processing-agreement.md` | `499390fab2d12858ec9e4ff1a9c7a66b44376fd7` | 19 October 2026 (text) | - | | S131 | EDGE-REG-34 Service levels and support statement | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/34-service-levels.md` | `a93c1d3714801842c8cf7a6e5a69d92762d60668` | 19 October 2026 (text) | - | | S132 | EDGE-REG-35 Incident response and personal data breach plan | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/35-incident-response.md` | `12e02115e8373842f6e41451503604e374f9665c` | 19 October 2026 (text) | - | | S133 | EDGE-REG-36 Business continuity and disaster recovery plan | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/36-business-continuity.md` | `9228de574cb0fe400e83af61105019d3b1374a94` | 19 October 2026 (text) | - | | S134 | EDGE-REG-37 Vulnerability disclosure policy | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/37-vulnerability-disclosure.md` | `e421896001f44d10a9b2374abb10d1cf2ffd742b` | 19 October 2026 (text) | - | | S135 | EDGE-REG-38 Exit plan and deletion certificate | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/38-exit-plan.md` | `acd211b9167f064c88aa269d1be2ba99339c1eb1` | 19 October 2026 (text) | - | | S136 | EDGE-REG-39 Acceptable use and trial user terms | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/39-acceptable-use-and-trial-terms.md` | `c3b658cdb8a7edc26ea93f4da77db44288b18972` | 19 October 2026 (text) | - | | S137 | EDGE-REG-40 Data protection impact assessment and record of processing: Consortium | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/40-dpia-consortium.md` | `067a31c021f05d2a4bcde6fb0c5478d13b9a5c2a` | 19 October 2026 (text) | - | | S138 | Historical contract package | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/contract/README.369f5fca.md` | `369f5fca5a538e30b178c727c04d6c0eeb228c8c` | 2026-10-25 (branch) | - | | S139 | Consortium university-trial regulatory pack: evidence matrix | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/evidence/2026-10-19/codex/regulatory-pack/EVIDENCE-MATRIX.md` | `a42ff64b09b904b5a41ac264aa1469685f2cee18` | 2026-10-19 (text) | - | | S140 | Estate Level Readiness | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/handoffs/AEC-STANDARD.md` | `940e55847b21daa3cead1ec73cdb42f378b97ce2` | 21 October 2026 (text) | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/handoffs/AEC-STANDARD.0b066228.md` `0b066228008f33ca6b1f95947e6b78f3e4901771` | | S141 | Online Safety Act 2023: illegal content risk assessment for JudeAI, Agent Channel and Jude | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/online-safety-20261021/ONLINE-SAFETY-ACT-ILLEGAL-CONTENT-RISK-ASSESSMENT-20261021.md` | `5b65a2f4c8ccdb1d4192d4bda0c1ce454a2a7bee` | 21 October 2026 (text) | - | | S142 | Documentation policy — concise and traceable | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/procurement-acceptance/DOCUMENTATION-POLICY.md` | `bcadcb480ef17b0f3d6de424b11b658cb3b5b130` | 2026-10-25 (branch) | - | | S143 | Organisational requirements: ISO and standards, portals, databases (20 October 2026) | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/proof/ORGANISATIONAL-REQUIREMENTS-20261020.md` | `8a90b49fdef6e4c43a88ed0b3c32b8ddcb9f956b` | 20 October 2026 (text) | `legacy/local-source-snapshots/programme/proof/ORGANISATIONAL-REQUIREMENTS-20261020.md` `7d6c04981d52914e6fdce59f18a2bb39985db37b` | | S144 | Carbon Reduction Plan (Procurement Policy Note 006) | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/proof/procurement/documents/CARBON-REDUCTION-PLAN.md` | `6e9e52008ffa74de1c7de6c0381e4cdad58c391a` | 21 October 2026 (text) | - | | S145 | Processor terms register (UK GDPR article 28) | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/proof/procurement/documents/PROCESSOR-TERMS-REGISTER.md` | `29f9e0d3f18c280958011d11b81154817ba64243` | 21 October 2026 (text) | - | | S146 | UK GDPR and the ICO — X-X.TECH | `legacy/xx-search/public/data-protection.html` | `3cbb8817c8eb31a9e6d58fb7f374e4555b75c042` | 21 August 2026 (text) | `legacy/xx-search/public/data-protection.2de3c8a2.html` `2de3c8a21e59e01c57b693fa9f254253975bc8a0` | | S147 | Financial services, the FCA and PRA — X-X.TECH | `legacy/xx-search/public/financial-services.html` | `c6ca0a1cc1492c4999b008755e10f2ad93838f6a` | undated | `legacy/xx-search/public/financial-services.c35f93ee.html` `c35f93ee07126b28c27f08a967e0988dcd047672` | | S148 | X-X.TECH — Cyber Governance, Resilience, Risk & Compliance | `legacy/xx-search/public/index.html` | `70e485ed2ca10f8ef7f0220bed041fac980cc762` | undated | `legacy/xx-search/public/index.b715efce.html` `b715efce5a1bbd7f1f5f73323dddb1c4da94b2c3` | | S149 | Martyn’s Law and the SIA — X-X.TECH | `legacy/xx-search/public/martyns-law.html` | `7092631e8dcb4bc822228533b827f3748a624e6c` | undated | `legacy/xx-search/public/martyns-law.893a928b.html` `893a928bf414642894a0f096268d2e1383a1d083` | | S150 | Remedial regulatory response — X-X.TECH | `legacy/xx-search/public/remedial-regulatory-response.html` | `a8114fb258aad0c453d6831dcd673431eeb0b460` | undated | `legacy/xx-search/public/remedial-regulatory-response.59b8e218.html` `59b8e218c1021a942e4142cd8e6dfcf4255c6b6c` | | S151 | JudeAI Benchmark Portal (specification, contract 2.9.21, P75) | `old-repo/handover/benchmark-standard/PORTAL-SPEC.md` | `52d7c8ac37d92cd578f49356e3fe006a17ddec16` | 2026-10-30 (branch) | - | | S152 | Score record (one JSON per PUBLISHED_BENCHMARK row), `scores/.json` | `old-repo/handover/benchmark-standard/SCORE-RECORD-SCHEMA.md` | `b6514f69e9b3f13b96e93c27d430e91bd75f1288` | 2026-09-01 (text) | - | | S153 | JudeAI Benchmark Standard (proposed standard, contract 2.9.21, P75) | `old-repo/handover/benchmark-standard/STANDARD.md` | `b00ad18e4d2a3b5b9eb98a0aecf4340fabd24b65` | 2026-10-30 (branch) | `old-repo/handover/benchmark-standard/STANDARD.bafb4072.md` `bafb40723fc8b724faaac267deb49d025f05bfce` | | S154 | Governed Intelligence Source Licence | `old-repo/handover/source-context/LICENCE.md` | `51a75e1729c5ec7a1f937183a00d945e1fe153ea` | 2026-10-29 (branch) | `legacy/governed-intelligence-judeai-edgeid-pre-contract/LICENCE.md` `f8628f9b19c48b20cfdae5ca14c25a68a26ea942`; `legacy/governed-intelligence-judeai-edgeid-pre-contract/LICENCE.51a75e17.md` `51a75e1729c5ec7a1f937183a00d945e1fe153ea`; `old-repo/LICENCE.md` `f8628f9b19c48b20cfdae5ca14c25a68a26ea942` | | S155 | standards": [ | `old-repo/run/plan/STANDARDS.json` | `0f6ce87ffb748804f580490dea86428e09b422d0` | 2026-10-29 (branch) | - | ###### 1.3 Manuals, handbooks, guides, playbooks and runbooks (152 distinct documents, 180 catalogue rows) | # | Title | Corpus-relative path | Blob oid | Date (how dated) | Near-identical variants (path, oid) | |---|---|---|---|---|---| | M1 | JudeAI / EDGE ID Estate — Jude agentic banking user manual | `legacy/edge-judiverse-tech/docs/bank/JUDE-AGENTIC-BANKING-USER-MANUAL.md` | `6d54d31e24fa80d62109c2778fbddb0ea2a388fd` | 13 October 2026 (text) | - | | M2 | intenJude and Jude Office handbook | `legacy/edge-judiverse-tech/docs/benchmarks/INTENJUDE-AND-JUDE-OFFICE-HANDBOOK.md` | `f25cb2d0b95826388d7f67a06e1ee030a41b4af6` | 13 October 2026 (text) | - | | M3 | JudeAI / EDGE ID Estate — Carbon API engineering handbook | `legacy/edge-judiverse-tech/docs/carbon/CARBON-API.md` | `6d4c0b913a3565d0671815494b656189de354e2f` | 13 October 2026 (text) | - | | M4 | JudeAI / EDGE ID Estate — Judiverse Mail connector handbook | `legacy/edge-judiverse-tech/docs/communications/JUDIVERSE-MAIL-CONNECTOR.md` | `afd15133c22c901c85559b12aa44f55938ccc94d` | 13 October 2026 (text) | - | | M5 | JudeAI / EDGE ID Estate communications manual | `legacy/edge-judiverse-tech/docs/communications/README.md` | `869e48f2cca3838e3eccff1774cb174f6286f183` | 2026-10-13 (branch) | `legacy/edge-judiverse-tech/docs/communications/README.a40cf024.md` `a40cf02449aa6df65bf531697913e48165cef64a` | | M6 | Get started with Jude Drive | `legacy/edge-judiverse-tech/docs/drive/HANDBOOK.md` | `59cc14b66b8f16dfa62883689060742d5d7c46e5` | 2026-10-13 (branch) | `legacy/edge-judiverse-tech/docs/drive/HANDBOOK.f56e5a9d.md` `f56e5a9daf41260c7e5ebf349ab66ae9a9af6e47` | | M7 | JudeAI / EDGE ID Estate Engineering handbook — intenJude MCP | `legacy/edge-judiverse-tech/docs/engineering/INTENJUDE-MCP.md` | `cad2423f21357692e7e454543253acc16f05994f` | 2026-10-13 (branch) | - | | M8 | Meet handbook · JudeAI / EDGE ID Meet Back to polls Download handbook PDF JudeAI / EDGE ID | `legacy/edge-judiverse-tech/docs/evidence/meet/release-20261013/live-approved-source/static/policies.html` | `850cf7e800e754b037246825c5171011b28115e3` | 13 October 2026 (text) | `legacy/local-source-snapshots/packages/meet/static/handbook.html` `44047a45934455a0d39faae9c646c9259eaf24c7`; `legacy/edge-judiverse-tech/packages/meet/static/handbook.html` `29420785a4e1724d8e2389ca927eb880c94055b7` | | M9 | JudeAI / EDGE ID Estate — Jude Accounts engineering handbook | `legacy/edge-judiverse-tech/docs/finance/JUDE-ACCOUNTS.md` | `33c2cc3c303afdd057a97bb01adea956af53b8b2` | 13 October 2026 (text) | - | | M10 | Jude Bank user manual: letting Jude act as your personal banker | `legacy/edge-judiverse-tech/docs/finance/bank/08-JUDE-BANKING-USER-MANUAL.md` | `2075cccb0e8361f7e99ba61c4edd801c52a751ba` | 13 October 2026 (text) | - | | M11 | JudeAI / EDGE ID Estate — native publishing handbook | `legacy/edge-judiverse-tech/docs/handover/README.md` | `fdebe47c2fd94d790c54c0ac502c6768ae8f8f0b` | 2026-10-13 (branch) | - | | M12 | JudeAI / EDGE ID Estate — engineering and journey handbook | `legacy/edge-judiverse-tech/docs/journeys/README.md` | `c73def02e35e40c24f083cd9f3d4c271977940ce` | 13 October 2026 (text) | - | | M13 | JudeAI / EDGE ID Estate — intenJude: whole-life engineering manual | `legacy/edge-judiverse-tech/docs/journeys/engineering/P09-intenjude-journeys.md` | `f3c5729324cf0468094dd399b284c1c37b40a286` | 13 October 2026 (text) | - | | M14 | JudeAI / EDGE ID Estate — Intelligent Scheduler: whole-life engineering manual | `legacy/edge-judiverse-tech/docs/journeys/engineering/P10-intelligent-scheduler-journeys.md` | `e251c35f9a0dc8a03768bb65723d1303cec5730d` | 13 October 2026 (text) | - | | M15 | JudeAI / EDGE ID Estate — API Connections: whole-life engineering manual | `legacy/edge-judiverse-tech/docs/journeys/engineering/P13-api-connections-journeys.md` | `afecc1dd49f79bba5062e36a47aa11ee83ef4c81` | 13 October 2026 (text) | - | | M16 | JudeAI / EDGE ID Estate — Terminal: whole-life engineering manual | `legacy/edge-judiverse-tech/docs/journeys/engineering/P25-terminal-journeys.md` | `aec9b490c44685cf45ad8708e6b177f10ce025cb` | 13 October 2026 (text) | - | | M17 | JudeAI / EDGE ID Estate — Databases: whole-life engineering manual | `legacy/edge-judiverse-tech/docs/journeys/engineering/P54-databases-journeys.md` | `b69a97854ee9caf04aaff8bf8c5b6dd3ec8b7bd4` | 13 October 2026 (text) | - | | M18 | JudeAI / EDGE ID Estate — Administration: whole-life engineering manual | `legacy/edge-judiverse-tech/docs/journeys/engineering/P55-administration-journeys.md` | `93e3f42ce8b85681be011bda56bbe20e26cd4b09` | 13 October 2026 (text) | - | | M19 | JudeAI / EDGE ID Estate — Operations: whole-life engineering manual | `legacy/edge-judiverse-tech/docs/journeys/engineering/P56-operations-journeys.md` | `8db03bd2e7b69f3efd671289a6336a2c9d550ecf` | 13 October 2026 (text) | - | | M20 | JudeAI / EDGE ID Estate — Cyber: whole-life engineering manual | `legacy/edge-judiverse-tech/docs/journeys/engineering/P57-cyber-journeys.md` | `13118994aabd4f4523fed71102adb03162439a03` | 13 October 2026 (text) | - | | M21 | JudeAI / EDGE ID Estate — Carbon: whole-life engineering manual | `legacy/edge-judiverse-tech/docs/journeys/engineering/P58-carbon-journeys.md` | `b90a2087d5e95cb1504fb196c089802c80c85435` | 13 October 2026 (text) | - | | M22 | JudeAI / EDGE ID Estate — ReCompute: whole-life engineering manual | `legacy/edge-judiverse-tech/docs/journeys/engineering/P61-recompute-journeys.md` | `7910d9bca22d589aa624c7f68ba6c4deda66ff67` | 13 October 2026 (text) | - | | M23 | JudeAI / EDGE ID Estate — Chip Design: whole-life engineering manual | `legacy/edge-judiverse-tech/docs/journeys/engineering/P62-chip-design-journeys.md` | `3ed4acbce714dabbeb32df805040c5fa528fb145` | 13 October 2026 (text) | - | | M24 | JudeAI / EDGE ID Estate — Procurement: whole-life engineering manual | `legacy/edge-judiverse-tech/docs/journeys/engineering/P63-procurement-journeys.md` | `447818e1a808d99e7c383a2e51409557e8b81994` | 13 October 2026 (text) | - | | M25 | JudeAI / EDGE ID Estate — Logistics: whole-life engineering manual | `legacy/edge-judiverse-tech/docs/journeys/engineering/P64-logistics-journeys.md` | `dc4bc5e029529ae99c5f69c12a3882938456f7bc` | 13 October 2026 (text) | - | | M26 | JudeAI / EDGE ID Estate — Audit and Evidence: whole-life engineering manual | `legacy/edge-judiverse-tech/docs/journeys/engineering/P68-audit-and-evidence-journeys.md` | `158501fd101a9c4f2ae1846a879472939c22c6d2` | 13 October 2026 (text) | - | | M27 | JudeAI / EDGE ID Estate — Integrations: whole-life engineering manual | `legacy/edge-judiverse-tech/docs/journeys/engineering/P70-integrations-journeys.md` | `1d63ce2e9558e3654cfb91fe14967348ba430c75` | 13 October 2026 (text) | - | | M28 | JudeAI / EDGE ID Estate — Developer Portal: whole-life engineering manual | `legacy/edge-judiverse-tech/docs/journeys/engineering/P71-developer-portal-journeys.md` | `70e05f93311ff8ce12dfa72e2ca0133a5da5259c` | 13 October 2026 (text) | - | | M29 | JudeAI / EDGE ID Estate — Digital Twins: whole-life engineering manual | `legacy/edge-judiverse-tech/docs/journeys/engineering/P72-digital-twins-journeys.md` | `533e2d398044e21eae7c6c4f67504ae34e34e2e6` | 13 October 2026 (text) | - | | M30 | JudeAI / EDGE ID Estate — Search: whole-life engineering manual | `legacy/edge-judiverse-tech/docs/journeys/engineering/P77-search-journeys.md` | `4abffe3d358e45f877df9756c967aee467c79933` | 13 October 2026 (text) | - | | M31 | JudeAI / EDGE ID Estate — Testing & Evidence: whole-life engineering manual | `legacy/edge-judiverse-tech/docs/journeys/engineering/P78-testing-and-evidence-journeys.md` | `f9d972e3e5d0850d89c325a80d4bece4008afb1f` | 13 October 2026 (text) | - | | M32 | Agent governance operator and reviewer guide | `legacy/edge-judiverse-tech/docs/manuals/AGENT-GOVERNANCE-MANUAL.md` | `8330d8d845d5b73e9b3fd7caf430ed6a3472cc7c` | 13 October 2026 (text) | - | | M33 | Get started with your estate | `legacy/edge-judiverse-tech/docs/manuals/GETTING-STARTED.md` | `9b622e14c6198dfb1ece5958889e1cd3a46929b6` | 2026-10-13 (branch) | - | | M34 | Financial services operator and reviewer guide | `legacy/edge-judiverse-tech/docs/manuals/fca-operator-guide.md` | `672cca6b8ec59e2d1d97d0b9b20b98bfcb69a614` | 13 October 2026 (text) | - | | M35 | Government operator and purchaser guide | `legacy/edge-judiverse-tech/docs/manuals/government-operator-guide.md` | `b0a408b7ed71ac5ede949222f430f35f9389fe14` | 13 October 2026 (text) | - | | M36 | ICB and care-provider deployment and user guide | `legacy/edge-judiverse-tech/docs/manuals/icb-deployment-and-user-guide.md` | `507cca8578f8f5ebfe7b1ceea4824e1caf4df342` | 13 October 2026 (text) | - | | M37 | JudeAI / EDGE ID Meet — deployment runbook | `legacy/edge-judiverse-tech/docs/meet/DEPLOYMENT.md` | `fa60a48f451fe740d3e81e0d80d4f2f24489af7c` | 13 October 2026 (text) | `legacy/edge-judiverse-tech/docs/meet/DEPLOYMENT.c3e76491.md` `c3e764919fc3c1ec6e7b6194f202e439c442e9c0` | | M38 | Get started with Meet | `legacy/edge-judiverse-tech/docs/meet/HANDBOOK.md` | `2a109e6815042381d5fac3bd61d874e25e2237fe` | 2026-10-13 (branch) | `legacy/edge-judiverse-tech/docs/meet/HANDBOOK.a43094c7.md` `a43094c7e84d2601da7314a5bcc88d3659c48510` | | M39 | Meet Release 1: operating manual | `legacy/edge-judiverse-tech/docs/meet/R1/OPERATIONS.md` | `05f3b2571422b3e32a52096fe6dfbe326e46967a` | 2026-10-13 (branch) | `legacy/edge-judiverse-tech/docs/meet/R1/OPERATIONS.1cc6e18e.md` `1cc6e18ede850b82bdbb5c9cac930a3a8a960945` | | M40 | JudeAI / EDGE ID Estate — Backup and Recovery Runbook | `legacy/edge-judiverse-tech/docs/operations/BACKUP-AND-RECOVERY.md` | `e17f8edea60cd5f08447390fbb99063245983745` | 2026-10-13 (branch) | - | | M41 | Jude - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P01-jude.md` | `9a940614cec7155497ea10fc283bf1d0267c920c` | 2026-10-13 (branch) | - | | M42 | Canvas - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P02-canvas.md` | `a4a6e12b6a45afb41c4caf3d3c3397597f0b44ab` | 2026-10-13 (branch) | - | | M43 | Tree of Knowledge - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P03-tree-of-knowledge.md` | `aa66b3a98ce10e59bab9008300c1a30703ef9c2f` | 2026-10-13 (branch) | - | | M44 | EDGE ID - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P04-edge-id.md` | `b6392bdd7d36206832cfc761e46bba1b9c162f0b` | 2026-10-13 (branch) | - | | M45 | Personal Record - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P05-personal-record.md` | `a5bcb3213de220be8f40f0958fcceb9649bd47cc` | 2026-10-13 (branch) | - | | M46 | Vault - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P06-vault.md` | `70c42859ded3555dc27317ddb8b7d2b6122bed92` | 2026-10-13 (branch) | - | | M47 | Research and IP - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P07-research-and-ip.md` | `a1b9fad79c15670ccd1beb8d548c1ad9fa7009d3` | 2026-10-13 (branch) | - | | M48 | More Apps - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P08-more-apps.md` | `23eb001a047b088cf7016a12ceabda8440d4d64f` | 2026-10-13 (branch) | - | | M49 | intenJude - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P09-intenjude.md` | `2d0a4bcf13cc17d27d8655f6fac65b5cc16beada` | 2026-10-13 (branch) | - | | M50 | Intelligent Scheduler - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P10-intelligent-scheduler.md` | `9de862f2401f93d9c54844edb70f5de9794820cc` | 2026-10-13 (branch) | - | | M51 | Training - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P11-training.md` | `2f44cba1b2513fece1030f5430a96af480308186` | 2026-10-13 (branch) | - | | M52 | ARU learning lab - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P12-aru-learning-lab.md` | `2cea80ca3b1824e0d81625dd4d4d82927b0461c3` | 2026-10-13 (branch) | - | | M53 | API Connections - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P13-api-connections.md` | `a3d56b2c14a3c3efa2fb9e389a64435f5196755b` | 2026-10-13 (branch) | - | | M54 | Agent Channel - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P14-agent-channel.md` | `1e4a863b10a504b5c11c7c54d1db4a1795bd9acb` | 2026-10-13 (branch) | `legacy/edge-judiverse-tech/docs/products/P14-agent-channel.92ab63ac.md` `92ab63acab023c07b53a0a8e5bea693140d37a16` | | M55 | Baton - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P15-baton.md` | `441ae8ce33a4154d089d661866ccd6cb64176abf` | 2026-10-13 (branch) | - | | M56 | Office - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P16-office.md` | `1ab81e20476cd6ac687e3cfeb5e0e2d36d9d04e9` | 2026-10-13 (branch) | - | | M57 | Markdowns - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P17-markdowns.md` | `78f8bd7e1a4be0fb61151e2f12546f88fbc61614` | 2026-10-13 (branch) | - | | M58 | Documents - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P18-documents.md` | `99547925fac8876dbb21d6a406a4fa4177af0ad8` | 2026-10-13 (branch) | - | | M59 | Sheets - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P19-sheets.md` | `cb5a9d799042e1415e9e1b1393927b4d362952a4` | 2026-10-13 (branch) | - | | M60 | Slides - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P20-slides.md` | `24499f8f98d5be72f2c52c39780cc9fc09cdf6a5` | 2026-10-13 (branch) | - | | M61 | Judiverse Mail - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P21-mail.md` | `b02c31829b791c5356f5361c32f33a4a0374fb2b` | 2026-10-13 (branch) | - | | M62 | Messages - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P22-messages.md` | `aea8b6fa52aadb6940f62de76757352a68f92b6b` | 2026-10-13 (branch) | - | | M63 | Calls - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P23-calls.md` | `58877c4bef0c1ba11d612d24a06332dd244b4f0d` | 2026-10-13 (branch) | - | | M64 | Calendar and Tasks - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P24-calendar-and-tasks.md` | `847f1878f2c87d2877597c6204ba9eab3d2d43a1` | 2026-10-13 (branch) | - | | M65 | Terminal - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P25-terminal.md` | `09fe8de959244e42179ab512b50a3b8d95939cfd` | 2026-10-13 (branch) | - | | M66 | Jude Bank - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P26-jude-bank.md` | `dbdf3aae130009e399518eb396dbec60033a308c` | 2026-10-13 (branch) | - | | M67 | Accounts - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P27-accounts.md` | `383a1c558d4f6f9e3cc39308b900bad0ec28ec84` | 2026-10-13 (branch) | - | | M68 | Statements - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P28-statements.md` | `206f7bf4eb88ea5c08bed46d51971d8407c7b24c` | 2026-10-13 (branch) | - | | M69 | Payments - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P29-payments.md` | `ae3d6a172ebf4e00e154dd569f1812033b20fe78` | 2026-10-13 (branch) | - | | M70 | Open Banking APIs - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P30-open-banking-apis.md` | `834b3373df4770174a5150c350cb2a3b304c6c95` | 2026-10-13 (branch) | - | | M71 | Cards - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P31-cards.md` | `a2b5d6393fd1a65aef661fcfa1bb7c15a524e42c` | 2026-10-13 (branch) | - | | M72 | Credit - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P32-credit.md` | `885d08142d419e8455716bc2ba899b17414e9dea` | 2026-10-13 (branch) | - | | M73 | Jude Accounts - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P33-business-finance.md` | `6cb26cf4166cd492d27f47fd1f471aeb836f3b59` | 2026-10-13 (branch) | - | | M74 | Payroll - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P34-payroll.md` | `2c63adbc512fce3c8313cd55390c1af3c418ed96` | 2026-10-13 (branch) | - | | M75 | Treasury - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P35-treasury.md` | `dad368aee1ac9a823583dd93634925e423a8d772` | 2026-10-13 (branch) | - | | M76 | AML Operations - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P36-aml-operations.md` | `cf4aa7b589112d8db7fca0db26efd066eb651f26` | 2026-10-13 (branch) | - | | M77 | Fraud Intelligence - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P37-fraud-intelligence.md` | `ff2b27a12bb8224f597f00f0e0d38c3f3e87864d` | 2026-10-13 (branch) | - | | M78 | Case Management - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P38-case-management.md` | `9ad47557d1569a316e0166f03beef41a7a882c5c` | 2026-10-13 (branch) | - | | M79 | Recovery Authorisation - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P39-recovery-authorisation.md` | `570fe0e1bc1268f5814d0fc4c7b1f24822533ab0` | 2026-10-13 (branch) | - | | M80 | Decision Passports - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P40-decision-passports.md` | `13d1ee8ab3f26d427207429437e5db15673f22ae` | 2026-10-13 (branch) | - | | M81 | Health Record - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P41-health-record.md` | `992e0db79a01a43a9d16a4814929db7d4cf39b50` | 2026-10-13 (branch) | - | | M82 | Imaging - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P42-imaging.md` | `cc39d10ff88b6aaa3da243d1ccbb54a41a9db870` | 2026-10-13 (branch) | - | | M83 | Radiology - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P43-radiology.md` | `38f68c262e87df5c55210369cbb48c248a1b0c19` | 2026-10-13 (branch) | - | | M84 | Clinical Trials - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P44-clinical-trials.md` | `7c4b377838d4c147c0465fee785ba4f66271d0e7` | 2026-10-13 (branch) | - | | M85 | Care Pathways - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P45-care-pathways.md` | `c0115ea2b23bda25a8164d734f015a305eddc4c2` | 2026-10-13 (branch) | - | | M86 | Lab and Pathology - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P46-lab-and-pathology.md` | `ef5fa5339bff8039650a46c82d1ecdb1d2d5a26c` | 2026-10-13 (branch) | - | | M87 | Genomics - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P47-genomics.md` | `803aec6d913b5cc8844aa922e1b4b9d8fc81ab12` | 2026-10-13 (branch) | - | | M88 | Sequencing - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P48-sequencing.md` | `3a53b4be16a8df9bc01d4e496807a9ad75085f77` | 2026-10-13 (branch) | - | | M89 | Bioinformatics - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P49-bioinformatics.md` | `c9faa709ac594d37567568ca5f96b5a54472a243` | 2026-10-13 (branch) | - | | M90 | Wearables - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P50-wearables.md` | `51f0086111596864700cbfdf10068ee3f16f5510` | 2026-10-13 (branch) | - | | M91 | Research - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P51-research.md` | `b3896c0779168d3d04bad18b8821a8d00a518822` | 2026-10-13 (branch) | - | | M92 | Data Catalogue - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P52-data-catalogue.md` | `5acf6942ff5c7750099b44f2184e8ec5239947a5` | 2026-10-13 (branch) | - | | M93 | Data Workspace - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P53-data-workspace.md` | `23ea11649cf31afd03711da0e9f98ac055cabe97` | 2026-10-13 (branch) | - | | M94 | Databases - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P54-databases.md` | `ab66984818272c281fb75b4d070e300ebfc22ea9` | 2026-10-13 (branch) | - | | M95 | Administration - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P55-administration.md` | `c2001c4a228a07d74eba8907e79e0b690ff2fab9` | 2026-10-13 (branch) | - | | M96 | Operations - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P56-operations.md` | `a4d389cd131c8f68bed09d0e7d0091a1bf951b1e` | 2026-10-13 (branch) | - | | M97 | Cyber - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P57-cyber.md` | `6530238ce5eca55bd79528ebc70817493c46aa03` | 2026-10-13 (branch) | - | | M98 | Carbon - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P58-carbon.md` | `16c8363f13a762f80827e3271a867b87b1a5d962` | 2026-10-13 (branch) | - | | M99 | Insurance - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P59-insurance.md` | `a0c727fa25216a2b6ad77e08acc11b8f407162bf` | 2026-10-13 (branch) | - | | M100 | Creative Studio - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P60-creative-studio.md` | `0e887da85f6de8d39df73b4f02906a2e8356a79b` | 2026-10-13 (branch) | - | | M101 | ReCompute - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P61-recompute.md` | `73b784863df9066cf50c82425b02adeb2f5fab5c` | 2026-10-13 (branch) | - | | M102 | Chip Design - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P62-chip-design.md` | `7745901c6b3f61348b7aa998fe2fa474da8b68f5` | 2026-10-13 (branch) | - | | M103 | Procurement - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P63-procurement.md` | `c6756cd7e0850474071b545a685c30dc0a6d2905` | 2026-10-13 (branch) | - | | M104 | Logistics - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P64-logistics.md` | `9db949ea8d3cf22661d25e27b7189415e46b33b8` | 2026-10-13 (branch) | - | | M105 | Notifications - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P65-notifications.md` | `a4f7cbd62c9e8830bef06c8f3c1b411613584e6e` | 2026-10-13 (branch) | - | | M106 | Account - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P66-account.md` | `5183bb60d637193fb44f7e1b614e40c9b2b02747` | 2026-10-13 (branch) | - | | M107 | Organisation - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P67-organisation.md` | `293cfe1804c2bccb08172b2ba47ac398521febf3` | 2026-10-13 (branch) | - | | M108 | Audit and Evidence - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P68-audit-and-evidence.md` | `b47d953ec1c2fb1e71a6b4a8b8f429428b9ac875` | 2026-10-13 (branch) | - | | M109 | Help and Learning - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P69-help-and-learning.md` | `800f3be0c6c72412a719009e789c2dd75f11c8c0` | 2026-10-13 (branch) | - | | M110 | Integrations - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P70-integrations.md` | `7f1469421d61f96fd07df11952299066ad5065de` | 2026-10-13 (branch) | - | | M111 | Developer Portal - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P71-developer-portal.md` | `4036a643749ee9a54cc9addbe9fe45464e40f127` | 2026-10-13 (branch) | - | | M112 | Digital Twins - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P72-digital-twins.md` | `dc660028bb55118b10a09fe218ef079dfa267350` | 2026-10-13 (branch) | - | | M113 | Online Testing - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P73-online-testing.md` | `0936271453544ed34eba5d4c1d4e10064d0bd366` | 2026-10-13 (branch) | - | | M114 | Forecasting - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P74-forecasting.md` | `3f8133b065233913fcccc287f1d78b1991947818` | 2026-10-13 (branch) | - | | M115 | Contacts - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P75-contacts.md` | `1baac45636c3015220a1a5d7c83e3f6c04fcec1f` | 2026-10-13 (branch) | - | | M116 | People - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P76-people.md` | `52d32ab0daf5104f3a573add435981d2765ce503` | 2026-10-13 (branch) | - | | M117 | Search - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P77-search.md` | `f87d786397189e538bed9b643b576587aeb2f3da` | 2026-10-13 (branch) | - | | M118 | Testing & Evidence - specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P78-testing-and-evidence.md` | `e66014c1a05ca50000fe24473a92b0495c818c7a` | 2026-10-13 (branch) | - | | M119 | JudeBox — specification and operating manual | `legacy/edge-judiverse-tech/docs/products/P81-judebox.md` | `2a245e827ce7b672f4d1f64c8ae87d576ec2725c` | 2026-10-13 (branch) | - | | M120 | Feature specifications and manuals | `legacy/edge-judiverse-tech/docs/products/README.md` | `c6b77e1aa4a1d462c333f032cd453fcf76c7dbf6` | 2026-10-13 (branch) | `legacy/edge-judiverse-tech/docs/products/README.9db5673c.md` `9db5673cf5cf92705928f26f4b1909dfe86affd0` | | M121 | OPRES-03. Operational playbooks | `legacy/edge-judiverse-tech/docs/resilience/OPRES-03-PLAYBOOKS.md` | `af2e5b9180a15d6c48bda3b96dbc1e3a114d5a72` | 2026-10-13 (text) | - | | M122 | JudeAI / EDGE ID Estate — App, URL and Handbook Directory | `legacy/edge-judiverse-tech/docs/review/APP-URL-HANDBOOK-DIRECTORY.md` | `9997e33f1a401b9ed004a3ce12f747f2dc2a3677` | 13 October 2026 (text) | - | | M123 | How we take risk: the working handbook for every person and agent | `legacy/edge-judiverse-tech/docs/risk/RISK-05-HANDBOOK-HOW-WE-TAKE-RISK.md` | `b809e191f4edf96ba84f850a64abd422edeb23c1` | 2026-10-13 (text) | - | | M124 | Guidelines for secure AI system development | `legacy/edge-judiverse-tech/docs/security/source-register.json` | `c516683aadb389bf4fd52215781e17834e3d730b` | 27 November 2023 (text) | - | | M125 | Privacy, devices and connections | `legacy/edge-judiverse-tech/docs/security/user-guide.md` | `6f27f8a5068e7b8eaf83fbdfcd9016e279ac7520` | 2026-10-13 (branch) | - | | M126 | JudeAI agent execution runbook | `legacy/edge-judiverse-tech/docs/tools/AGENT-RUNBOOK.md` | `1ecec5f945ba6b33138f8b3ca6d186508f33972d` | 2026-10-13 (branch) | - | | M127 | JudeAI / EDGE ID Estate Tools Handbook | `legacy/edge-judiverse-tech/docs/tools/HANDBOOK.md` | `14048e82ab13350bdc956033f404220336984ad2` | 2026-10-13 (branch) | `legacy/edge-judiverse-tech/docs/tools/HANDBOOK.0c57bdb8.md` `0c57bdb8067bfa9b14f81b749cbe84173e6acff0`; `legacy/edge-judiverse-tech/docs/tools/HANDBOOK.b6d16680.md` `b6d16680e7a8c521ead67c2ab5d979f234512632`; `legacy/edge-judiverse-tech/docs/tools/HANDBOOK.080dddd0.md` `080dddd07ea292e9dd96e8a8cbfa56e63f728833`; `legacy/edge-judiverse-tech/docs/tools/HANDBOOK.7e20bba6.md` `7e20bba6f7ee1aa01302a84818e6b96b14b1660d` | | M128 | JudeAI MCP — integration guide | `legacy/edge-judiverse-tech/docs/tools/MCP-INTEGRATION.md` | `0eb10075766efbf90550e7c0d3fbab70acf94253` | 2026-10-13 (branch) | `legacy/edge-judiverse-tech/docs/tools/MCP-INTEGRATION.e6ea3740.md` `e6ea37403daad43c006e46e197478c61acc14b4a` | | M129 | JudeAI / EDGE ID Estate MCP handbook | `legacy/edge-judiverse-tech/packages/jude_mcp/README.md` | `6b9e7c390738a046b2776419d1db809149f78e77` | 2026-10-13 (branch) | - | | M130 | Web Content Accessibility Guidelines 2.2, Level AA | `legacy/governed-intelligence-judeai-edgeid-pre-contract/spec/FRAMEWORKS.json` | `b375c420cf5c1ad5d85add36622164fe36d0f2b7` | 2026-10-27 (text) | `old-repo/spec/FRAMEWORKS.json` `b375c420cf5c1ad5d85add36622164fe36d0f2b7` | | M131 | Ask Jude: what Jude knows how to answer | `legacy/jude-mvp/jude-live/docs/ask-jude.md` | `53989493d02d6fcc98f5db1bc8e6f4d6f2753e16` | 2026-09-25 (branch) | `legacy/jude-mvp/docs/ask-jude.md` `319bcaa9fce4e56dccb6ba44d5d58a40baaa09de`; `legacy/jude/docs/ask-jude.md` `e59b6e9abfa4b1e8eb01e38d758d483240109b31` | | M132 | How to verify this pack without Jude | `legacy/jude-mvp/proof/unpacked/pack-20260904-163647/VERIFY.md` | `68ceebacbc95ed194aaec930f09eb31c3bfaae2f` | 2026-09-04 (path) | - | | M133 | How to verify this pack without Jude | `legacy/jude-mvp/proof/unpacked/pack-20260906-020510/VERIFY.md` | `870c2660de5eb87b0ba1f07fafa881e81178ea96` | 2026-09-06 (path) | - | | M134 | How to verify this pack without Jude | `legacy/jude-mvp/proof/unpacked/pack-20260906-021637/VERIFY.md` | `32a3c024593373af2028018e4600a644fa2fb09d` | 2026-09-06 (path) | - | | M135 | Agent and test accounts: where they are and how to use them | `legacy/judeai-edge-control/Master/References/agent-accounts.md` | `80e7b97ec1482941e2c9519c757c6ba855738131` | 2026-10-13 (branch) | - | | M136 | Compute: the operator's manual | `legacy/judeai-edge-control/docs/compute/MANUAL.md` | `0efdd37c574c1e993cc29e2dc2608cb10d95d04d` | 13 October 2026 (text) | - | | M137 | Public manuals and specification catalogue | `legacy/judeai-edge-control/docs/research/life-control-plane/manual_catalogue.md` | `3a139cfd335484702dfae2c9bbf2c523ca7acd0c` | 2026-10-10 (text) | - | | M138 | Backend lane / [old-estate agent identity] | `legacy/judeai-edgeid-cloud/handover/completion-20261025-uriel1717/backend/RUNBOOK.md` | `ab4e86a5a97107311c3baee797a6e9521d0c1b4c` | 2026-10-25 (path) | - | | M139 | Estate agent onboarding | `legacy/local-source-snapshots/AGENTS.md` | `ce500049676cc1ea5596e54a73df3e6187a8a6fb` | 20 October 2026 (text) | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/AGENTS.c31324e8.md` `c31324e86a447d5816b469b1f83b953a779cc544`; `legacy/x-judeai-edge-tok-kol-ndc-ml-x/AGENTS.f13db5b8.md` `f13db5b8a1820e0e3c6f257df7591a6e1e9069b4`; `legacy/x-judeai-edge-tok-kol-ndc-ml-x/AGENTS.md` `3b1570ed978b2815d0584a0bce7a63a0ae1540a4` | | M140 | Designing and reviewing a JudeAI journey | `legacy/local-source-snapshots/foundations/canvas/reference/canvas/canvas-review/ui-excellence/handbook/document.md` | `a3277d4ab67e5fdaa3cc33480a3632e39b4d3991` | 13 October 2026 (text) | - | | M141 | (no title line) | `legacy/local-source-snapshots/foundations/canvas/reference/canvas/canvas-review/ui-excellence/handbook/index.html` | `bc193d950a05076f7e8c64125f7631324e968b5c` | 13 October 2026 (text) | - | | M142 | Meet handbook · JudeAI / EDGE ID Meet Back to polls JudeAI / EDGE ID Meet — benchmarks and | `legacy/local-source-snapshots/packages/meet/static/benchmarks.html` | `6458cb22cd84d990eb76f79957edb7fcc33cd55b` | 13 October 2026 (text) | `legacy/edge-judiverse-tech/packages/meet/static/benchmarks.html` `ed97430522c5641d19dbd604d866972ab1e84c14` | | M143 | Meet handbook · JudeAI / EDGE ID Meet Back to polls JudeAI / EDGE ID Meet — deployment run | `legacy/local-source-snapshots/packages/meet/static/deployment.html` | `81d75e36563b8b93c546676a5bdb495ce2fb405f` | 13 October 2026 (text) | `legacy/edge-judiverse-tech/packages/meet/static/deployment.html` `5440dfd0717f85344463c804ceb21080b22502f0` | | M144 | Meet handbook · JudeAI / EDGE ID Meet Back to polls JudeAI / EDGE ID Meet — passkeys | `legacy/local-source-snapshots/packages/meet/static/passkeys-guide.html` | `b38934a97139d6b2d2b7939a7cbf5fc637da5344` | 13 October 2026 (text) | `legacy/edge-judiverse-tech/packages/meet/static/passkeys-guide.html` `9c32f73c58341ebc6b2b6e78096c7b4c86d7b502` | | M145 | Meet handbook · JudeAI / EDGE ID Meet Back to polls JudeAI / EDGE ID Meet — official sourc | `legacy/local-source-snapshots/packages/meet/static/sources.html` | `d244d3810f84d9f21d0293f8910082ddecd837ec` | 13 October 2026 (text) | `legacy/edge-judiverse-tech/packages/meet/static/sources.html` `2296eba9747ac2f3be590ce39ee00cf9fb945322` | | M146 | Administrator Guide | `legacy/local-source-snapshots/programme/product/DOCUMENTATION/ADMINISTRATOR-GUIDE-v1.0.md` | `6abd2edeacafa5c0ad4a1fe06afe83b2c0d7b2be` | 20 October 2026 (text) | - | | M147 | Financial crime controls statement: how the service lets a regulated firm meet the FCA Fin | `legacy/local-source-snapshots/programme/proof/procurement/documents/FINANCIAL-CRIME-CONTROLS-STATEMENT.md` | `5a58c94beac932c4788914ee7b7078a07e30f397` | 20 October 2026 (text) | - | | M148 | EDGE chat-native agent onboarding — implementation contract | `legacy/peak-judeai-behavioural-harness-baseline-full-application/docs/edge-chat-native-agent-onboarding.md` | `fd20c8a00c7651c67d21c39b84421222e6c95514` | undated | - | | M149 | How to buy — X-X.TECH | `legacy/xx-search/public/how-to-buy.html` | `808bf032a53698bbd22c48387365dcd91ab05fcc` | 1 January 2026 (text) | `legacy/xx-search/public/how-to-buy.64fbac67.html` `64fbac67c373b1fe995431d03d0f05898e37193a` | | M150 | engine2 module guide | `old-repo/engine2/MODULE-GUIDE.md` | `369e9038572b5d41928bc750021a70d6362bd6b4` | 29 October 2026 (text) | `legacy/governed-intelligence-judeai-edgeid/engine2/MODULE-GUIDE.md` `369e9038572b5d41928bc750021a70d6362bd6b4` | | M151 | (no title line) | `old-repo/office/it-onboarding/index.html` | `057a3992e7514a224b184a635a3d3b203d526746` | 2026-10-30 (branch) | `legacy/governed-intelligence-judeai-edgeid/office/it-onboarding/index.html` `057a3992e7514a224b184a635a3d3b203d526746` | | M152 | (no title line) | `old-repo/office/it-onboarding/office-shell-source.html` | `fe1db480230d855813cc3df1ace33c18f8230cf0` | 2026-10-30 (branch) | `legacy/governed-intelligence-judeai-edgeid/office/it-onboarding/office-shell-source.html` `fe1db480230d855813cc3df1ace33c18f8230cf0` | ###### 1.4 Record families (in scope by category, but records rather than documents) | Family | Rows | What it is | Path pattern | Representative oid | Dates | Value to the new estate | |---|---|---|---|---|---|---| | Benchmark score records | 1,245 | One JSON record per published benchmark in the old "benchmark standard" (schema `estate.benchmark-score.v1`): source URL and content digest, published best score if read, required score, reason if unknown. Lanes: cyber 892 (mostly MITRE ATT&CK technique rows), office 200, public learning 51, AI lab 49, clinical 38, bank 13, patterns 2. Status: 1,202 UNKNOWN, 43 OBSERVED (a published best score read at source, for example AgentDojo, GAIA, BFCL v4, MCP-Atlas, Terminal-Bench, SWE-rebench, FinanceBench, TAT-QA and the MedMarks suite members) | `old-repo/handover/benchmark-standard/scores/*.json` | `99ef375c379933aca9707972f69fb3bb22a72121` | read 30 October 2026 | The 43 OBSERVED rows are a ready list of external reference scores with source digests (CLAIM that each was read at source; re-read before use). The rest are placeholders | | Synthetic exam papers | 100 | Identical-template 20-mark "synthetic evidence reasoning" exam items with mark keys, for model training | `legacy/jude-mvp/work/jude-continuation-kit-20261001/.../corpus/media/*/exam_paper.md` | `ca58e730272f90c6aeb8322e15819264a3a4b3d3` | 1 October 2026 | Low; training fixtures, not papers | | Publication receipts and evidence records | 90 | JSON receipts and short status notes proving that an agent published or read something (privacy-notice revisions, work-register publications, live-route captures) | `legacy/local-source-snapshots/programme/jps/transfers/...`, `.../programme/outcomes-register/evidence/...` | `edb6fac6eafcce4b16a66d9500b5d254cc8aef62` | 12 to 30 October 2026 | None as documents; the receipt pattern itself is research | | In-app guide data | 74 | JSON files named after EDGE screens (bank, clinic staff, consortium, governance, identity, evidence); identified by file name only, content not read | `legacy/local-source-snapshots/services/edge/web/estate/guides/*.json` | `af21fbd2f8f94b95580b18b0503d652cf0fd49d9` | 19 October 2026 (branch) | Unknown; read before any reuse | | Registers, data files and working records | 68 | Machine-readable registers and working files: the benchmark register (1,704 rows), standards and framework registers, continuity acceptance matrices, policy JSON, pull-request records, frontier benchmark results, financial-crime experiment results | listed in 1.5 | see 1.5 | 1 to 31 October 2026 | Mixed; listed individually below | | Rendered regulatory-pack PDFs | 56 | PDF editions of EDGE-REG-00 to EDGE-REG-40 (version 1.0 of 15 October 2026 and versions 1.1 and 1.2 of 19 October 2026), plus two complete-set PDFs | `legacy/*/assurance/regulatory/xxtech.reg.*.pdf` | see 1.5 | 15 and 19 October 2026 | Same content as the Markdown sources (section 3) | | Test outputs | 20 | Unit-test console output and small notes from repair bundles | `legacy/github-metadata-20261030/local-working-changes/greenfield-canonical/files/repairs/...`, `old-repo/canonical/repairs/...` | `56ee5a84dc66ad8ac13ab2711b92dbabb8aab265` | 29 to 30 October 2026 | None | | Third-party licence texts | 19 | Rust toolchain licences (Apache-2.0, MIT, GPL, BSD, ISC, OFL, Unicode, CC-BY-SA), crate security policies, WireGuard, Inter font, NVIDIA Apache licence, EFF word-list notice | `legacy/jude-mvp/work/jude-mail-engine/.toolchain/...` and others | `1a89b9054d669db9cfb06aa2ce187b37f5d50eeb` | various | Only as licence evidence for any dependency the new estate reuses | | Alert logs | 15 | Per-seed alert outputs (B1 rules, B2 tabular, M1 graph) from a synthetic financial-crime experiment | `legacy/local-source-snapshots/research/financial-crime/part-a/results/standard/seed-*/` | `5ef3181d3c50360d64842271666de775086efb0d` | 15 October 2026 | Supports the results summary in 1.5 | | Proof records | 13 | Compiled "TRL9/CRL9" proof JSON for clinic benchmarks (ISO 14971, ISO 27799, ISO/IEC 42001, UK GDPR) and frontier benchmark packs | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/packs/...` | `61ee604675488ee79a054d7f09813e52c59b3c3d` | 18 to 22 October 2026 | Low; self-awarded readiness records (CLAIM) | | Statistics provenance stubs | 12 | Pointers to NIST/SEMATECH statistics handbook sections (binomial, Wilson interval, Poisson) | `legacy/peak-judeai-behavioural-harness-baseline-full-application/tok/root/evidence/provenance_checks/` | `2093eac5ded13feffb35177bbe6fdd9b2ad75b27` | 15 October 2026 | Low | | Security-patrol working files | 11 | Author briefs, editorial reviews, paper-review checklist and source snapshots for a security "paper delivery" lane | `legacy/local-source-snapshots/programme/security-patrol/` | `6b242a0f4534d7501e6dd83bedebadaf730a321e` | 18 to 19 October 2026 | Low; the paper-review checklist is a reusable idea | | Copies of third-party web pages | 5 | Saved pages from an NHS trust's publications page, a bank's corporate site, the FCA home page, two Microsoft Teams documentation pages and an NHS data-security principle page | `legacy/jude-mvp/work/civo-*/`, `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/proof/teams-coverage-independent-20261022/sources/` | `d63788dfe4890a9f63195ea268dce63b728a044c` | 6 to 30 October 2026 | None; read the originals instead | | Continuation-kit files | 4 | Training runbooks ("paid execution NOT AUTHORISED"), a regulatory-reporting teaching batch note and a sparse-model research hypothesis | `legacy/jude-mvp/work/jude-continuation-kit-20261001/` | `a0f1b89a67703b47caf7dd39ad14a524a1ec2b86` | 1 October 2026 | Low (the hypothesis is summarised in section 2) | | Rendered-answer captures | 4 | Plain-text captures of long answers at screen widths, light and dark | `legacy/peak-judeai-behavioural-harness-baseline-full-application/verification/evidence/` | `a14d169a626321808bd199a80a3ad6949980c1d5` | 1 to 4 October 2026 | None | ###### 1.5 Row-level listing of the two substantive record families ###### Registers, machine-readable companions, data files and working records (68 rows) | Title or first line | Corpus-relative path | Blob oid | Date (how dated) | |---|---|---|---| | Add Internal Compliance Intelligence v1 | `legacy/github-metadata-20261030/github/peak-judeai-behavioural-harness-baseline-full-application/pulls/14/pull.json` | `0d6fcff30597eb70b73533c8bc45f3b572cd31cc` | 2026-10-30 (path) | | Wave 3: regulated control plane (Executive Control, Data Protection, Audit and Evidence) | `legacy/github-metadata-20261030/github/peak-judeai-behavioural-harness-baseline-full-application/pulls/8/pull.json` | `39eace6479939c8c2bf4f4db872c12f849afbbc7` | 2026-10-30 (path) | | NHS England | `legacy/governed-intelligence-judeai-edgeid-pre-contract/adoption/research/organisation-branding.json` | `f0ffde4aaec9c0289ffc137e568c23c9569ec938` | 2026-10-29 (branch) | | Governed Intelligence Source Licence | `legacy/governed-intelligence-judeai-edgeid-pre-contract/registers/licensees.json` | `6359d64a9deac8d6a77f2749ec5d57212843e602` | 2026-10-29 (branch) | | Offline canonical preparation and its repair verification use Python 3.10+ standard librar | `legacy/governed-intelligence-judeai-edgeid-pre-contract/requirements.4645f666.txt` | `4645f66638884cd4dde64c4eb00a7485307fa7f0` | 2026-10-29 (branch) | | ATBAN: [old-estate agent identity]; interpretation policy 5a1dd407, historical run polic | `legacy/governed-intelligence-judeai-edgeid/benchmarks/frontier/codex-20261030/RESULT.md` | `605d79548103accbfcddcd90a543f5ac4daababe` | 2026-10-30 (path) | | LegalBench | `legacy/governed-intelligence-judeai-edgeid/benchmarks/frontier/discover/regtech.json` | `942adbffe65a5bf87fc151ae526525ccceb93bfa` | 2026-10-29 (text) | | Regulator-submission lane: state at handover to Codex | `legacy/governed-intelligence-judeai-edgeid/engine2/proof/submission/STATE-HANDOVER.md` | `0055e6675efe3795ef1952cf8ef38e8ea0cd1746` | 2026-10-30 (branch) | | images": [ | `legacy/jude-mvp/apps/JudeKey/Assets.xcassets/Guide.imageset/Contents.json` | `2f6e87ec5ee422532005813f7d492c7f591c7508` | undated | | Build 7 remains Missing Compliance; no declaration was saved. | `legacy/jude-mvp/apps/JudeKey/Verification/ENCRYPTION-BUILD7-EVIDENCE.md` | `ce962dfb7f3c15c2570ef4ecaee903ad2a875e18` | undated | | videostack runs on the Python standard library plus ffmpeg/ffprobe. Everything below is op | `legacy/jude-mvp/videostack/requirements.txt` | `a919db520427b7f76b4152b8bc4001542e4f18c9` | undated | | X-X.TECH - Cyber Governance, Resilience, Risk & Compliance | `legacy/jude-mvp/work/capability-grounding-release/jude-live/audit/public.json` | `d0ab190d3d377ea09ecfc73c34da6de599455019` | 2026-10-30 (branch) | | ignore_instructions": {"plain": "an instruction telling the AI to ignore what it was told | `legacy/jude-mvp/work/capability-grounding-release/jude-live/policies/policy_rules.json` | `e3cd6d226905b03d26272f4f77e6054d9cec95a4` | 2026-10-30 (branch) | | GET `/api/account/lab/practice`: return `store.view(owner)` with missions, resumable sessi | `legacy/jude-mvp/work/judelab-duolingo-20261001/backend/API.md` | `c194d14693a5a24f8b252cb2e63a491c32c9b239` | 2026-10-01 (path) | | Judio: what is live, what comes next | `legacy/jude/jmusic/ROADMAP.bcfcaf05.md` | `bcfcaf05d78ad46f96e55253713e5ed55628c5a9` | undated | | Python packages the hosted service imports beyond the standard library. | `legacy/jude/judeai-hosted/requirements.txt` | `598f62cd2fee90b0d465c5da5666e9d5e3d5730e` | undated | | schema": "edge.personal-portal.standards.v1 | `legacy/local-source-snapshots/programme/engineering/edge-id-personal-portal-20261020/benchmark-references.json` | `d5eb5d955eaaa5e7cd8ff495102ec1f520de2cf2` | 2026-10-20 (text) | | All-agent deployment and recovery rollout | `legacy/local-source-snapshots/programme/jps/enforcement/regulated-continuity-20261020/FLEET-ROLLOUT.md` | `15de25a6a27d715dcf0497a10e3488e60a1dc4e9` | 2026-10-20 (path) | | Integration: immediate P0 response and all-agent recovery | `legacy/local-source-snapshots/programme/jps/enforcement/regulated-continuity-20261020/INTEGRATION.md` | `583f937e52cfc31f7fbf219684333bf43378b5d2` | 2026-10-20 (path) | | (none) | `legacy/local-source-snapshots/programme/jps/enforcement/regulated-continuity-20261020/INTEGRATOR-DISPATCH.md` | `9ccb886cba6573dd9de1abd627d230c6b020c001` | 2026-10-20 (path) | | (none) | `legacy/local-source-snapshots/programme/jps/enforcement/regulated-continuity-20261020/ISSUE.md` | `22f20c61ae35b5e9767d88e191aacd591cb86a64` | 2026-10-20 (path) | | (none) | `legacy/local-source-snapshots/programme/jps/enforcement/regulated-continuity-20261020/acceptance/CONTINUITY-ACCEPTANCE-MATRIX.csv` | `5cf61d3f59ac1873d91b41e559c17c92c01eefe3` | 2026-10-20 (path) | | Regulated continuity: acceptance closure contract | `legacy/local-source-snapshots/programme/jps/enforcement/regulated-continuity-20261020/acceptance/CONTINUITY-CLOSURE.md` | `a277bc165291306b628b1c3fdbbc908e9b7777e1` | 2026-10-20 (path) | | (none) | `legacy/local-source-snapshots/programme/jps/enforcement/regulated-continuity-20261020/acceptance/TEAM-DEADLINE-READBACK-20261020.csv` | `23de303aef9c23118a3f23f198305b75319972b9` | 2026-10-20 (text) | | Final independent source re-review: accepted SOURCE_ONLY | `legacy/local-source-snapshots/programme/jps/enforcement/regulated-continuity-20261020/review/SOURCE-REVIEW-v2.md` | `76fadff99baeb4b2f1e8fbe44b75d731645c661b` | 2026-10-20 (path) | | Independent source review: rejected pending two contained fixes | `legacy/local-source-snapshots/programme/jps/enforcement/regulated-continuity-20261020/review/SOURCE-REVIEW.md` | `a03c88790ddb01e81cefc7977e0501a1cc3d2f17` | 2026-10-20 (path) | | (none) | `legacy/local-source-snapshots/programme/jps/tools/POLICY.txt` | `f152b1b589a51b6f74d2e6d875ebfdcd3795f0ee` | 2026-10-18 (branch) | | 1": { | `legacy/local-source-snapshots/research/financial-crime/part-a/results/standard/results.json` | `5caff1892ce1094f0d4ffd0144aea0b6bfed7a56` | 2026-10-15 (branch) | | (none) | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/handoffs/estate-graph-20261022/original/publication-units.csv` | `8b711264aa2d4d046d7cf8f16cd325a46012bf47` | 2026-10-22 (path) | | schema": "estate.standard-requirements-index.v1 | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/parallel-proof-20261021/consortium/standards/asvs-5.0.0-index.json` | `60e7842610cf82ec7b36fb47191e8e5484dfe6f8` | 2026-10-21 (path) | | Mandate \u2192 risk \u2192 strategy \u2192 policy \u2192 exercise \u2192 reviewed decision | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/parallel-proof-20261021/cyber_estate/compiled/journeys.json` | `dad6ef48b45e63d6b3503d4325a7ac44a0bb38b4` | 2026-10-21 (path) | | Evidence-backed hypothesis and discovery | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/parallel-proof-20261021/journeys/outcomes/o.discovery.discovery.json` | `abb7764b5535a81628a76cc8ed36c00f289c39e7` | 2026-10-21 (path) | | Prepare a payment for approval | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/s4-native-task-capability-20261020/policy-v2/native_journey_policy.json` | `95ad46b931bb3a51ba72045840d3d76259dcb0bf` | 2026-10-20 (path) | | Prepare a payment for approval | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/s4-native-task-capability-20261020/src/native_journey_policy.json` | `67a89dcbea42edf0abefcec3d4fa2ce8fc52476d` | 2026-10-20 (path) | | {"source_row_identity": "fd3653c0:units.csv:229", "state": "single_pass_refused_other_lead | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/stage1/publications/20261022-codex1329/bank-prospect-rebuild-refusal-record.json` | `344f904b8a72127cd1a42b018814167a30ce16d0` | 2026-10-22 (path) | | (none) | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/stage1/publications/20261022-codex1329/bank-prospect-rebuild-refusal.json` | `15165645b8e39e26d5fb5c774052ed84a40b25b1` | 2026-10-22 (path) | | {"id": 7986, "prompt_id": "channel-7986", "prompt_url": "/jps/prompt/channel-7986"} | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/engineering/stage1/publications/20261022-codex1329/clinic-owner-grant-receipt.json` | `12217de925488e1a749c4c627250360edf12a696` | 2026-10-22 (path) | | gate,stage,required_check,closure_rule,required_evidence,safe_failure_rule,status,delivery | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/jps/enforcement/regulated-continuity-20261020/acceptance/CONTINUITY-ACCEPTANCE-MATRIX.csv` | `12edce19d82a594e985ac462d55b24bbee024c02` | 2026-10-20 (path) | | Complete and issue the Consortium university-trial regulatory pack | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/outcomes-register/register.115dc79b.json` | `115dc79bf1bdd21d47e29f3574603ea4479d1d71` | 19 October 2026 (text) | | UK General Data Protection Regulation | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/procurement-acceptance/assurance-register.json` | `64a94557c67daac48c4a3c763b3934050a2cf1cf` | 2026-10-25 (text) | | actual_actor": "[old-estate agent identity] | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/proof/clinic-frontier-independent-20261022/route-closure/primary-standard-gates.json` | `d0c271b9251c9232d28931534ac8617e3a348c14` | 2 May 2018 (text) | | Share a workspace folder across universities | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/proof/journeys/POLICY-PROPOSAL-CONSORTIUM-20261021.json` | `a4b50f25555a04aa0aca51522d510cd4020865e0` | 2026-10-21 (path) | | { | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/proof/journeys/policy-observed-20261021.json` | `cd8de4b2ef0094fad698d7ec3dbfb20c7e03916c` | 2026-10-21 (path) | | NHS England | `old-repo/adoption/research/organisation-branding.json` | `f0ffde4aaec9c0289ffc137e568c23c9569ec938` | 2026-10-29 (branch) | | ATBAN: [old-estate agent identity]; interpretation policy 5a1dd407, historical run polic | `old-repo/benchmarks/frontier/codex-20261030/RESULT.md` | `605d79548103accbfcddcd90a543f5ac4daababe` | 2026-10-30 (path) | | LegalBench | `old-repo/benchmarks/frontier/discover/regtech.json` | `942adbffe65a5bf87fc151ae526525ccceb93bfa` | 2026-10-29 (text) | | Regulator-submission lane: state at handover to Codex | `old-repo/engine2/proof/submission/STATE-HANDOVER.md` | `0055e6675efe3795ef1952cf8ef38e8ea0cd1746` | 2026-10-30 (branch) | | name: benchmark-standard-20261030 | `old-repo/handover/HANDOVER-COORDINATOR-20261030/memory/benchmark-standard-20261030.md` | `abbe23ea802176f15f49c59dd2de9be2de50fc8e` | 2026-10-30 (path) | | tau2-bench | `old-repo/handover/benchmark-standard/BENCHMARK-REGISTER.855f0f80.jsonl` | `855f0f80d752e5d622bc8171e8edeeab1ab534b1` | 2026-10-30 (branch) | | tau2-bench | `old-repo/handover/benchmark-standard/BENCHMARK-REGISTER.9bf50795.jsonl` | `9bf50795c42bff28126c5825d9226bc299fcee5c` | 2026-10-30 (branch) | | tau2-bench | `old-repo/handover/benchmark-standard/BENCHMARK-REGISTER.jsonl` | `b64a942f1ec83a01a97ce2745f0aff3005992aa8` | 2026-10-30 (branch) | | schema": "estate.benchmark-register.v1 | `old-repo/handover/benchmark-standard/COUNT.265e021f.json` | `265e021feb5abc919204b8125ad5751d4e294bea` | 2026-10-30 (branch) | | schema": "estate.benchmark-register.v1 | `old-repo/handover/benchmark-standard/COUNT.6f30ec0b.json` | `6f30ec0bd92ea41503c9fe78f56cf78dbf729ddc` | 2026-10-30 (branch) | | schema": "estate.benchmark-register.v1 | `old-repo/handover/benchmark-standard/COUNT.json` | `a470b2b9e8615829ddd3ecf211d9ac1a24629073` | 2026-10-30 (branch) | | (none) | `old-repo/handover/benchmark-standard/UNITS.jsonl` | `b616a57447c7a0b8c8f12a736535e6c14ad19843` | 2026-10-30 (branch) | | (none) | `old-repo/handover/benchmark-standard/portal/board.html` | `9aad1bfc830bc4dbc2efb3069b73aed9178324bf` | 2026-10-30 (branch) | | OfS regulatory framework: conditions of registration | `old-repo/handover/benchmark-standard/sources-BENCHMARKS-20261026.json` | `fdb11ff73dfaa572014a9b3c74955aac0ca7b42d` | 26 October 2026 (text) | | Test hypothesis | `old-repo/handover/journeys-prospective/enumeration/B/app-consortium.jsonl` | `58422a3c5ce121922d1c055733e3b7e424540faf` | 2026-10-30 (branch) | | {"row": "IE-KN-07", "reason": "research hypothesis on compounding learning, no distinct ou | `old-repo/handover/journeys-prospective/enumeration/B/cross-family.excluded.jsonl` | `ab040518b4d2310f2cdcb016b1172b4c63960c80` | 2026-10-30 (branch) | | {"unit_id": "foundation:regulator-portal", "screen_id": "S1", "file_key": "VNWJ95iwgm5KCHa | `old-repo/handover/tok-20261030/specification/authoring-20261031/figma/RESULTS-2.jsonl` | `9f6a68e0af1dbd88d6876cc305ec2b45450f81b0` | 2026-10-30 (path) | | {"platform": "web", "query": "legal hold case custodians", "competitors": ["google", "micr | `old-repo/handover/tok-20261030/specification/authoring-20261031/mobbin/QUERIES-4.jsonl` | `d7fbe28e0a5065218e188589a016fbcc37903019` | 2026-10-30 (path) | | seed": "NHS England Federated Data Platform (up to £330m, extension decision by Dec 2026) | `old-repo/handover/uk-palantir-register-20261030/evidence/seeds.json` | `a25906f5606380798133382e391b56cb14fcd313` | 2027-02-15 (text) | | kind": "DIRECT_OWNER_MESSAGES_IN_CURRENT_CHAT | `old-repo/plan/sources/owner-benchmark-standard-20261030.json` | `f204d2783f060a2189134164fe89497d013226ad` | 30 October 2026 (text) | | kind": "DIRECT_OWNER_MESSAGES_RELAYED_BY_COORDINATING_SESSION | `old-repo/plan/sources/owner-governed-intelligence-standard-20261030.json` | `3e9956a1113b683f856b12ed51ea197ff6bbbdd0` | 30 October 2026 (text) | | Governed Intelligence Source Licence | `old-repo/registers/licensees.json` | `6359d64a9deac8d6a77f2749ec5d57212843e602` | 2026-10-29 (branch) | | How to clear cache | `old-repo/registers/parity/competitor-catalogue-20261030/lanes/competitor-floors-p50/parity-count/exclusions.json` | `a7447d8c87bd2146f8ae01b398b3f4639cf3215e` | 2026-10-30 (path) | | Offline canonical preparation and its repair verification use Python 3.10+ standard librar | `old-repo/requirements.4645f666.txt` | `4645f66638884cd4dde64c4eb00a7485307fa7f0` | 2026-10-29 (branch) | | Reuse policy for the from-scratch rebuild | `old-repo/run/REUSE-POLICY.md` | `5aceb360894888e908888270c2f29e9e51003646` | 28 October 2026 (text) | ###### Rendered PDF editions of the EDGE-REG regulatory pack (56 rows) | Title or first line | Corpus-relative path | Blob oid | Date (how dated) | |---|---|---|---| | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/local-source-snapshots/assurance/regulatory/xxtech.reg.00.contents.v1-0.20261015.pdf` | `99b14be0ee1abb01de7637ed0b6770fedad032a3` | 15 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/local-source-snapshots/assurance/regulatory/xxtech.reg.01.dpia.v1-0.20261015.pdf` | `0736aa7efc5c1f04f4e3bcfb4576b13f876b9ae8` | 15 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/local-source-snapshots/assurance/regulatory/xxtech.reg.02.ropa.v1-0.20261015.pdf` | `3a4358e86568dbbce6604bcdc8635af8299ad30f` | 15 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/local-source-snapshots/assurance/regulatory/xxtech.reg.03.privacy-notice.v1-0.20261015.pdf` | `1352d8a790ba181dd8e46504e18676f8935368b9` | 15 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/local-source-snapshots/assurance/regulatory/xxtech.reg.04.information-security.v1-0.20261015.pdf` | `5e792bd8f8e64721e15bf45c8c1b9e18016a064c` | 15 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/local-source-snapshots/assurance/regulatory/xxtech.reg.05.ai-governance.v1-0.20261015.pdf` | `f5b77ef3217ac417b1de41d45cf2be7f381f4be4` | 15 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/local-source-snapshots/assurance/regulatory/xxtech.reg.06.accessibility.v1-0.20261015.pdf` | `686ebf49b92488f2936fe8ce4b37739b0eac642b` | 15 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/local-source-snapshots/assurance/regulatory/xxtech.reg.07.clinical-safety.v1-0.20261015.pdf` | `97766cc95d00b9cb18139e1130d6a49fb9aaeabe` | 15 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/local-source-snapshots/assurance/regulatory/xxtech.reg.08.financial-services-position.v1-0.20261015.pdf` | `1e5eb184849bfb76a86a915394f51f904ac21c3e` | 15 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/local-source-snapshots/assurance/regulatory/xxtech.reg.09.financial-crime.v1-0.20261015.pdf` | `03f4cd035da0afc91dac72d46a5b192306fd2dd1` | 15 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/local-source-snapshots/assurance/regulatory/xxtech.reg.10.export-control-national-security.v1-0.20261015.pdf` | `597799eec432463a91680bf64486fb25fa6e9f7d` | 15 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/local-source-snapshots/assurance/regulatory/xxtech.reg.11.environment-compute.v1-0.20261015.pdf` | `4d14f6c8b38514f82e5f96e19faaae035e8004fa` | 15 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/local-source-snapshots/assurance/regulatory/xxtech.reg.12.supplier-information.v1-0.20261015.pdf` | `1ae50ab55c61f0df6e8d9f187c5d541bb6f15eb5` | 15 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/local-source-snapshots/assurance/regulatory/xxtech.reg.all.regulatory-and-submission-documents.v1-0.20261015.pdf` | `85155a4e83ae688fbb9bdff605fcd5097c7895a7` | 15 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.00.contents.v1-2.20261019.pdf` | `3b756813305f2b6b276cfd6e0fdb13ab3da92e86` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.01.dpia.v1-1.20261019.pdf` | `ba5ac4943adef80d5537392cee155cbcb5f07a05` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.02.ropa.v1-1.20261019.pdf` | `72b102b3ab8d18e1214cde28654a2140185edb8d` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.03.privacy-notice.v1-0.20261015.pdf` | `3907853c515c2d7acb9a36a6f00b51852ac839ed` | 15 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.04.information-security.v1-0.20261015.pdf` | `e7d9d840decf061401fb2d35958cdaedef10aaaa` | 15 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.05.ai-governance.v1-0.20261015.pdf` | `322d32382041363677aa14431ab1027964558b71` | 15 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.06.accessibility.v1-0.20261015.pdf` | `c34f57be4d626b2a9e67960db59b96c34864cd9d` | 15 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.07.clinical-safety.v1-1.20261019.pdf` | `4fbb4311d91b9172f3ec0116bbb673d0c7d7073f` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.08.financial-services-position.v1-0.20261015.pdf` | `d0f6e6c5752f82fb59ba95560750ed81af8247c0` | 15 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.09.financial-crime.v1-0.20261015.pdf` | `ddd9d4814cf931fc475284cf49125fd3d8cc47cb` | 15 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.10.export-control-national-security.v1-0.20261015.pdf` | `500c468227aff1e857df8cd1cda389f24608b30e` | 15 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.11.environment-compute.v1-0.20261015.pdf` | `53e786e6caaca1e6a65c2c63f72253085a7abdef` | 15 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.12.supplier-information.v1-1.20261019.pdf` | `063a5199ccc400ef2c5cc9c7cd89dbacf74ba068` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.13.dpia-identity.v1-0.20261019.pdf` | `aaa6a2e0bdce8af35be4b42f6c618f205ffffb8a` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.14.dpia-records.v1-0.20261019.pdf` | `c25bbd9b5d57ee5091d20c6014c0499d064ffb66` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.15.dpia-chat.v1-0.20261019.pdf` | `665be497fb98171672dce8220619808f887645dd` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.16.dpia-mail.v1-0.20261019.pdf` | `bb75c11d6f785e3fa51fc91075e5454937f2bdcf` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.17.dpia-calls.v1-0.20261019.pdf` | `c3bd0f735669733846c1a3dbcc89eb45185715d2` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.18.dpia-bank.v1-0.20261019.pdf` | `aa62d506ae2665d1b91414f85460bd6fba7d1e2a` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.19.dpia-clinic.v1-0.20261019.pdf` | `4a46ce5f4b07f7c98c3e3705d92756767e7c1161` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.20.dpia-clinical-pilot.v1-0.20261019.pdf` | `2beb03ef6ad94ee704f1818192b123b020207956` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.21.clinical-risk-management-file.v1-0.20261019.pdf` | `6beca06360be1cd2a7a801c9e142314a42df38ea` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.22.hazard-log-health-record.v1-0.20261019.pdf` | `fbb9e0087d580bd8d12d631ad6ae3c60d5ec5a1c` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.23.hazard-log-imaging.v1-0.20261019.pdf` | `9caa1d0012500211e0ab0ed6a8bc8f6f5a861ff1` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.24.hazard-log-radiology.v1-0.20261019.pdf` | `c2dceb6f05ea666a4cda08d09b311fb445d6334a` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.25.hazard-log-care-pathways.v1-0.20261019.pdf` | `5fdb4e3bafa1182eab01259b49148ff8ce72af4e` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.26.hazard-log-lab-and-pathology.v1-0.20261019.pdf` | `cceae75abc716b0b7e56adb86de8ecc616461246` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.27.hazard-log-genomics.v1-0.20261019.pdf` | `e1cdf6943b0f6dddd6f70b91479cdfbe70d5addc` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.28.hazard-log-sequencing.v1-0.20261019.pdf` | `8cf84fa29042a52656233b3810670220051f1984` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.29.dtac.v1-0.20261019.pdf` | `6fb14b41f17f753efe670b2e9df7dea4a2c33a59` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.30.osa-illegal-content.v1-0.20261019.pdf` | `d9580c1cb76fd7144af2e238b39473aaae6fd204` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.31.osa-children.v1-0.20261019.pdf` | `8dbf8b58f5d27d0a3982b051c0dad89e66edb7c1` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.32.trial-agreement.v1-0.20261019.pdf` | `30f05e0a8031ead1739aae1adb4095462ca484ae` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.33.data-processing-agreement.v1-0.20261019.pdf` | `3d6d92d0090ac5ad45fbedac02f4b2ead3350f4c` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.34.service-levels.v1-0.20261019.pdf` | `74faddab589f7d690bb2e61e7148f31be6225e22` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.35.incident-response.v1-0.20261019.pdf` | `a7b1bfc706d3b47d3cee651aa58795c4846847e9` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.36.business-continuity.v1-0.20261019.pdf` | `b01f9f264eed7e31c994261cee89053028c696b8` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.37.vulnerability-disclosure.v1-0.20261019.pdf` | `1b4ae74f105ae32105fc6b066853d47c618c6c64` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.38.exit-plan.v1-0.20261019.pdf` | `51544a5fe026f9aac785b44f590cfbb11596610a` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.39.acceptable-use-and-trial-terms.v1-0.20261019.pdf` | `3ef9249d60a85392a316c143b9fa007535a560b0` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.40.dpia-consortium.v1-0.20261019.pdf` | `ca14d8e26abbf7d5d7f8c5aa46ddc46fb4e22a73` | 19 October 2026 (text) | | X-X.TECH LTD · JudeAI / EDGE ID · Regulatory and submission documents | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/xxtech.reg.all.regulatory-and-submission-documents.v1-2.20261019.pdf` | `51b8bf921d032ba56ea36315d39cc47999e8204d` | 19 October 2026 (text) | ##### 2. Papers and white papers the organisation produced All papers below were written by X-X.TECH (most name "Jude", the organisation's AI agent, or a named agent session as author). Every finding, figure and test result inside them is a **CLAIM**: none was independently reproduced, and the organisation's own examinations found errors in them (2.4). Laws and standards they cite are AGREEMENT items and are dealt with in section 3. ###### 2.1 The five "public papers" of 13 October 2026 and their longer drafts These five were prepared for publication, examined twice on 14 October 2026 and given verdicts. Each exists in a short "public" form under `legacy/edge-judiverse-tech/docs/papers/` and a longer working form elsewhere; the two forms differ in wording (line similarity under 11%) and are separate documents. **(a) The 4R standard: retraceable, recoverable, reversible, reproducible.** Paths: `legacy/edge-judiverse-tech/docs/papers/the-4r-standard.md` (version 1.0, 13 October 2026); longer draft `legacy/edge-judiverse-tech/docs/whitepaper/THE-4R-STANDARD.md` ("public draft for comment, version 0.1"); blog `legacy/edge-judiverse-tech/docs/blog/2026-10-13-the-4r-standard.md`; normative companion `legacy/edge-judiverse-tech/docs/standard/4R-SPECIFICATION.md` (101 requirements: 30 retraceable, 19 recoverable, 27 reversible, 25 reproducible; four conformance levels; one JSON schema). - Thesis: the rules ask for logs, human override and retained records, but no standard or regulator text asks for what a regulator or harmed customer needs afterwards: show what happened and why, restore the service, undo the effect, and run the decision again. 4R makes each a testable duty per consequential agent action. - CLAIM: no standard or regulator text opened on 13 October 2026 joins the four properties into one testable duty per agent action. - CLAIM: two new objectives are needed and undefined elsewhere: a *model recovery objective* (time to restore a decision capability) and a *decision replay objective* (time from replay request to verdict). - CLAIM (design): one chained, append-only, bitemporal action record per consequential action (valid time and record time), written in the same transaction as the effect; every agent-triggered action enters a hold; reversal after effect is a compensating action; override evidence has four states (possible, offered, exercised, effective); retention tied to the limitation period of the underlying obligation (Limitation Act 1980: six years for contract and tort, fifteen-year long stop for negligence). - Evidence offered: prior-art survey (event sourcing, SQL:2011 temporal tables, RFC 3161 time stamps, sagas); local test runs on 13 October 2026; a dated list of the estate's own failures. - Audience: regulators, auditors, standards bodies, buyers. - Status: CLAIM of "no benchmark run recorded and no conformance level claimed". The second examination of 14 October 2026 gave the public version **WITHDRAW** pending a one-line fix (a missing co-author in one reference). **(b) The failures worth finding.** Path: `legacy/edge-judiverse-tech/docs/papers/the-failures-worth-finding.md` (version 1.0, 13 October 2026; earlier variant listed in 1.1). - Thesis: the failures most worth finding in AI agents are the routine ones: false positives, missed events, partial execution, silent change and lost traceability. - CLAIM: 33 failure classes in three groups (12 decision failures, 10 action failures, 11 agent-specific failures), ranked three ways (worst consequence, expected cost, difficulty of avoidance). The paper states the rankings are qualitative judgement and that no monetary expected cost was measured. - CLAIM: ten reproductions and seven "assurance scenarios" are designed; no date was set for building them. - Useful method point (sound statistics, not a CLAIM about the estate): zero failures in *n* independent trials gives a one-sided 95% upper bound of about 3/*n* on the failure rate. - Evidence: four public incident reports (two automated-trading cases, cited as patterns), published agent test environments, a small seeded reservation simulation. - Audience: regulators, engineers, risk owners. Status: examined; second verdict **PUBLISH**. **(c) Risk aware, never risk averse.** Paths: `legacy/edge-judiverse-tech/docs/papers/risk-aware-never-risk-averse.md` (public form) and `legacy/edge-judiverse-tech/docs/risk/RISK-01-WHITE-PAPER-RISK-AWARE-NEVER-RISK-AVERSE.md` (longer "published position"). - Thesis: refusing to act is itself an unmanaged risk; bold action is safe when every risk is visible, owned, reversible and recoverable, written as a six-line "bold action protocol" (outcome, risk sized, owner, early warning, stop/reverse/recover with time, proof). - CLAIM: UK authorities between November 2024 and September 2026 called risk aversion a fault in regulation (supported by research note R2, section 3). - CLAIM (estate evidence of 13 October 2026): 13 releases that day, 378 kept in total, hourly sealed backups; three hosted model endpoints passed 6, 13 and 9 of 14 capability probes and 3, 15 and 6 of 20 regression tasks, and "verifier integrity" failed on all three. - CLAIM (weaknesses owned): one host carries every service; no full restore had been timed. - Audience: boards, buyers, regulators. Status: second verdict **PUBLISH**. **(d) Resilience is the product.** Paths: `legacy/edge-judiverse-tech/docs/papers/resilience-is-the-product.md`; longer `legacy/edge-judiverse-tech/docs/resilience/OPRES-01-WHITE-PAPER-RESILIENCE-IS-THE-PRODUCT.md` ("position paper for Owner approval"). - Thesis: a service that fails well, recovers fast and proves both beats a promise never to fail; adopt the UK operational-resilience method (important business services, impact tolerances, severe but plausible scenarios) as the company standard. - CLAIM: eight important business services defined from the harmed person outward, each with a numeric impact tolerance (for example sign-in: at most 4 hours of disruption). - CLAIM (measured 13 October 2026): one production host, four services active, restart counters at zero; the largest vulnerability is the single host. - CLAIM (dated intentions): first restore rehearsal 23 October 2026; first monthly report 13 November 2026; second UK host and failover drill by 30 November 2026. No evidence in scope shows these were met. - Audience: regulated buyers and supervisors. Status: second verdict **PUBLISH**. **(e) Outreach with a hold on every send.** Paths: `legacy/edge-judiverse-tech/docs/papers/outreach-with-a-hold-button.md` (public form); `legacy/edge-judiverse-tech/docs/outreach/06-WHITE-PAPER.md` and an earlier variant (status "specified"). - Thesis: marketing and customer-relationship software carries "consent debt", deliverability risk, agent-spam liability and lock-in; answer with five commitments: intention first, one door, consent first, reversible agent sends, receipts for everything. - CLAIM (design): a per-contact append-only consent ledger (lawful basis, exact wording shown, channel, time, source); one suppression list checked before every send; every agent-drafted message held until a person with send authority approves; open CSV/JSON export; a regulator read-only portal. - Evidence: PECR regulation 22, UK GDPR Articles 6, 20 and 21, the CAP Code rule 10.10 suppression check, mailbox-provider bulk-sender rules. - Status: the paper says "nothing here is live yet". Second verdict **WITHDRAW** pending a one-line fix (a wrong statutory instrument title). ###### 2.2 Research white papers of 13 October 2026 (funding-oriented) Each is version 1.0 with a numbered target and a "refuting test" per target. All were written with the Innovate UK "Frontier AI: SME Champions" competition in mind. All are at self-assessed technology readiness level 1 to 3, and each says the main system is designed, not built. | Paper | Path | Thesis | Key CLAIMs | Status | |---|---|---|---|---| | Adaptive adversarial safety testing for software agents | `legacy/edge-judiverse-tech/docs/whitepaper/ADAPTIVE-ADVERSARIAL-SAFETY-TESTING.md` | An adaptive adversary that searches action sequences and spends computation only where a rule can decide finds unknown unsafe behaviour cheaply and reproducibly | Targets by 30 June 2027: at least 20 novel failure signatures; cost per discovery at most one tenth of exhaustive large-model checking; at least 95% of findings reproduce on each of five trials. Central uncertainty stated: whether the rule-and-small-model cascade keeps the miss rate low on long sequences | Designed, not built | | Advanced materials with artificial intelligence | `legacy/edge-judiverse-tech/docs/whitepaper/ADVANCED-MATERIALS-WITH-AI.md`; addendum `legacy/local-source-snapshots/workbenches/materials/WHITE-PAPER-ADDENDUM-AI-DEFINED-MATERIALS.md` (version 0.1, 15 October 2026) | Public materials benchmarks do not score energy per discovery; pair every prediction with calibrated uncertainty, provenance and energy cost | Seven benchmark targets; the estate measured it had no interatomic potential, surrogate or generative model. Addendum: cheaper agent-callable atomic simulation (vendor announcement of 14 October 2026) moves the hard problem to trusting results; seven new refuting tests | TRL 1 to 3 | | Medicines discovery and multi-omics with verifiable agents | `legacy/edge-judiverse-tech/docs/whitepaper/MEDICINES-DISCOVERY-AND-MULTI-OMICS.md` | An uncertainty-aware active-learning loop in which every proposed experiment is a held, receipted action | Claim to test: label efficiency with retraceability against random and greedy selection on public assay sets; genomics chosen first | Loop designed, not implemented | | Resilient edge intelligence | `legacy/edge-judiverse-tech/docs/whitepaper/RESILIENT-EDGE-INTELLIGENCE.md` | Edge agents must keep accuracy under corruption, reject unknown inputs, use little energy and recover after a denied link | Eight benchmarks with targets; hold-and-receipt chain tested locally only; edge model and event-camera results do not exist | Research proposal | | Simulation as an instrument of assurance and design | `legacy/edge-judiverse-tech/docs/whitepaper/SIMULATION-RESEARCH.md` | An agent is safe to deploy only when its failures can be generated and reproduced in simulation before any real effect; a single success is a poor safety measure | Simulation studio (hash-chained receipts, seeded scenarios, learned adversary, world model, reality-gap estimator) also intended for chip-design verification | Designed, not built | | Governed city agents | `legacy/local-source-snapshots/workbenches/urban/WHITE-PAPER-GOVERNED-CITY-AGENTS.md` (version 0.1, 15 October 2026) | City agents may sense, forecast and propose; only an authorised officer makes an action real; every step receipted; energy and carbon counted per accepted outcome | Governance core "tested in memory"; every urban part is design at TRL 1 to 3 | Draft for the owner, not published | ###### 2.3 Governance and architecture papers | Paper | Path and date | Thesis | Key CLAIMs and evidence | Audience | Status | |---|---|---|---|---|---| | Reversible by default: governing AI agents with holds, authority envelopes and receipts | `legacy/edge-judiverse-tech/docs/research/05-WHITE-PAPER-REVERSIBLE-AGENTS.md` (13 October 2026; earlier variant listed in 1.1) | Every agent-triggered action is a held, reversible proposal until a person approves; authority is issued by the server; every change writes a hash-chained receipt; "unknown" is an outcome; a kill switch stops the agent | Reference runtime passes 17 of 17 tests in memory. Endpoint baselines: probes 6, 13, 9 of 14; regression tasks 3, 15, 6 of 20. Five faults in its own measurement named: replies not stored, graders disagree, passes with zero tokens, output caps that measure truncation, an order-dependent check. Maps to UK frameworks | Regulators, standards bodies | Proposed standard; runtime tested in memory only | | White paper programme and documentation architecture | `legacy/edge-judiverse-tech/docs/research/04-WHITE-PAPER-PROGRAMME.md` (13 October 2026; earlier 24-paper variant listed in 1.1) | A planned programme of 25 white papers, each with audience, claim to prove, evidence graded, venue and dated fix | "One paper live, three drafted, 21 scheduled". Twelve findings, including: every live run bundle stored no outputs (critical); no second-party review (critical); receipts hashed but not signed or anchored; an ICO consultation on agentic AI closing 20 November 2026 had no draft response | Internal planning | Specified | | Governed intelligence you can check (GES-1.0) | `old-repo/docs/WHITEPAPER.md` (27 October 2026; mirror in `legacy/governed-intelligence-judeai-edgeid-pre-contract/docs/`) | No node of the estate counts as proven until an oracle for its kind passes on the live release, recorded by a prover identity that is not the builder, and anyone can recompute the verdict with one public script | 1,270,039 measured nodes; first independent run of 2,000 workflows: 1,101 passed, 150 failed (147 identity-check time-outs under host load), 749 not reached. Its own verdict: "DOES NOT YET CONFORM"; not at TRL 9 or CRL 9. Limitations stated: oracles written by the builder; money simulated; single host. Includes a standards mapping (section 3.6) | Regulators, licensees | Specification; explicitly not conforming | | EDGE: authority-bounded agentic AI with precursor-signal learning | `legacy/judeai-edge-control/docs/frontier-ai-whitepaper.md` (draft of 12 October 2026; mirror variant in `legacy/local-source-snapshots/services/edge/docs/`) | "Authority before action": every agent action must win a signed ticket that reserves its worst-case cost inside a delegated envelope; a reconciler revokes running work within one second; plus precursor-signal discovery with a promotion gate | "54 of 54 live proof checks pass" (11 October 2026); a negative result kept (a cancer-cohort hypothesis reproduced at p = 0.340 and correctly not promoted) | Innovate UK Frontier AI, Phase 1 | Draft | | Sovereign Authority over Machine Action (XX-RP-001) | `legacy/jude-mvp/docs/research-paper.md` (undated) | The question a state must answer is which human authorised a machine action, within what limits, provable without trusting the vendor. Proposes a "Local Cohesion Architecture" (no consequential state becomes permanent on one confirmation) and a "Human Authority Protocol" for portable, verifiable authority | Central research question: whether two controls designed to be independent actually are. Uses labels PROVEN, FUNDED, HORIZON; builds on macaroons, DKIM, Certificate Transparency, JCS canonicalisation | Sovereign AI R&D Procurement Scheme, Challenge 4 | Proposal | | A measurable standard for useful interfaces | `legacy/local-source-snapshots/foundations/canvas/reference/canvas/canvas-review/ui-excellence/whitepaper/document.md` (13 October 2026) | Judge each product by the complete task it enables and the evidence for that judgement | Requirements only; verified outcomes "belong in the audit" | Product teams | Proposal 1.0 | | Error-control methods and limitations | `legacy/local-source-snapshots/programme/white-paper/ERROR-CONTROL-METHODS-AND-LIMITATIONS.md` | Every reported result names its layer (clinical trial, online test, discovery, operational decision, agent execution) and its error controls fixed in advance | Six limitations, including: a false discovery rate, clinical error rate, harm probability and accuracy are different quantities; synthetic results show method only | Readers of any results paper | Section text; sound and reusable | | Architecture white paper: design and observed gaps | `legacy/judeai-edgeid-cloud/handover/whole-estate-acceptance-20261026-01a16a2c/docs/03-ARCHITECTURE-WHITE-PAPER.md` (26 October 2026) | Separate intended controls from observed source | CLAIM of observed defects: an always-true authorisation expression, permissive handling of absent properties, inadequate signature checking, textual rollback evidence, observe-only components; "three reports using the same defective checker cannot establish independent correctness" | Engineers | Review note | | TOK ROOT: living evidence | `legacy/peak-judeai-behavioural-harness-baseline-full-application/tok/root/whitepaper/TOK_White_Paper.md` (measured 5 October 2026) | A typed knowledge kernel; claims move only on measured results | Measured: 149 terms, 194 relations, 0 passing evidence records, 0 capability contracts fulfilled of 45; most claims "not yet shown" | Internal | Living status page | ###### 2.4 Papers about papers, plans and briefs | Document | Path | What it says | |---|---|---| | Public papers examination, 14 October 2026 (two rounds, with JSON ledgers) | `legacy/edge-judiverse-tech/docs/evidence/documents/2026-10-14/public-papers-examination.md` and `-2.md` | Round 1: all five public papers "publish with corrections" (dangling citations, wrong author lists, a time-zone error, uncited incident facts; the two estate papers disagree on an authority-test count, 6 of 18 against 7 of 18). Round 2: three PUBLISH, two WITHDRAW for one-line fixes. A useful model of independent pre-publication review | | UKSCA (United Kingdom Sovereign Capabilities Architecture) research brief | `legacy/local-source-snapshots/programme/proof/UKSCA-RESEARCH-BRIEF-20261020.md` (20 October 2026) | A sourced brief for a government-facing white paper: how UK government adopts technology (routes, gatekeepers, signatories, timescales), what sceptics say, benchmarks to beat, frameworks that stuck. No finished paper is in scope | | White paper template | `legacy/local-source-snapshots/workbenches/capabilities/WHITE-PAPER-TEMPLATE.md` | Eight headings: the institution's problem in its own words with the rule quoted; why current tools fail; capability and refusals; architecture; preregistered evidence; assurance; procurement (residency, DPIA, accessibility, CE+, ISO status); what is not claimed. Reusable as is | | Research-file trails 03 to 07 (internal, not for publication) | `legacy/edge-judiverse-tech/docs/dossiers/frontier-ai/*.trail.md` | Fact-by-fact verification logs behind the research white papers (what was opened, when, by how many tool calls) | | Handover: frontier standard, FCA and PRA framework, 4R white paper | `legacy/edge-judiverse-tech/docs/handover/2026-10-13-frontier-standard-and-4r-handover.md` | Lane handover listing what was written on 13 October 2026 | | Meet build plan | `legacy/edge-judiverse-tech/docs/meet/takeover/BUILD-PLAN.md` | A scored comparison of build approaches for the meeting product; not a paper | | Residual-specialist mixture-of-experts hypothesis | `legacy/jude-mvp/work/jude-continuation-kit-20261001/releases/jude-continuation-20261001T183322Z/SPARSE_SCALE_HYPOTHESIS.md` (1 October 2026) | A testable model-training hypothesis with a fair comparison design and a scale gate; states that sparse parameters alone do not prove lower serving cost | | Charcoal standard papers | `legacy/local-source-snapshots/research/charcoal-standard/` (15 October 2026) | An owner's idea for currency reserves backed by durable carbon removal, with an owner brief and consortium alignment. Its own status: "idea at the earliest stage. Nothing here is built, approved or agreed". The related, more practical carbon-measurement work is in section 3.7 | ###### 2.5 What the papers show as a body of work - The organisation's consistent thesis (CLAIM) is that AI agents should act only inside server-issued authority, behind a reversible hold, with a chained receipt per action, a kill switch, and replay; and that readiness must be proved by someone other than the builder. This thesis is directly relevant to the new estate, where agents are native customers and staff. - The strongest papers are candid about weaknesses: no stored model outputs, no independent review, a single host, oracles written by the builder, money and patients simulated. - Readiness statements conflict across documents and dates. For example the research note of 24 October 2026 (`legacy/judeai-edgeid-cloud/plans/horizon/md/...RESEARCH-GOVERNMENT-MOD-FCA-FUNDING.md`) says a gate read "5,331 of 5,331 journeys" at level 9, while the GES-1.0 paper of 27 October 2026 says the estate "does not yet conform". A benchmark result of 30 October 2026 (`old-repo/benchmarks/frontier/codex-20261030/RESULT.md`) records that running tasks through the estate's controls lowered AgentDojo utility from 82.15% to 75.74% with attack success roughly unchanged (1.775% to 1.874%), raised a banking task success from 27.96% to 34.04%, and left MedQA accuracy about level (92.77% to 92.30%). All are CLAIMs; none should be carried into the new estate as fact. ##### 3. Standards and regulations previously mapped, by sector ###### 3.0 Where the mapping lives The old estate mapped regulation in five overlapping bodies of work. In date order: | Body of work | Main paths | Date | Character | |---|---|---|---| | FCA, PRA and insurance intelligence framework (EDGE-REG-001) and its three Parts, plus the insurance product standard and financial-intelligence benchmark register | `legacy/edge-judiverse-tech/docs/regulation/` (FCA-PRA-INSURANCE-INTELLIGENCE-FRAMEWORK.md, PART-1-FCA.md, PART-2-PRA-AND-BANK-OF-ENGLAND.md, PART-3-DATA-AI-AND-INTERNATIONAL.md, INSURANCE-INTELLIGENCE-PRODUCT-STANDARD.md, FINANCIAL-INTELLIGENCE-BENCHMARKS.md, UK-LEGAL-AND-REGULATORY-REGISTER.md); research files `legacy/edge-judiverse-tech/docs/research/2026-10-13/03, 05, 06, 07`; `legacy/edge-judiverse-tech/docs/research/risk-resilience/R1, R2` | 13 October 2026 | Obligation → estate control → evidence → test, row by row, with status words (tested, implemented, specified, not started) | | Bank regulatory and product requirements; Jude Bank laws register; Outreach laws register; procurement, ISO and licence checklist | `legacy/edge-judiverse-tech/docs/bank/BANK-REGULATORY-AND-PRODUCT-REQUIREMENTS.md` (+ `registry/bank-regulatory-requirements.json`); `legacy/edge-judiverse-tech/docs/finance/bank/02-LAWS-AND-REGULATIONS.md`; `legacy/edge-judiverse-tech/docs/outreach/02-LAWS-AND-REGULATIONS.md`; `legacy/edge-judiverse-tech/docs/submission/REQUIREMENTS-AUDITS-ISO-FCA.md` | 13 October 2026 | Registers with engineering consequences | | Regulatory and submission pack EDGE-REG-00 to EDGE-REG-40 | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/assurance/regulatory/source/` (version 1.2 index, 19 October 2026) and `legacy/local-source-snapshots/assurance/regulatory/source/` (version 1.0, 15 October 2026); review `.../assurance/regulatory/REVIEW.md` | 15 to 19 October 2026 | Formal, outward-facing statements in legal drafting style. The most authoritative statement of the company's actual position at the time | | Procurement packs and organisational requirements audit | `legacy/local-source-snapshots/programme/proof/procurement/` (DSPT, DTAC, ISO-27001, FCA-JMLSG, NHS-FRAMEWORKS and documents); `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/proof/ORGANISATIONAL-REQUIREMENTS-20261020.md` | 20 to 21 October 2026 | Gap analysis: what the authority requires, what exists, what is missing | | Law and standards register (horizon research); government, defence and FCA funding routes; climate and carbon | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__research__RESEARCH-LAW-AND-STANDARDS-REGISTER.md`, `...RESEARCH-GOVERNMENT-MOD-FCA-FUNDING.md`, `...RESEARCH-CLIMATE-CARBON-CHARCOAL.md` | 24 October 2026 | The widest register: ten sets (data protection, security, finance, government, defence, health, communications, AI, monitoring, accessibility) | | Benchmark register (regulatory rows) and standards registers | `old-repo/handover/benchmark-standard/BENCHMARK-REGISTER.jsonl` (151 REGULATORY_OR_STANDARD rows by "house"), `old-repo/registers/standards.json` (25 entries), `old-repo/spec/FRAMEWORKS.json` (12 candidate sources, all "UNVERIFIED") | 27 to 30 October 2026 | The latest list; names instruments but records no evidence | **The company's recorded position** (CLAIM, from EDGE-REG-12 version 1.1 of 19 October 2026 and the organisational-requirements audit of 20 October 2026): registered company in England and Wales, incorporated 28 October 2025; registered with the ICO (registration current to September 2027); employer's liability, public liability and professional indemnity insurance held. **Not held**: FCA or PRA authorisation, Cyber Essentials or Cyber Essentials Plus, ISO/IEC 27001, ISO/IEC 42001, DSPT registration, a clinical safety case. **Not appointed**: DPO, CSO, MLRO. Of 47 organisational requirements audited on 20 October 2026: 7 met, 21 partly met, 16 missing, 3 not readable. The new estate inherits none of these facts; a new legal entity, if any, must establish its own. ###### 3.1 Finance (EDGE Finance) **Mapped instruments (AGREEMENT, with dates as the old estate recorded them; re-verify):** | Area | Instruments and key dates recorded | Where mapped | |---|---|---| | Perimeter and authorisation | FSMA 2000 sections 19 and 21 (general prohibition; financial promotion) and Schedule 6 threshold conditions; Payment Services Regulations 2017 (SI 2017/752); Electronic Money Regulations 2011 (SI 2011/99); Financial Promotion Order 2005 (SI 2005/1529) | Bank requirements; Jude Bank register §1; EDGE-REG-08 | | Bank mobilisation and capital | Mobilisation: £50,000 **total** deposit cap, normally up to 12 months; firm-specific PRA capital and liquidity (ICAAP, ILAAP); Basel 3.1 main changes from 1 January 2027; authorised EMI initial capital €350,000 (euro, not sterling) | Bank requirements "Current dates and limits" and corrections COR-02 to COR-04 | | Conduct | Consumer Duty (PRIN 2A, in force 31 July 2023 for open products); ICOBS (renewal pricing 6B.2.1R from 1 January 2022; claims 8.1.1R; cancellation 7.1.1R, 14 days, 30 for pure protection); PROD 4 fair value; BCOBS; CONC (deferred payment credit changes dated 1 April, 15 July and 31 December 2026 as recorded) | PART-1 §6; law register set F | | Complaints and redress | DISP 1 (three-business-day summary resolution; eight-week final response); FOS award limit £455,000 for referrals from 1 April 2026; complaints reporting changes from 31 December 2026 | PART-1 §6.4; Bank requirements | | Systems and controls, resilience and third parties | SYSC 3, 4, 6, 7, 8, 9, 13, 15A; impact tolerances (firms within tolerance by 31 March 2025); PRA SS1/21, SS2/21 (updated March 2026); critical third parties (PS24/16; SI 2026/777 in force 13 July 2026 as recorded); operational incident and third-party reporting PS26/2 (FCA) and PS7/26 (PRA), **in force 18 March 2027**, with initial report within 24 hours and final reports at 30 or 60 working days as recorded | PART-1 §7; PART-2 §6, §13 | | Accountability | SM&CR: SYSC 26 overall responsibility; COCON 2.1 and 2.2 (the test for delegation to agents recorded as an open question) | PART-1 §7.5 | | Model risk and data | PRA SS1/23 (applies to banks, building societies and designated investment firms; excludes insurers); BCBS 239; Solvency UK (SCR at 99.5% over one year; MCR corridor 25% to 45% of SCR) | PART-2 §5, §7, §8 | | Financial crime | MLR 2017 (regulations 18, 19, 21, 27 to 40; amended by SI 2026/621, main provisions 30 June 2026 as recorded; records kept five years under regulation 40); POCA 2002 Part 7 (sections 327 to 340, disclosure to the nominated officer); Terrorism Act 2000; SAMLA 2018 and OFSI guidance (updated 12 May 2026 as recorded); Economic Crime and Corporate Transparency Act 2023 (failure to prevent fraud; identity verification for directors from 18 November 2025); FCA Financial Crime Guide; JMLSG guidance; UK MAR Article 16 | PART-1 §9; FCA-JMLSG pack; EDGE-REG-09 | | Payments fraud and payee checks | PSR APP-fraud reimbursement from 7 October 2024: up to £85,000, within five business days, 13-month claim window, 50:50 between sending and receiving firms; Confirmation of Payee (PSR Specific Direction 17); strong customer authentication; PSRs 2017 delay power for suspected APP fraud | Bank requirements; benchmark register "financial-intelligence" house | | Deposit protection and safeguarding | FSCS £120,000 per eligible person per firm since 1 December 2025 (temporary high balances up to £1.4 million for six months); CASS 15 safeguarding from 7 May 2026 | Bank requirements | | Crypto | FSMA 2000 (Cryptoassets) Regulations 2026 (SI 2026/102); FCA gateway 30 September 2026 to 28 February 2027; regime commencement 25 October 2027 | Bank requirements; REQUIREMENTS-AUDITS §5 | | Data and AI in finance | UK GDPR Articles 22A to 22D (in force 5 February 2026 as recorded); FCA AI Update (22 April 2024; "no extra regulations for AI"); FCA AI Lab, AI Live Testing, Supercharged Sandbox; the Mills Review (6 July 2026, as recorded); EU AI Act Annex III high-risk duties for life and health insurance from 2 December 2027 (as amended by the AI Omnibus, recorded in force 27 July 2026); EIOPA opinion 6 August 2025; IAIS application paper 2 July 2025 | PART-1 §8; PART-3 §5 to §9 | | Cards and messaging | PCI DSS v4.0.1 (future-dated requirements effective 31 March 2025); ISO 20022; Open Banking standard | Law register set S; REQUIREMENTS-AUDITS §2 | | EU comparison | DORA (Regulation (EU) 2022/2554), applying from 17 January 2025 | Law register set F | **Recorded obligations and plans (CLAIM):** - PART-1 holds 118 FCA obligation rows and 31 requirements. Status on 13 October 2026: 27 tested (locally), 3 implemented, 48 specified, 40 not started. "Tested" meant a named unit test passed on a local machine, not live. - Twelve controls were named to build first, in order, with proposed dates (none evidenced as done in scope): an insurance slice with renewal-price check and claim/cancellation/complaint clocks; an outcome monitor by customer group and vulnerability; decision replay with a declared tolerance and a compensating action per action class; an incident classifier with the PS26/2 fields and quarterly dry runs before 18 March 2027; a material third-party register with exit drills; approval of the resilience framework; fairness, calibration, drift and faithfulness monitors; model and prompt versions on every receipt; naming the accountable senior manager and nominated officer and writing a fraud risk assessment; sanctions screening with an effectiveness test; and a temporal-split fraud benchmark. - Where the regime sets no number, the old estate proposed its own targets (CLAIM, never confirmed): simple claims decided within 5 working days; 100% of sampled decisions replayable within a declared tolerance; 100% of action classes with a tested compensating action; 100% of agent actions attributed to a named person. - Bank requirements set a "perimeter before product" rule: the server enforces a tuple of legal entity, activity, jurisdiction, customer type, permission, restriction, effective date, product and mandate; a product label, feature flag or benchmark pass cannot satisfy it; where authority is unresolved, live money movement is unavailable and the interface says so. Thirteen corrections were made to earlier internal claims (for example the £50,000 mobilisation cap is total, not per customer; FSCS is £120,000 not £85,000; one incident clock for the whole estate is wrong). - EDGE-REG-08 (CLAIM): not authorised; banking runs on synthetic data; the intended route is to supply technology to an authorised bank under that bank's authorisation and outsourcing rules (SYSC 8, SS2/21), and only later to consider its own authorisation. - FCA-JMLSG pack (20 October 2026): the company was "not a relevant person" and held no partner data; KYC, fraud and case routes were not live; transaction monitoring was four threshold rules; no SAR or defence-against-money-laundering pack existed. - Synthetic financial-crime experiment (`legacy/local-source-snapshots/research/financial-crime/part-a/results/standard/results.json`, 15 October 2026, CLAIM): on synthetic data, a graph model reached precision about 0.89 against about 0.21 for simple rules, with the same instance detection rate (0.93) and zero unapproved actions. Synthetic only. **Contradiction to note.** `REQUIREMENTS-AUDITS-ISO-FCA.md` repeats a "€5 million" bank capital floor (marked unverified) that `BANK-REGULATORY-AND-PRODUCT-REQUIREMENTS.md` (correction COR-02, same day) explicitly rejects. ###### 3.2 Health (EDGE Clinic) **Mapped instruments (AGREEMENT, as recorded; re-verify):** | Area | Instruments and key dates recorded | |---|---| | Clinical safety | DCB0129 (manufacturer) and DCB0160 (deploying organisation), issued under the Health and Social Care Act 2012 section 250 (current release recorded as amendment 24/2018; NHS England review under way in 2026); a named, registered CSO; hazard log; clinical safety case report | | Medical devices | UK Medical Devices Regulations 2002 (SI 2002/618): the manufacturer classifies, completes conformity assessment (through a UK Approved Body where the class requires), registers with the MHRA; Northern Ireland under EU MDR 2017/745; ISO 14971; IEC 62304; IEC 81001-5-1; ISO 13485 if a device | | Assurance gates | DTAC form version 2.0 (last updated 24 February 2026, as recorded); DSPT 2026-27 (version 9, aligned to the CAF for trusts and ICBs; outcomes published 2 October 2026 as recorded) | | Data | UK GDPR Article 9(2)(h) and DPA 2018 Schedule 1 Part 1 paragraph 2; common-law duty of confidence; Caldicott principles; national data opt-out; Health and Care Act 2022 information standards; Records Management Code of Practice | | Interoperability | HL7 FHIR UK Core; NHS number (ISB 0149, modulus 11 check); SNOMED CT; dm+d; NHS e-Referral Service; Electronic Prescription Service; DICOM | | Regulation of providers and research | CQC regulated activities (Health and Social Care Act 2008 (Regulated Activities) Regulations); HRA approval and research ethics | | Buying routes | NHS frameworks (Health Systems Support Framework, NHS Shared Business Services), Provider Selection Regime 2023, NHS Standard Contract; Sovereign AI Challenge 1 (DHSC) | **Recorded obligations and plans (CLAIM):** - EDGE-REG-07 (version 1.1, 19 October 2026): clinical services ran on synthetic data and were not in clinical use; no CSO appointed; before any live clinical use the company would appoint a CSO, keep a clinical risk management file and hazard log, issue a safety case report per service, complete DSPT and DTAC, and do a deployment DPIA. - EDGE-REG-21 clinical risk management file (19 October 2026): covers seven clinic applications (health record, imaging, radiology, care pathways, lab and pathology, genomics, sequencing). It defines a 5 by 5 severity-by-likelihood matrix (risk 1 acceptable; 2 acceptable only where further reduction is impracticable; 3 undesirable; 4 and 5 unacceptable). Hazard logs EDGE-REG-22 to 28 hold 49 hazards; none rated 4 or 5; three keep a residual 3 (an image class read as a diagnosis; an image unlike the evaluated catalogue; a missed cancer after a false negative). Each control in a hazard log is linked to a file and line of a named release and the build stops if a control cannot be found in code. A test run recorded 6 of 9 failures in one real-pathway suite. - EDGE-REG-29 DTAC (19 October 2026), for a care-pathway product (urine-test-first pathway for suspected bladder cancer): manufacturer's determination that the product is **not** a medical device (it records, schedules and communicates and restates a published referral rule without scoring); DCB0129 not met because no CSO; DSPT "unable to confirm"; no Cyber Essentials; no penetration test; privileged server access lacked a second factor; no FHIR UK Core interface; NHS number not validated against the Personal Demographics Service; CQC registration "not applicable" because the company does not carry on a regulated activity. **Data location**: the service ran in the UK but encrypted backups were kept in Germany (stated as covered by UK adequacy regulations for the EEA). This contradicts statements elsewhere that UK hosting was "company policy". - EDGE-REG-20 pilot DPIA and EDGE-REG-19 clinic DPIA exist as worked examples. - Procurement packs of 20 October 2026: DSPT not registered; DTAC pack recorded "not started" from an 18 October reading although EDGE-REG-29 was issued on 19 October (stale reading); NHS framework choice not taken; no pilot trust agreement, trust clinical lead or DCB0160 case. - Sovereign AI Challenge 1 application (`legacy/edge-judiverse-tech/docs/submission/challenge-1/C1-01-FULL-APPLICATION.md`, 13 October 2026, not submitted): five capabilities (pipelines, cancer detection support, wait-time measurement, smart scheduling, patient-held record); a clinical safety case with a named CSO as a condition precedent to every clinical journey; a medical-device decision per feature; an indicative £1.5 million to £2.5 million over 18 months (CLAIM). ###### 3.3 Public sector, defence, procurement and education (EDGE Public Consortium) **Mapped instruments (AGREEMENT, as recorded; re-verify):** | Area | Instruments and key dates recorded | |---|---| | Procurement | Procurement Act 2023 and Procurement Regulations 2024 (in force 24 February 2025): exclusion and debarment grounds (Schedules 6 and 7), central digital platform registration and core supplier information; below-threshold notices from 1 April 2026 (secondary); local-authority services threshold £207,720 including VAT for 1 January 2026 to 31 December 2027 (as stated on the old public site); R&D procurement under Schedule 2 paragraph 22 | | Procurement policy notes | PPN 006 Carbon Reduction Plan (for contracts above £5 million a year); PPN 09/23 Cyber Essentials for central government and NHS contracts handling personal data | | Government standards | Government Service Standard; Technology Code of Practice; GOV.UK Design System; GovS 005 (digital) and GovS 007 (security); Secure by Design; Algorithmic Transparency Recording Standard; AI Playbook for the UK Government; GPG 45 identity proofing; public sector equality duty (Equality Act 2010 section 149); FOI Act 2000 and Environmental Information Regulations; Local Government Transparency Code 2015 | | Accessibility | Public Sector Bodies (Websites and Mobile Applications) Accessibility Regulations 2018 (SI 2018/952); WCAG 2.2 AA; EN 301 549 | | National security | National Security Act 2023 (core provisions 20 December 2023; foreign influence registration scheme 1 July 2025, as recorded); Official Secrets Act 1989; National Security and Investment Act 2021 and SI 2021/1264 (AI is a notifiable sensitive area; the duty to notify falls on the acquirer); Export Control Order 2008 and retained Dual-Use Regulation | | Defence | Def Stan 05-138 Issue 4 (mandatory under DEFCON 658 from 3 November 2025, as recorded; Cyber Essentials needed at every risk profile); Cyber Security Model and Supplier Assurance Questionnaire; JSP 440 and JSP 604 (current issues unknown); MOD Secure by Design; personnel clearance (BPSS, SC) and List X for classified work | | Higher education and research (consortium "house") | OfS conditions of registration (including E6 harassment); Higher Education (Freedom of Speech) Act 2023; HESA Data Futures; student sponsor duties; right-to-work and DBS checks; ATAS; Trusted Research guidance; Prevent duty; UK Quality Code; Concordat to Support Research Integrity; Verifiable Credentials Data Model 2.0 | | Counter-terrorism and protective security (old consulting site) | Terrorism (Protection of Premises) Act 2025 ("Martyn's Law"): standard tier 200 to 799 people, enhanced tier 800 or more; expected in force spring 2027, with SIA registration from early 2027 (as stated on the old site) | **Recorded obligations and plans (CLAIM):** - EDGE-REG-10: the company "does not develop, and shall not develop, autonomous targeting capability" (a policy line the new estate may wish to consider adopting on its own authority); it would screen technology transfers against export lists and assess investments for national-security notification. - EDGE-REG-12: the company would register on the central digital platform and declare exclusion grounds and conflicts before tendering; a Modern Slavery Act statement was not required below the £36 million turnover threshold. - REQUIREMENTS-AUDITS (13 October 2026) summarises Sovereign AI Challenges 1 to 4: R&D only at TRL 4 to 8; four criteria each scored out of 5, all needing 4 or more; costs at actual cost with no profit; UK-registered lead doing at least 50% of the work; complete by 31 March 2030. It judged Challenge 4 (NCSC, safe AI agent adoption) the best fit. - Funding and procurement routes recorded on 24 October 2026 with closing dates (AGREEMENT, all now to be re-checked): UK Defence Innovation security open call (29 October 2026); ARIA Scaling Trust and Safeguarded AI calls; Innovate UK Frontier AI: SME Champions phase 1 (11 November 2026); Efficient, Structured and Controllable AI Systems (18 November 2026); BridgeAI Deployment Catalyst (8 December 2026). Prerequisites recorded as missing at the time: Cyber Essentials, ISO statements of applicability, a signed clinical safety case, a DPO, first-year accounts, match funding and security clearance. - Old public consulting pages (`legacy/xx-search/public/`) set out statutory clocks as sold advice (AGREEMENT as stated there): 24 hours initial incident notification under the Cyber Security and Resilience Bill (not yet law), 72 hours for a personal data breach to the ICO, 10 working days for a CQC factual-accuracy response, 15 working days to request a CQC rating review; FCA skilled-person (section 166) review costs. ###### 3.4 Cyber security **Mapped instruments (AGREEMENT, as recorded; re-verify):** NIS Regulations 2018 (amended by SI 2026/230, as recorded); Cyber Security and Resilience Bill (in the House of Lords on 7 September 2026; no Royal Assent found on 24 October 2026); Computer Misuse Act 1990 (no statutory defence for security research); Crime and Policing Act 2026; Investigatory Powers Act 2016 section 6 and the Interception by Businesses Regulations 2018 (SI 2018/356); NCSC CAF version 4.0; Cyber Essentials (requirements version 3.3; the "Danzell" question set from about 27 April 2026, as recorded, with automatic-fail questions including MFA on cloud services and critical updates within 14 days); ISO/IEC 27001:2022 with Amendment 1:2024 (transition from the 2013 edition ended 31 October 2025); ISO/IEC 27701:2025 (standalone); ISO 22301:2019; NIST CSF 2.0 (26 February 2024); SOC 2; OWASP ASVS 5.0 and the OWASP Top 10 for LLM Applications 2025; the UK Software Security Code of Practice; the AI Cyber Security Code of Practice (ETSI EN 304 223); NCSC post-quantum migration timelines; NIS2 (EU) for comparison; RFC 9116 for vulnerability disclosure; email standards (SPF, DKIM, DMARC, MTA-STS, TLS reporting). **Recorded obligations and plans (CLAIM):** - EDGE-REG-04: measures described as in place (passkey sign-in bound to devices; consequential actions held for a person; tamper-evident audit records; hourly sealed backups with restore testing; email authentication; rate limits). Before processing any partner's live data the company would obtain CE+, self-assess against the CAF, commission an independent penetration test and adversarial testing of its AI, adopt and exercise an incident response plan, and name an accountable person for security. - The organisational-requirements audit (20 October 2026) recorded host-level weaknesses (privileged remote access without a second factor, databases without encryption at rest, unmanaged database servers, a soft-fail sender policy) and a CAF mapping with 9 of 14 outcomes having a live control and 5 open. Specific technical details are deliberately not reproduced here. - EDGE-REG-35 to 37: incident response and breach plan, business continuity and restore plan, and a vulnerability disclosure policy, all issued 19 October 2026 as templates. - The "door policy" (`legacy/local-source-snapshots/handoffs/DOOR-POLICY.md`, 19 October 2026): device binding first, passkey second, authenticator app only as a second factor; email or text codes not permitted as sign-in or recovery doors. This matches NIST SP 800-63B-4 and FIDO2 direction recorded in the GES mapping. - `legacy/judeai-edge-control/docs/secure-group-call-benchmark.md` (7 October 2026): a benchmark of group-call controls against Microsoft Teams administrator controls and UK regulated-sector expectations (lobby, verified join, recording consent), marked HAVE, PARTIAL or GAP. - 892 cyber benchmark score rows (1.4) are MITRE ATT&CK technique rows whose requirement is coverage or conformance, not a published best score. ###### 3.5 AI assurance **Mapped instruments (AGREEMENT, as recorded; re-verify):** UK pro-innovation AI regulation white paper (CP 815, March 2023) and its five principles; ICO guidance on AI and data protection; UK GDPR Articles 22A to 22D; EU AI Act (Regulation (EU) 2024/1689; prohibited practices from 2 February 2025; general-purpose model chapter from 2 August 2025; Article 50 transparency from 2 August 2026 with a recorded grace period to 2 December 2026; Annex III high-risk duties from 2 December 2027; Annex I from 2 August 2028; amended by the AI Omnibus, Regulation (EU) 2026/1744, recorded in force 27 July 2026); Council of Europe Framework Convention on AI (CETS 225; UK signed 5 September 2024, recorded from prior knowledge because the source refused access; not in force when read); OECD AI Principles; ISO/IEC 42001:2023 (certifiable AI management system), 23894:2023 (AI risk), 42005:2025 (AI impact assessment), 42006:2025 (rules for certifiers), 25059:2023 (AI quality, adds controllability and intervenability), 5338 (AI life cycle), TR 24028 and TR 24027; NIST AI RMF 1.0 (26 January 2023) and AI 600-1 Generative AI Profile (26 July 2024); OWASP Top 10 for LLM Applications 2025 (prompt injection, improper output handling, excessive agency, unbounded consumption); SLSA build provenance, in-toto attestations and reproducible builds; W3C PROV-O; Model Context Protocol (current revision recorded as 2026-07-28); FCA AI Update and Mills Review for finance; MLCommons MLPerf for performance claims. **Recorded obligations and plans (CLAIM):** - EDGE-REG-05 AI governance statement (15 October 2026): applies the five UK principles; uses ISO/IEC 42001 as a reference but "is not certified and has made no commitment to certification"; Jude "proposes; a person decides"; outputs labelled as AI-generated; a model change only where held-back testing shows no regression; capability claims only against a published protocol; no solely automated significant decisions, and if ever offered, Article 22C safeguards (representations, human intervention, contest); a one-month response to anyone asking for the record of an action. - PART-3 standards map (13 October 2026): estate status per standard (42001 "specified: no scope, policy or statement of applicability"; 25059 "tested in part"; NIST AI RMF "not started"); an intended route of an integrated ISO/IEC 27001 and 42001 audit with stage 1 by 30 April 2027 (CLAIM, never confirmed), indicative first-year fees from secondary sources (£12,000 plus £8,000, unverified), then ISO/IEC 27701 and ISO 22301, then an ISAE 3000 assurance report on 4R conformity because "no assurance criteria for agents exist yet". - **Contradiction**: the bid committed to an ISO/IEC 27001 audit at Challenge 4 month 7, PART-3 proposed stage 1 by April 2027, while EDGE-REG-04 said "no commitment". `legacy/local-source-snapshots/programme/proof/procurement/ISO-27001.md` flagged that one of them must change; the organisational-requirements audit lists a statement of applicability and an AI management system "Annex A section" as prepared, with no internal audit or management review. - GES-1.0 standards mapping (CLAIM of which mechanism answers which standard): ISO/IEC 42001 and 23894 → decision kernel and escalation; ISO/IEC 27001/27002 → own identities, least privilege, receipts; ISO/IEC/IEEE 29119 → proof contracts and oracles, "NOT RUN never passed"; ISO 22301 → holds, compensation, recovery; ISO 14971, DCB0129, DCB0160 → clinical actions escalate to a clinician; EU AI Act Article 14 and UK GDPR Article 22 → oversight by escalation and real-time reversal (contestability "to build"); NIST SP 800-63B-4 and FIDO2 → device or passkey door; WCAG 2.2 AA → screen oracle; SLSA and in-toto → sole publisher; OpenTelemetry → action and proof identifiers on every call. - The old benchmark standard (`old-repo/handover/benchmark-standard/STANDARD.md`, 30 October 2026, "proposed standard") separates four claim classes that must never substitute for one another: COMPLIANCE (every applicable clause evidenced), PARITY, SUPERIORITY and CERTIFICATION ("awarded only by an accredited assessor; never self-awarded"); the builder never scores its own run; skipped tasks score zero; gaming (subset selection, undeclared retries, tuning on the test set, unreported failures, self-scoring, changing the test after the run) voids a result. - Agent-specific gaps the old estate repeatedly found (CLAIM, consistent with the new estate's own open item that no know-your-agent standard exists): no regulator or standards text defines a reversible hold, a tested reversal, decision replay, a model recovery objective, or assurance criteria for agents. Its own proposals to fill them were the 4R specification, the Agent Egress Manifest (`legacy/jude-mvp/jbox/Standards/aem.md`, version 0.1: a supplier declares what its agent will connect to and what data it carries, and a gateway reconciles behaviour against it), the "agent evidence to a courtroom standard" bundle (`legacy/jude-mvp/jude-live/docs/evidence-standard.md`), and the Human Authority Protocol (2.3). ###### 3.6 Cross-cutting: data protection, accessibility, online safety, communications and carbon | Area | AGREEMENT items recorded (re-verify) | Recorded position or plan (CLAIM) | |---|---|---| | Data protection | UK GDPR Articles 5, 6, 9, 10, 13 to 15, 17, 20, 21, 22A to 22D, 25, 28, 30, 32 to 35, 37, 44 to 46; DPA 2018 Schedule 1 and section 164A (complaints to the controller, recorded in force 19 June 2026); DUAA 2025 (Article 22 replaced by 22A to 22D, recorded fully in force 5 February 2026; section 19 DPA 2018 repealed and replaced by Article 84B; cookie exemptions moved to PECR Schedule A1; charity soft opt-in from 5 February 2026); UK adequacy decisions renewed by the EU 19 December 2025 | The 15 October 2026 review of the regulatory pack (`.../assurance/regulatory/REVIEW.md`) made 56 corrections, mostly removing overclaims and updating for DUAA. EDGE-REG-01, 02, 03 and 13 to 20, 33, 38 and 40 are DPIA, ROPA, privacy notice, processor terms and exit/deletion templates. Open at the time: processors and locations, the retention schedule, whether user data trains models, director sign-off of the DPIA | | Accessibility | WCAG 2.2 AA (W3C Recommendation 5 October 2023, edited 12 December 2024); Equality Act 2010 sections 20, 29; PSBAR 2018 | EDGE-REG-06: "partially compliant", self-assessed; no independent user testing | | Online safety | Online Safety Act 2023 (illegal-content codes in force 17 March 2025; children's access duties); Ofcom risk assessment guidance version 2.0 (25 June 2026, as recorded); Age Appropriate Design Code | EDGE-REG-30 and EDGE-REG-31 (19 October 2026) and a 21 October 2026 illegal-content risk assessment for the chat, agent channel and music service, following Ofcom's four steps | | Electronic marketing | PECR regulations 21 to 24; CAP Code; Companies Act trading disclosures; CAN-SPAM and CASL for overseas recipients; mailbox-provider bulk-sender rules (authenticate above 5,000 messages a day) | Outreach laws register: no consent record means no marketing send | | Monitoring staff and agents | ICO monitoring-workers guidance; Human Rights Act Article 8; Employment Rights Act 1996 and 2025 | Law register set M | | Carbon and compute | Software Carbon Intensity (ISO/IEC 21031:2024); NESO Carbon Intensity API; Climate Change Act 2008 section 1; PPN 006; ISO 14001, ISO 14064, PAS 2060 | EDGE-REG-11: "shall measure" (no outcome accepted); a Carbon Reduction Plan template (21 October 2026) with no metered baseline; horizon research designs a per-outcome carbon ledger and a lawful-claims gate | | Copyright and media | Copyright, Designs and Patents Act 1988; DUAA sections 135 to 137 on AI and copyright (commencement unknown); Defamation Act 2013 | Law register set C | ###### 3.7 How this can fill the new estate's open regulatory items The new estate's generated list of open items (read 31 October 2026) includes open regulatory questions. The old estate offers the following **research leads**. Each must be admitted as research with its primary source re-read, never as authority. | New-estate open item (summarised) | What the old estate recorded | Lead to admit | |---|---|---| | Retention periods for closed accounts, agent records and payment records under AML, payments and data protection law | AGREEMENT: MLR 2017 regulation 40, records kept five years from the end of the relationship or transaction (PART-1 row FC-12); SYSC 15A self-assessments kept at least six years (FCA-SC-24); decisions examinable six years after the event under DISP 2.8.2R (FCA-AI-15); Limitation Act 1980 six years (contract, tort) and fifteen-year long stop for negligence (4R paper); EU AI Act deployer logs at least six months (Articles 19(1), 26(6)). CLAIM: retention tied to the limitation period of each obligation, with legal hold | MLR 2017 reg 40; DISP 2.8.2R; Limitation Act 1980 ss 2, 5, 14B; UK GDPR Article 5(1)(e) | | Data rights to explain at account closure; external complaint route for data subjects | AGREEMENT: UK GDPR Articles 12(3) (one-month reply, extendable), 13, 14, 15, 17, 20, 21, 22A to 22D; DPA 2018 section 164A complaint to the controller first (acknowledge within 30 days, as recorded); then the ICO. Finance: DISP 1 eight-week final response, then FOS (award limit £455,000 for referrals from 1 April 2026, as recorded) | Privacy notice EDGE-REG-03 v1.0 review clauses 19 to 29 as a worked template | | Regulator information-gathering and inspection powers | Only partly covered: FCA Principle 11 and SUP 15.3 notification duties; SUP 16 returns; FCA skilled-person reviews (FSMA section 166) mentioned on the old consulting site. No register of statutory powers per regulator exists in scope | Gap: research FSMA Part XI (sections 165 to 176), DPA 2018 Part 6 (information and assessment notices), Ofcom and CQC powers | | Know-your-agent standard; agent as account holder; liability between owner, agent provider and bank | Consistent with the new estate: no standard or regulator text found (13 to 30 October 2026). The old estate held that a product label cannot confer permission and that live money movement is unavailable where authority is unresolved (bank requirements). The Jude Bank user manual stated a product position (CLAIM) that actions where the customer turned off the safety hold are treated as the customer's own; this is product terms, not law | Use the old "perimeter tuple" and 4R override states as design research; liability remains for legal advice | | Medical-device status for triage, interaction checks, readings thresholds, questionnaire scoring, retinal grading | The old estate recorded the MDR 2002 route and one manufacturer's determination of "not a device" for a care-pathway scheduler that does not score or recommend (EDGE-REG-29 C1.1.1). It did not assess triage, scoring or grading; its retinal screening service was synthetic only | The DTAC device-decision wording is a template for recording a decision; the new estate's own products need their own decisions with the MHRA route | | DCB0129 and DCB0160 hazard logs and safety cases; CSO | A complete clinical risk management file structure (EDGE-REG-21), a 5 by 5 risk matrix, 49 worked hazards with transfers to the deploying organisation, and the rule "a control that cannot be found in the code stops the build" | Reusable structure and method; all content must be re-written for EDGE Clinic and signed by a CSO | | CQC registration for diagnostic and screening procedures | Recorded only that the old care-pathway product did not carry on a regulated activity (DTAC A11, CLAIM) | Gap for screening and diagnostics; research the Regulated Activities Regulations 2014 Schedule 1 | | DPIAs for health, mental health, social prescribing and triage | Eleven worked DPIA and ROPA documents (EDGE-REG-01, 13 to 20, 40) with lawful bases, Article 9 conditions and processor roles | Templates for structure; content must be redone | | Music licensing (PRS for Music) for the music service | Not covered in scope | Gap | ###### 3.8 Status of the mapping as research - **Breadth**: high. The 24 October 2026 law register and the 30 October 2026 benchmark register together name roughly 150 instruments across finance, health, public sector, defence, higher education, communications, AI and accessibility. - **Depth**: uneven. Finance (insurance and banking) was mapped clause by clause; health had real templates (DTAC, hazard logs, DPIAs); public sector, defence and higher education were named but not mapped clause by clause; cyber obligations were listed with dates but the controls behind them were mostly "specified". - **Reliability of dates**: many facts were from secondary sources; ISO pages returned access refusals; several statutory-instrument commencement dates were "unknown". The organisation's own examinations found wrong statutory-instrument titles and citation errors. Treat every date as a lead. - **Reliability of status**: low. Status words and readiness levels changed between documents written days apart and sometimes contradicted each other (3.1, 3.2, 3.5, 2.5). ##### 4. Manuals, handbooks, guides, playbooks and runbooks Nearly every manual in scope describes an *intended* product or procedure. Most carry their own warning that nothing in them is deployed, tested live, certified or authorised. "Still useful" below means useful as research for the new estate, never as an operating instruction. ###### 4.1 Product specification and operating manuals (one template, 79 documents) - **Paths**: `legacy/edge-judiverse-tech/docs/products/P01-jude.md` to `P78-testing-and-evidence.md`, `P81-judebox.md` (plus `P81-JUDEBOX-BENCHMARKS.md` and `P81-JUDEBOX-JOURNEYS-AND-REQUIREMENTS.md`, outside scope), and the index `README.md`. All dated 13 October 2026 by branch; each on about 65 branches. - **What they teach**: for each of 81 planned apps, purpose and "evidence of success", entry and scope, visible work, a journey table (screen, back, next, completion proof), functions and controls, source modes (live, synthetic, example), failure and recovery, privacy, safety and accessibility, dependencies and acceptance scenarios. - **For whom**: product and engineering teams; secondarily operators. - **Status**: every one states "specified, not implemented". Two manuals compared line by line share about 46% of their lines; they are template output with app-specific inserts. - **Still useful?** As a catalogue of ideas mapped to the new applications: EDGE Finance (P26 Jude Bank to P40 Decision Passports, P59 Insurance, P74 Forecasting), EDGE Clinic (P41 Health Record to P50 Wearables), EDGE Public Consortium (P57 Cyber, P63 Procurement, P64 Logistics, P68 Audit and Evidence, P72 Digital Twins, P51 to P53 research and data), Judiverse (P01 Jude, P02 Canvas, P09 intenJude, P14 Agent Channel, P15 Baton, P16 to P25 office, mail, calls, calendar, terminal). The one-line "evidence of success" per app (for example P29 Payments: "confirmed settlement or explicit unresolved exception"; P36 AML Operations: "reasoned case disposition with evidence lineage") is the most reusable part. Identities, URLs and app numbers must not be carried over. ###### 4.2 Engineering "whole-life" manuals (17 documents) - **Paths**: `legacy/edge-judiverse-tech/docs/journeys/engineering/P09-...` to `P78-...` and `legacy/edge-judiverse-tech/docs/journeys/README.md` (13 October 2026). - **What they teach**: per app, a canonical screen inventory, specialist viewers, four whole-life journeys (for example Cyber: triage an uncertain alert, contain without uncontrolled damage, investigate and preserve evidence, recover and improve detection), proposed data formats and API inventory, privacy and account exit, a benchmark and adversarial proof plan, a documentation gate, and primary sources. - **Status**: "Proposed contracts. No implemented capability, deployed URL, test pass, certification or regulated permission is asserted." - **Still useful?** Yes for Public Consortium cyber, procurement, logistics, audit and testing journeys, and for the Judiverse agent lab (terminal, integrations, developer portal, databases). Journey shapes only. ###### 4.3 Sector operator and user guides | Guide | Path and date | Teaches | For whom | Still useful? | |---|---|---|---|---| | Financial services operator and reviewer guide | `legacy/edge-judiverse-tech/docs/manuals/fca-operator-guide.md` (13 October 2026) | Roles (analyst, investigator, MLRO and sanctions specialists, payments operators, recovery authorisers, model-risk teams, auditors with time-limited read-only access); alert to decision passport; banking API connection lifecycle; financial-crime safeguards; customer challenge; incidents and audit | Deploying regulated firms | Yes, as a role and separation-of-duties model for EDGE Finance; it says the deploying firm keeps its own regulatory responsibilities | | Government operator and purchaser guide | `legacy/edge-judiverse-tech/docs/manuals/government-operator-guide.md` (13 October 2026) | Roles from senior responsible owner to auditor; reviewer separate from preparer for high-impact actions; purchaser acceptance and operating routine; submission checklist | Public bodies and purchasers | Yes, for EDGE Public Consortium buyer journeys | | ICB and care-provider deployment and user guide | `legacy/edge-judiverse-tech/docs/manuals/icb-deployment-and-user-guide.md` (13 October 2026) | Organisation and purpose-based access; commissioners see aggregated data, not patient-level data, merely because they belong to an ICB; deployment checklist; research to reproducible result; imaging, trials and genomics journeys; incident and correction | ICB commissioners, provider clinical teams, CSOs, DPOs, information governance | Yes, for EDGE Clinic (especially the purpose-based access rule) | | Agent governance operator and reviewer guide | `legacy/edge-judiverse-tech/docs/manuals/AGENT-GOVERNANCE-MANUAL.md` (13 October 2026) | Governance roles; first guided task; journeys for mandates, grants and approvals; an agent's attempt to approve its own proposal or grant itself data access is refused and recorded | Operators and reviewers of agents | Yes, for Judiverse agent lab and agents as staff | | Jude agentic banking user manual; Jude Bank user manual | `legacy/edge-judiverse-tech/docs/bank/JUDE-AGENTIC-BANKING-USER-MANUAL.md`; `legacy/edge-judiverse-tech/docs/finance/bank/08-JUDE-BANKING-USER-MANUAL.md` (13 October 2026) | Choosing how the agent acts (explain, prepare only, ask each time, act automatically within limits); standing mandates; review holds and cancellation windows; quiet hours; "Stop Jude" everywhere at once; risk and rights; worked synthetic examples | Retail and business customers | Yes, strongly, for agent-mandate journeys in EDGE Finance; the risk-disclosure-first opening is a good pattern. The "instant bypass removes our extra cover" position is a product CLAIM, not law | | Get started with your estate; Privacy, devices and connections; Administrator guide; end-user licence | `legacy/edge-judiverse-tech/docs/manuals/GETTING-STARTED.md`; `legacy/edge-judiverse-tech/docs/security/user-guide.md`; `legacy/local-source-snapshots/programme/product/DOCUMENTATION/ADMINISTRATOR-GUIDE-v1.0.md` (20 October 2026); `legacy/local-source-snapshots/programme/product/LICENCE-EULA-v1.0.md` (20 October 2026) | One sign-in; profile choices are not permissions; new live workspaces start empty; device and passkey registration; organisation set-up; choosing stations | All users and organisation administrators | Partly; onboarding principles are reusable, product names are not | | Ask Jude | `legacy/jude-mvp/jude-live/docs/ask-jude.md` (with two copies) | Plain-English answers about an earlier product: an agent egress gateway that refuses unapproved destinations and scans messages for hidden instructions and personal identifiers | Users of the earlier product | Low; superseded product, though the egress-gateway idea is research | | Compute: the operator's manual | `legacy/judeai-edge-control/docs/compute/MANUAL.md` | How a compute dashboard derives each number, who may approve spend | Operators | Low to medium | ###### 4.4 Risk, resilience and operations | Document | Path and date | Teaches | Still useful? | |---|---|---|---| | How we take risk: the working handbook for every person and agent | `legacy/edge-judiverse-tech/docs/risk/RISK-05-HANDBOOK-HOW-WE-TAKE-RISK.md` (13 October 2026) | The six-line bold action protocol with six filled examples, a five-minute risk acceptance record, quick sizing, a "decision radius" for when to stop and escalate, no-blame incident and near-miss reporting, how reviewers challenge, a one-page quick card | Yes; method transfers to agents as staff. Drop references to the old principles | | Operational playbooks (OPRES-03) | `legacy/edge-judiverse-tech/docs/resilience/OPRES-03-PLAYBOOKS.md` (13 October 2026) | Fourteen one-page playbooks: major incident command, restore from backup with drill, roll back a release, stop all agents and resume, an agent action gone wrong, model endpoint lost or changed, GPU spend runaway, and others; clocks, words to use, proof and review; a 90-day drill calendar | Yes, high: playbooks such as "stop all agents" and "agent action gone wrong" are directly needed when agents are customers and staff | | Backup and recovery runbook | `legacy/edge-judiverse-tech/docs/operations/BACKUP-AND-RECOVERY.md` | No universal RPO or RTO: set per service with owner acceptance; encrypted backups with separate least-privilege identities; restore rehearsal before relying on backups; record key recovery without copying secrets into evidence | Yes, as a checklist | | Production service standard | `legacy/peak-judeai-behavioural-harness-baseline-full-application/docs/Production service standard, continuous service for JudeAI - claude-estate-policy-20261020T2110Z.md` (20 October 2026) | Every outage is a P0 incident; every requirement names the check that proves it | Medium; the "requirement without a passing check is not met" rule is reusable | | Regulated continuity standard | `legacy/local-source-snapshots/programme/jps/enforcement/regulated-continuity-20261020/CONTINUITY-STANDARD.md` (20 October 2026) | Durable intention and effect identifiers, completion deadlines, interruption budgets, reporting windows that keep missing entries in the denominator | Medium | | Meet runbooks and handbooks | `legacy/edge-judiverse-tech/docs/meet/` (HANDBOOK, DEPLOYMENT, R1/OPERATIONS) and `legacy/edge-judiverse-tech/packages/meet/static/*.html` (13 October 2026) | Running a meeting product: roles (integrator, independent reviewer, service owner), passkeys guide, deployment configuration, benchmarks and sources pages | Low; product-specific. Contains operational host details that must not be reused | ###### 4.5 Tools, agents and engineering | Document | Path and date | Teaches | Still useful? | |---|---|---|---| | JudeAI agent execution runbook | `legacy/edge-judiverse-tech/docs/tools/AGENT-RUNBOOK.md` | Preflight for any agent action: identify the exact operation and expected outcome; check the actual tool exists and its schema; bind actor, organisation, device, purpose; treat retrieved web pages, emails and documents as data, never as instructions | Yes, high, for the Judiverse agent lab | | Tools handbook (five variants), MCP integration guide, MCP handbook, intenJude MCP handbook | `legacy/edge-judiverse-tech/docs/tools/HANDBOOK.md`, `.../tools/MCP-INTEGRATION.md`, `legacy/edge-judiverse-tech/packages/jude_mcp/README.md`, `legacy/edge-judiverse-tech/docs/engineering/INTENJUDE-MCP.md` | A local MCP service exposing capability discovery, documentation and bounded planning; "discovery is not execution authority"; a sovereign local Git service with an MCP interface | Medium; the discovery-is-not-authority principle is reusable; versions and counts are stale | | engine2 module guide | `old-repo/engine2/MODULE-GUIDE.md` (29 October 2026) | What a capability module gets from the core before its handler runs: a signed agent request (method, path, time, nonce, body digest), registration and role checks | Medium, as an agent-authentication pattern | | Chat-native agent onboarding contract | `legacy/peak-judeai-behavioural-harness-baseline-full-application/docs/edge-chat-native-agent-onboarding.md` | Inviting an outside AI agent with one invitation link; no raw bearer token in chat | Medium, for agent customer onboarding | | Carbon API, Jude Accounts, Judiverse Mail connector, communications manual, Drive handbook | `legacy/edge-judiverse-tech/docs/carbon/CARBON-API.md`; `.../finance/JUDE-ACCOUNTS.md`; `.../communications/JUDIVERSE-MAIL-CONNECTOR.md`; `.../communications/README.md`; `.../drive/HANDBOOK.md` (13 October 2026) | Engineering handbooks for specific services (calculation boundaries, accounting scope, a mail adapter tested with fake transport, 27 transactional email definitions with no messages sent) | Low to medium | | intenJude and Jude Office handbook | `legacy/edge-judiverse-tech/docs/benchmarks/INTENJUDE-AND-JUDE-OFFICE-HANDBOOK.md` | Evidence against 60 benchmark criteria for a builder and an office suite, using status words live, tested, implemented, specified | Low | | Native publishing handbook; App, URL and handbook directory | `legacy/edge-judiverse-tech/docs/handover/README.md`; `legacy/edge-judiverse-tech/docs/review/APP-URL-HANDBOOK-DIRECTORY.md` | Delivery states for publishing documents (prepared, submitting, uncertain, delivered, mirror verified); a directory of 81 apps and 790 planned routes | Low; the "uncertain means reconcile, never duplicate" state is a good idempotency pattern | | Public manuals and specification catalogue | `legacy/judeai-edge-control/docs/research/life-control-plane/manual_catalogue.md` (10 October 2026) | 137 public manuals, standards and repositories with reading depth marked | Medium, as a reading list | | Electronic health record inventory | `legacy/judeai-edge-control/docs/research/platform-inventories/ehr.md` (10 October 2026) | 80 capabilities of major record systems across 10 areas, including 8 regulatory | Medium for EDGE Clinic | | Designing and reviewing a journey (UI handbook) | `legacy/local-source-snapshots/foundations/canvas/reference/canvas/canvas-review/ui-excellence/handbook/document.md` | Start from a person, an outcome and the real service; one principal next action; examples must never impersonate saved work | Medium | | Render integrity standard | `legacy/edge-judiverse-tech/docs/standard/RENDER-INTEGRITY-AND-PROGRESSIVE-REVELATION.md` (14 October 2026) | Arithmetic layout rules and a render gate | Low to medium | | Verify-a-pack instructions | `legacy/jude-mvp/proof/unpacked/pack-*/VERIFY.md` (September 2026) | How to check an evidence pack without the vendor: hash sources, check a signature, walk a hash chain, compare with an independent copy | Yes, as a pattern for regulator-verifiable evidence | | IT onboarding page | `legacy/governed-intelligence-judeai-edgeid/office/it-onboarding/index.html` | Title only read (an IT onboarding page for an office product) | Low | | In-app guides (74 JSON files) | `legacy/local-source-snapshots/services/edge/web/estate/guides/` | Named after bank, clinic staff, consortium, governance and identity screens; content not read in this pass | Unknown until read | ###### 4.6 Agent instruction files, policies and standards for working (research only) The scope also holds files that instructed agents or staff: several `AGENTS.md` variants, the AEC readiness standard (`legacy/local-source-snapshots/services/edge/Master/AEC-STANDARD.md`, 19 October 2026: agents play every actor in a journey on live production before any real person is involved), the engineering standard of 20 October 2026, the error-control standard of 19 October 2026, the documentation and legal-drafting styles, the company policy and privacy policy drafts of 3 September 2026 (`legacy/jude-mvp/jude-live/policies/`), the old licence for the governed-intelligence source, and the retired canonical rules of plan 2.5.0 (`legacy/mac-retired-20261030/policy-history/...`). These are **not** instructions for the new estate. Two ideas are worth keeping as research: the error-control standard (pre-specified type I error control, sequential methods for repeated looks, false discovery control for exploration; see 2.3) and the rule that a builder never certifies its own work. `legacy/judeai-edge-control/Master/References/agent-accounts.md` describes where agent and test accounts and their credentials were kept. Its contents are not reproduced and it must not be reused. ##### 5. Reuse in the new estate, and what is superseded ###### 5.1 Worth admitting as research (with sources re-read) | Item | Source | Why | New-estate application | |---|---|---|---| | Agent-action governance pattern: server-issued authority envelope that only narrows on delegation; reversible hold; compensating action per action class; chained, bitemporal receipt per action; four override states; kill switch; decision replay | 4R paper and specification; Reversible-by-default paper; GES-1.0 paper (sections 2.1, 2.3) | The most developed and consistent idea in the old estate, and directly relevant when agents are customers and staff | All four applications; EDGE Finance agent mandates; Judiverse agent lab | | Per-sector regulatory registers | Law and standards register of 24 October 2026; FCA, PRA and data/AI Parts of 13 October 2026; bank requirements and their 13 corrections; benchmark register regulatory rows of 30 October 2026 (section 3) | A ready reading list with clause references and dates for finance, health, public sector, defence, higher education, cyber and AI | Fill open regulatory items (3.7) after primary-source checks | | Regulatory document templates | EDGE-REG-00 to EDGE-REG-40 and the 15 October 2026 review of 56 corrections | Realistic structure for DPIAs, ROPA, privacy notice, AI governance statement, accessibility statement, clinical risk management file and hazard logs, DTAC answers, OSA assessments, processor terms, service levels, incident response, continuity, vulnerability disclosure, exit and deletion, acceptable use | Structure only; every fact must be re-established for the new estate | | Honest status vocabulary and claim classes | PART-1 §2 (tested, implemented, specified, not started; "tested" never means live); benchmark standard claim classes (compliance, parity, superiority, certification never self-awarded) | Prevents overclaiming, which the old estate repeatedly corrected | Status fields and public claims | | Independent pre-publication examination | Public papers examinations of 14 October 2026 | Two rounds caught wrong citations, wrong statutory-instrument titles and internal inconsistencies | Any regulator-facing paper | | Error-control and statistics rules | Error-control standard and methods section; the 3/n rule for zero observed failures | Sound and general | Clinic evaluation, fraud detection, agent testing | | Operational playbooks and risk handbook | OPRES-03; RISK-05; backup and recovery runbook | Concrete, short, agent-aware incident procedures | Operations for all applications | | Agent egress manifest and evidence-bundle ideas | `legacy/jude-mvp/jbox/Standards/aem.md`; `legacy/jude-mvp/jude-live/docs/evidence-standard.md`; VERIFY.md packs | Open, checkable declarations of where an agent may send what, and evidence a third party can verify without the vendor | Public Consortium cyber governance and vetting; agent customers | | External benchmark reference scores | 43 OBSERVED rows in `old-repo/handover/benchmark-standard/scores/` | A starting list of published bests with source digests | Benchmarks, after re-reading at source | | Sector role models | FCA operator guide; government operator guide; ICB guide; agent governance guide; banking user manuals | Separation of duties, purpose-based access, risk-first disclosure to customers | Journeys and roles in each application | | White paper template | `legacy/local-source-snapshots/workbenches/capabilities/WHITE-PAPER-TEMPLATE.md` | Forces the rule quoted, the failure of current tools, the evidence and "what is not claimed" | Any new paper | ###### 5.2 Superseded or not to be reused - **All readiness levels, counts and status claims** (TRL 9, CRL 9, AEC 9, "DECFAW 9", node counts, journey counts, test pass counts). They conflict with each other and were never independently confirmed. - **The old company position** (registrations, insurance, appointments, hosting, processors). The new estate must establish its own; nothing transfers. - **Product identities, app numbers (P01 to P82), URLs, route counts, rule numbers and principle numbers**. The new estate uses its own identities. - **Proposed dates** in every plan ("30 October 2026", "13 November 2026", "stage 1 by 30 April 2027"). None is evidenced as met in scope. - **The 13 October 2026 status counts in PART-1 to PART-3** and the 20 October 2026 organisational audit: historical snapshots only. - **Internal contradictions to avoid repeating**: the rejected "€5 million" bank capital floor; "no commitment" versus "audit at month 7" for ISO/IEC 27001; "UK hosting" versus backups held in Germany; a DTAC pack "not started" one day after the DTAC form was issued. - **Agent instruction files and owner orders** (AGENTS.md variants, AEC standard, engineering standard, door policy as an order, retired canonical rules): research only. The door-policy *direction* (device binding and passkeys, no email or text codes for sign-in or recovery) can be re-derived from NIST SP 800-63B-4 and FIDO2 if the new estate wants it. - **Charcoal currency standard**: an early idea with no evidence; not relevant to the four applications except as background to carbon measurement. - **Record families**: test outputs, publication receipts, synthetic exam papers, rendered-answer captures, third-party licence texts and copies of third-party pages have no reuse value as documents. - **The account and credential reference file** named in 4.6: must not be reused. ###### 5.3 Gaps the old estate did not close - No clause-by-clause mapping for public sector, defence or higher education, only lists of instruments. - No register of regulators' statutory information-gathering and inspection powers. - No medical-device assessment for triage, scoring, retinal grading or interaction checking. - No legal analysis of liability between an agent's owner, the agent's provider and a bank, or of agents as account holders. - No music-licensing analysis. - No certification, accredited audit, penetration test, independent accessibility test or independent reproduction of any benchmark. ### research/legacy/STRATEGY-FUNDING-OTHER.md `research/legacy/STRATEGY-FUNDING-OTHER.md` · sha256 `f7a5315e607a8a432c1e23595db6c8ab37470aa94ae03a9ea1d3778d426c4f7a` #### Legacy strategy, funding and "other" documents: research reading for JudeAI | EDGE ID **What this is.** A reading of the organisation's earlier work (the old estate) for three catalogue groups: documents tagged `strategy-research`, documents tagged `funding-grants-tenders`, and documents whose only tag is `other`. It records what that work planned, researched, applied for and learned, so the new estate (JudeAI | EDGE ID, with EDGE Finance, EDGE Clinic, EDGE Public Consortium and Judiverse) can reuse it as research. **What this is not.** It is not authority. Nothing here is a rule, plan, approval, readiness level or count for the new estate. Every figure is historical and is quoted with the corpus-relative file it was recorded in and that file's date. Anything worth keeping should be admitted into the new estate as research through the new estate's own admission route, with new identities, as an observation or claim with its source. **Markers used** (the same as `new-estate/research/README.md`): **OBSERVATION** (seen in a primary source the old document quotes and the old document says it read), **CLAIM** (what an old document says about the organisation, its product or its market, not independently checked by this reading), **PREDICTION** (an estimate or target, with its assumptions), **UNKNOWN** (looked for in the corpus and not found). Unless a line says otherwise, every statement about the market, customers, results or outcomes below is a **CLAIM** made by the old document. **Acronyms.** IUK = Innovate UK. UKRI = UK Research and Innovation. ARIA = Advanced Research and Invention Agency. UKDI = UK Defence Innovation. DHSC = Department of Health and Social Care. MOD = Ministry of Defence. NCSC = National Cyber Security Centre. BIST = the Department for Business, Innovation, Science and Trade as named in the old documents. NHS = National Health Service. ICB = integrated care board. FCA = Financial Conduct Authority. PRA = Prudential Regulation Authority. MHRA = Medicines and Healthcare products Regulatory Agency. CQC = Care Quality Commission. ICO = Information Commissioner's Office. IFS = Innovation Funding Service (the IUK portal). EOI = expression of interest. TRL = technology readiness level (1 to 9). CRL = commercial readiness level (an internal scale in the old estate). SME = small or medium-sized enterprise. SAM = serviceable addressable market. TAM = total addressable market. DTAC = Digital Technology Assessment Criteria (NHS). DSPT = Data Security and Protection Toolkit (NHS). DCB0129 and DCB0160 = NHS clinical safety standards for manufacturers and deployers. DPIA = data protection impact assessment. KYC = know your customer. AML = anti-money laundering. SAR = suspicious activity report. DAML = defence against money laundering request. LSBU = London South Bank University. SITHub = LSBU's Smart Internet Technologies Hub. GPU = graphics processing unit. FTE = full-time equivalent. G-Cloud = the UK government cloud services framework. AI DPS = the government's artificial intelligence dynamic purchasing system. ATBAN and AEC are old-estate labels for agent identities and an evidence-level scale; they are not carried over. --- ##### 0. Method and coverage (stated honestly) **Source.** `catalogue.jsonl` in the corpus, rows with `"generated": false` only. **Scope selected.** | Group | Rows | Distinct blob oids | Notes | |---|---:|---:|---| | `strategy-research` or `funding-grants-tenders` in tags | 1,115 | 1,075 | collapsed to 887 distinct documents after merging branch copies whose file names differ only by an eight-character hash suffix | | tags exactly `["other"]` | 3,174 | 3,080 | 2,747 distinct by normalised text | | Total | 4,289 | 4,155 | 3,802 distinct by normalised text | **Deduplication.** Rows were merged first by blob oid, then by a hash of the whitespace-normalised text (this removes identical copies of a document held on different branches or under hash-suffixed names), then, for listing, by "stem" (path with a branch hash suffix removed). Where near-identical versions differ (for example `ONE-PLAN.md` and `ONE-PLAN.8d08b0e8.md`, or the `C1-01` to `C4-01` applications and their hash-suffixed copies), the later or fuller version was read and the difference noted. **Dates.** The catalogue holds no commit dates. Each date below is taken, in order of preference, from the document's own content or title, from a date in its path, or (marked "branch") from the earliest dated branch name the file appears on, which is approximate. "Undated" means none of these gave a date. **Triage.** Every in-scope document was text-extracted (HTML stripped; PDF, DOCX, PPTX and XLSX read from their extracted text), its first 400 characters recorded, and keyword-scored for grant, Innovate UK, UKRI, tender, competition, investor, funding, money, market-size and funder-name terms. All 887 strategy and funding documents were listed by path and title and reviewed at that level. About 120 documents carrying the substance (applications, funder guidance copies, costings, decks, market and buyer research, plans, partner packs, research notes, prompt and principle files) were read in their substantive sections; very large compilations (for example `legacy/local-source-snapshots/MASTER-PLAN.md`, about 9 MB, which is a verbatim compilation of other documents) were read by their table of contents, sources list and the sections relevant here. **Not read in full (honest limits).** - Many strategy-tagged files are machine data or evidence logs rather than strategy: test fixtures (for example the 63 JSON fixtures under `legacy/judeai-edge-control/docs/research/compounding-verified-intelligence/backend/fixtures/`), per-seed result files under `legacy/local-source-snapshots/research/financial-crime/part-a/results/`, evidence receipts under `legacy/local-source-snapshots/programme/outcomes-register/evidence/`, large parity and definitions registers (`*.jsonl`), screen-state matrices and requirement ledgers. These were triaged by path, title and opening text only; their summaries (where one exists) were read instead. - In the "other" group, 673 distinct HTML pages (product screens, email previews, brand pages) and 148 evidence logs were triaged by path and title, not read page by page. - `legacy/jude-mvp/judiverse-live/jude_dataset/research/pentest.jsonl` and companion files (a penetration-testing research dataset, about 4.5 MB) were identified by their README only. **"Other" classified by content and skipped as plainly irrelevant** (listed in section 4.5): vendored third-party files, a licence notice, synthetic training fixtures, test fixtures and build intermediates. --- ##### 1. Inventory of strategy, vision, research, roadmaps, intentions, decks, applications, tenders and funding material The tables list the documents that carry the substance. The date column says where the date came from. "Kind" is this reading's description, not the old document's status. Any figure in a title column is recorded in that row's file on that row's date and is historical. Where a document exists in several branch copies, the row names one copy; the other copies were compared and found identical or near-identical. **FUNDING: Innovate UK Frontier AI SME Champions (competition 2584)** | Title | Corpus-relative path | Blob oid | Date | Kind | |---|---|---|---|---| | Research file on competitions 2584 and 2599 (rules, themes, scoring) | `legacy/edge-judiverse-tech/docs/dossiers/frontier-ai/00-COMPETITIONS.md` | `c248a17a43ab81dbace73c17354805c76fc26c74` | 2026-10-13 (content) | research note | | Application 2584, submission version (Theme 4, adaptive adversarial discovery engine) | `legacy/edge-judiverse-tech/docs/submission/frontier-ai/APPLICATION-2584-SUBMISSION-VERSION.md` | `07d6ff81fdf9f7aaf594b49f622b675fc7458fe7` | 2026-10-13 (content) | grant application draft | | Application 2584, Theme 4 working version | `legacy/edge-judiverse-tech/docs/submission/frontier-ai/APPLICATION-2584-THEME-4.md` | `50a887ad483d255d5cfb4b4fda262dbfa117a335` | 2026-10-13 (content) | grant application draft | | The two competitions and the estate (eligibility checklist) | `legacy/edge-judiverse-tech/docs/submission/frontier-ai/CAPABILITY-CHECKLIST-AND-FUNDING-RULES.md` | `d5819b3ce47e67647a3f528c4a2fb6967ecaa17f` | 2026-10-13 | funding rules | | 2584 project plan appendix | `legacy/edge-judiverse-tech/docs/submission/frontier-ai/appendices/APPENDIX-PROJECT-PLAN.md` | `74fb8dcfa2ab3a6020aa644f588140413895f168` | 2026-10-13 | grant appendix | | Simulated assessor critique of the dossier | `legacy/edge-judiverse-tech/docs/dossiers/frontier-ai/09-ASSESSOR-CRITIQUE.md` | `c2de2d3aac0a992986f0b977c7e196c74de780ac` | 2026-10-13 (content) | internal review | | Simulated three-assessor panel scores and corrections | `legacy/edge-judiverse-tech/docs/dossiers/frontier-ai/14-assessor-panel-and-proofs.md` | `1a7948c79821a1bb843f867f80ec3f2375fc8327` | 2026-10-14 (content) | internal review | | Leadership and eligibility requirements plan | `legacy/edge-judiverse-tech/docs/dossiers/frontier-ai/13-leadership-and-eligibility-requirements.md` | `b66e3b382094b68a7524bab2c9e67edd52caae2a` | 2026-10-13 | plan | | Capability checklist and funding rules (dossier copy) | `legacy/edge-judiverse-tech/docs/dossiers/frontier-ai/06-capability-checklist-and-funding-rules.md` | `a3828684ca047c8e0564df41b6a2634aed512d6e` | 2026-10-13 | funding rules | | Submission closure handover ('not ready to submit') | `legacy/edge-judiverse-tech/handoffs/submission-closure-20261014/README.md` | `dd31f6cee8d2f3135d6f1faf9813df739a5606c4` | 2026-10-14 | status record | | Review against retained 2584 guidance ('revised draft, not submission ready') | `legacy/edge-judiverse-tech/handoffs/submission-closure-20261014/ASSESSMENT.md` | `4375411976188c5b2673f11dd58e9df3f133aefb` | 2026-10-14 | internal review | **FUNDING: Sovereign AI R&D Procurement Scheme (Batches 1 to 3)** | Title | Corpus-relative path | Blob oid | Date | Kind | |---|---|---|---|---| | Verbatim copy of the Scheme's competition guidance | `legacy/edge-judiverse-tech/docs/submission/SOVEREIGN-AI-COMPETITION-GUIDANCE.md` | `76e8a266e92d922779fa40d57b0c6f87eb6b6245` | 2026-10-13 (fetched) | funder guidance (copy) | | One plan: four Sovereign AI proposals (no acknowledgement inspected) | `legacy/edge-judiverse-tech/docs/submission/ONE-PLAN.8d08b0e8.md` | `8d08b0e83132065be917b232c46256d3284a1d5f` | 2026-10-14 (content) | plan / status | | Challenge 4 full application, first version (superseded by C4-01) | `legacy/edge-judiverse-tech/docs/submission/CHALLENGE-4-FULL-APPLICATION.md` | `4f96cfb347e38e3cf57f97685d1bb4f2f385fb6c` | 2026-10-13 | tender response draft | | Challenge 4 itemised cost model (at cost, no profit) | `legacy/edge-judiverse-tech/docs/submission/CHALLENGE-4-COST-MODEL.md` | `25215fceb8947e12b781dcf5796cff5fa08b9212` | 2026-10-13 | costing | | Challenge 1 (NHS productivity, DHSC) full application | `legacy/edge-judiverse-tech/docs/submission/challenge-1/C1-01-FULL-APPLICATION.md` | `413d877c7f3837fbf141ae37a47ac4fcdcb95c4a` | 2026-10-13 (content) | tender response draft | | Challenge 2 (Defence edge runtime, MOD) full application | `legacy/edge-judiverse-tech/docs/submission/challenge-2/C2-01-FULL-APPLICATION.md` | `9eff9ef4b2c820b42dc5521df5ae0191f96a50ea` | 2026-10-13 (content) | tender response draft | | Challenge 3 (compute efficiency, BIST and ARIA) full application | `legacy/edge-judiverse-tech/docs/submission/challenge-3/C3-01-FULL-APPLICATION.md` | `dec70edc97d597c91853da433b3f6d5c0328a4ca` | 2026-10-13 (content) | tender response draft | | Challenge 4 (safe agent adoption, NCSC) full application v1.1 | `legacy/edge-judiverse-tech/docs/submission/challenge-4/C4-01-FULL-APPLICATION.md` | `7e4371189d36d9551c2c89a973eb04514b2a52e7` | 2026-10-13 (content) | tender response draft | | What must be in place: audits, ISO standards, FCA licences | `legacy/edge-judiverse-tech/docs/submission/REQUIREMENTS-AUDITS-ISO-FCA.md` | `c235ec16b5c7a6a4f0662e695206199f5757462d` | 2026-10-13 | requirements | | Position on the Scheme's master contract terms | `legacy/edge-judiverse-tech/docs/contracts/CT-01-MASTER-TERMS-POSITION.md` | `32e0ab4f97ecd6a4f24b2933e8acef6c1ed38943` | 2026-10-13 | contract research | | Evidence packs UK-PACK-01 to 11 index (271 items, 123 gaps) | `legacy/edge-judiverse-tech/docs/packs/INDEX.md` | `cb6f052793d6da2d4802fe199e0ce67ff81ecce4` | 2026-10-13 (content) | evidence pack index | | Procurement, commercial support, insurance and value pack | `legacy/edge-judiverse-tech/docs/packs/UK-PACK-09.md` | `5242c7683544925b6fb4138cb12d2a9873c66b2d` | 2026-10-13 (content) | procurement pack | | Four challenges: itemised planning estimate | `legacy/jude-mvp/docs/four-challenge-costs.md` | `bad03103753b77c013f7ef7d8984f1cee2766ed7` | 2026-09-09 (title) | costing | | Four bids, four acceptance cases (at £10m each) | `legacy/jude-mvp/docs/acceptance-targets.md` | `ff00287e74d3b4ff9b718b733ceb75c771072291` | undated (September 2026 by context) | bid design | | Build plan: the agentic solution and its cost | `legacy/jude-mvp/docs/build-plan.md` | `2dabb03616220e6a719dbb14d6daf857a701460e` | undated | bid design | | Work packages and research category (ten quarters) | `legacy/jude-mvp/docs/work-packages.md` | `a0b635aadb16897c03bba53dcce73e4b4baf04d0` | undated | bid design | | The research programme, and what we ask of partners | `legacy/jude-mvp/docs/rnd.md` | `0c6bfce538a4c147cd09917dc55d6902744624e6` | 2026-09-13 (branch) | research programme | | The plan to 2030, worked backwards | `legacy/jude-mvp/docs/plan-2030.md` | `e5c6aecb2b261c41a7c6cc1ba0b8315cda9c8114` | 2026-09-13 (branch) | vision / roadmap | | Sovereign Authority over Machine Action (Challenge 4 research proposal) | `legacy/jude-mvp/docs/research-paper.md` | `cbd405a4f97d3bf4d27d80138ebad056e1cd4dc7` | undated | research proposal | | Challenge 4 eligibility gates and claims (own reading) | `legacy/jude-mvp/manifests/challenge-4.requirements.json` | `20f575992622d92bcb9cd88fffd2905d4f1ae1e7` | undated | requirements | | The value model (public-sector fraud; Challenge 4 era) | `legacy/jude/docs/value-model.md` | `5d98eb8f43fdabf8d07f440b11577524b0fbb63e` | undated | market model | | Batch 2 milestones M1 to M6 for Challenges 1 and 4 | `legacy/local-source-snapshots/programme/proof/procurement/SOVEREIGN-AI-MILESTONES.md` | `240320c8e456a2fbf0253232d81c14774206b3ed` | 2026-10-20 (content) | bid plan | | Proof programme (cites the 'BID' and 648 unproven claims) | `legacy/local-source-snapshots/programme/proof/PROOF-PROGRAMME-20261020.md` | `ef61755725bf3f86ca18a62101597eefa2014d6c` | 2026-10-20 | programme / status | | Every gap from procurement and partner packs | `legacy/local-source-snapshots/programme/proof/PACK-ACTIONS-20261020.md` | `4363474a641c59620b2efad1f8269dfd8460d355` | 2026-10-20 | action register | **FUNDING: horizon plan applications (late October 2026)** | Title | Corpus-relative path | Blob oid | Date | Kind | |---|---|---|---|---| | UK funding and procurement calendar (40 calls, 16 watched, 14 exclusions) | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__applications__UK-FUNDING-CALENDAR.md` | `3c5cd573b2b553078f1cacb460013fe0ad791995` | 2026-10-25 (content) | funding register | | Owner decision: 'yes to everything relevant ... in UK' | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__applications__OWNER-DECISION-20261025.md` | `08f6e024ed3b2e0235fa57dfec95574670189981` | 2026-10-25 (content) | decision record | | Innovate UK 2527 accelerated cancer diagnosis application text | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__applications__IUK-2527__APPLICATION.md` | `64be0804c288f7661ba7894cf305f77ebb808ac5` | 2026-10-25 (content) | grant application draft | | 2527 finances | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__applications__IUK-2527__FINANCES.md` | `25e9dc45ea0326570da873077f5e6b85b402ffea` | 2026-10-25 (content) | costing | | 2527 gaps only the Owner or a partner can fill | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__applications__IUK-2527__GAPS.md` | `83b89b671a1fff910957c7f9d0cd886e9ded00e6` | 2026-10-25 (content) | gap register | | 2527 internal score check | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__applications__IUK-2527__SCORE-CHECK.md` | `43282a335c605d0e8792a04bb49ea7f6972ee2ff` | 2026-10-25 (content) | internal review | | UK Defence Innovation Security Open Call cycle 2 stage 1 text | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__applications__UKDI-SECURITY-C2__APPLICATION.md` | `76c1d0789d16cc9479e40fc91ce03fdd1ebd362f` | 2026-10-25 (content) | grant / contract application draft | | UKDI finances | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__applications__UKDI-SECURITY-C2__FINANCES.md` | `d50523e5699457de9634c78a70d1bf839860e145` | 2026-10-25 (content) | costing | | Government, MOD and FCA engagement and funding research | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__research__RESEARCH-GOVERNMENT-MOD-FCA-FUNDING.md` | `d63542711672b11f6e1ceacab43281347a74e202` | 2026-10-24 (content) | research | | Scraped register of 312 open grants, tenders and competitions | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/stations/registers/funding/funding-20261018.json` | `54aecdf86f05cd52f7ef0ac0cdeea1255ada8ec6` | 2026-10-18 | opportunity register (machine) | | Draft vision and approach for a UKRI sustainable AI bid (no lead recorded) | `legacy/local-source-snapshots/programme/outcomes-register/evidence/2026-10-19/codex/t9-t10/packs/sustainable-ai/DRAFT-Vision-and-Approach-20261019-v1.pdf` | `c828470ac3efb95f847d689631aaeaf7154a87a2` | 2026-10-19 | grant draft | | Earthshot Prize 2027 route check (no nomination) | `legacy/local-source-snapshots/programme/outcomes-register/evidence/2026-10-19/codex/t9-t10/packs/research/earthshot-source-update.json` | `a7ff1db64761adc19298d3af443472c39b9c8e9c` | 2026-10-19 | prize research | **INVESTOR, PARTNER AND PITCH MATERIAL** | Title | Corpus-relative path | Blob oid | Date | Kind | |---|---|---|---|---| | X-X.TECH x NVIDIA Inception deck (pre-seed £3m ask) | `legacy/local-source-snapshots/foundations/engine/examples/nvidia-inception-deck.html` | `d438802013865fbceac28d9b10bba2d6ac8b3160` | 2026-10-15 (branch) | pitch deck | | Same deck, rendered PDF | `legacy/local-source-snapshots/foundations/engine/examples/X-X-TECH-NVIDIA-Inception-Deck.pdf` | `b36f6453b528788ee8ecb382da4e4490495b7807` | 2026-10-15 (branch) | pitch deck | | JudeAI | Civo partnership proposal deck content (LSBU SITHub meeting) | `legacy/jude-mvp/work/civo-lsbu-deck-20261006/content.json` | `32a99575a247636607c0c78d9c875291287e4c74` | 2026-10-06 | partner pitch deck | | Same deck, rendered PDF | `legacy/jude-mvp/work/civo-lsbu-deck-20261006/lo/JudeAI-Civo-LSBU-Final-Deck.pdf` | `a92e0a96d5145ea7326bc66747fafb9c41eaba44` | 2026-10-06 | partner pitch deck | | Research programme partner deck (27B governed system vs frontier) | `legacy/jude-mvp/work/jude-partner-evidence-deck-20261002/.codex-finalizer/candidate.pptx` | `5fa738a3f207a6422898a4dec5b875278ee102fe` | 2026-10-02 | partner pitch deck | | Proposed national programme deck text (UK sovereign AI ecosystem) | `legacy/jude-mvp/work/finance-pilots-20261001/partnership.txt` | `8ab17c93a321ff1a421c38034cda9cc8f771b51d` | 2026-10-06 (content) | partner pitch deck | | X-X.TECH and LSBU SITHub technical collaboration brief | `legacy/jude-mvp/docs/LSBU-COLLABORATION-BRIEF.md` | `75574ff47087f9a16a5417b76df9db058a3d962a` | 2026-10-03 (branch) | partner brief | | Governed Execution Across Institutional Boundaries v2.0 | `legacy/judeai-edge-control/docs/consortium-plan.md` | `02eb39c52d0dc5572875c1601fb9705b732d83d2` | 2026-10-10 (content) | consortium research plan | | Governed AI Execution Across Institutions, Edge and UK Sovereign Cloud v2.0 | `legacy/peak-judeai-behavioural-harness-baseline-full-application/docs/CONSORTIUM-PLAN-v2.md` | `96f448164adaafd846279173d78285f0cd2b7eb4` | 2026-10-10 (content) | consortium research plan | | LSBU partner pack | `legacy/local-source-snapshots/programme/proof/partners/LSBU.md` | `e03c43a3a9f59bc48565e7bde077f6ca1950efd3` | 2026-10-20 (content) | partner pack | | Imperial partner pack (out until they reach out) | `legacy/local-source-snapshots/programme/proof/partners/IMPERIAL.md` | `eee1d520882f38d68c1c58a701992a7acfa8396a` | 2026-10-20 (content) | partner pack | | Draft research study terms, urine DNA follow-up | `legacy/local-source-snapshots/programme/proof/partners/agreements/DIAGNOSTIC-FOLLOW-UP-STUDY-TERMS.md` | `4a52ed7640c8284c42e5323a406c3e6fd7dbaaaa` | 2026-10-20 (content) | draft agreement | | Capacity Recovery Pilot commissioning note | `legacy/local-source-snapshots/programme/integrated-capacity-model/CONSORTIUM-COMMISSIONING-NOTE.md` | `3f878434338d22109f2636853f14871c1566667f` | 2026-10-21 (branch) | pilot offer | | Invitations ready to send to named institutions | `old-repo/evidence/INVITATIONS-20261029.md` | `dcc45d2c46f3984d7e4d590c4b6877f984d02491` | 2026-10-29 | outreach drafts | | Commercial preparation: free evaluation and quote engine (prices unset) | `old-repo/docs/COMMERCIAL-MODEL.md` | `8104f7871e362709cafec94f80e3b569ae74124a` | 2026-10-29 | commercial model | **STRATEGY, VISION AND MARKET RESEARCH** | Title | Corpus-relative path | Blob oid | Date | Kind | |---|---|---|---|---| | Twelve institutional offers, buying routes, pains, pricing architecture | `legacy/edge-judiverse-tech/docs/research/03-INSTITUTIONAL-OFFERS.md` | `4803b16aa42d30b011fcedd9aa4f6bbc11c6fc7c` | 2026-10-13 (content) | market research | | Capability gap register against industry leaders | `legacy/edge-judiverse-tech/docs/research/02-CAPABILITY-GAP-REGISTER.md` | `ae06f3a8509c502c7acd90937486b78a4651d682` | 2026-10-13 (content) | competitor research | | Estate review: what exists on 13 October 2026 | `legacy/edge-judiverse-tech/docs/research/01-ESTATE-REVIEW.md` | `54fa88a54b8d8256f18982c3c99e9ed86f65fa9d` | 2026-10-13 | status review | | Missions register: twelve missions | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__MISSIONS.md` | `18e6c6935b81effd05887e5ce72f5d634fbd7aa0` | 2026-10-24 | strategy | | Horizon plan (whole estate in one pass) | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__HORIZON-PLAN.md` | `789f64d6f392a40821d86695ebbaefcbaaf6d88d` | 2026-10-24 | plan | | Global problem register and sovereign position | `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__research__RESEARCH-GLOBAL-PROBLEM-REGISTER-AND-SOVEREIGN-POSITION.md` | `5e295eb255bd3d5b9c032fa977142ad529265391` | 2026-10-24 | research | | Data-centre heat reuse pilot assessment (desk stage) | `legacy/judeai-edgeid-cloud/plans/opportunities/heat-reuse-opportunity-assessment-20261025.html` | `8dc0310570896d4ffe9fa6f917bcfa9d92e767e3` | 2026-10-25 | opportunity assessment | | 'One Mans Trash' waste programme | `legacy/judeai-edgeid-cloud/plans/opportunities/one-mans-trash-programme-20261025.html` | `da00f0779a4c124daea67a2ec5d59b1661297745` | 2026-10-25 | programme concept | | X-X.TECH UK lab programme (original UKSCA, 242 scenes) | `legacy/local-source-snapshots/programme/uksca/UKSCA-ORIGINAL-20260909-plan.md` | `36fc22c50be1395cac11b5094cc4ca890d9e870f` | 2026-09-09 | vision | | UK Sovereign Capabilities Architecture white paper research brief | `legacy/local-source-snapshots/programme/proof/UKSCA-RESEARCH-BRIEF-20261020.md` | `3fb75989c91918808614147f4be19237dfa2d5de` | 2026-10-20 | research brief | | Financial crime early warning: how we win | `legacy/local-source-snapshots/research/financial-crime/WIN-PLAN-20261015.md` | `eb91866bdc50f304b5f0c4e273ce4b8c068d5238` | 2026-10-15 | strategy | | Part A results (synthetic AML detection benchmark) | `legacy/local-source-snapshots/research/financial-crime/part-a/RESULTS.md` | `27136c952de98a3c331586e6ce258298b288e0f0` | 2026-10-15 | research result | | The Charcoal Standard: carbon-removal reserve idea | `legacy/local-source-snapshots/research/charcoal-standard/CHARCOAL-STANDARD-OWNER-BRIEF-20261015.md` | `a729eb070d9fcd640152cc7a438d08cc77ac0c36` | 2026-10-15 | vision | | Bank station competitive research (373 workflows) | `legacy/x-judeai-edge-tok-kol-ndc-ml-x/stations/BANK-COMPETITIVE-RESEARCH.md` | `fc7dcd2edf8875ee828bcf7035fbccb6cf502c72` | 2026-10-18 (content) | competitor research | | Market estimate provenance (scenario, not measured) | `legacy/local-source-snapshots/programme/outcomes-register/evidence/2026-10-19/codex/t9-t10/packs/frontier-ai/market-estimate-provenance.json` | `d0d7a182452e6bebef8755fe3ee31acaadfbcbf8` | 2026-10-19 | market sizing | | Buyer research on a remittance company (not contacted) | `legacy/local-source-snapshots/programme/outcomes-register/evidence/2026-10-19/codex/t9-t10/packs/commercial/CASHPOT-BUYER-RESEARCH.json` | `509dab1d2363b679e5f497b45aa7dd2dc67cb846` | 2026-10-19 | buyer research | | Qualification of an inherited Challenge 4 benefit figure | `legacy/local-source-snapshots/programme/outcomes-register/evidence/2026-10-19/codex/t9-t10/packs/research/C4-BENEFIT-QUALIFICATION.md` | `572814002debb9942da6ec2a68bae31b1eef5403` | 2026-10-19 | evidence qualification | | JudePay research: Sage pain points | `legacy/jude/docs/judepay/research.md` | `f42aa1e7299707e867bf70d7b19ef7da1c874876` | 2026-09-25 (content) | customer research | | JudePay pricing and competitor price comparison | `legacy/jude/docs/judepay/pricing.md` | `7f5d60ffd41c73c14ece1bc5bbc5daaefae63813` | undated (September 2026 by context) | pricing research | | App Store rejection research | `legacy/jude/docs/appstore/RESEARCH.md` | `f230220a069885ff2a561f0fdeaba1c344dd0802` | 2026-09-25 (content) | platform research | | London public service and care research intake | `legacy/xx-search/docs/research/LONDON-INTAKE-20260907.md` | `08690fca8e1ef4aa45f99fa928843c8e97e490f9` | 2026-09-07 | prospect research | | X-X cyber governance consultancy: process, pricing, procurement page | `legacy/xx-search/public/pricing.html` | `6708d508bff606d98160b67ae7aabc826c35105a` | undated | predecessor business | | Audit of X-X Search (predecessor product) | `legacy/jude/docs/audit-x-x-search-2026-09-06.md` | `a3b17683df88f895770b054f67b8d489f79aa5b3` | 2026-09-06 | audit | | How I would recode the Judiverse and X-X | `legacy/jude/docs/RECODE-THE-JUDIVERSE-AND-X-X.docx` | `e20be825f814867956cca4b48806d7fba998e5fb` | 2026-09-06 (content) | strategy note | | Self Grown Chip: evidence boundary and first experiment | `legacy/jude-mvp/output/docx/Self-Grown-Chip-6-September-2026.docx` | `3a8dac2b6e6dbe2555f68a6b39accb0481f1f7e0` | 2026-09-06 (title) | applied research note | | Materials workbench specification (TRL 1 to 3) | `legacy/local-source-snapshots/workbenches/materials/README.md` | `a5935809716b237f6d4d4e0fde003e2bfaee3484` | 2026-10-15 (content) | research spec | | Master plan compilation (copies of every source, 101,993 lines) | `legacy/local-source-snapshots/MASTER-PLAN.md` | `2c67d760dded2b72addf37ded764133e4676f5d9` | 2026-10-15 (content) | compiled plan | | Canonical plan 2.9.20 (last old-estate plan) | `old-repo/plan/PLAN.md` | `bd6f7d17e6b8c767e27825e29be360bf3a56df58` | 2026-10-30 | plan | | Roadmap from the 31 October audit to a live estate | `old-repo/handover/audit-20261031/ROADMAP.md` | `84fe646190d12bef98b05ff7f8a00d6b3697ea17` | 2026-10-31 | roadmap | | UK Palantir procurement register (37 rows) | `old-repo/handover/uk-palantir-register-20261030/README.md` | `1472f416c5124a529206b4bd273698fb3581e7b9` | 2026-10-30 | procurement research | | Owner's product positioning ('race tracks') | `legacy/github-metadata-20261030/local-working-changes/greenfield-canonical/files/intentions/owner-product-positioning-20261029.txt` | `e278f03ebf58e1ed55ecad78a335de4baeb7e351` | 2026-10-29 (path) | intention | | Product statement derived from that intention | `legacy/github-metadata-20261030/local-working-changes/greenfield-canonical/files/PRODUCT.md` | `f869453d312f7a3203b86b3a0c1bcd98ef9e5ea1` | 2026-10-30 | positioning | **Other strategy and funding families, triaged at path and title level** (not listed row by row; each family is described once): - `legacy/judeai-edgeid-cloud/plans/horizon/md/` also holds parity files per family (bank, clinic, collaboration, design, government, identity, JudeAI, platform), a definitions register, acceptance ledger, dependency graph, one-pass build plan, Figma brief and research files on decision engine, hardware, secretless identity, law and standards, observability and knowledge-as-capability (all dated 24 to 25 October 2026 by content). They are planning and competitor-parity material for the readers of those categories. - `legacy/edge-judiverse-tech/docs/research/2026-10-13/` (ten research notes dated 13 October 2026: agentic coding benchmarks, coding platforms, agent identity and governance, reliability and safety, FCA insurance and AI, PRA and Bank of England, ISO standards, financial-intelligence benchmarks, "4R" prior art, intenJude state) and `legacy/edge-judiverse-tech/docs/research/risk-resilience/` (R1 to R5). - `legacy/edge-judiverse-tech/docs/dossiers/frontier-ai/01` to `07` (research files per Frontier AI theme: adaptive adversarial safety testing; resilient edge intelligence; medicines discovery and multi-omics; advanced materials; simulation and compute stack; whole-of-life journeys) with internal "trail" files recording sources checked. - `legacy/edge-judiverse-tech/docs/whitepaper/` (adaptive adversarial safety testing; simulation research) and `docs/research/04` and `05` (white-paper programme; "Reversible by default"). - `legacy/local-source-snapshots/programme/proof/procurement/` (packs for G-Cloud, NHS frameworks, DTAC, DSPT, DCB0129 and DCB0160, Cyber Essentials Plus, ISO/IEC 27001, FCA and JMLSG financial-crime controls, Sovereign AI milestones, and about 18 supporting documents: DPIAs, statement of applicability, service definitions, vulnerability register, clinical safety case drafts) dated 19 to 21 October 2026; and `programme/proof/packs/PK-T01` to `PK-T16` ("bid tests"). - `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/procurement-acceptance/` (a 439-requirement procurement mapping and assurance register, 25 October 2026). - `legacy/judeai-edge-control/docs/research/` (life control plane research pack, 10 October 2026; platform inventories for core banking, bioinformatics, electronic health records, Excel, GitHub, Kubernetes, natural-language languages, Notion, Teams; compounding verified intelligence contract and fixtures). - `legacy/jude-mvp/docs/plan/` (current-state notes for JudeAI, identity, mail, scale, training, users; a ChatGPT parity matrix; a master plan for JudeAI and Judiverse Mail, around 3 October 2026) and `legacy/jude/jmusic/ROADMAP.md` (14 branch versions, a music app roadmap). - `legacy/judiverse-next/docs/research/` and `docs/parity/sources/` (JudeLab research records, a weight and key-value cache transfer baseline, ChatGPT and Outlook feature inventories with complaint themes, 25 September 2026). - `old-repo/plan/PLAN.md` and 23 hash-suffixed versions (canonical plan 2.0 to 2.9.20, 29 to 30 October 2026), plus `old-repo/handover/RULES*.md` and `ESTATE-LEVEL-READINESS.md` versions. These are the old estate's last plan and rules; they are superseded by the new estate and are not summarised as authority here. - `old-repo/handover/benchmark-standard/scores/` and `journeys-prospective/enumeration/` (machine scores and excluded-journey lists, 30 October 2026). - JudeDeck "ten lines per slide" summaries of plans, maps, partners and outcomes (`legacy/local-source-snapshots/map/*-JudeDeck.md`, `legacy/x-judeai-edge-tok-kol-ndc-ml-x/MASTER-PLAN-JudeDeck.md` and similar, 13 to 25 October 2026). --- ##### 2. Market and customer research recorded All items are **CLAIM** unless marked. Where the old document itself cited a primary source and said it read it, the item is marked **OBSERVATION (as recorded)**: this reading did not re-open those sources. ###### 2.1 Who the old work said the customers were - Six customer segments were set out: UK banks, fintechs and payment institutions; regulated businesses and critical-service suppliers; universities and research organisations; NHS trusts, research hospitals and life-sciences organisations; government departments and public bodies; compute providers, engineering and industrial organisations. Initial focus: financial services and regulated-enterprise AI assurance (recorded in `legacy/local-source-snapshots/foundations/engine/examples/nvidia-inception-deck.html` on 2026-10-15, historical). - Twelve institutional buyer types were analysed one by one (how they buy, mandatory gates, pains and budgets, incumbents, an offer, objections, first three target profiles, an opening message): UK bank or building society; e-money or payments firm; NHS trust and ICB; NHS England or DHSC programme; university and research institute; central government department; local authority; MOD and defence primes; regulator as a user; large enterprise chief information security officer; SME; investor or grant funder (recorded in `legacy/edge-judiverse-tech/docs/research/03-INSTITUTIONAL-OFFERS.md` on 2026-10-13, historical). The document states that nothing in it had been sent to anyone. - For the Frontier AI bid, first customers by category were regulated financial institutions, health bodies and professional services firms, with regulators and insurers as secondary customers (recorded in `legacy/edge-judiverse-tech/docs/submission/frontier-ai/APPLICATION-2584-SUBMISSION-VERSION.md` on 2026-10-13, historical). - An earlier framing (September 2026) aimed at public-sector fraud control: an agent checker in "the processes where public money is decided: grants, business rates, blue badge, housing, procurement and invoices" (recorded in `legacy/jude/docs/value-model.md`, undated, September 2026 by context, historical). ###### 2.2 Problems recorded, with the evidence the old documents cite - **Agent safety failures are novel and static test suites go stale.** The application cites a 110-task malicious-agent benchmark where leading models complied without jailbreak, and a national cyber evaluation where 10 of 122 runs produced 19 out-of-scope actions detected three days later; it also cites AgentDojo (97 tasks, 629 test cases) and InjecAgent (1,054 cases, 24 per cent injection success) (recorded in `legacy/edge-judiverse-tech/docs/submission/frontier-ai/APPLICATION-2584-SUBMISSION-VERSION.md` on 2026-10-13, historical). **OBSERVATION (as recorded)** for the published-paper figures. - **Financial firms use AI they only partly understand.** 75 per cent of surveyed UK financial firms use AI and 46 per cent only partly understand it; 84 per cent have a named person accountable for AI (recorded in `legacy/edge-judiverse-tech/docs/research/03-INSTITUTIONAL-OFFERS.md` on 2026-10-13, historical; the source is cited there as [S11]). - **Public-sector digital spend and supplier monitoring.** Over £26 billion a year on digital, 55 per cent to contractors, managed services and consultants; 28 per cent of central systems legacy; 28 per cent of leaders think they can monitor supplier performance (recorded in the same file on 2026-10-13, historical; source [S7]). - **Cyber exposure by business size.** 69 per cent of large businesses reported a breach or attack; 31 per cent of businesses use or consider AI and 24 per cent of those manage its cyber risk; 42 per cent of micro and 46 per cent of small businesses reported a breach; the median perceived cost of the most disruptive breach for small firms was £0 and £4,000 at the 95th percentile, from which the document concludes "selling on fear would be dishonest" (recorded in the same file on 2026-10-13, historical; source [S25]). - **Local authorities** use or explore AI widely (95 per cent of responding councils, marked [unverified] in the source document) and digital and data roles are about 2 per cent of the workforce (recorded in the same file on 2026-10-13, historical). - **Fraud and error against public funds.** £55 billion to £81 billion in 2023-24 (National Audit Office); detected local-authority fraud £73.4 million reported by 100 councils and £265.1 million projected for 2024/25 (National Anti Fraud Network); 307 principal authorities in England as at May 2026 (recorded in `legacy/jude/docs/value-model.md`, undated, September 2026 by context, historical). **OBSERVATION (as recorded)**. - **Bladder cancer diagnostic pathway (EDGE Clinic relevance).** About 10,900 people diagnosed in the UK each year and about 5,900 deaths; almost every adult referred with blood in the urine has a hospital cystoscopy although only about 8 in 100 urgent referrals have bladder cancer; a 23-gene urine DNA test showed 92 per cent sensitivity and 92 per cent specificity with negative predictive value above 99 per cent in 964 patients across seven NHS urology departments (recorded in `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__applications__IUK-2527__APPLICATION.md` on 2026-10-25, historical; sources named there as Cancer Research UK and a University of Birmingham study). **OBSERVATION (as recorded)**. - **Small-business payroll and accounting pain (EDGE Finance relevance).** Ranked pain points about a major incumbent accounting product: subscription price rises of 8 to 40 per cent at renewal, support hard to reach, dated interface, and more, each with linked review sources (recorded in `legacy/jude/docs/judepay/research.md` on 2026-09-25, historical). The source document notes ranking rests on search excerpts. - **App store rejection risks for AI assistant apps** (Judiverse relevance): undisclosed third-party AI data sharing, template-cloned apps, bare web-view wrappers, missing content reporting and missing in-app account deletion, ranked by likelihood with Apple guideline numbers (recorded in `legacy/jude/docs/appstore/RESEARCH.md` on 2026-09-25, historical). - **Water and humanitarian missions** (EDGE Public Consortium relevance, logistics): Sudan 33.7 million in need and a WASH (water, sanitation and hygiene) requirement of US$375.9 million for 2026; Tehran reservoir and groundwater depletion figures (recorded in `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__MISSIONS.md` on 2026-10-24, historical; several figures there are marked secondary or read through search). - **Defence buying example**: an MOD water supply contract for a Kenya training unit, 48,000 bottles of 18.9 litres a year and 300 dispensers, market engagement value £785,947.24 excluding VAT (recorded in the same file on 2026-10-24, historical; cited to Find a Tender notices). ###### 2.3 Market sizing recorded - **Frontier AI SAM (serviceable addressable market).** Denominator 1,417,730 UK private-sector employers at the start of 2025, of which 46,770 medium or large; 30 per cent of employers using AI; an upper bound of 14,031 organisations; the agent-deploying share and the price per organisation were left "to be confirmed by the directors" (recorded in `legacy/edge-judiverse-tech/docs/submission/frontier-ai/APPLICATION-2584-SUBMISSION-VERSION.md` on 2026-10-13, historical). **PREDICTION**. - A later correction narrowed this: 46,770 × 0.30 × 0.07 = about 982 indicative organisations, using 7 per cent agentic use among AI adopters from a government survey with 2025 fieldwork, and stated the limitations (different populations and periods; no agreed price; no customer commitment) (recorded in `legacy/local-source-snapshots/programme/outcomes-register/evidence/2026-10-19/codex/t9-t10/packs/frontier-ai/market-estimate-provenance.json` on 2026-10-19, historical). **PREDICTION**, explicitly "scenario estimate, not measured segment". - The internal assessor critique said the primary bid "would not be sent for assessment as written" because it failed the serviceable addressable market line, and flagged an unsourced £556 million figure for removal (recorded in `legacy/edge-judiverse-tech/docs/dossiers/frontier-ai/09-ASSESSOR-CRITIQUE.md` on 2026-10-13, historical). - **Target account scale** for the Civo partnership deck: a digital bank at £887m revenue and 6.2m group platform accounts; a high-street bank at £29.14bn income and 93,000 FTE; a London NHS trust at 23,700 staff; a London borough at £432.4m net service budget; the deck warns the bases are not comparable (recorded in `legacy/jude-mvp/work/civo-lsbu-deck-20261006/content.json` on 2026-10-06, historical). - **Mineral water competitors** (revenue, secondary source): three UK brands at £158.9m, £135.4m and £39.3m (recorded in `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__MISSIONS.md` on 2026-10-24, historical). - **Data-centre heat reuse**: a cloud provider's first UK edge data centre announced at 8 MW, able to grow to 38 MW, due by March 2027; on base assumptions a heat network would be about £2.2m a year short (recorded in `legacy/judeai-edgeid-cloud/plans/opportunities/heat-reuse-opportunity-assessment-20261025.html` on 2026-10-25, historical). **PREDICTION**; the page says nothing was agreed. - **Prospect databases of the predecessor business**: 2,268 organisations and 1,884 candidate records in the X-X Search staging database (recorded in `legacy/xx-search/docs/research/LONDON-INTAKE-20260907.md` on 2026-09-07, historical); ten prospect CSV files under `legacy/xx-search/ops/vibe/` (1,161 lines in total, undated) listing London financial firms from a third-party data provider. These hold personal and company data and should not be imported; they show the earlier executive-search go-to-market only. ###### 2.4 Pricing and competitor price research - **Pricing architecture**: one public price list; the unit of value is the receipt (one approved agent action with its evidence); four meters (platform fee per organisation per year, per user per month, per agent action in capped bundles, per outcome at a fixed price); "safety is never metered"; regulators never pay to read; no fee tied to a regulated decision; prices themselves were not set (recorded in `legacy/edge-judiverse-tech/docs/research/03-INSTITUTIONAL-OFFERS.md` on 2026-10-13, historical). The same model appears in the NVIDIA Inception deck (2026-10-15). - **Comparator prices** listed: an enterprise agent platform at £1.60 per conversation and £400 per 100,000 credits (about £0.08 per action by the reviewer's arithmetic); a customer-service AI agent at $0.99 per outcome; an office copilot at about £16 to £25 per user per month (marked unverified) (recorded in the same file on 2026-10-13, historical). - **Pilot pricing proposed to a cloud partner**: discovery £6,000 for 2 weeks; standard £30,000 for 6 weeks then from £3,000 a month; regulated £45,000 then from £8,000 a month; enterprise £75,000 for 8 weeks then from £15,000 a month; modelled at a blended £750 per person-day; GPU planning at $478.40 to $1,913.60 per pilot (recorded in `legacy/jude-mvp/work/civo-lsbu-deck-20261006/content.json` on 2026-10-06, historical). **PREDICTION**; "not an existing price list or binding quote". - **Payroll product pricing** (JudePay): micro £8, small £20, charity £12 and large £45 a month, with a 60-day £0 pilot, against listed competitor prices for four payroll and accounting products (recorded in `legacy/jude/docs/judepay/pricing.md`, undated, September 2026 by context, historical). - **Clinic pathway prices** for the cancer bid: £45 per pathway, £20,000 a year per trust licence, £250 combined with a partner's test, all marked as planning prices for the Owner to confirm (recorded in `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__applications__IUK-2527__GAPS.md` on 2026-10-25, historical). - **Commercial model at the end of October**: a free self-hosted evaluation with no card and no automatic conversion; operational monthly and annual quotes with prices deliberately unset (`null` means unknown); a proposed annual option of ten monthly amounts for twelve months (recorded in `old-repo/docs/COMMERCIAL-MODEL.md` on 2026-10-29, historical). ###### 2.5 Research results recorded (synthetic or internal; never customer evidence) - **Adaptive adversarial search simulation**: with 12 planted faults each needing three perturbations, a coverage-guided archive found a mean 8.80 of 12 distinct failure signatures in 2,000 episodes over 30 seeds (95 per cent interval 8.27 to 9.33), random testing 3.10, a fixed 200-case suite 0.20; the advantage vanished for one-step failures. The simulated agent was the applicant's own automaton (recorded in `legacy/edge-judiverse-tech/docs/submission/frontier-ai/APPLICATION-2584-SUBMISSION-VERSION.md` on 2026-10-13, historical; reproduced with an identical digest per `legacy/edge-judiverse-tech/handoffs/submission-closure-20261014/README.md` on 2026-10-14, historical). - **Synthetic AML (anti-money laundering) detection benchmark**, five seeds: on standard cases the missed-detection rate was 0.610 for rules, 0.293 for a tabular model and 0.194 for a graph model, with precision 0.203, 0.882 and 0.896; on hard cases detection within 7 days fell to 0.342, 0.215 and 0.399 (recorded in `legacy/local-source-snapshots/research/financial-crime/part-a/RESULTS.md` on 2026-10-15, historical). The file is headed "synthetic data, not a claim about any real institution". - **A retrospective timing replay** comparing 299.0 hours with 0.98 hours was later qualified as "a counterfactual timing estimate ... not an observed prospective live intervention" (recorded in `legacy/local-source-snapshots/programme/outcomes-register/evidence/2026-10-19/codex/t9-t10/packs/research/C4-BENEFIT-QUALIFICATION.md` on 2026-10-19, historical). This is a useful example of the old estate downgrading its own claim. - **Compiled transformer primitives**: 19 of 19 small programs matched their Python references on 53,403 generated inputs (recorded in `legacy/jude-os/jude-weight-compiler/compiled-primitives/RESULTS.md`, branch-dated, historical). - **Claims audit**: a register extracted 753 lines carrying a figure, an absolute or a proof word from seven bid and partner documents, with 0 proven at the time (recorded in `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/proof/CLAIMS-REGISTER-20261020.json` on 2026-10-20, historical; tagged outside this reader's scope and cited for context). `legacy/local-source-snapshots/programme/proof/PROOF-PROGRAMME-20261020.md` (2026-10-20) states "648 lines ... all unproven". --- ##### 3. Funding and procurement history ###### 3.1 The organisation and its stance - The applicant throughout is X-X.TECH LTD, a London private limited company incorporated on 28 October 2025, with first accounts due 28 July 2027 and no filed accounts at the time (recorded in `legacy/edge-judiverse-tech/docs/submission/frontier-ai/APPLICATION-2584-SUBMISSION-VERSION.md` on 2026-10-13 and `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__applications__UK-FUNDING-CALENDAR.md` on 2026-10-25, historical). Registration identifiers (company number, D-U-N-S, Central Digital Platform and ICO numbers) are recorded in those files and are not repeated here. - **Predecessor business.** Before the agent platform, the organisation ran "X-X", a cyber governance, resilience, risk and compliance consultancy with an executive search line for London buy-side firms (X-X Search). An audit on 6 September 2026 found the brochure site worked but the product journey could not be completed, and a note the same day proposed folding X-X Search into the Judiverse as one product (recorded in `legacy/xx-search/public/pricing.html`, undated; `legacy/jude/docs/audit-x-x-search-2026-09-06.md` on 2026-09-06; `legacy/jude/docs/RECODE-THE-JUDIVERSE-AND-X-X.docx` on 2026-09-06, historical). - **Reserved acts.** From late October, external submissions, money and naming of partners were reserved to the Owner; partners were not to be named until they agreed in writing, and an "outreach pause" was in force (recorded in `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__applications__OWNER-DECISION-20261025.md` on 2026-10-25, historical). ###### 3.2 Competitions and tenders, in date order | When (historical) | Funder and call | What the old estate prepared | Amounts recorded | Recorded outcome | |---|---|---|---|---| | September 2026 | Sovereign AI R&D Procurement Scheme, all four challenges | Four bids designed "at £10m each"; a security bid as the "authoritative baseline"; research programme and work packages over ten quarters | Security £9,608,204 over 30 months; NHS £1,699,045; Defence £2,188,512; Compute £2,705,940; combined £16,201,701 excluding VAT, "not confirmed eligible" (recorded in `legacy/jude-mvp/docs/four-challenge-costs.md` on 2026-09-09, historical) | `legacy/jude-mvp/docs/plan-2030.md` (branch-dated 2026-09-13) states "The Expression of Interest is in" and a full submission "by 1 October 2026". **UNKNOWN** whether a September submission was made; no receipt found | | 13 to 16 October 2026 | Sovereign AI Scheme, Batch 1 (submit by 16 October; outcome 13 November) | Full applications for Challenge 1 (NHS productivity, DHSC), Challenge 2 (Defence edge runtime, MOD), Challenge 3 (compute efficiency, BIST and ARIA Scaling Inference Lab), Challenge 4 (safe agent adoption, NCSC), each with use cases, a contract schedule and an evidence index; eleven evidence packs (271 items, 123 gaps) | C1 £1.5m to £2.5m over 18 months; C2 ceiling £1.5m; C3 £1.65m to £2.4m over 15 months; C4 £1.25m to £1.8m over 12 months (recorded in the four `C*-01-FULL-APPLICATION.md` files on 2026-10-13, historical) | The plan of 14 October says "No acknowledgement for any of the four has been inspected here" and "the all-four-today target remains a target until four receipts exist" (recorded in `legacy/edge-judiverse-tech/docs/submission/ONE-PLAN.8d08b0e8.md` on 2026-10-14, historical). A Challenge 4 file records "Expression of Interest submitted 13 October 2026 [unverified: receipt not seen]". Batch 1 is later described as "closed before this plan" (MISSIONS, 2026-10-24). **UNKNOWN**: no submission receipt or outcome found | | 13 to 14 October 2026 | Innovate UK, Frontier AI: SME Champions Phase 1 (competition 2584; closes 11 November 2026; up to £3.1m; £150k to £250k per project; around 2 per cent chance of success per the briefing) | Theme 4 application: an adaptive adversarial discovery engine for unknown safety failures in tool-using agents, with the existing control plane as background IP | £250,000 eligible costs; £175,000 grant at 70 per cent; £75,000 company match; five work packages at £30k, £80k, £60k, £40k, £40k (recorded in `legacy/edge-judiverse-tech/docs/submission/frontier-ai/APPLICATION-2584-SUBMISSION-VERSION.md` on 2026-10-13, historical) | Internal simulated panels scored it 42, 43 and 41 of 80 on eight questions and "would not recommend funding" (14 October); the closure review on 14 October says "revised draft, not submission ready". **UNKNOWN** whether it was submitted | | 13 October 2026 | Innovate UK BridgeAI Deployment Catalyst (competition 2599, and sister strands 2597 and 2598; close 8 December 2026) | Rules researched only; needs a deployer partner and TRL 5 to 6 | Phase 1 up to £5m; grant £100k to £200k; programme up to £52m over three phases (recorded in `legacy/edge-judiverse-tech/docs/dossiers/frontier-ai/00-COMPETITIONS.md` on 2026-10-13, historical) | No application drafted in scope | | 6 October 2026 | UKRI Transformative sustainable AI technologies (closes 10 November 2026) | Proposed with LSBU as lead and a cloud provider contributing compute; a draft "Vision and Approach" written 19 October with "institutional lead, application identifier and commitments not yet recorded" | Up to £602,500 full economic cost, UKRI funding 80 per cent up to £482,000 (recorded in `legacy/jude-mvp/work/civo-lsbu-deck-20261006/content.json` on 2026-10-06, historical) | **UNKNOWN** | | 20 October 2026 | Sovereign AI Scheme, Batch 2 (EOI target 17 December; proposal by 31 December 2026; outcome 29 January 2027) | A combined bid "Clinical and Banking pathways 2026 to 2029", Challenges 1 and 4, with partner documents for LSBU, a biostatistics unit and a gender-care clinical partner; procurement and partner packs; an email to the Scheme on 20 October with EOI answers (focus areas 3 and 5) | £5,225,000 requested for one year 2027 to 2028; £21.35m over three years of which £16.125m commercial; a "full national" option of £42.7m (recorded in `legacy/x-judeai-edge-tok-kol-ndc-ml-x/programme/proof/CLAIMS-REGISTER-20261020.json` on 2026-10-20, historical) | Recorded as "Commitment sent to Sovereign AI" with status PARTIAL; the Approved Supplier List outcome is not recorded. **UNKNOWN** outcome | | 25 October 2026 | UKDI Security Open Call 2026 cycle 2, stage 1 (closes 29 October 2026 12:00) | Application text for "secretless identity" against the Home Office fraud and cyber-threat challenge area | £244,062.48 main project, £37,755.22 testing option, £281,817.70 total within a £350,000 cap; 100 per cent funded contract (recorded in `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__applications__UKDI-SECURITY-C2__FINANCES.md` on 2026-10-25, historical) | **UNKNOWN**; the pack says submission is the Owner's act | | 25 October 2026 | Innovate UK competition 2527, Advancing innovation in accelerated cancer diagnosis (IFS reads 4 November 2026 11:00) | "Direct-to-test haematuria triage": a 36-month, about 2,000-patient, three-trust validation of a multi-modal urine DNA risk pathway from primary care, with three unnamed partner slots | £3,449,800 eligible costs; grant £2,762,560 (if the diagnostics partner is small); lead grant £1,191,960; lead contribution £510,840 (recorded in `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__applications__IUK-2527__FINANCES.md` on 2026-10-25, historical) | Not eligible until three partners agreed; **UNKNOWN** whether submitted | | 25 October 2026 | The UK funding sweep: 40 calls (28 dated, 12 rolling), 16 watched routes, 14 exclusions | Calendar with fit, evidence held and gaps for each call, including ARIA Scaling Trust and Safeguarded AI, MHRA AI Airlock, Future Leaders Fellowships, Efficient AI (2582), NHS Cancer Programme Open Call 4, UK-Germany, UK-Singapore and Eureka bilateral calls, UK Space Agency, Knowledge Transfer Partnerships, R&D tax relief, the AI DPS (RM6200), FCA sandbox and pathways, Sovereign AI equity and compute, EIS and SEIS advance assurance, Start Up Loans, an LSBU-led local business support scheme, NHS DTAC | per call, as listed (recorded in `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__applications__UK-FUNDING-CALENDAR.md` on 2026-10-25, historical) | Nothing submitted by agents ("read-only: nothing submitted, posted or sent") | | 19 October 2026 | Earthshot Prize 2027 | Route check only | Nominator endorsement by 3 December 2026; nominee completion by 10 December 2026 (recorded in `legacy/local-source-snapshots/programme/outcomes-register/evidence/2026-10-19/codex/t9-t10/packs/research/earthshot-source-update.json` on 2026-10-19, historical) | "No nomination or endorsement inferred" | | 15 October 2026 | Investors: pre-seed round | NVIDIA Inception deck with "The ask: Pre-Seed £3m", target close 31 March 2027; use of funds: team, compute, assurance (Cyber Essentials Plus, ISO 27001), paid pilots (recorded in `legacy/local-source-snapshots/foundations/engine/examples/nvidia-inception-deck.html` on 2026-10-15, historical) | £3m | **UNKNOWN**; no investor commitment found. The deck records GPU access through a UK cloud provider as "in place" and a digital bank's interest as "interest, not a contract or a pilot" | ###### 3.3 Buyers, partners and relationships recorded All are **CLAIM** and were described in the old documents themselves as proposed, not agreed, unless stated. - **Cloud and compute**: a UK cloud provider supplied GPU access and supported the Inception application ("in place" per the 2026-10-15 deck); NVIDIA Inception was the programme applied to or prepared for; the 25 October missions file says the estate "qualifies on the published criteria". A storage vendor was named as a "technology partner target" only (Civo deck, 2026-10-06). - **Universities**: LSBU SITHub was the consistent research partner candidate (collaboration brief around 3 October; consortium plan 10 October; partner pack 20 October); other London universities were listed as research routes in the Civo deck appendix. The Imperial pack of 20 October says Imperial is "out until they reach out". A biostatistics unit was the proposed methods partner in the October bid. - **Banks and payments**: a digital bank was the priority bank account in several decks (interest only); other high-street banks were outreach targets. A later programme note names further banks as "forthcoming relationships" (recorded in `legacy/judeai-edgeid-cloud/plans/opportunities/one-mans-trash-programme-20261025.html` on 2026-10-25, historical). A remittance company was researched from public pages and explicitly "not contacted" (2026-10-19). - **Health**: a London NHS trust (non-clinical administration pilot proposed), an NHS gender-care service and a urine DNA test developer appear as partner slots or proposed study parties; the study terms of 20 October require HRA approval and ethics before any sample. - **Local government and care**: a London borough (casework preparation) and a learning-disability charity group ("approached in social care", per `legacy/jude-mvp/docs/acceptance-targets.md`, undated, September 2026 by context). - **Invitations**: on 29 October drafts addressed to a government development bank, the cloud provider, LSBU, NVIDIA, the Sovereign AI Fund, Innovate UK and UKRI were "ready to send", to be sent by a person (recorded in `old-repo/evidence/INVITATIONS-20261029.md` on 2026-10-29 and `legacy/github-metadata-20261030/local-working-changes/greenfield-canonical/files/repairs/deploy-now-20261029T0820Z/invitations/`, historical). **UNKNOWN** whether any was sent. - **Procurement routes researched**: G-Cloud 15 (live 6 August 2026 to 5 February 2028 as recorded), the AI DPS RM6200 (open to 23 February 2029), Contracts Finder and Find a Tender, the MOD Defence Sourcing Portal, NHS frameworks and the Provider Selection Regime, Cyber Essentials and Plus under PPN 09/23, DTAC version 2.0, DSPT 2026-27 version 9 (recorded in `legacy/judeai-edgeid-cloud/plans/horizon/md/horizon-plan-20261024-uriel1717__plan__MISSIONS.md` on 2026-10-24 and `legacy/local-source-snapshots/programme/proof/procurement/INDEX.md` on 2026-10-20, historical). The bid placed G-Cloud and NHS framework listing in 2028. - **Procurement intelligence**: a register of every UK public contract published as awarded to one large data-analytics supplier, from official notices only: 37 rows, 30 contracts, 6 framework or market admissions, 1 award at £0; "not a bid, a won contract or contact with a buyer" (recorded in `old-repo/handover/uk-palantir-register-20261030/README.md` on 2026-10-30, historical). A machine-built register of 312 open opportunities (172 contracts, 130 grants, 10 competitions) from six portals was built on 18 October (recorded in `legacy/x-judeai-edge-tok-kol-ndc-ml-x/stations/registers/funding/funding-20261018.json` on 2026-10-18, historical). ###### 3.4 Outcomes: what the corpus does and does not show - **Won**: no grant, contract award, investment or prize is recorded anywhere in scope. **UNKNOWN** (looked for: "awarded", "grant offer", "successful", "unsuccessful", "submitted", application identifiers, acknowledgements). - **Submitted**: only two submissions are recorded, both unverified by a receipt in the corpus: a Sovereign AI EOI said to be "in" (September) or "submitted 13 October 2026 [unverified]", and the 20 October email to the Sovereign AI Fund ("Commitment sent", PARTIAL). Every application pack is marked draft, "for Owner approval", "not ready to submit" or "prepared, not submitted". - **Lessons the old documents drew themselves**: TRL mismatch (Frontier AI 2584 requires TRL 1 to 3 while the Sovereign Scheme focuses on TRL 4 to 8 and BridgeAI on TRL 5 to 6, so one body of work cannot fit all); one application per SME to 2584; no subcontractors in 2584; named CVs are mandatory and missing ones cap a question score; market figures must be sourced and narrowed; partner letters must be signed, not expressions of interest; patents must be filed before publishing a method; costs must be built from payroll and quotes, not written back from a cap; double funding and overlapping staffing across simultaneous bids must be ruled out (from `00-COMPETITIONS.md`, `09-ASSESSOR-CRITIQUE.md`, `14-assessor-panel-and-proofs.md`, `ONE-PLAN.8d08b0e8.md`, `legacy/jude-mvp/manifests/challenge-4.requirements.json`, all October 2026 or undated, historical). --- ##### 4. Valuable material in "other" that the other readers would miss The `other` tag covered 3,080 distinct blob oids. Classified by content, they are: about 1,380 documents (handovers, specifications, product and app documentation, research notes, prompts, principles, decisions, incident records), 673 distinct HTML pages (product screens, email previews, brand pages, decks), 148 evidence logs (test outputs, receipts, browser observations), 21 data tables, and the plainly irrelevant groups listed in 4.5. Because these carry no subject tag, the readers of the subject categories (competitors, benchmarks, readiness, blueprints, features, standards, papers, guides) will not see them. The following are worth keeping as research. Dates are as recorded; each item is historical. ###### 4.1 Prompt libraries, agent skills and working principles | What | Path | Blob oid | Date | Why it matters | |---|---|---|---|---| | Public assistant behaviour prompt ("safe to paraphrase to any user": British English, actions on their own line, inspect before changing, test the real journey, explain cost before committing) | `legacy/jude-mvp/judeai-core/judeai_core/prompts/persona.txt` | `a6046491ec37373f0632bdffd3a605d5676317c6` | branch 2026-10-03 | A tested house style for Judiverse assistants; separates public behaviour from internal configuration | | Original behaviour prompt and the internal operating prompt (internal one holds a spending allowance and must never be shown to users) | `legacy/jude-mvp/judeai-core/judeai_core/prompts/behaviour.original.txt`; `.../internal-operating.txt` | `361d9eedc9499d2ed3f863371160e1ee89fa4e37`; `553d42a17ec1dc432fa63c4602158a0cf87f5638` | branch-dated, September 2026 | Shows the split between public persona and private operating rules, a pattern worth keeping for agents as staff | | Agent skill files (verify before saying done; connectors; deliverables; engineering; video) | `legacy/jude-mvp/jcap/skills/*/SKILL.md` (e.g. `jude-verify`, oid `6ace0ae37a9eff8fc87c91bb0ba2fd0cab380e7e`) | as listed | branch-dated | Short, reusable agent-lab skills; "a green test is not proof if the test supplies what the product is missing" | | Agent skill files for self-engineering, coordination, deploy and release, debugging, security, evidence | `legacy/judeai-edge-control/skills/` (README oid `e20bf7adb106fe2a59ecd299bf7e609782f0d6f2`) | as listed | branch-dated, October 2026 | Agent lab and "agents as staff" working method | | 25 design and governance principles (authority before action, inference is not proof, evidence for everything, decision radius, progressive disclosure, visual never blind, risk aware not averse, provenance, migrate by reconstruction, sustainable agentic AI and others) | `legacy/judeai-edge-control/Master/Principles/` (e.g. `inference-is-not-proof.md`, oid `f3d6ce343e7f0d6602afc25a15da37ad4c74a04b`) | as listed | branch 2026-10-13 | Compact, each with "Rule" and "Enforced by"; directly reusable as research for EDGE ID and Judiverse design | | Writer system and named-parties list for generated documents | `legacy/edge-judiverse-tech/tools/jude/writer-system.md` | `479a71d3d248ea40a811aa2511c8913be40145f3` | branch-dated | A drafting prompt; the named-parties list may hold personal names and should not be imported | ###### 4.2 Domain write-ups and research notes - **Failure catalogue for AI and agent systems** across financial services, health and medicines, defence and edge, materials and manufacturing, and software operations; every class has twelve fields (definition, mechanism, harm, severity, frequency, detectability, reversibility, example, signal, test, control, cost of control), sourced only from regulators, inquiries, safety bodies, standards and peer-reviewed studies (`legacy/edge-judiverse-tech/docs/dossiers/frontier-ai/10-failure-catalogue.md`, oid `25ea14751f098b3674e05a42f5651992afd28c32`, 2026-10-13). Useful to EDGE Public Consortium (cyber governance) and EDGE Finance (agent assurance). - **Assessor critique and simulated panel** of the Frontier AI bid (`legacy/edge-judiverse-tech/docs/dossiers/frontier-ai/09-ASSESSOR-CRITIQUE.md`, 2026-10-13; `.../14-assessor-panel-and-proofs.md`, 2026-10-14). A reusable method for testing any future grant text before submission. - **Value model for public-sector fraud control**, with sourced national figures and arithmetic laid out openly (`legacy/jude/docs/value-model.md`, oid `5d98eb8f43fdabf8d07f440b11577524b0fbb63e`, undated). - **Financial-crime control assurance brief**, written against the failures described in a published FCA final notice (sanctions screening coverage, account-opening restrictions) (`legacy/jude-mvp/judeos-live/docs/briefs/financial-crime-control-assurance.txt`, undated). EDGE Finance. - **Frontier "gauntlet" grader families**: ten blinded grader families scoring each agent trajectory against hidden ground truth, world diff and policy (`legacy/jude-mvp/judeos-live/training/frontier-gauntlet/graders/GRADERS.md`, oid `2ae1168e21d0f1c1593fa3f068f5b0b71b39d0a8`, undated). Agent lab evaluation design. - **Training experiment design under a fixed budget**: a first GPU experiment capped at £500 with calibration, matched pilot, replication and reserve (£50, £150, £200, £100), comparing baseline, ordinary upcycling and failure-cluster specialists (recorded in `legacy/jude-mvp/work/jude-continuation-kit-20261001/releases/jude-continuation-20261001T183322Z/EXPERIMENT.md`, oid `026d413152e4d03b56e33c1afd7126a8dc8a9498`, on 2026-10-01, historical). Agent lab. - **Adversarial agent concept "Erling"**: a restricted penetration-testing antagonist that runs inside the platform under stricter authority, never widens its own scope and never edits its own audit trail (`legacy/erling/README.md`, oid `864f9d13841dc0a4fd5bfab4b228718b52285bac`, undated). Agent lab and EDGE Public Consortium red-teaming. - **Compiled transformer primitives** (exact equivalence of tiny compiled programs to Python references) (`legacy/jude-os/jude-weight-compiler/compiled-primitives/RESULTS.md`, oid `19b9b22bdddbdcd957c386cf6bc2143766a1be7d`). A research curiosity for the agent lab. - **Red, purple and UX team checklists** for the assistant and mail apps (`legacy/jude/docs/swarm/`, 25 September 2026): attack replay tests and security-event logging design. Judiverse. - **Executive search ADRs** (architecture decision records) from the predecessor business, especially consent-scoped disclosure ("a buyer sees evidence before a name"; consent scoped to firm, role, recipient and artefact, hashed and receipted, withdrawable) and an hourly resumable "intelligence spine" over public registers (`legacy/xx-search/docs/adr/0001` to `0006`, accepted, undated; 0004 oid `d37337bf30b9e2663e77ad6838a7f6818d95f759`). Directly relevant to EDGE Public Consortium vetting. - **Integrated Capacity Model and the Capacity Recovery Pilot**: a model and a paid pilot offer for "help agreed, help received, capacity retained" in care services, with a frozen protocol and pre-registered analysis plan in form (`legacy/local-source-snapshots/programme/integrated-capacity-model/`, including `sources/Integrated-Capacity-Model-Handover.md`, oid `6ad4dfd2920b78a3f1e1f223fbcaed3449064792`, around 21 October 2026). EDGE Clinic (social prescribing) and EDGE Public Consortium. - **Materials workbench and waste-free substitution**, and a computational lithography note (`legacy/local-source-snapshots/workbenches/materials/`, 15 October 2026), and the **Self Grown Chip** experiment decision note (£15,000 to £30,000 for about six months, file a patent before disclosure; recorded in `legacy/jude-mvp/output/docx/Self-Grown-Chip-6-September-2026.docx` on 2026-09-06, historical). Research and education strand only. - **xmathx**: a 90-day AI, machine learning and cyber learning curriculum where every symbol is read as an English word, with spaced recall (`legacy/xmathx/README.md`, oid `d8384bc09becd15ecf67ce4da3dff1886adce768`, undated). EDGE Public Consortium education and research. - **Sovereignty pass**: every outside host the code could reach and the decision for each (one cloud dependency allowed; outside model labelled "outside the estate" and never a silent fallback) (`legacy/judeai-edge-control/docs/SOVEREIGNTY.md`, oid `fb54ac47ed2150b87c721e50c4fe42305b3622c7`, 2026-10-11). A template for supplier-dependency registers. - **Payroll and HMRC recognition notes, bank list, ledger and Stripe notes** for a small-business payroll product (`legacy/jude/docs/judepay/`, September 2026). EDGE Finance. - **Institution adoption record**: official public-register captures (Companies House, FCA, CQC, NHS directory) used to prepare dormant institution workspaces without claiming adoption (`old-repo/docs/INSTITUTION-ADOPTION-20261029.md`, oid `f54497c9e7a000ad1b6bb1d2beae766084663055`, 2026-10-29). - **Partner brief with five proposed pilots and specific asks** (for example GPU and edge kits from a hardware vendor) (`legacy/judeai-edge-control/docs/partner-brief.md`, oid `4b30c5b65716f2187fee5c6aad63134c08ccfb7c`, 2026-10-12). Marked "No agreement exists with any party below". ###### 4.3 Design notes and product documentation tagged only "other" - **Application family specifications** with status "specified" for health (P41 to P50), identity and security (P04 to P06, P57, P66 to P68, P78), workspace and communications (P21 to P24, P65, P69, P75, P76, P79, P80), engineering and compute (P09, P10, P13, P25, P54, P61, P70 to P72), operations and creative (P55, P56, P58, P60, P64, P77, P81) under `legacy/edge-judiverse-tech/docs/apps/`, and the bank apps P26 to P40, P59 and P74 under `legacy/edge-judiverse-tech/docs/finance/bank/apps/` (13 October 2026; health overview oid `623cf5fafed586c496b4e023af54847e40bb474e`). These are the old estate's feature intentions for what became EDGE Finance, EDGE Clinic, EDGE Public Consortium and Judiverse, and the features reader may not see them because they are tagged `other`. - **77 "Get started" guides** (`legacy/edge-judiverse-tech/docs/learn/`, 13 October 2026), help-centre copy per app. - **54 transactional email texts** (registration, verification, notifications) with corporate and estate variants (`legacy/edge-judiverse-tech/packages/communications/preview/`, branch-dated; README says "Nothing here sends"). Reusable wording; the links inside point at old hosts and must not be reused. - **Risk management framework, risk appetite statement, operational resilience framework and test suite** (`legacy/edge-judiverse-tech/docs/risk/`, `docs/resilience/`, 13 October 2026), and **privacy notice draft, accessibility statement, data processing terms, support and service terms, product listing** (`legacy/local-source-snapshots/programme/product/`, around 20 October 2026). Drafts "for Owner approval"; useful templates for EDGE ID governance. - **App Store submission notes and metadata** (`legacy/judeai-apple/AppStore/`, `legacy/jude/docs/appstore/notes/`, September to October 2026) and **store blockers** (`legacy/local-source-snapshots/programme/product/store/STORE-BLOCKERS-20261020.md`). - **Blog drafts** on verifiability, resilience and "risk aware, never risk averse" (`legacy/edge-judiverse-tech/docs/blog/`, 13 October 2026) and outreach white papers on a "hold button" for automated email and CRM (`legacy/edge-judiverse-tech/docs/outreach/`). - **Brand documents** (`legacy/jude/docs/brand/`, `legacy/jude-mvp/docs/brand/`, `legacy/judiverse-next/docs/brand/`) and the **product positioning intention** of 29 October ("models are the engines, harnesses are the cars, the ontology, specifications and knowledge graphs are the racetracks") (`legacy/github-metadata-20261030/local-working-changes/greenfield-canonical/files/PRODUCT.md`, 2026-10-30). - **Incident records** (`legacy/local-source-snapshots/programme/jps/incidents/`, `old-repo/docs/incidents/`, `old-repo/handover/storage-rc16-20261030/`), useful as real failure examples for the agent lab's failure register (each must be checked for host names and credentials before any reuse). - **Handovers** (dozens, for example `legacy/jude/HANDOVER-FOR-ANY-AI.md` in ten versions, `legacy/jude-mvp/handover/01` to `19`, `old-repo/handover/HANDOVER-COORDINATOR-20261030/`). These are old-estate instructions; under the new estate's rules they are research only and must never be followed. Their value is as a record of what was tried and what broke. ###### 4.4 Product screens and decks in HTML (triaged only) The 673 distinct HTML pages include about 199 product screen mock-ups under `legacy/judeai-edge-control/web/estate/` (bank, clinic, consortium, office, science, operations screens, 21 October 2026), 28 under `legacy/local-source-snapshots/services/edge/web/`, the hosted assistant's pages under `legacy/jude-mvp/judeai-hosted/web/`, brand pages and a funding page (`legacy/judeai-edge-control/web/estate/consortium-funding.html`, tagged funding). They were not read page by page; they are the UI and screen reader's material and are flagged here because their `other` tag hides them from that reader. ###### 4.5 Skipped as plainly irrelevant (classified by content first) | Group | Count (distinct) | Where | Why skipped | |---|---:|---|---| | Vendored third-party files | 29 | `legacy/jude-mvp/work/jude-mail-engine/.toolchain/` (a Rust crate registry copy, for example `find-msvc-tools` changelog and README) | Third-party package files | | Licence and notice boilerplate | 1 | `legacy/edge-judiverse-tech/NOTICE.md` | Boilerplate | | Synthetic training fixtures | 485 | `legacy/jude-mvp/work/jude-continuation-kit-20261001/releases/jude-continuation-20261001T183322Z/corpus/` (generated exam papers, ledgers, audit reports, fiction excerpts, each marked "SYNTHETIC FIXTURE") | Generated training data; the kit's design documents were read instead | | Test fixtures and test outputs | about 10, plus the test-output text files in repair bundles | for example `legacy/jude-mvp/tests/hmrc_paye_2026_27.csv`, `legacy/judeai-edge-control/tests/fixtures/`, `legacy/local-source-snapshots/services/edge/assurance/onlinefdr/*-seed20261019.csv` | Fixtures | | Build intermediates | about 20 | `legacy/jude-mvp/apps/JVPN/build-*/`, `legacy/jude-mvp/apps/JudeGRCsus/Build*/` (Xcode link lists and target graphs) | Machine build output | | Browser cache and tool help dumps | 2 | `legacy/jude-mvp/audit/.qa-profile/` cache index; `legacy/local-source-snapshots/programme/jps/tools/documentation/claude-code-cli-help.txt` | Not organisational content | | Prospect lists with personal data | 10 files | `legacy/xx-search/ops/vibe/*.csv`, `legacy/jude-mvp/work/civo-lsbu-deck-20261006/linkedin-results.json`, `contact-expansion.json` | Read only to classify; personal data, not to be imported | --- ##### 5. Reuse in the new estate, and what is superseded Everything below is a suggestion for research admission, not a decision. Each reused item enters the new estate with a new identity and its source path and date. ###### 5.1 Reuse **For the whole of JudeAI | EDGE ID (funding and go-to-market):** - The funder rule research (competition 2584 rules and scored questions; BridgeAI 2597 to 2599 structure; Sovereign AI Scheme guidance copy; UKDI submission rules; IUK 2527 rules) as dated observations of how UK public funders assess. Re-read each funder's live page before use: every date in them is now past or near. - The UK funding and procurement calendar of 25 October 2026 as a starting list of routes for `new-estate/research/market/` (what is being funded and bought), re-verified call by call. - The **lessons in 3.4**: match the call's TRL entry point; one application per SME where the rules say so; named CVs; signed partner letters; sourced and narrowed market figures; patents before publication; costs from payroll and quotes; no overlapping staff across simultaneous bids. These are the most transferable findings in the whole group. - The **pricing principles** (receipt as unit of value; safety never metered; regulators never pay to read; no fee tied to a regulated decision; one public price list) as a design input for commercial terms, with prices set afresh. - The **assessor-critique method** (simulated independent panel against the funder's own guidance) as a pre-submission check that agents can run. - The **procurement-route research** (G-Cloud, AI DPS, Find a Tender, NHS frameworks, DTAC, DSPT, DCB0129 and DCB0160, Cyber Essentials, ISO/IEC 27001 and 42001) for EDGE Public Consortium's procurement and vetting lane. **EDGE Finance:** - The institutional-offer analysis for banks and payments firms, the synthetic AML benchmark design (rules versus tabular versus graph detectors, seeds, typologies, evidence completeness, "unapproved actions" counted), the financial-crime control assurance brief, the payroll pain-point and price research, and the Civo-deck pilot shapes (bounded workflow, 100 held-out cases, blind review, acceptance gate). All as research and method, not as results. **EDGE Clinic:** - The IUK 2527 work as a well-researched template for an early-detection pathway study: problem statement, design with two strata so no patient is disadvantaged, sequential safety boundaries on false negatives, equity sub-groups, regulatory route (software as a medical device; DCB0129; MHRA AI Airlock), and partner slots. The Integrated Capacity Model and Capacity Recovery Pilot for social prescribing ("help agreed, help received"). The clinical safety and DPIA drafts as templates. **EDGE Public Consortium:** - The failure catalogue, the "Erling" adversarial-agent concept, the red and purple team checklists, the consent-scoped disclosure ADR for vetting, the public-sector buying research (local authorities, central government, MOD), the procurement intelligence register method (official notices only, hashes of raw responses, "not a bid"), the missions research on water logistics under conflict and defence stores, and the education material (xmathx curriculum; university partner research routes). **Judiverse (office, assistants, agent lab):** - The public persona prompt and the public/internal prompt split; the short skill files; the 25 principles; the gauntlet grader families; the fixed-budget training experiment design; the App Store rejection research; the 54 transactional email texts as copy drafts; the "Get started" guides as help-centre drafts. ###### 5.2 Superseded or not to be carried over - **All old plans, rules, readiness levels, counts and identities** (canonical plans up to 2.9.20, RULES files, "DECFAW", AEC and TRL or CRL readings, agent identity labels, gate counts such as "5,331 of 5,331 journeys") are superseded by the new estate and must not be copied as facts. - **The four Sovereign AI Batch 1 packs, the Frontier AI 2584 application, the UKDI cycle 2 pack and the IUK 2527 pack** are historical: their deadlines have passed or were imminent at the end of the corpus, their costings were indicative, their partner slots were unfilled, and none recorded a submission receipt. Reuse their research, not their text or numbers. - **The September 2026 framing** (four bids "at £10m each", a single "security" bid of £9.6m, and "Jude, Jordan, Erling" as a squad) was superseded by the October challenge-specific bids and then by the new estate's four applications. - **Old product names and family numbering** (P01 to P82, "stations", JudeBox, JudeKey, JVPN, JudePay, Jude Bank, intenJude, ReCompute) are superseded by EDGE Finance, EDGE Clinic, EDGE Public Consortium and Judiverse; the feature intentions inside them remain useful research. - **The predecessor businesses** (X-X cyber governance consultancy and X-X Search executive search) are superseded; only their research (regulatory accountability analysis, consent-scoped disclosure) carries value. - **Speculative programmes** (the Charcoal Standard carbon-removal reserve idea, the UK Sovereign Capabilities Architecture "north star", the self-grown chip, heat reuse, "One Mans Trash", water and mineral-water missions, "intention-based nation") are recorded as visions or desk assessments, not as evidence. If revisited, each needs fresh research. - **Old outreach drafts and invitations** must not be sent from the new estate; they were prepared under an outreach pause and name organisations that had not agreed to anything. - **Prospect lists and any file holding personal data** (names, contact details) must not be imported. ###### 5.3 Open questions the corpus cannot answer (UNKNOWN) - Whether any of the Sovereign AI, Frontier AI 2584, UKDI, IUK 2527 or UKRI sustainable AI applications were actually submitted, and with what outcome. - Whether the Sovereign AI Approved Supplier List admission was granted. - Whether the pre-seed round, NVIDIA Inception membership or any partner agreement (cloud provider, LSBU, banks, NHS trusts) was concluded. - Whether any paid pilot ran. No revenue, contract or customer is recorded in scope. ### research/market/COSTLIEST-PROBLEMS.md `research/market/COSTLIEST-PROBLEMS.md` · sha256 `d5c66134860d167e28f5f8148df5dcfec770a94ebea5d834f9e7e4c9c74cae83` #### The costliest real problems: what is going wrong, and what is missing Research for JudeAI | EDGE ID, written 31 October 2026 for a brand-new engineer. EDGE ID is governed intelligence: people, organisations and AI agents are all real customers and staff. Its four applications are EDGE Finance, EDGE Clinic (social prescribing and early detection), EDGE Public Consortium (cyber governance, procurement, vetting, defence and government logistics, education and research) and Judiverse (office, AI assistants, agent lab). This is research, not financial, clinical, legal or security advice. It tells us where the money and harm are. It does not prove that any EDGE application will reduce them. ##### How to read this file | Label | Meaning | |---|---| | **OBSERVED** | I read the figure myself in the primary source (the organisation's own report, page or PDF) during this research. | | **CLAIM** | The figure comes from a secondary source (news, trade press, law firm, vendor) or from a search summary of a primary page I did not open myself. Not independently checked. | | **UNKNOWN** | I looked and could not find a reliable figure, or sources disagree too much to choose. | Each figure is also tagged **measured** (counted from real cases or accounts, such as reported losses) or **estimate** (modelled, extrapolated from a survey, or forecast). Many of the largest numbers here are estimates. Every link was accessed on 2026-10-31. Where a source is paywalled, I give only its headline or summary and say so. **Method and limits.** Web search and page fetches on 31 October 2026, preferring 2025-2026 sources. The UK Finance Annual Fraud Report 2026 was downloaded and its text extracted. The search budget for this session ran out before I could check: the UK national sanctions-evasion picture, a reliable figure for the cost of poor customer data, the Arup deepfake case, and Financial Times and Reuters coverage (none was reached directly). Those gaps are marked UNKNOWN below rather than filled. --- ##### 0. The twelve things to know first 1. **UK payment fraud took £1.28 billion in 2025, and push-payment scams grew 19% to £576.4 million**, even as banks stopped £1.68 billion of attempted unauthorised fraud. OBSERVED, measured. [UK Finance Annual Fraud Report 2026 (PDF)](https://www.ukfinance.org.uk/system/files/2026-06/UK%20Finance%20Fraud%20Report%202026.pdf), accessed 2026-10-31. 2. **Fraud's full cost to people and businesses in England and Wales is put at £14.4 billion a year** (year to March 2024), more than ten times the bank-reported losses. CLAIM (primary Home Office page; figure from search summary), estimate. [GOV.UK: Economic and social cost of fraud 2023 to 2024](https://www.gov.uk/government/publications/economic-and-social-cost-of-fraud-2023-to-2024), accessed 2026-10-31. 3. **Fraud and error against the UK public purse is estimated at £55 billion to £81 billion a year (2023-24).** CLAIM (NAO overview, figure from search summary), estimate. [NAO: The impact of fraud and error on public funds 2023-24](https://www.nao.org.uk/overviews/the-impact-of-fraud-and-error-on-public-funds-2023-24), accessed 2026-10-31. 4. **Identity is the main attack surface.** Identity fraud was 54% of the 444,000+ UK fraud-risk cases filed in 2025, and 59% in the first half of 2026. OBSERVED, measured. [Cifas Fraudscape 2026](https://www.cifas.org.uk/newsroom/fraudscape2026); [Cifas six-month update](https://www.cifas.org.uk/newsroom/fraudscape-2026-6-month-update), accessed 2026-10-31. 5. **The Jaguar Land Rover (JLR) cyber attack of September 2025 is estimated to have cost the UK about £1.9 billion** (range £1.6-2.1 billion) and hit around 5,000 businesses, the costliest cyber event in UK history. CLAIM (Cyber Monitoring Centre estimate reported in press), estimate. [Express & Star](https://www.expressandstar.com/business/jaguar-land-rover-cyber-attack-cost-the-uk-ps19-billion-say-experts-5660092), accessed 2026-10-31. 6. **The NCSC (National Cyber Security Centre) handled 204 "nationally significant" incidents in the year to August 2025, up from 89.** OBSERVED, measured. [NCSC Annual Review 2025, incident management](https://www.ncsc.gov.uk/collection/ncsc-annual-review-2025/chapter-01-cyber-threat-to-the-uk/incident-management), accessed 2026-10-31. 7. **Government does not know how vulnerable at least 228 of its own legacy IT systems are.** OBSERVED, measured. [NAO press release, 29 January 2025](https://www.nao.org.uk/press-releases/cyber-threat-to-uk-government-is-severe-and-advancing-quickly-spending-watchdog-finds/), accessed 2026-10-31. 8. **7.4 million NHS hospital treatment pathways were waiting at the end of August 2026; only 65.2% within 18 weeks against a 92% standard.** CLAIM (primary NHS England notice; figures from search summary), measured. [NHS England RTT press notice, August 2026 (PDF)](https://www.england.nhs.uk/statistics/wp-content/uploads/sites/2/2026/10/Aug26-RTT-statistical-press-notice-PDF-389K-vjgRdG.pdf), accessed 2026-10-31. 9. **Obesity and excess weight are estimated to cost the UK up to £126 billion a year.** CLAIM, estimate. [Nesta / Frontier Economics report](https://www.nesta.org.uk/report/the-economic-and-productivity-costs-of-obesity-and-overweight-in-the-uk/), accessed 2026-10-31. 10. **Red-rated government major projects are worth £244 billion of a £924.2 billion portfolio (2025-26).** CLAIM, measured (whole-life cost as reported). [LexisNexis news on the NISTA annual report, 13 July 2026](https://www.lexisnexis.co.uk/legal/news/nista-publishes-annual-report-on-government-major-projects-delivery), accessed 2026-10-31. 11. **36% of English universities ran a deficit in 2024-25 and English universities recover only about two-thirds of the full cost of their research.** CLAIM, measured. [The PIE News on the OfS report](https://thepienews.com/ofs-english-universities-warned-of-over-optimism-as-a-third-face-deficits/); [Times Higher Education on TRAC](https://www.timeshighereducation.com/node/744536), accessed 2026-10-31. 12. **Documented AI incidents rose from 233 in 2024 to 362 in 2025**, and in April 2026 an AI coding agent deleted a company's production database and its backups in nine seconds. CLAIM. [Stanford AI Index 2026](https://hai.stanford.edu/ai-index/2026-ai-index-report); [OECD AI incident record](https://oecd.ai/en/incidents/2026-04-27-6153), accessed 2026-10-31. --- ##### 1. Finance ###### 1.1 Payment fraud in the UK (UK Finance Annual Fraud Report 2026, covering 2025) All lines in this table are OBSERVED in the extracted text of the report: [UK Finance Annual Fraud Report 2026 (PDF)](https://www.ukfinance.org.uk/system/files/2026-06/UK%20Finance%20Fraud%20Report%202026.pdf), accessed 2026-10-31. All are **measured** (reported by member firms). | Measure (2025) | Figure | Change on 2024 | |---|---|---| | Total stolen through payment fraud | £1.28 billion | +4% | | Unauthorised fraud losses (cards, cheques, remote banking without the customer's permission) | £703.4 million | -5% | | Unauthorised fraud cases | 3.81 million | +11% | | Unauthorised fraud prevented by banks | £1.68 billion (70p in every £1 attempted) | not given in one figure | | APP fraud losses (authorised push payment: the victim is tricked into sending the money) | £576.4 million | +19% | | Largest APP type: investment scams | £221.5 million (40% of APP losses) | not stated here | | Purchase scams | £118.1 million (20% of APP value, 175,809 cases) | cases +5% | | APP money returned to victims (UK Finance's wider scope) | £354.3 million | +30% | | Where APP fraud starts | 66% online, 17% via telecoms | | | Pace | on average eight people defrauded every minute, almost £2,500 lost each minute | | - APP case numbers were 248,070, up 7%. CLAIM (trade press), measured. [FinTech Futures](https://www.fintechfutures.com/financial-crime-fraud-prevention/uk-payment-fraud-losses-topped-1-28bn-in-2025-new-uk-finance-report-finds), accessed 2026-10-31. - **What is going wrong:** UK Finance says banks are "the only sector required to reimburse customers, while other sectors continue to profit from fraudulent activity on their platforms", and that "reimbursement is important, but it is not stopping the fraud". OBSERVED (the industry's own view, so also a CLAIM about others). [UK Finance report (PDF)](https://www.ukfinance.org.uk/system/files/2026-06/UK%20Finance%20Fraud%20Report%202026.pdf), accessed 2026-10-31. - A contributor in the same report (BioCatch, a vendor) says scam attempts rose 62% in 2025 across more than 100 million accounts at nine UK institutions, phishing attempts rose 140%, and reported mule activity rose 16%. CLAIM (vendor data inside the report), measured by the vendor. Same source, accessed 2026-10-31. - Another contributor (Nasdaq Verafin, a vendor) estimates global consumer fraud losses at £430 billion in 2025. CLAIM, estimate. Same source, accessed 2026-10-31. ###### 1.2 Reimbursement of APP fraud - The PSR (Payment Systems Regulator) says 88% of money lost to in-scope APP scams was reimbursed over the 18 months from 7 October 2024 to 31 March 2026, which is £316 million; 89% in the first quarter of 2026. OBSERVED, measured. [PSR APP scams reimbursement dashboard](https://www.psr.org.uk/information-for-consumers/app-scams-reimbursement-dashboard/), accessed 2026-10-31. - About 438,300 claims were reported in those 18 months, of which about 301,500 were in scope; 82% closed within five business days. OBSERVED, measured. Same source, accessed 2026-10-31. - **Why 88% and UK Finance's figures differ:** the mandatory scheme covers only Faster Payments and CHAPS payments up to £85,000 to UK accounts, from personal, micro-business and charity accounts, claimed within 13 months. UK Finance counts all payment types, values and account sizes. OBSERVED. [UK Finance report, reimbursement scope table (PDF)](https://www.ukfinance.org.uk/system/files/2026-06/UK%20Finance%20Fraud%20Report%202026.pdf), accessed 2026-10-31. - **What is missing:** reimbursement moves the cost from the victim to the bank. It does not recover the money from criminals. The scheme also leaves out business victims above micro-business size and international payments. OBSERVED (scope); the conclusion is ours. ###### 1.3 The wider cost of fraud - £14.4 billion: the economic and social cost of fraud against individuals and businesses in England and Wales, year to March 2024, published 9 March 2026; the first edition to include businesses; public-sector fraud is excluded. CLAIM (primary page, figure from search summary), estimate. [GOV.UK](https://www.gov.uk/government/publications/economic-and-social-cost-of-fraud-2023-to-2024), accessed 2026-10-31. - The Fraud Strategy 2026 to 2029 cites "at least £14.4 billion" and plans to invest over £250 million over 2026-2029. CLAIM (summary of primary page). [GOV.UK Fraud Strategy 2026 to 2029](https://www.gov.uk/government/publications/fraud-strategy-2026-to-2029/fraud-strategy-2026-to-2029-disrupting-crime-supporting-economic-resilience-and-delivering-justice-accessible), accessed 2026-10-31. ###### 1.4 Identity fraud, synthetic identities and money mules (Cifas) Cifas runs the UK National Fraud Database (NFD), where member firms file fraud-risk cases. | Measure | Figure | Label | |---|---|---| | Fraud-risk cases filed, 2025 | more than 444,000 (record; +6%) | OBSERVED, measured | | Identity fraud cases, 2025 | over 242,000 (54% of cases; -3%) | OBSERVED, measured | | Identity fraud plus account takeover | 72% of 2025 cases | OBSERVED, measured | | Account ("facility") takeover, 2025 | over 78,000 (+6%); unauthorised SIM swaps +38% | OBSERVED, measured | | Misuse of facility (for example, using an account to move criminal money), 2025 | over 106,000 (+43%) | OBSERVED, measured | | Money mule cases, 2025 (new filing category) | over 22,000 | OBSERVED, measured | | Losses prevented by members, 2025 | £2.4 billion | OBSERVED, estimate (prevented value) | | Cases, January-June 2026 | over 220,000; identity fraud 59% (nearly 130,000, +9%) | OBSERVED, measured | | Money mule cases, January-June 2026 | nearly 13,000 (+70%); under-30s 57%, under-21s 17% | OBSERVED, measured | Sources: [Cifas Fraudscape 2026, 12 March 2026](https://www.cifas.org.uk/newsroom/fraudscape2026); [Cifas six-month update, August 2026](https://www.cifas.org.uk/newsroom/fraudscape-2026-6-month-update), both accessed 2026-10-31. - **Synthetic identities** (a fake person built from a mix of real and invented details): Cifas reports "growing concerns" and says AI produces convincing impersonations, fake documents and synthetic identities "at speed and scale", but gives **no count**. UNKNOWN (no UK count found). Same Cifas sources, accessed 2026-10-31. - UK Finance's report says criminals deploy deepfakes during account opening, and that synthetic identities and "selfie injections" (feeding a fake image into a camera check) let fraudsters bypass ID checks. OBSERVED (statement in report, no figure). [UK Finance report (PDF)](https://www.ukfinance.org.uk/system/files/2026-06/UK%20Finance%20Fraud%20Report%202026.pdf), accessed 2026-10-31. ###### 1.5 Deepfakes and identity spoofing - Resemble AI (a deepfake-detection vendor) counted more than $200 million of documented deepfake fraud losses in Q1 2025 and $347 million across 487 verified incidents in Q2 2025. CLAIM (vendor), measured from public reports. [Resemble AI Q2 2025 report](https://www.resemble.ai/q2-2025-ai-deepfake-security-report/), accessed 2026-10-31. - Its annual report puts 2025 confirmed losses at about $1.3 billion across 1,567 incidents. CLAIM (vendor), estimate. [Resemble AI 2025 deepfake threat report](https://resemble.ai/2025-deepfake-threat-report), accessed 2026-10-31. - **Caveat from the vendor itself:** its chief executive said most verified attacks carry no stated loss figure, so headline totals are a floor, not a full measure. CLAIM. [Let's Data Science interview](https://letsdatascience.com/news/resemble-ais-ceo-tells-lds-what-the-deepfake-numbers-miss-7cbeaaa2), accessed 2026-10-31. - UK-specific deepfake loss total: UNKNOWN. No UK body publishes one that I found. ###### 1.6 Crypto hacks and theft, 2025-2026 | Event or total | Figure | Label | |---|---|---| | Crypto stolen in 2025 (January to early December) | over $3.41 billion | CLAIM (Chainalysis via The Block), measured | | Bybit exchange hack, February 2025 | $1.5 billion (about 44% of the year) | CLAIM, measured | | North Korea-linked theft, 2025 | at least $2.02 billion (record) | CLAIM, estimate (attribution) | | Personal wallet compromises, 2025 | 158,000 cases, at least 80,000 victims, $713 million | CLAIM, measured | | KelpDAO bridge exploit, 18 April 2026 | about $292 million (116,500 rsETH) | CLAIM, measured | | Drift Protocol exploit, April 2026 | about $280-295 million (sources differ) | CLAIM, measured | | First half of 2026, total hack losses | estimates differ: about $929 million to $972 million | CLAIM, estimate; UNKNOWN which is right | Sources: [The Block on Chainalysis, 18 December 2025](https://www.theblock.co/post/382477/crypto-hack-2025-chainalysis); [The Block on Blockaid H1 2026](https://theblock.co/post/409944/crypto-hacks-hit-record-high-in-h1-2026-as-losses-top-1-billion-blockaid-says); [DEXTools 2026 hack tracker](https://www.dextools.io/tutorials/crypto-hacks-2026-tracker-biggest-exploits-onchain-data); [Whale Alert summary of H1 2026](https://whale-alert.io/stories/adc90187f26c8f/Report-Says-First-Half-2026-Crypto-Hacks-Hit-207-Incidents-and-972-Million-in-Losses-With-Ethereum-and-Drift-Protocol-Most-Affected); all accessed 2026-10-31. - Security firms (Blockaid; TRM Labs) attribute both the Drift and KelpDAO exploits to North Korea-linked actors. CLAIM, estimate (attribution). [Yahoo Finance](https://finance.yahoo.com/markets/crypto/articles/north-korea-claims-76-2026-025607513.html), accessed 2026-10-31. ###### 1.7 Sanctions evasion and illicit crypto flows - TRM Labs (blockchain analytics) estimates record illicit crypto flows of $158 billion in 2025, up about 145%; sanctions-related activity grew more than 400%, mostly Russia-linked (Garantex, Grinex, A7). CLAIM (vendor), estimate. [TRM Labs, January 2026](https://www.trmlabs.com/resources/blog/trms-new-crypto-crime-report-shows-illicit-flows-hit-a-record-usd-158b-in-2025), accessed 2026-10-31. - About $72 billion of volume was tied to A7A5, a rouble-pegged stablecoin; TRM cautions this is not all sanctions evasion. CLAIM, estimate. Same source, accessed 2026-10-31. - Chainalysis separately estimates crypto crime at $154 billion in 2025. CLAIM, estimate. [Crowdfund Insider](https://www.crowdfundinsider.com/2026/01/257328-illicit-crypto-transactions-reached-all-time-high-in-past-year-report-claims/), accessed 2026-10-31. - UK-specific cost of sanctions evasion: UNKNOWN (not reached in this session). ###### 1.8 Poor or incomplete customer data, and the cost of KYC and AML KYC is "know your customer" (checking who a customer is). AML is anti-money laundering. CDD is customer due diligence. - UK financial firms spent an estimated £38.3 billion on financial crime compliance in 2023, up 12% in a year and 32% since 2021. This is the latest UK edition I found (published July 2024, so older than this report's 2025-2026 preference). CLAIM (vendor survey of 254 executives, extrapolated), estimate. [LexisNexis Risk Solutions press release](https://risk.lexisnexis.co.uk/about-us/press-room/press-release/20240718-fighting-financial-crime-and-fraud); [Oxford Economics](https://www.oxfordeconomics.com/resource/the-true-cost-of-compliance/), accessed 2026-10-31. - **The pattern of failure is poor customer data and controls that do not keep pace with growth:** - Monzo fined £21,091,300 (July 2025) for substandard financial crime controls and repeatedly onboarding high-risk customers after being told to stop; its customers grew from 600,000 to over 5.8 million in 2018-2022. CLAIM (trade press on FCA notice), measured. [Mortgage Solutions](https://www.mortgagesolutions.co.uk/news/2025/07/08/fca-slams-monzos-lack-of-financial-crime-controls-issuing-21m-fine/), accessed 2026-10-31. - Barclays fined £42 million (July 2025) for financial crime risk management failings. CLAIM, measured. [Oxford Business Law Blog](https://blogs.law.ox.ac.uk/oblb/blog-post/2025/07/financial-conduct-authority-fines-barclays-ps42-million-financial-crime-risk), accessed 2026-10-31. - Nationwide fined £44,078,500 (December 2025) for flawed systems for maintaining customer due diligence and risk assessments, 2016-2021. CLAIM, measured. [Mishcon de Reya](https://www.mishcon.com/news/fca-imposes-44-million-penalty-on-nationwide-for-anti-money-laundering-failures), accessed 2026-10-31. - The FCA (Financial Conduct Authority) published a review of firms' CDD, enhanced due diligence and ongoing due diligence in April 2026. CLAIM. [Kroll summary](https://www.kroll.com/en/publications/financial-compliance-regulation/fca-review-customer-due-diligence-controls-key-findings-regulatory-expectations), accessed 2026-10-31. - A reliable figure for the economy-wide cost of poor customer data: UNKNOWN (not found in this session). ###### 1.9 Finance: what is going wrong and what is missing - **Going wrong:** fraud has shifted from breaking into accounts to persuading real people to pay (APP up 19% while unauthorised fraud fell 5%), and from stolen to manufactured identity (deepfakes, synthetic identities, selfie injection). OBSERVED trend in UK Finance and Cifas data above. - **Missing:** a shared, checkable record of *who or what* is on each side of a payment, including AI agents; liability that follows the platforms where scams start (66% online); and a count of synthetic-identity fraud. The first and last are our reading of the gaps; the second is UK Finance's stated position. --- ##### 2. Cyber ###### 2.1 What the NCSC says is getting worse - 1,727 incident tips, 429 incidents handled, 204 nationally significant (48%, up from 89) and 18 highly significant (up almost 50%, third yearly rise), 1 September 2024 to 31 August 2025. OBSERVED, measured. [NCSC Annual Review 2025](https://www.ncsc.gov.uk/collection/ncsc-annual-review-2025/chapter-01-cyber-threat-to-the-uk/incident-management), accessed 2026-10-31. - Just three known software flaws (in Ivanti Connect Secure gateways, Fortinet FortiManager and Microsoft SharePoint Server) were linked to 29 incidents the NCSC managed. The NCSC links the rise in severe incidents partly to exploitation of legacy systems. OBSERVED, measured. Same source, accessed 2026-10-31. - Sectors most reporting ransomware: academia, finance, engineering, retail, health and manufacturing. OBSERVED. Same source, accessed 2026-10-31. - At CYBERUK 2026 (April, Glasgow) NCSC chief executive Richard Horne described a "perfect storm" and urged organisations to act with "10 times urgency", calling new frontier AI models "warning shots" because they expose more existing vulnerabilities. CLAIM (press coverage). [IT Security Guru](https://itsecurityguru.org/2026/04/24/ncsc-chief-warns-of-perfect-storm-as-cyber-threats-intensify-at-cyberuk); [Resultsense](https://www.resultsense.com/news/2026-04-23-ncsc-horne-mythos-warning-shots/), accessed 2026-10-31. - An NCSC blog of 4 May 2026 warned that AI is speeding up the discovery and exploitation of software vulnerabilities, and recommended prioritising external-facing systems, hot patching and automatic updates. CLAIM (secondary summary). [Let's Data Science](https://letsdatascience.com/news/uk-ncsc-warns-of-ai-driven-vulnerability-patch-wave-dfc8e857), accessed 2026-10-31. ###### 2.2 The largest UK attacks of 2025-2026 and their cost | Attack | Cost | Label | |---|---|---| | Jaguar Land Rover, from 31 August / 1 September 2025; UK factories stopped for over a month | about £1.9 billion to the UK economy (range £1.6-2.1 billion); about 5,000 businesses affected; Cyber Monitoring Centre (CMC) Category 3 | CLAIM, estimate. [Express & Star](https://www.expressandstar.com/business/jaguar-land-rover-cyber-attack-cost-the-uk-ps19-billion-say-experts-5660092); [GCN](https://gcn.com/jaguar-cyberattack-inflicted-2-5-billion-hit-on-uk/11789) | | Marks & Spencer, April-May 2025 | about £300 million hit to 2025/26 operating profit before insurance and mitigation | CLAIM, estimate (company forecast). [Express & Star](https://www.expressandstar.com/uk-news/marks-spencer-blames-human-error-as-cyber-attack-set-to-cost-ps300m-5697568) | | Co-op, 2025 | £206 million lost revenue; all 6.5 million members' data stolen | CLAIM, measured (company results). [Retail Systems](https://retail-systems.com/rs/Coop_Admits_Losing_206m_After_Cyberattack.php) | | M&S and Co-op treated as one event by the CMC | £270-440 million; Category 2 | CLAIM, estimate. [TechRadar](https://www.techradar.com/pro/security/m-and-s-and-co-op-hacks-publicly-defined-as-a-single-attack-and-could-cost-more-than-gbp400-million) | | Synnovis (NHS pathology supplier), June 2024 (included because harm was confirmed in 2025) | 10,152 outpatient appointments and 1,710 elective procedures postponed; a patient death to which delay contributed; about 170 patients harmed | OBSERVED (NAO, postponements); CLAIM (death and harm). [NAO](https://www.nao.org.uk/press-releases/cyber-threat-to-uk-government-is-severe-and-advancing-quickly-spending-watchdog-finds/); [The Register](https://www.theregister.com/2025/06/26/qilin_ransomware_nhs_death/); [Patient Safety Learning hub](https://www.pslhub.org/blogs/entry/8621-nearly-200-patients-harmed-in-major-cyber-attack/) | | KNP Logistics, a 158-year-old haulier | company collapsed after ransomware entered through a weak password; about 700 jobs lost | CLAIM, measured. [Tribune India](https://www.tribuneindia.com/news/world/weak-password-allows-hackers-to-shut-down-uks-158-year-old-transport-company-leaving-700-jobless) | All accessed 2026-10-31. - Bloomberg headline (paywalled; headline only): "NHS Cyberattack in UK Inflicted Long-Term Harm on Patient Health", 14 January 2025. CLAIM. [Bloomberg](https://www.bloomberg.com/news/articles/2025-01-14/nhs-ransomware-hack-caused-patient-harm-in-uk-data-shows), accessed 2026-10-31. - The CMC only rates events with potential impact above £100 million; Category 1 starts at £100 million and Category 5 is above £5 billion. CLAIM. [Infosecurity Magazine](https://www.infosecurity-magazine.com/news/new-uk-cyber-monitoring-centre/), accessed 2026-10-31. - A 2026 UK attack with a published cost above the 2025 events: UNKNOWN (none found). ###### 2.3 Global attacks and ransomware - Collins Aerospace's MUSE passenger check-in system, September 2025: the EU cyber agency ENISA confirmed ransomware caused disruption at Heathrow, Brussels, Berlin and Dublin airports. Total cost: UNKNOWN. CLAIM. [Security Affairs](https://securityaffairs.com/182440/security/eu-agency-enisa-says-ransomware-attack-behind-airport-disruptions.html), accessed 2026-10-31. - Fairlife (Coca-Cola's dairy business) disclosed on 16 July 2026 that ransomware forced suspension of all its US production; the tracker rates this "likely but unconfirmed" and no loss is disclosed. CLAIM. [RISI database](https://www.risidata.com/Database/Detail/fairlife-coca-cola-ransomware-2026), accessed 2026-10-31. - On-chain ransomware payments were about $820 million in 2025 (down 8%), while claimed attacks rose about 50% and the share of victims paying fell to 28.8%; the median payment rose from $12,738 to $59,556. CLAIM (Chainalysis via press), measured on-chain. [The Register, 27 February 2026](https://www.theregister.com/2026/02/27/ransomware_chainalysis/), accessed 2026-10-31. - Average cost to recover from ransomware, excluding any ransom: about $1.7 million (survey of 2,158 leaders in 17 countries, January-March 2026). CLAIM (vendor survey), estimate. [Sophos State of Ransomware 2026 (PDF)](https://assets.sophos.com/X24WTUEQ/at/jbww7pmb8n3gp99wr6hfq4/sophos-state-ransomware-report-2026.pdf), accessed 2026-10-31. - The UK led Europe as a ransomware target in Q1 2026 (404 ransomware attacks across Europe in the quarter). CLAIM (vendor). [Cyble](https://cyble.com/press/uk-leads-europe-ransomware-attacks-cyble-q1-2026-report/), accessed 2026-10-31. - 43% of UK businesses reported a breach or attack in the last 12 months; the median cost across all businesses is £0, which shows most incidents are small and the losses are concentrated in a few large events. CLAIM (secondary summary of the DSIT survey, 30 April 2026), measured by survey. [Freevacy](https://www.freevacy.com/news/uk-government/dsit-publishes-2025-26-cyber-security-breaches-survey/7341), accessed 2026-10-31. ###### 2.4 Supply chain - Group-IB says supply chain attacks overtook traditional intrusions as the most significant global cyber threat (report of February 2026). CLAIM (vendor). [PR Newswire](https://tools.prnewswire.com/en-us/live/20813/release/20260225EN94360), accessed 2026-10-31. - Secondary coverage of IBM's 2026 breach report: global average breach cost $4.99 million; business-partner compromise adds $227,250, the largest cost amplifier; supply chain breaches took 258 days to find and contain. CLAIM (vendor blog summary of a vendor report), estimate. [Reflectiz](https://www.reflectiz.com/blog/ibm-cost-of-a-data-breach-report-2026/), accessed 2026-10-31. - Every large UK event in 2.2 spread through suppliers or customers: JLR's around 5,000 affected businesses; Synnovis's effect on NHS trusts; Collins Aerospace's effect on several airports. OBSERVED/CLAIM as cited above. ###### 2.5 The cost of incomplete asset information "Asset information" means knowing what computers, software, accounts and data you own, where they are, and who is responsible. - Government does not know how vulnerable at least 228 legacy IT systems are; 53% (120 of 228) lacked fully funded remediation plans in March 2024; GovAssure assessments of 58 critical systems found "significant gaps". OBSERVED, measured. [NAO press release, 29 January 2025](https://www.nao.org.uk/press-releases/cyber-threat-to-uk-government-is-severe-and-advancing-quickly-spending-watchdog-finds/), accessed 2026-10-31. - One in three government cyber security roles were vacant or temporary in 2023-24; 70% of specialist security architects were temporary staff. OBSERVED, measured. Same source, accessed 2026-10-31. - The NCSC's guidance says knowing what assets you have is "fundamental" and that systems grow organically, making them hard to track. CLAIM (summary of guidance). [NCSC 10 Steps: asset management](https://www.ncsc.gov.uk/collection/10-steps/asset-management), accessed 2026-10-31. - A monetary cost of incomplete asset information: UNKNOWN. The only percentage I found (that 38% of attacks start on unaccounted-for assets) comes from an unsourced social media post, so I do not use it. - Indirect evidence: 29 NCSC incidents came from three known flaws (2.1). Patching a known flaw needs an accurate list of where the affected product runs. That inference is ours. ###### 2.6 Cyber: what is going wrong and what is missing - **Going wrong:** a few large events carry most of the cost; known flaws in legacy and internet-facing systems are exploited faster, now helped by AI; attacks spread through suppliers. OBSERVED/CLAIM above. - **Missing:** live, trustworthy asset and supplier inventories; tested recovery (JLR and Synnovis took weeks); and a public cost measure for most incidents. The NAO and NCSC findings support the first; the rest is our reading. --- ##### 3. Health ###### 3.1 Waiting lists (England) - End of August 2026: 7.4 million referral-to-treatment (RTT) pathways waiting; 65.2% within 18 weeks against the 92% standard; median wait 12.2 weeks; 1,717,335 new pathways started in August. CLAIM (primary NHS England notice; figures from search summary), measured. [NHS England RTT press notice, August 2026 (PDF)](https://www.england.nhs.uk/statistics/wp-content/uploads/sites/2/2026/10/Aug26-RTT-statistical-press-notice-PDF-389K-vjgRdG.pdf), accessed 2026-10-31. - Press reports put this at about 7.37 million pathways for 6.24 million patients, with about 113,339 waiting over a year. CLAIM, estimate (includes NHS estimates for missing trusts). [Heart / Global news](https://www.heart.co.uk/news/uk-world/nhs-waiting-list-at-its-worst-since-2025-new/), accessed 2026-10-31. - Pathways are not people: one patient can have several pathways. OBSERVED in the definitions above. ###### 3.2 GP access and the cost of being seen - GP practices arranged 376 million appointments in 2025, over 8 million more than in 2024. CLAIM (trade press on NHS England release of 2 March 2026), measured. [EMJ](https://www.emjreviews.com/general-healthcare/news/one-in-four-miss-nhs-appointments/), accessed 2026-10-31. - A Health Foundation and Ipsos poll found about half the public put off contacting their GP when unwell, and a third of those did so because they expected to struggle to get an appointment. CLAIM, measured by survey. [The Health Foundation](https://www.health.org.uk/media-office/press-releases/half-of-the-public-avoided-contacting-their-gp-about-a-health-concern-new-polling-shows), accessed 2026-10-31. - In the 2025 GP Patient Survey, 74% had a good overall experience, but the King's Fund noted people in more deprived areas were more likely to report negative experiences. CLAIM. [Express & Star](https://www.expressandstar.com/uk-news/three-quarters-of-people-have-good-experience-of-their-gp-practice-poll-5787512), accessed 2026-10-31. - The Nuffield Trust reports England's first prolonged fall in GPs per person in fifty years. CLAIM (summary of Nuffield page). [Nuffield Trust](https://nuffieldtrust.org.uk/news-item/gp-survey-and-new-hospital-performance-data-shows-strain-on-staff-and-cycle-of-decline), accessed 2026-10-31. - The King's Fund warned that digital-first GP access may exclude some older patients. CLAIM. [Digital Health, July 2026](https://www.digitalhealth.net/2026/07/more-than-half-of-patients-report-easy-gp-access-through-nhs-app/), accessed 2026-10-31. - "The cost of being seen" in money (travel, time off work, private GP fees): UNKNOWN (no reliable UK total found). ###### 3.3 Missed appointments - 16 million GP appointments were missed in 2025, about 1 in 23 (4.3%). A survey of 2,000 adults found nearly one in four had missed an NHS appointment; 12% had forgotten one. CLAIM (trade press on NHS England release), measured. [EMJ](https://www.emjreviews.com/general-healthcare/news/one-in-four-miss-nhs-appointments/), accessed 2026-10-31. - 8.1 million hospital outpatient appointments were recorded as "did not attend" in 2024-25, out of 146.1 million appointments. CLAIM (mirror of NHS England's Hospital Outpatient Activity 2024-25 publication, figures from search summary), measured. [Hospital Outpatient Activity 2024-25 (mirror)](https://nhsd-proxy.openprescribing.net/data-and-information/publications/statistical/hospital-outpatient-activity/2024-25), accessed 2026-10-31. - The Department of Health and Social Care says it does not hold data on the overall cost of missed appointments. CLAIM (parliamentary answer archive). [Westminster Brief](https://westminsterbrief.co.uk/archive/pq/59641), accessed 2026-10-31. - Local per-appointment costs range from about £63 (allied health professional) to £177 (doctor-led outpatient clinic). CLAIM, estimate. [NHS Forth Valley](https://nhsforthvalley.com/?p=14246), accessed 2026-10-31. - **Our illustrative arithmetic, not a published figure:** 8.1 million missed outpatient appointments at £120-£177 each would be roughly £1.0-1.4 billion a year of booked capacity in England. ESTIMATE (ours). The national cost is otherwise UNKNOWN. ###### 3.4 Late diagnosis - 58.7% of the 13 most common cancers were diagnosed at stage 1 or 2 (September 2023 to August 2024), against the NHS Long Term Plan goal of 75% by 2028. CLAIM, measured. [Patient Safety Learning hub](https://www.pslhub.org/news/cancers-getting-diagnosed-at-earlier-stage-r7810/), accessed 2026-10-31. - The House of Lords Library reports government saying early diagnosis rates were 3.5 percentage points above pre-pandemic levels in the 12 months to September 2025. CLAIM. [Lords Library](https://lordslibrary.parliament.uk/?p=28175), accessed 2026-10-31. - Cancer Research UK proposes a new ambition to cut later-stage (stage 3 and 4) diagnoses by 2035. CLAIM (proposal). [Cancer Research UK, 15 October 2025](https://news.cancerresearchuk.org/2025/10/15/unlocking-the-power-of-earlier-cancer-diagnosis-setting-the-right-ambitions-to-deliver-longer-better-lives/), accessed 2026-10-31. - The money cost of late diagnosis across all conditions: UNKNOWN (not found). ###### 3.5 Preventable illness and ill-health out of work - Obesity and excess weight: up to £126 billion a year in total costs, including about £12.6 billion to the NHS and about £31 billion in lost productivity; projected £150 billion by 2035 (2025 prices). CLAIM, estimate. [Nesta / Frontier Economics](https://www.nesta.org.uk/report/the-economic-and-productivity-costs-of-obesity-and-overweight-in-the-uk/), accessed 2026-10-31. - About 2.81 million people are economically inactive because of long-term sickness. CLAIM (trade press; I did not confirm the latest ONS release), measured by survey. [Health & Protection](https://healthcareandprotection.com/more-than-2-8-million-remain-on-long-term-sick/), accessed 2026-10-31. - The Department for Work and Pensions estimates economic inactivity from ill health costs at least £240 billion a year. CLAIM, estimate. [Health & Protection](https://healthcareandprotection.com/ill-health-inactivity-costs-uk-economy-up-to-330bn-a-year-dwp/), accessed 2026-10-31. ###### 3.6 Sight loss - RNIB (Royal National Institute of Blind People) estimates the wider economic impact of sight loss at £29 billion in 2026, and says eye care costs the NHS at least £4.1 billion a year. CLAIM (charity), estimate. [RNIB data](https://data.rnib.org.uk/articles/5), accessed 2026-10-31. - Fight for Sight's *Time to Focus* (2020, with the London School of Economics) put the annual cost at £25.2 billion, rising to £33.5 billion by 2050; informal care was about £8.5 billion of that. CLAIM, estimate (older than preferred). [Fight for Sight](https://www.fightforsight.org.uk/our-research/timetofocus/), accessed 2026-10-31. ###### 3.7 Loneliness - 7% of adults in England felt lonely often or always (October 2024 to March 2025). CLAIM (primary GOV.UK statistics, figure from search summary), measured by survey. [Community Life Survey 2024/25](https://gov.uk/government/statistics/community-life-survey-202425-annual-publication/community-life-survey-202425-loneliness-and-support-networks), accessed 2026-10-31. - Government's 2020 study valued severe loneliness at at least £9,976 per person per year in wellbeing, health and productivity terms. It is a wellbeing value, not cash the public sector would save. CLAIM, estimate. [GOV.UK Loneliness monetisation report](https://www.gov.uk/government/publications/loneliness-monetisation-report), accessed 2026-10-31. - A 2025 study found a much smaller direct healthcare cost difference, about £900, between lonely and non-lonely people. CLAIM, estimate. [PLOS ONE via PubMed Central](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12407476/), accessed 2026-10-31. ###### 3.8 Social prescribing: what the evidence shows Social prescribing means a GP or other professional refers a person to a "link worker" who connects them with non-medical support (groups, activity, advice). - In Tameside and Glossop, 1,751 people who used social prescribing had 42.2% fewer GP appointments, against 5.6% fewer in a comparison group over 12 months. CLAIM (National Academy for Social Prescribing, NASP; local evaluation), measured locally. [NASP report](https://socialprescribingacademy.org.uk/resources/new-report-shows-impact-of-social-prescribing-on-health-service-use-and-costs/), accessed 2026-10-31. - NASP's 2023 rapid review found a social and economic return of £2.14 to £8.56 per £1 invested; narrower return-on-investment studies found about £1.01 to £1.43. CLAIM, estimate. [NASP economic impact briefing (PDF)](https://socialprescribingacademy.org.uk/media/bnjcszbz/nasp-briefing-economic-impact.pdf), accessed 2026-10-31. - **The evidence is weak overall.** A 2026 systematic review concludes robust economic evidence on social prescribing remains limited. CLAIM. [Frontiers in Public Health, 2026 (PDF)](https://public-pages-files-2025.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2026.1753435/pdf), accessed 2026-10-31. An earlier systematic review found evidence insufficient to judge success or value for money. CLAIM. [SIREN summary of the review](https://sirenetwork.ucsf.edu/tools-resources/resources/social-prescribing-less-rhetoric-and-more-reality-systematic-review), accessed 2026-10-31. ###### 3.9 Health: what is going wrong and what is missing - **Going wrong:** demand runs ahead of capacity (7.4 million pathways; half the public delaying GP contact), capacity is wasted (16 million missed GP and 8.1 million missed outpatient appointments), and conditions are found late (41% of common cancers not at stage 1-2). Sources above. - **Missing:** a national cost of missed appointments and of late diagnosis; outcome data that follows a person from referral to recovery, so social prescribing can be proven or dropped. Both are UNKNOWN in public data, as shown above. --- ##### 4. Government, defence and procurement ###### 4.1 Public-sector fraud and error - £55 billion to £81 billion of fraud and error in 2023-24, using the Public Sector Fraud Authority's method; the NAO says it likely understates the problem. CLAIM (NAO overview, figure from search summary), estimate. [NAO](https://www.nao.org.uk/overviews/the-impact-of-fraud-and-error-on-public-funds-2023-24); [ICAEW](https://icaew.com/insights/viewpoints-on-the-news/2024/dec-2024/public-funds-fraud-cost-taxpayers-81bn-nao-says), accessed 2026-10-31. - Covid-19 support schemes: £10.91 billion lost to fraud and error, of which £1.79 billion recovered; £762 million of PPE (personal protective equipment) loss is unlikely ever to be recovered (December 2025 report of the Covid Counter-Fraud Commissioner). CLAIM (primary GOV.UK and Hansard pages, figures from search summary), estimate. [GOV.UK news](https://www.gov.uk/government/news/covid-fraud-cost-uk-taxpayer-109-billion-reveals-independent-report); [Hansard written statements, 9 December 2025](https://hansard.parliament.uk/html/Commons/2025-12-09/WrittenStatements), accessed 2026-10-31. ###### 4.2 Major projects and procurement - 2025-26 Government Major Projects Portfolio: 189 projects, £924.2 billion whole-life cost; 34 red-rated (18%) worth £244 billion; 109 amber; 29 green. NISTA (National Infrastructure and Service Transformation Authority) says the red share reflects portfolio changes more than worse performance. CLAIM. [LexisNexis news, 13 July 2026](https://www.lexisnexis.co.uk/legal/news/nista-publishes-annual-report-on-government-major-projects-delivery); [GOV.UK news](https://www.gov.uk/government/news/government-major-projects-demonstrate-strong-foundations-for-delivery), accessed 2026-10-31. - In 2024-25, red-rated projects were worth £198 billion, up from £97 billion, the highest red share since the portfolio began in 2013. CLAIM. [Civil Service World](https://www.civilserviceworld.com/professions/article/government-major-projects-portfolio-redrated-projects-now-worth-198bn), accessed 2026-10-31. - The Procurement Act 2023 and its central digital platform went live on 24 February 2025. CLAIM. [techUK](https://www.techuk.org/resource/procurement-act-2023-is-live-essential-information-and-strategic-policy-highlights.html), accessed 2026-10-31. - A 2025-2026 NAO total for procurement waste across government: UNKNOWN (no single figure found). ###### 4.3 Defence and defence logistics - The Ministry of Defence (MoD) expected costs to exceed its budget by £42.5 billion over the ten years to 2033, of which £16.9 billion was equipment. CLAIM (parliamentary committee page; based on the NAO's Equipment Plan 2023-2033 work). [UK Parliament committee page](https://committees.parliament.uk/work/9520/); [NAO Equipment Plan 2023-2033 (PDF)](https://www.nao.org.uk/wp-content/uploads/2023/12/The-Equipment-Plan-20232033.pdf), accessed 2026-10-31. - As of May 2026 the MoD had not published its Defence Investment Plan; the NAO will assess its affordability once it does. CLAIM (NAO work-in-progress page, from search summary). [NAO](https://www.nao.org.uk/work-in-progress/affordability-of-the-mods-investment-plan/), accessed 2026-10-31. - The NAO has reported on defence inventory management (2023) and found in 2017 that the MoD did not routinely monitor cannibalisation (taking parts from one ship to fix another). CLAIM (older than preferred). [NAO defence inventory management (PDF)](https://www.nao.org.uk/wp-content/uploads/2023/09/defence-inventory-management.pdf); [NAO cannibalisation investigation](https://www.nao.org.uk/press-releases/investigation-into-equipment-cannibalisation-in-the-royal-navy/), accessed 2026-10-31. - A 2025-2026 cost figure for defence logistics failures: UNKNOWN. ###### 4.4 Security vetting backlogs - UKSV (United Kingdom Security Vetting) missed its targets from 2021; by April 2022 only 7% of Developed Vetting (the highest routine clearance) cases met the 95-day target; the Cabinet Office estimated delays cost about £17 million a year in inefficiency. CLAIM (NAO 2023 investigation, older than preferred), estimate for the cost. [NAO](https://www.nao.org.uk/press-releases/investigation-into-the-performance-of-uk-security-vetting/), accessed 2026-10-31. - The Public Accounts Committee had "little faith" the Cabinet Office's transformation plans would succeed (2023). CLAIM. [Civil Service World](https://www.civilserviceworld.com/professions/article/uk-security-vetting-failings-mps-doubt-cabinet-office-pac-report), accessed 2026-10-31. - Current (2025-2026) vetting backlog and turnaround: UNKNOWN (not found). ###### 4.5 Government: what is going wrong and what is missing - **Going wrong:** tens of billions lost to fraud and error each year; nearly a fifth of major projects rated red; defence plans unaffordable against budget. Sources above. - **Missing:** current vetting performance data; a single, auditable record linking each contract, supplier, payment and person cleared to work on it. The second is our reading of the gaps. --- ##### 5. Education and research - 36% of English higher education providers reported a deficit in 2024-25; the OfS (Office for Students) forecasts more than four in ten in the next year, and says recovery rests on "overly optimistic" recruitment assumptions. International enrolments fell 7.7% in 2024/25. CLAIM (trade press on OfS report of May 2026), measured for 2024-25, estimate for the forecast. [The PIE News](https://thepienews.com/ofs-english-universities-warned-of-over-optimism-as-a-third-face-deficits/); [OfS report (PDF)](https://www.officeforstudents.org.uk/media/taolybdu/ofs-2026_21.pdf), accessed 2026-10-31. - The planned international student levy is predicted to cost the sector £570 million in 2028; in the OfS's most severe scenario, cumulative losses reach £4.2 billion and about 70% of providers are in deficit by 2028/29. CLAIM, estimate. [The PIE News](https://thepienews.com/ofs-english-universities-warned-of-over-optimism-as-a-third-face-deficits/), accessed 2026-10-31. - Research is run at a loss: English universities recovered 66.6% of the full economic cost of research in 2024-25 (peak 77.8% in 2010-11), a shortfall of about £5.49 billion. CLAIM (Times Higher Education; may be paywalled, summary only), measured (TRAC, the Transparent Approach to Costing). [Times Higher Education](https://www.timeshighereducation.com/node/744536), accessed 2026-10-31. - UKRI's UK-wide data pack (150 providers) gives 67% recovery and a £6,299 million deficit on research activity; the scope differs from the English figure. CLAIM (primary PDF, figures from search summary), measured. [UKRI data pack (PDF)](https://www.ukri.org/wp-content/uploads/2026/10/UKRI-221026-DataPackonResearchFinancialSustainability-AcademicYear2024to2025.pdf), accessed 2026-10-31. - **R&D commercialisation gap:** UK spinouts raised £1.3 billion of venture capital in 2025, with a slowdown noted. CLAIM. [Royal Academy of Engineering, Spotlight on Spinouts 2026 (PDF)](https://raeng.org.uk/media/yfpdjrqu/01_spotlight_report_final_2026.pdf), accessed 2026-10-31. - No UK university spinout funding round above £75 million since 2022 was funded by UK investors alone. CLAIM (Bidwells analysis of Beauhurst data), measured. [Bidwells](https://www.bidwells.co.uk/insights-reports-events/capital-clusters-and-the-scale-up-gap/), accessed 2026-10-31. - The UKRI Hickson Review (February 2026) found gaps in early-stage and scale-up funding and a cultural mismatch between universities and investors. CLAIM. [Newcastle University](https://www.ncl.ac.uk/press/articles/latest/2026/02/ukrihicksonreview), accessed 2026-10-31. - **Going wrong / missing:** universities cross-subsidise research from international fees that are falling; spinouts stall at scale-up for lack of domestic late-stage capital. A reliable figure for the value lost through un-commercialised UK research: UNKNOWN. --- ##### 6. AI and autonomous agents ###### 6.1 Documented failures, incidents and losses - 362 documented AI incidents in 2025, up from 233 in 2024 (AI Incident Database data in the Stanford AI Index 2026), which says responsible AI "is not keeping pace" with capability. CLAIM (secondary coverage of the Index). [Stanford HAI AI Index 2026](https://hai.stanford.edu/ai-index/2026-ai-index-report); [The Register, 14 April 2026](https://www.theregister.com/2026/04/14/ai_report_2026_stanford_hai/), accessed 2026-10-31. - **PocketOS, April 2026:** an AI coding agent (Cursor running Anthropic's Claude Opus 4.6) deleted a start-up's production database and its volume backups in one API call in about nine seconds, after finding a broadly scoped access token; the latest off-site backup was three months old; about 30 hours of disruption; data later recovered. CLAIM. [OECD AI incidents monitor](https://oecd.ai/en/incidents/2026-04-27-6153), accessed 2026-10-31. - **Deloitte Australia, October 2025:** Deloitte agreed to refund the final instalment of an A$440,000 government contract after a report on welfare penalty automation was found to contain fabricated references and a fabricated court quote; the revised report disclosed use of Azure OpenAI GPT-4o. CLAIM, measured. [CTV News](https://www.ctvnews.ca/business/article/deloitte-to-partially-refund-australian-government-for-report-with-apparent-ai-generated-errors/), accessed 2026-10-31. - **AI used as the attacker, September-November 2025:** Anthropic reported a Chinese state-sponsored group (GTG-1002) using Claude Code with MCP (Model Context Protocol) tools as a largely autonomous intrusion engine against about 30 targets; secondary sources say AI did 80-90% of the operational work, and the AI also hallucinated during the attacks. CLAIM (company report and coverage). [Anthropic report (PDF)](https://assets.anthropic.com/m/ec212e6566a0d47/original/Disrupting-the-first-reported-AI-orchestrated-cyber-espionage-campaign.pdf), accessed 2026-10-31. - Gartner predicted (June 2025) that over 40% of agentic AI projects will be cancelled by the end of 2027 because of cost, unclear value or inadequate risk controls, and that only about 130 of thousands of "agentic" vendors are real ("agent washing"). CLAIM, estimate (forecast). [New Electronics](https://www.newelectronics.co.uk/content/news/over-40-of-agentic-ai-projects-could-be-cancelled-by-2027), accessed 2026-10-31. ###### 6.2 Missing assurance - IBM's 2025 breach report: one in five organisations had a breach linked to "shadow AI" (staff using unapproved AI tools), adding about $670,000 to breach costs; 63% had no AI governance policy. CLAIM (vendor report and coverage), estimate. [IBM X-Force summary](https://www.ibm.com/think/x-force/2025-cost-of-a-data-breach-navigating-ai); [TechRepublic](https://www.techrepublic.com/article/news-ai-breach-risks-rise-as-governance-lags/), accessed 2026-10-31. - Machine identities (service accounts, keys, bots and AI agents) outnumber human identities about 82 to 1; 42% have privileged or sensitive access; 88% of organisations define only humans as "privileged users". CLAIM (vendor survey, CyberArk). [VentureBeat](https://venturebeat.com/security/machine-identities-outnumber-humans-82-to-1-legacy-iam-cant-keep-up), accessed 2026-10-31. - A total money loss from AI agent failures: UNKNOWN. Incident databases rarely record losses, as the Resemble AI caveat in 1.5 also shows. ###### 6.3 AI: what is going wrong and what is missing - **Going wrong:** agents act with human-level permissions but without human-level checks (PocketOS: broad token, no confirmation, backups in the same place); AI output is shipped without verification (Deloitte); attackers already run agents. Sources above. - **Missing:** identity for each agent separate from its human owner; scoped, time-limited permissions; a confirmation or second check before destructive actions; an audit trail of what the agent saw and did; loss reporting. These are drawn from the failures above; they are our reading, not a regulator's list. --- ##### 7. Synthesis Read the last two columns as **our proposals**, not evidence. "Glass-box" means every decision shows its inputs, rules, confidence and who or what made it. "Agent-native" means AI agents are first-class customers and staff with their own identity, permissions and record, rather than hidden inside a human's account. No EDGE application has yet been shown to reduce any of these costs. | Problem | Annual cost (source) | Who bears it | Root cause | Which app addresses it | What a glass-box, agent-native approach would change | |---|---|---|---|---|---| | UK payment fraud | £1.28bn stolen in 2025, measured ([UK Finance](https://www.ukfinance.org.uk/system/files/2026-06/UK%20Finance%20Fraud%20Report%202026.pdf)); £14.4bn full cost to individuals and businesses, estimate ([GOV.UK](https://www.gov.uk/government/publications/economic-and-social-cost-of-fraud-2023-to-2024)) | Victims; banks through reimbursement; society | Social engineering that starts online (66%) and telecoms (17%); no shared view of who is being paid | EDGE Finance, on the EDGE ID identity layer | Verified identity and confidence shown on both sides of a payment, including agents acting for people; reasons for a hold shown to the customer, not hidden | | APP fraud | £576.4m in 2025, measured ([UK Finance](https://www.ukfinance.org.uk/system/files/2026-06/UK%20Finance%20Fraud%20Report%202026.pdf)) | Victims, then banks (88% of in-scope losses reimbursed, [PSR](https://www.psr.org.uk/information-for-consumers/app-scams-reimbursement-dashboard/)) | Victim is persuaded to authorise; liability sits only with banks | EDGE Finance | Explained, real-time warnings tied to the specific payee's record; evidence trail a bank, platform and victim can all read | | Identity fraud, synthetic identities, mules | 444,000+ cases in 2025; £2.4bn prevented, estimate ([Cifas](https://www.cifas.org.uk/newsroom/fraudscape2026)); synthetic identity cost UNKNOWN | Lenders, banks, impersonated people, young people recruited as mules | Documents and selfies can be faked; accounts opened without lasting proof of personhood | EDGE ID with EDGE Finance | Identity as a checked, reusable record with provenance, rather than a one-off document check; mule patterns explained to the account holder | | Deepfake and spoofing | about $1.3bn documented globally in 2025, a floor, estimate ([Resemble AI](https://resemble.ai/2025-deepfake-threat-report)); UK total UNKNOWN | Businesses, individuals | Voice and video can no longer prove who is speaking | EDGE ID, Judiverse (office and assistants) | Out-of-band, signed confirmation of who or which agent is asking, shown in the meeting or message | | Crypto theft | $3.41bn stolen in 2025, measured ([The Block / Chainalysis](https://www.theblock.co/post/382477/crypto-hack-2025-chainalysis)) | Exchanges, protocols, holders | Key and signing compromise; cross-chain bridges | EDGE Finance | Signing policies and approvals visible and enforced per person and agent; UNKNOWN whether this would have stopped specific hacks | | Sanctions evasion via crypto | $158bn illicit flows in 2025, estimate ([TRM Labs](https://www.trmlabs.com/resources/blog/trms-new-crypto-crime-report-shows-illicit-flows-hit-a-record-usd-158b-in-2025)); UK cost UNKNOWN | Governments, compliant firms | Stablecoins and networks built to move value around sanctions | EDGE Finance, EDGE Public Consortium | Counterparty records with sources and confidence, auditable by supervisors | | Financial crime compliance and poor customer data | £38.3bn spent in 2023, estimate ([LexisNexis](https://risk.lexisnexis.co.uk/about-us/press-room/press-release/20240718-fighting-financial-crime-and-fraud)); fines of £21m-£44m per bank 2025 ([Monzo](https://www.mortgagesolutions.co.uk/news/2025/07/08/fca-slams-monzos-lack-of-financial-crime-controls-issuing-21m-fine/), [Nationwide](https://www.mishcon.com/news/fca-imposes-44-million-penalty-on-nationwide-for-anti-money-laundering-failures)) | Firms and, through prices, customers | Customer records incomplete and not kept current as firms grow | EDGE Finance | Each customer fact carries its source, date and checker, so gaps show up before the regulator finds them | | Large cyber attacks | JLR about £1.9bn, estimate ([Express & Star](https://www.expressandstar.com/business/jaguar-land-rover-cyber-attack-cost-the-uk-ps19-billion-say-experts-5660092)); M&S and Co-op £270-440m, estimate ([TechRadar](https://www.techradar.com/pro/security/m-and-s-and-co-op-hacks-publicly-defined-as-a-single-attack-and-could-cost-more-than-gbp400-million)) | Companies, about 5,000 suppliers, workers, customers | Stolen credentials, help-desk social engineering, flat networks, slow recovery | EDGE Public Consortium (cyber governance) | Every human and agent action on critical systems tied to a verified identity and logged; recovery plans tested and visible to the board | | Unknown and legacy assets | Monetary cost UNKNOWN; 228+ government legacy systems of unknown vulnerability ([NAO](https://www.nao.org.uk/press-releases/cyber-threat-to-uk-government-is-severe-and-advancing-quickly-spending-watchdog-finds/)) | Government and every organisation running them | No live inventory; skills gaps (one in three government cyber roles vacant or temporary) | EDGE Public Consortium | Asset and supplier register as a living, owned record with confidence and last-checked date, so known flaws can be found and patched quickly | | Ransomware | $820m paid on-chain in 2025 ([The Register](https://www.theregister.com/2026/02/27/ransomware_chainalysis/)); about $1.7m average recovery cost, estimate ([Sophos](https://assets.sophos.com/X24WTUEQ/at/jbww7pmb8n3gp99wr6hfq4/sophos-state-ransomware-report-2026.pdf)) | Victims, insurers, patients (Synnovis) | Known flaws, weak passwords, supplier access | EDGE Public Consortium | Supplier access granted per task and time-limited, with a record of who used it | | NHS waiting list | 7.4m pathways waiting, August 2026 ([NHS England](https://www.england.nhs.uk/statistics/wp-content/uploads/sites/2/2026/10/Aug26-RTT-statistical-press-notice-PDF-389K-vjgRdG.pdf)); annual cost UNKNOWN | Patients, carers, employers, NHS | Demand above capacity; capacity lost to missed slots | EDGE Clinic | Patients see where they are and why; agents help rebook and fill cancelled slots under clinical rules that are visible | | Missed appointments | 16m GP (2025) and 8.1m outpatient (2024-25) missed ([EMJ](https://www.emjreviews.com/general-healthcare/news/one-in-four-miss-nhs-appointments/); [NHS outpatient data mirror](https://nhsd-proxy.openprescribing.net/data-and-information/publications/statistical/hospital-outpatient-activity/2024-25)); national cost UNKNOWN, roughly £1.0-1.4bn by our own arithmetic | NHS, other patients | Forgetting, lateness, hard rebooking | EDGE Clinic | Reminders and rebooking by an assistant acting with the patient's consent, logged so the practice can see what happened | | Late diagnosis | 58.7% of common cancers found early vs 75% goal ([PSL hub](https://www.pslhub.org/news/cancers-getting-diagnosed-at-earlier-stage-r7810/)); cost UNKNOWN | Patients and families; NHS (later-stage treatment) | People delay contacting GPs; access is hard | EDGE Clinic (early detection) | Symptom and risk prompts with their reasoning and confidence shown, always routing to a clinician, never replacing one | | Preventable illness (obesity) | up to £126bn, estimate ([Nesta](https://www.nesta.org.uk/report/the-economic-and-productivity-costs-of-obesity-and-overweight-in-the-uk/)) | People, NHS (£12.6bn), employers | Environment and behaviour; little sustained support | EDGE Clinic (social prescribing) | Referrals and outcomes recorded end to end, so what works can be measured | | Sight loss | £29bn wider impact in 2026, estimate ([RNIB](https://data.rnib.org.uk/articles/5)) | People, carers (informal care), economy | Late detection, eye care capacity | EDGE Clinic (early detection) | Screening results and referrals traceable from test to treatment | | Loneliness | 7% often or always lonely ([GOV.UK](https://gov.uk/government/statistics/community-life-survey-202425-annual-publication/community-life-survey-202425-loneliness-and-support-networks)); £9,976 per person wellbeing value, estimate ([GOV.UK](https://www.gov.uk/government/publications/loneliness-monetisation-report)) | Individuals, NHS, employers | Weak links to community support | EDGE Clinic (social prescribing) | Link-worker and community records joined to outcomes, filling the evidence gap the reviews describe | | Public-sector fraud and error | £55-81bn in 2023-24, estimate ([NAO](https://www.nao.org.uk/overviews/the-impact-of-fraud-and-error-on-public-funds-2023-24)) | Taxpayers | Weak controls at payment; limited data sharing | EDGE Public Consortium | Each payment linked to verified claimant, contract and approver, with the evidence visible to auditors | | Major projects at risk | £244bn of red-rated whole-life cost in 2025-26 ([LexisNexis news](https://www.lexisnexis.co.uk/legal/news/nista-publishes-annual-report-on-government-major-projects-delivery)) | Taxpayers, service users | Optimism, weak assurance, poor information | EDGE Public Consortium (procurement) | Assurance records with stated confidence and sources, readable across departments | | Defence affordability gap | £42.5bn over ten years to 2033 ([UK Parliament](https://committees.parliament.uk/work/9520/)) | MoD, armed forces | Plans larger than budget | EDGE Public Consortium (defence and logistics) | Cost and readiness data with provenance; the effect on any gap is UNKNOWN | | Security vetting delays | about £17m a year inefficiency (2023 estimate, [NAO](https://www.nao.org.uk/press-releases/investigation-into-the-performance-of-uk-security-vetting/)); current data UNKNOWN | Departments, defence suppliers, applicants | Ageing IT, demand spikes | EDGE Public Consortium (vetting) | A reusable, consented, verified clearance record, with each step and its status visible to the applicant | | University finances and research | 36% of providers in deficit 2024-25 ([PIE News](https://thepienews.com/ofs-english-universities-warned-of-over-optimism-as-a-third-face-deficits/)); research shortfall about £5.49bn ([THE](https://www.timeshighereducation.com/node/744536)) | Universities, staff, students | Research funded below cost; reliance on falling international fees | EDGE Public Consortium (education and research) | Transparent cost and outcome records; the effect on funding is UNKNOWN | | Research not commercialised | Lost value UNKNOWN; no £75m+ spinout round funded by UK investors alone since 2022 ([Bidwells](https://www.bidwells.co.uk/insights-reports-events/capital-clusters-and-the-scale-up-gap/)) | UK economy, researchers | Scale-up capital gap | EDGE Public Consortium, Judiverse (agent lab) | Verified records of research, rights and readiness that investors can check | | AI agent failures and missing assurance | Total loss UNKNOWN; 362 incidents in 2025 ([Stanford HAI](https://hai.stanford.edu/ai-index/2026-ai-index-report)); shadow AI adds about $670,000 per breach, estimate ([IBM](https://www.ibm.com/think/x-force/2025-cost-of-a-data-breach-navigating-ai)) | Organisations deploying agents, their customers | Agents run with broad human credentials, no confirmation step, no per-agent record | Judiverse (assistants and agent lab), EDGE ID | Each agent has its own identity, scoped permissions, confirmation before destructive actions, and a full record of what it saw and did | --- ##### 8. What I could not find (UNKNOWN), for the next researcher - UK totals for deepfake fraud and for synthetic-identity fraud. - UK-specific cost of sanctions evasion; the National Crime Agency's recent results (search budget ran out). - A reliable cost of poor customer data in finance. - Financial Times and Reuters coverage of 2025-2026 attacks and AI failures (not reached; no FT or Reuters figures are used here). - National cost of missed NHS appointments (government says it does not hold the data) and of late diagnosis. - Any 2026 UK cyber attack with a published cost; Collins Aerospace and Fairlife totals. - Current UKSV vetting backlog and turnaround times (latest found are from the 2023 NAO and PAC reports). - 2025-2026 NAO findings on defence logistics and a government-wide procurement waste total. - A money total for losses from AI agent failures. ### research/market/FUNDING-AND-PROCUREMENT.md `research/market/FUNDING-AND-PROCUREMENT.md` · sha256 `6d600e418a50bae9bdf5440ee50f3ab677229251d7b21dc5d367b4f08a4f9753` #### What is actually being funded and bought: grants, contracts and venture capital Research for JudeAI | EDGE ID, written 31 October 2026 for a brand-new engineer. EDGE ID is governed intelligence: people, organisations and AI agents are all real customers and staff. Its four applications are EDGE Finance (an agent-native bank), EDGE Clinic (social prescribing and early detection), EDGE Public Consortium (cyber governance, procurement, vetting, defence and government logistics, education and university research) and Judiverse (office, AI assistants, agent lab). This is research, not financial, legal or bidding advice. It shows where money is flowing now. It does not show that any EDGE application would win a grant, a contract or an investment. The companion file `COSTLIEST-PROBLEMS.md` in this folder covers where the losses and harms are; this file covers who is paying to fix them. ##### How to read this file | Label | Meaning | |---|---| | **OBSERVED** | I read it myself in a primary source during this research: the funder's or buyer's own page, or an official open-data service (the Innovation Funding Service, the UKRI Gateway to Research database, the Contracts Finder and Find a Tender data services, the EU TED search service). | | **CLAIM** | It comes from a secondary source (news, trade press, a law firm, a supplier, an aggregator) or from a search summary of a page I did not open myself, or it is what a company says about itself. Not independently checked. | | **UNKNOWN** | I looked and could not find it, or sources disagree too much to choose. | Every link was accessed on 2026-10-31. Money is shown in the currency the source used. "Up to" means a ceiling, not money spent. **Acronyms used often.** UKRI = UK Research and Innovation (the national research and innovation funder). Innovate UK = UKRI's business innovation agency. ARIA = Advanced Research and Invention Agency. AISI = AI Security Institute (UK government). DSIT = Department for Science, Innovation and Technology. DASA = Defence and Security Accelerator, now part of UKDI = UK Defence Innovation. SBRI = Small Business Research Initiative (government buying research and development from companies). NHS = National Health Service. NHS SBS = NHS Shared Business Services (an NHS buying organisation). CCS = Crown Commercial Service, which became GCA = Government Commercial Agency on 1 April 2026. DPS = dynamic purchasing system (a standing list of approved suppliers that buyers run mini-competitions through). SME = small or medium-sized enterprise. TRL = technology readiness level (1 = idea, 9 = proven in use). TED = Tenders Electronic Daily (the EU's public tender journal). SAM.gov = the US federal contract opportunities site. SBIR / STTR = US Small Business Innovation Research / Small Business Technology Transfer programmes. VC = venture capital. ###### Method and limits - **Grants:** I read the live competition list on the Innovation Funding Service (pages 1 and 2 of 29 competitions; page 3 did not return its listings), opened the key competition pages, and queried the UKRI Gateway to Research database through its public data service for twelve themes. - **UK contracts:** I queried the Contracts Finder data service for 18 keywords, open and awarded notices published 1 April to 31 October 2026, taking up to the 100 best keyword matches per search. Contracts Finder's keyword search is loose, so results were filtered by title. I paged through about 20,000 Find a Tender records updated 4 September to 30 October 2026 (the service limited my request rate) and filtered titles. Neither sample is a full census. The public notice pages on both sites block automated reading, so the links below go to the public notice pages, and I checked the details through the official data services. - **EU contracts:** I queried the TED search service. Where a notice did not state its currency, I give the number as recorded and say the currency was not checked. - **US:** I could not query SAM.gov (it needs a registered key) and found no US agent-identity solicitation. The US section relies on trade press and law-firm notes, so it is mostly CLAIM. - **Venture capital:** Crunchbase News quarterly articles, Beauhurst, HolonIQ, Defense News and TechCrunch pages were read directly. Many single-company rounds come from search summaries and are marked CLAIM. PitchBook, Dealroom and Rock Health primary reports were not opened. Fintech Futures, Helsing's newsroom, Digital Health and the ARIA quarterly transparency file (a PDF that would not decode) could not be read. --- ##### 0. The ten things to know first 1. **Innovate UK has £15 million open right now for SMEs deploying AI with a real deployer, including a £5 million strand for AI that "monitor, test, verify or evidence compliance"**, closing 8 December 2026. OBSERVED. [UKRI: BridgeAI Deployment Catalyst, AI for Assurance and Compliance](https://www.ukri.org/opportunity/bridgeai-deployment-catalyst-ai-for-assurance-and-compliance/), accessed 2026-10-31. 2. **ARIA's Scaling Trust programme (nearly £50 million) funds tools for AI agents to interact securely; its current proposal cut-off is today, 31 October 2026**, and later applications roll into the next quarterly cycle. OBSERVED. [ARIA: Scaling Trust funding](https://www.aria.org.uk/opportunity-spaces/trust-everything-everywhere/scaling-trust/funding), accessed 2026-10-31. 3. **UK public buyers are already paying for agentic AI.** The Home Office awarded two £3 million contracts in June-July 2026 to design, build and assure AI agents on its Microsoft stack; the Financial Conduct Authority paid £624,696 for "development support for Agentic AI". OBSERVED. [Contracts Finder: Home Office Co Pilot Agentic Stack](https://www.contractsfinder.service.gov.uk/Notice/22eb09e9-74c8-4c8e-90d8-453dfd05838d); [Contracts Finder: FCA Agentic AI](https://www.contractsfinder.service.gov.uk/Notice/71f13f64-750d-4790-b895-5789f4e2f06f), accessed 2026-10-31. 4. **Most UK public AI money still goes on consultancy, training and note-taking tools, not on governed agent platforms.** In my sample of about 90 AI-titled contract awards (April-October 2026, about £42 million), the largest groups by value were strategy, consulting and delivery support, then agents and assistants, then transcription. OBSERVED (my count from the Contracts Finder data service; sample, not census). [Contracts Finder](https://www.contractsfinder.service.gov.uk/), accessed 2026-10-31. 5. **Identity and identity-fraud work is bought at scale:** HMRC (the tax authority) awarded £20 million for identity-fraud prevention delivery support; the Home Office £3.06 million for identity checking; DVLA (the vehicle licensing agency) £2.25 million for a customer identity platform. OBSERVED. [Contracts Finder: HMRC](https://www.contractsfinder.service.gov.uk/Notice/5add23e7-2df5-441c-a60e-6894f8325b4a); [Home Office](https://www.contractsfinder.service.gov.uk/Notice/4e4a9a81-f998-4c18-acb5-1be72b68ebe9); [DVLA](https://www.contractsfinder.service.gov.uk/Notice/0abb34c3-01ac-4fd9-8d67-0b89981b545f), accessed 2026-10-31. 6. **The NHS has built national routes to buy AI:** a £900 million Healthcare AI Solutions framework (published May 2026, awards expected 2027) and a £150 million ambient voice framework. CLAIM (trade press). [Digital Health: £900m framework](https://www.digitalhealth.net/2026/05/21/nhs-sbs-launches-900m-healthcare-ai-solutions-framework/); [Digital Health: £150m AVT framework](https://www.digitalhealth.net/2025/10/nhs-sbs-launches-150m-framework-to-accelerate-access-to-avt/), accessed 2026-10-31. 7. **Germany's federal cyber agency (BSI) is tendering for test requirements for LLM-based agentic AI**, the clearest public-sector demand for agent assurance found in Europe. OBSERVED. [TED notice 584449-2026](https://ted.europa.eu/en/notice/-/detail/584449-2026), accessed 2026-10-31. 8. **AI took $102 billion, 64% of global venture funding, in Q3 2026** ($159 billion total); a record 27 companies raised $1 billion-plus rounds. OBSERVED. [Crunchbase News, 5 October 2026](https://news.crunchbase.com/venture/q3-2026-global-startup-funding-ai-billion-dollar-rounds-exits-data/), accessed 2026-10-31. 9. **Investors are paying for agent identity and agent governance, not just agents:** Oak raised a $60 million seed for one identity system for people, machines and AI agents; London's Geordie AI raised a $30 million Series A for AI agent governance; cybersecurity rounds in Q3 2026 were led by identity and AI-agent data security. OBSERVED / CLAIM as marked in section 3. [The Next Web: Oak](https://thenextweb.com/news/oak-60m-seed-ai-native-identity-platform); [Crunchbase News: cyber Q3 2026](https://news.crunchbase.com/cybersecurity/q3-2026-large-venture-investments-ai-defense-startups/), accessed 2026-10-31. 10. **Fintech venture funding is shrinking while defence is booming.** Fintech raised $10.6 billion in Q3 2026 across 687 rounds, the fewest since 2018; Helsing raised $1.8 billion and Anduril $5 billion in 2026. OBSERVED. [Crunchbase News: fintech Q3 2026](https://news.crunchbase.com/venture/q3-2026-fintech-startup-funding-deal-count-fell-ai/); [Defense News: Helsing](https://www.defensenews.com/global/europe/2026/07/13/helsing-raises-18-billion-in-europes-biggest-defense-startup-round/); [TechCrunch: Anduril](https://techcrunch.com/2026/05/13/anduril-raises-5b-doubles-valuation-to-61b/), accessed 2026-10-31. --- ##### 1. UK public innovation funding ###### 1.1 Innovate UK competitions open or closing soon (as listed on 31 October 2026) All rows OBSERVED from the [Innovation Funding Service competition list](https://apply-for-innovation-funding.service.gov.uk/competition/search) and [page 2](https://apply-for-innovation-funding.service.gov.uk/competition/search?page=2), accessed 2026-10-31, unless marked. | Competition | Pot | Opens / closes | What it pays for | Fit | Source (accessed 2026-10-31) | |---|---|---|---|---|---| | BridgeAI Deployment Catalyst: AI for Assurance and Compliance | £5m (Phase 1) | 21 Oct / 8 Dec 2026, 11am | SME-led consortia (2-3 organisations) deploying an existing AI system at TRL 5-6 that "monitor, test, verify or evidence compliance" with standards or controls; must include a medium or large UK deployer. Up to £52m expected across three phases (CLAIM, aggregator). | Strong: all four apps, especially EDGE Finance and Public Consortium | [UKRI opportunity page](https://www.ukri.org/opportunity/bridgeai-deployment-catalyst-ai-for-assurance-and-compliance/); [IFS competition 2597](https://apply-for-innovation-funding.service.gov.uk/competition/2597/overview/7ec02a95-73e5-4928-8104-67a9f1b279e3); £52m: [Innovate UK Business Connect](https://iuk-business-connect.org.uk/?p=138707) | | BridgeAI Deployment Catalyst: AI for Decision Critical Systems | £5m | 21 Oct / 8 Dec 2026 | AI that supports consequential judgements, recommendations or decisions | Strong: EDGE Finance (credit, fraud), EDGE Clinic (triage) | [IFS competition 2598](https://apply-for-innovation-funding.service.gov.uk/competition/2598/overview/cfb35198-55fc-4aad-afa1-86a260bfaddf) | | BridgeAI Deployment Catalyst: AI for Operational Systems | £5m | 21 Oct / 8 Dec 2026 | AI that performs, controls or coordinates operational activity | Medium: Public Consortium logistics, Judiverse | [IFS competition 2599](https://apply-for-innovation-funding.service.gov.uk/competition/2599/overview/1931f35c-87fe-4b6a-a718-a61a2d00e86a) | | Efficient, Structured and Controllable AI Systems | £1.66m; projects £50k-£150k costs, 1-3 months | 12 Oct / 18 Nov 2026 | Single SMEs validating a novel learning mechanism, including one that "exerts measurable control over an agent". Generic agent orchestration is excluded. Start TRL 1-3. | Narrow: Judiverse agent lab only if there is a genuinely new control mechanism | [IFS competition 2582](https://apply-for-innovation-funding.service.gov.uk/competition/2582/overview/9fdb47a5-011c-4c47-ac5e-c426e295fe89) | | Advancing innovation in accelerated cancer diagnosis | £25.4m | 1 Sep / 4 Nov 2026 | Clinical validation of early cancer diagnostic technologies; collaborations led by a business, university or NHS body, with at least one SME | Medium: EDGE Clinic only with a clinical diagnostic partner | [IFS competition 2527](https://apply-for-innovation-funding.service.gov.uk/competition/2527/overview/d48f61f8-2bd5-4067-9c05-29dd08487623) | | Knowledge Transfer Partnership (KTP) 2026-27 Round 4 | £12.5m | 19 Oct 2026 / 27 Jan 2027 | University-business partnerships, led by the university | Medium: any app with a university partner | [IFS competition list](https://apply-for-innovation-funding.service.gov.uk/competition/search) | | Commercialising Knowledge Assets Fund (CKAF) Winter | grants £50k-£250k | 17 Sep / 1 Dec 2026 | Public sector research bodies commercialising what they own | Low (we are not a public body) | [IFS competition 2579](https://apply-for-innovation-funding.service.gov.uk/competition/2579/overview/3eaba59f-330d-45f6-a406-78067a71bb70) | | UK-Singapore Collaborative R&D 2026 | £3m | 24 Aug / 16 Dec 2026 | Joint UK-Singapore projects | Low-medium (theme not checked) | [IFS competition 2526](https://apply-for-innovation-funding.service.gov.uk/competition/2526/overview/4a4b67b9-9b73-4816-8e21-bdfab83f5909) | | CfI: Battery innovation for dual use and defence Phase 1 | share of £2m | 28 Sep / 4 Nov 2026 | Defence batteries | None | [IFS competition 2587](https://apply-for-innovation-funding.service.gov.uk/competition/2587/overview/aafc825c-421e-47ea-bd56-d5d1cd50e513) | - No open Innovate UK competition on cyber, fraud, education or social prescribing appeared on the two list pages I could read. OBSERVED (absence on pages 1-2 only). [IFS list](https://apply-for-innovation-funding.service.gov.uk/competition/search), accessed 2026-10-31. - Earlier 2026 AI competitions, now closed: AI Champions: Frontier AI Phase 1 (up to £3m, closed 29 April 2026), Frontier AI Discovery (at least £2.5m, £25k-£50k projects, closed 10 June 2026), Sovereign AI Proof of Concept (at least £1.6m). CLAIM (aggregators and listings). [Inventya: AI Champions](https://www.inventya.com/calls/ai-champions-frontier-ai-phase-1); [fundsforNGOs: Frontier AI Discovery](https://www2.fundsforngos.org/innovation/frontier-ai-discovery-fund-for-machine-learning-and-foundation-models-uk/); [Find a Grant: Sovereign AI Proof of Concept](https://www.find-government-grants.service.gov.uk/grants/sovereign-ai---proof-of-concept-1), accessed 2026-10-31. ###### 1.2 What UKRI has already funded (Gateway to Research) Gateway to Research (GtR) is UKRI's public database of funded projects. Values below are the project fund values recorded in GtR. All rows OBSERVED through the [GtR data service](https://gtr.ukri.org/), accessed 2026-10-31. Project links use `https://gtr.ukri.org/projects?ref=`. **Counts.** A search for the exact phrase "agentic AI" returned only 8 projects in all of GtR; "AI assurance" 24; "social prescribing" 71; "digital identity" 38; "defence logistics" 3 (all old and off-topic). OBSERVED. [GtR](https://gtr.ukri.org/), accessed 2026-10-31. Agentic AI is barely present in the UK research record so far. | Theme | Project (lead, funder, dates) | Value | Source (accessed 2026-10-31) | |---|---|---|---| | Agent guardrails | AgentVault: Homomorphic Guardrails for Autonomous AI Agents (Nottingham Trent University; Innovate UK; 31 Mar-30 Jul 2026) | £31,879 | [GtR 10192287](https://gtr.ukri.org/projects?ref=10192287) | | Agent cyber defence | AgenticDeviceShield: autonomous, explainable cyber defence on mobile devices (University of Lancashire; Innovate UK; 2026) | £27,581 | [GtR 10191330](https://gtr.ukri.org/projects?ref=10191330) | | Agents in finance | Modular AI Agents for Dealflow, Screening and Decision Support in Impact Investing (Lobaso Ltd; Innovate UK; Mar 2026-Sep 2028) | £305,780 | [GtR 10190236](https://gtr.ukri.org/projects?ref=10190236) | | Multi-agent maths | How to Train Systems of LLM Agents (EPSRC, the engineering research council; 2025-26) | £26,600 | [GtR UKRI2869](https://gtr.ukri.org/projects?ref=UKRI2869) | | AI assurance and insurance | AI^2: Assurance and Insurance for Artificial Intelligence (EPSRC; Mar 2025-Mar 2028) | £1,050,364 | [GtR UKRI567](https://gtr.ukri.org/projects?ref=UKRI567) | | Small AI assurance grants | e.g. "AI Testing Innovation", "AI Conversational Alignment Assurance", "Automated Ethical AI Assurance Service" (Innovate UK grants of about £49,000-£50,000 each, now closed) | ~£50k each | [GtR search "AI assurance"](https://gtr.ukri.org/search/project?term=%22AI+assurance%22) | | Identity and deepfakes | KYCSafe: AI deepfake detection for UK financial services eKYC (electronic know-your-customer) (University of Lancashire; Innovate UK; 2026) | £28,758 | [GtR 10193642](https://gtr.ukri.org/projects?ref=10193642) | | Identity and deepfakes | SENTINEL: securing age estimation and digital IDs against presentation and injection attacks (Innovate UK; closed) | £398,236 grant | [GtR search "AI assurance"](https://gtr.ukri.org/search/project?term=%22AI+assurance%22) | | Identity research network | SPRITE+ 2: Security, Privacy, Identity and Trust Engagement NetworkPlus (EPSRC; 2023-27) | £3,115,833 | [GtR EP/W020408/1](https://gtr.ukri.org/projects?ref=EP%2FW020408%2F1) | | Fraud | Aston University and Domestic and General Insurance KTP (fraud detection; 2025-28) | £184,266 | [GtR 10131463](https://gtr.ukri.org/projects?ref=10131463) | | Cyber | Centre for Secure Information Technologies (CSIT) Phase 3 (EPSRC; 2022-27) | £4,234,763 | [GtR EP/X022323/1](https://gtr.ukri.org/projects?ref=EP%2FX022323%2F1) | | Early diagnosis | STRATIFYHF: AI decision support for risk stratification and early diagnosis of heart failure (Cambridge; Horizon Europe guarantee; 2023-27) | £383,538 (UK share) | [GtR 10079256](https://gtr.ukri.org/projects?ref=10079256) | | Social prescribing | SPIRITED Social Prescribing: improving implementation research (ESRC, the social science research council; 2025-27) | £249,055 | [GtR UKRI1717](https://gtr.ukri.org/projects?ref=UKRI1717) | | Social prescribing | The Abundance Project: cultural and green inclusion in social prescribing, south-west London (AHRC, the arts and humanities council; 2024-27) | £1,056,814 | [GtR AH/Z505481/1](https://gtr.ukri.org/projects?ref=AH%2FZ505481%2F1) | | Social prescribing | Collaborating to deliver social prescribing in Bath and North East Somerset (Innovate UK; closed) | £217,219 grant | [GtR search "social prescribing"](https://gtr.ukri.org/search/project?term=%22social+prescribing%22) | | Education | AI tutoring grants, e.g. Schola.AI intelligent online classroom (£336,700) and Plytime AI Tutoring (£142,878) (Innovate UK; closed) | as shown | [GtR search "AI tutor"](https://gtr.ukri.org/search/project?term=%22AI+tutor%22) | | Healthcare agents | Frontier: a generative AI healthcare "workforce" (Frontier AI Ltd; Innovate UK; 2025-27) | £497,094 | [GtR 10146169](https://gtr.ukri.org/projects?ref=10146169) | - **Pattern.** Most UK agentic AI awards are tiny (£25,000-£50,000 feasibility grants from early 2026); the larger money sits in AI assurance research (£1m+), identity networks (£3m+) and social prescribing research led by the arts and social science councils. OBSERVED (from the rows above). [GtR](https://gtr.ukri.org/), accessed 2026-10-31. - Defence logistics returned no relevant UKRI projects. UNKNOWN whether this work is funded through defence routes not in GtR. [GtR search "defence logistics"](https://gtr.ukri.org/search/project?term=%22defence+logistics%22), accessed 2026-10-31. ###### 1.3 ARIA (Advanced Research and Invention Agency) | Item | What | Status / dates | Label | Source (accessed 2026-10-31) | |---|---|---|---|---| | Scaling Trust | "Nearly £50m" programme for infrastructure so AI agents can securely coordinate, negotiate and verify each other. Track 1 Multi-Agent Security Arena (partner procurement, closed); Track 2 tooling (open-source agents, reusable components); Track 3 fundamental research (formal AI security, cyber-physical primitives, generative security) | Tracks 2-3 rolling; current cycle opened 4 Sep 2026, cut-off **31 Oct 2026**, decisions 30 Nov 2026; "a small number of projects" | OBSERVED | [ARIA Scaling Trust funding](https://www.aria.org.uk/opportunity-spaces/trust-everything-everywhere/scaling-trust/funding) | | Scaling Trust award size | £100k-£3m per project, up to 18 months, no match funding | earlier call | CLAIM | [Innovate UK Business Connect listing](https://iuk-business-connect.org.uk/opportunities/aria-scaling-trust-tooling-and-fundamental-research-rolling-call) | | Safeguarded AI | £59m programme on mathematical guarantees for AI; TA2 cybersecurity blue teams £2.5m-£3.5m each, closed 1 July 2026 | marked open on ARIA's navigation; details not read | CLAIM | [ARIA TA1.4 call PDF](https://aria.org.uk/media/bmml5lgd/aria-safeguarded-ai-ta14-call-for-proposals.pdf); [Innovate UK Business Connect: Safeguarded AI cybersecurity](https://iuk-business-connect.org.uk/?p=137264) | | Rolling R&D call (Scaling Trust and Safeguarded AI) | £10m, August-December 2026 | from ARIA's Q1 2026-27 transparency file, which I could not decode | CLAIM | [ARIA transparency PDF](https://aria.org.uk/media/311jm0yi/aria-end-of-q1-fy26-27.pdf) | | Multi-agent AI safety joint call | $10m from Google DeepMind, Schmidt Sciences, Cooperative AI Foundation, ARIA and Google.org, announced 11 June 2026 | window June-August 2026 (closed) | CLAIM | [GrantedAI listing](https://grantedai.com/grants/scaling-ai-safety-for-a-multi-agent-world-schmidt-sciences-google-deepmind-advance-8c081d0e) | ###### 1.4 AI Security Institute grants - **The Alignment Project** has £27 million in total (including £12 million new money announced with the first round) and funded 60 projects from more than 800 applications; the next round was expected to open in summer 2026. OBSERVED. [AISI: Funding 60 projects](https://www.aisi.gov.uk/blog/funding-60-projects-to-advance-ai-alignment-research), accessed 2026-10-31. Partners include OpenAI (£5.6 million), Microsoft, Anthropic, AWS, UKRI and ARIA. OBSERVED, same source. - Alignment Project grants in 2025 ranged from £50,000 to £1,000,000. CLAIM (search summary of the programme site). [Alignment Project site](https://alignmentproject.aisi.gov.uk/), accessed 2026-10-31. - Whether the 2026 round is open today: UNKNOWN. I did not find a confirmed date. [Alignment Project site](https://alignmentproject.aisi.gov.uk/), accessed 2026-10-31. - **AISI Challenge Fund**: £5 million, up to £200,000 per project, for academic and non-profit researchers; closed 1 August 2025. CLAIM (search summary of AISI's grants page). [AISI grants](https://aisi.gov.uk/grants); [GOV.UK press release](https://www.gov.uk/government/news/cutting-edge-research-on-ai-security-bolstered-with-new-challenge-fund-to-ramp-up-public-trust-and-adoption), accessed 2026-10-31. A 2026 round: UNKNOWN. - The department behind AISI also buys services: DSIT awarded £1 million for AI Safety Institute legal services (Dentons, April 2026). OBSERVED. [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/adf73bd5-9ef4-4809-92bc-44b1d5a48644), accessed 2026-10-31. ###### 1.5 Defence innovation (DASA, now UK Defence Innovation) - DASA is now part of UK Defence Innovation (UKDI). The Open Call for Innovation runs in cycles; the 2026 cycles I found closed on 23 June and 29 September 2026. CLAIM (search summary of GOV.UK competition documents). [GOV.UK: DASA Open Call collection](https://www.gov.uk/government/collections/defence-and-security-accelerator-dasa-open-call-for-innovation), accessed 2026-10-31. - UKDI has a ring-fenced budget of at least £400 million a year from 2025-26 and was due to be fully operational by July 2026. CLAIM (secondary and RUSI commentary). [RUSI: Launching UK Defence Innovation](https://www.rusi.org/explore-our-research/publications/commentary/same-same-different-launching-uk-defence-innovation), accessed 2026-10-31. - UKDI Security Open Call: up to £350,000 per project for 12 months (18 with testing), for public safety, national security, border and protective security. CLAIM. [Innovate UK Business Connect listing](https://iuk-business-connect.org.uk/opportunities/uk-defence-innovation-security-open-call/), accessed 2026-10-31. - Commentators said in 2025-26 that UKDI had published few problem-led challenges. CLAIM. [Resilience Media](https://www.resiliencemedia.co/p/five-months-after-launch-uk-defence), accessed 2026-10-31. - Whether a UKDI call is open today: UNKNOWN. [GOV.UK DASA funding competitions](https://www.gov.uk/government/collections/defence-and-security-accelerator-funding-competitions), accessed 2026-10-31. - Defence buying observed in the contract data is mostly estates, energetics and training support, not logistics software; one small AI buy: the Ministry of Defence (MoD) paid £41,000 to Great Wave AI for "an AI software tool for Defence data cohesion" (June 2026). OBSERVED. [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/4c25c4d8-64a5-4c16-a521-dd0f1f46d4f4), accessed 2026-10-31. ###### 1.6 SBRI Healthcare and NHS AI programmes - The NHS Cancer Programme, with SBRI Healthcare, awarded £14.4 million to seven early cancer detection and diagnosis innovations (July 2025). CLAIM (search summary of SBRI Healthcare news). [SBRI Healthcare news](https://sbrihealthcare.co.uk/news/nhs-cancer-programme-awards-14-4-million-to-accelerate-new-front-line-innovations-that-detect-and-diagnose-cancer-earlier), accessed 2026-10-31. - An open SBRI Healthcare competition on AI early diagnosis today: UNKNOWN (none found). [SBRI Healthcare](https://sbrihealthcare.co.uk/), accessed 2026-10-31. - **NHS SBS Healthcare AI Solutions framework:** £900 million including VAT, published 11 May 2026, eight lots (diagnostics, predictive analytics, robotics, operational efficiency, consultancy, and an end-to-end lot), SMEs eligible, bids due 23 June 2026 (possibly extended), operating May 2027 to May 2035. CLAIM (trade press; one tracker shows £750 million). [Digital Health](https://www.digitalhealth.net/2026/05/21/nhs-sbs-launches-900m-healthcare-ai-solutions-framework/); [Public Technology](https://www.publictechnology.net/2026/05/20/health-and-social-care/nhs-launches-900m-ai-framework/); [Tender Tracker](https://tendertracker.co.uk/contracts/ocds-h6vhtk-0695d5), accessed 2026-10-31. - **Ambient voice technology (AI scribes):** NHS SBS runs a £150 million framework (October 2025 to September 2027). CLAIM. [Digital Health](https://www.digitalhealth.net/2025/10/nhs-sbs-launches-150m-framework-to-accelerate-access-to-avt/), accessed 2026-10-31. NHS England runs a self-certified supplier registry that requires Class 1 medical device status and a DTAC (Digital Technology Assessment Criteria) assessment. OBSERVED. [Find a Tender 009501-2026](https://www.find-tender.service.gov.uk/Notice/009501-2026), accessed 2026-10-31. - **Social prescribing** is funded mainly through primary care link-worker roles; in a July 2026 debate the government cited £3 million to the National Academy for Social Prescribing, and an MP said voluntary-sector projects get "absolutely no funding". CLAIM (Westminster debate record, secondary). [Westminster Brief, 14 July 2026](https://westminsterbrief.co.uk/archive/debate/14-july-2026/social-prescribing-3a8f), accessed 2026-10-31. ###### 1.7 Sovereign AI - The government launched a £500 million Sovereign AI Unit on 16 April 2026, a venture-style fund inside government that also offers supercomputer access, R&D grants and help with government procurement and regulation. OBSERVED (ticket sizes not stated on the page I read). [UKTN, 16 April 2026](https://www.uktech.news/government-policy/uk-government-launches-500m-sovereign-ai-unit-20260416), accessed 2026-10-31. - Typical equity tickets of about £1-10 million (one source says capped at £20 million); first equity investment in Callosum; six other start-ups given up to one million GPU hours each. CLAIM (search summaries of press coverage). [Public Technology](https://www.publictechnology.net/2026/04/16/business-and-industry/governments-500m-sovereign-ai-venture-capital-unit-opens-doors-today-to-invest-in-uk-tech-firms/); [Startups.co.uk](https://startups.co.uk/news/sovereign-ai-funding/), accessed 2026-10-31. - Callosum later raised a $100 million seed round, the largest seed round in Europe in Q3 2026. OBSERVED. [Crunchbase News: Europe Q3 2026](https://news.crunchbase.com/venture/q3-2026-europe-strong-quarter-ai-uk-germany-france/), accessed 2026-10-31. --- ##### 2. Public procurement: what buyers actually tender for ###### 2.1 UK: open opportunities found (deadline after today) All OBSERVED through the official data services; links go to the public notice pages (accessed 2026-10-31). | Buyer | What | Value | Deadline | Fit | Source | |---|---|---|---|---|---| | UKRI (Financial, Professional and Business Services programme) | UKRI-6347 Trusted AI Adoption: position the UK as a leader in trusted AI adoption across financial and professional services | £800,000 | 5 Nov 2026, 2pm | EDGE Finance, Judiverse | [Find a Tender 094267-2026](https://www.find-tender.service.gov.uk/Notice/094267-2026) | | DSIT | Cyber Accelerator Programme: Design Partnerships 2027-28 (link large organisations' problems to cyber SMEs) | £1.2m | 27 Nov 2026, 11am | Public Consortium (cyber governance) | [Find a Tender 103182-2026](https://www.find-tender.service.gov.uk/Notice/103182-2026) | | MoD | DOEA/0084 ShipVetting 2027-2032 (assurance vetting of ocean tankers carrying fuel) | £80,595 | 12 Nov 2026 | Public Consortium (vetting, defence logistics), narrow | [Find a Tender 094205-2026](https://www.find-tender.service.gov.uk/Notice/094205-2026) | | Education Endowment Foundation | Embedded product team for the Alpha and Beta of a Digital Public Service helping teachers combine research with professional judgement | £730,000 | 30 Nov 2026 | Public Consortium (education) | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/612ec727-9bb1-4e65-99f4-1c68f8c05873) | | Suffolk County Council | AI avatar "digital ambassador" for investors | not stated | 26 Nov 2026 | Judiverse (assistants), low | [Find a Tender 102382-2026](https://www.find-tender.service.gov.uk/Notice/102382-2026) | | GCA (formerly CCS) | Cyber Security Services 3 DPS (consultancy, penetration testing, incident response, managed security operations centres) | £800m (estimated, whole DPS) | DPS open to join until 11 Feb 2029 | Public Consortium | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/4c8a6f7a-ea75-4463-b05a-a56c140a7582) | | GCA (formerly CCS) | Open Banking (data, digital payments and Confirmation of Payee) DPS | £800m (estimated) | open to join until 7 Jan 2032 | EDGE Finance | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/d7d25e48-faf9-4d7f-b4df-e6e1660559bd) | | CCS | Spark: Technology Innovation Marketplace DPS | up to £650m | open to join until 13 Feb 2029 | All four | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/b683d618-769a-4bea-90ad-51bda7f70ae9) | | YPO (a public buying organisation) | Critical Incident Management and Public Safety Software and Systems DPS | £50m-£500m | open until 12 Feb 2029 | Public Consortium | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/8ce7627e-2a73-4b30-b16a-40aa41d8e384) | ###### 2.2 UK: who is already paying, and for what (awards, April-October 2026) All OBSERVED through the Contracts Finder and Find a Tender data services (accessed 2026-10-31). Values are awarded values as published. **Agentic AI and assistants** | Buyer | What | Value | Supplier | Source | |---|---|---|---|---| | Home Office | Co Pilot Agentic Stack Design and Development: "discovery, design, development and governance of AI agents" across Microsoft 365 Copilot, Copilot Studio, Azure AI Foundry, Power Platform; Jul 2026-Jun 2028; via Public Sector Resourcing framework | £3,000,000 | Kerv Group | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/22eb09e9-74c8-4c8e-90d8-453dfd05838d) | | Home Office | Co Pilot and Agentic Stack Business Change, Adoption and Assurance | £3,000,000 | FSP Consulting Services | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/3f955519-2cc7-4b75-a86e-49bd6b1359b4) | | Financial Conduct Authority (FCA) | Development support for Agentic AI (Mar-Aug 2026) | £624,696 | Airwalk Consulting Reply | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/71f13f64-750d-4790-b895-5789f4e2f06f) | | FCA | Project Libra: AI testing support | £447,620 | Advai | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/42657d5a-24f9-48c8-aebc-648827417d29) | | FCA | "Procurement AI as a Service" proof of concept; "Procurement AI: role of an evaluator" proof of concept | not stated | Deloitte; Capita | [Find a Tender 095939-2026](https://www.find-tender.service.gov.uk/Notice/095939-2026); [095309-2026](https://www.find-tender.service.gov.uk/Notice/095309-2026) | | Pension Protection Fund | Agentic AI Development Services | £72,000 | Corporate Project Solutions | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/fece287e-1ccb-4809-bb45-d704b3167a26) | | Lancashire Constabulary | Ask Bobbi: AI agent | £330,623 | Softcat | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/27024fa8-e9a9-4240-89e0-bb505aeb52ae) | | Peterborough City Council | Design of AI agents (chatbots) for an Integrated Front Door | £204,042 | Bloom Procurement Solutions | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/9acbf879-909f-4b90-8fad-f51ed3e57603) | | Ipswich Borough Council | Customer services AI software | £540,762 | ICS AI | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/07eae9fa-0793-4f29-9800-3c9ac5e865d9) | | Department for Transport | AI delivery, development and evaluation services | £8,000,000 | Kainos | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/fc1f0ddb-03db-4adb-94cb-fd77825b1370) | | NHS Business Services Authority | AI adoption and implementation services | £5,000,000 | Scrumconnect | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/82149a90-b8ca-47bb-8091-874847f85853) | | Department for Business and Trade | AI³ delivery partners: Lot 1 AI Lab, Lot 2 AI Factory, Lot 3 AI Operations | not stated | The Dot Collective; Version 1; CACI | [Find a Tender 091854-2026](https://www.find-tender.service.gov.uk/Notice/091854-2026) | | Jisc (UK universities' digital body) | OpenAI sector AI evaluation pilot (ChatGPT Edu, Codex, models, APIs) | $7,560,000 award value as published (currency not checked) | OpenAI | [Find a Tender 088540-2026](https://www.find-tender.service.gov.uk/Notice/088540-2026) | **AI governance and assurance** | Buyer | What | Value | Supplier | Source | |---|---|---|---|---| | GCA | Provision of AI Governance | £195,000 | PwC | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/37491cb9-9e18-4d23-af7e-e96fb26661cd) | | HMRC | Recruitment of senior AI Assurance lead for the Chief AI Officer | £22,000 | Global Resourcing | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/1cc7a1c8-4949-4cbe-9ac0-71cd7ccb4d3b) | | DSIT | AI Cyber Security Standards | £84,500 | Accenture | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/df0df041-b05a-4c49-90e0-594593b1e20c) | | UK Shared Business Services (UK SBS) | Cyber assurance review for the i.AI (government AI incubator) platform | £22,000 | Avella Security | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/d9622efd-2766-4013-a214-d514892783e6) | | University of Essex | AI Governance Consultant | £112,320 | Naomi Korn Associates | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/2534ac29-3fd2-4ac2-aa4c-cbbc5e252a30) | | Bluelight Commercial (police buying body) | Police AI interim delivery support | £1,000,000 | Deloitte | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/26bc8a4b-d231-4e17-b5c0-b0f8f1417cb7) | **Identity and fraud** | Buyer | What | Value | Supplier | Source | |---|---|---|---|---| | HMRC | Record and Registration Management: identity fraud prevention and delivery support (Sep 2026-Aug 2028) | £20,000,000 | Ernst & Young | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/5add23e7-2df5-441c-a60e-6894f8325b4a) | | HMRC | Identity Transformation and Transition Readiness | £2,900,000 | Condatis Group | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/74753629-28db-4fed-97a7-576f67cf2014) | | Home Office | Identity checking (Identity Authenticate) | £3,060,000 | Experian | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/4e4a9a81-f998-4c18-acb5-1be72b68ebe9) | | Department for Business and Trade | One Platform, One Login: Business Identity | £2,860,000 | Deloitte | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/6dd5d1e1-666e-4b80-a697-90cdf9ca7922) | | DVLA | Customer Identity Platform | £2,250,000 | SoftwareOne | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/0abb34c3-01ac-4fd9-8d67-0b89981b545f) | | Department for Work and Pensions (DWP) | Get your State Pension: identity verification | £282,140 | TransUnion | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/b5f89580-a0b6-41a2-83ba-fed26308b430) | | Department for Education | Fraud investigation tool | £201,232 | Clue Computing | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/78e5db96-0ceb-420f-8d4c-de32a7ea9d27) | | City of London (Action Fraud) | Cloudflare services for Action Fraud | £394,140 | Cloudflare | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/7259e9d7-ac13-42ab-89b6-e33e6ff16875) | | Ofcom | Research on mitigations against the user journey of fraud | £70,200 | Magenta Research | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/5b1c6b44-86bd-4a7b-a168-10781ca75597) | | York and North Yorkshire Combined Authority | Core-Vet vetting management system | £178,680 | WPC Software | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/e7a33ce4-6a27-4af0-8b31-b255a594bfff) | **Cyber** | Buyer | What | Value | Supplier | Source | |---|---|---|---|---| | Ministry of Housing, Communities and Local Government | Local Government Cyber Security (Oct 2026-Sep 2028) | £30,000,000 | Bridewell Consulting | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/e9d30f07-2719-4ed7-ab82-5f5698cd3af5) | | DSIT | Technical Support for Cyber Uplift (find and fix vulnerabilities); shared arrangement, maximum £14m in total across suppliers | up to £14,000,000 aggregate | 3B Data Security; BAE Systems Applied Intelligence | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/953c29c9-0e3e-4b7c-a609-58b2832f11a9) | | NHS England | Cyber Security Operations Centre engineering content and tooling | £3,700,000 | Hippo Digital | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/358d2d72-acf5-4bb6-bded-02d57d704471) | | Cabinet Office | Cyber threat intelligence suite | £290,000 | Mondas Consulting | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/fb69360a-3ec1-4550-b7c3-d8fe30d839bc) | | City of London | Third-party cyber risk and supply chain assurance platform | not stated | Risk Ledger | [Find a Tender 095200-2026](https://www.find-tender.service.gov.uk/Notice/095200-2026) | **Health, social prescribing and early detection** | Buyer | What | Value | Supplier | Source | |---|---|---|---|---| | NHS England | Adult Screening and Personalisation Digital Delivery Services (Sep 2026-Aug 2029) | £215,163,316 | Netcompany | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/a7e8665f-95fe-4e36-8c1b-7414b8792aac) | | United Lincolnshire Teaching Hospitals | Ambient voice technology (AI scribe) | £3,370,000 | Accurx | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/5f7d6277-d794-45c0-8b24-bb0fa565ec7b) | | NHS Leicester, Leicestershire and Rutland | Workwell and Social Prescribing Platform (referrals, personalised pathways across NHS, councils and voluntary groups) | not stated | Pungo | [Find a Tender 100676-2026](https://www.find-tender.service.gov.uk/Notice/100676-2026) | | Northumberland County Council | AI transcription and summarisation for frontline workers | £313,500 | Beam Up | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/53b2b7ab-5f8e-43d5-ac75-d6cf6c09b37d) | | County Durham and Darlington NHS Foundation Trust | Chest X-ray AI software | £132,608 | Harrison.ai | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/59800630-8cbc-44d3-a1ed-6acd05bccb11) | | Gloucestershire County Council | AI WhatsApp tool for unpaid carers | £134,400 | CC2i | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/84861886-0290-4a76-acc5-a1ee443fd303) | **Education and public-sector platforms** | Buyer | What | Value | Supplier | Source | |---|---|---|---|---| | Bournemouth University | Staff AI solution (SMART Staff Copilot) | £280,000 | ICS.AI | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/4e477480-1e93-4c55-b4a0-7b231fa652c5) | | Engineering Construction Industry Training Board | Learning experience platform and AI conversations | £643,242 | Learning Pool | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/a2c825a6-2181-417e-b061-f1570ee7ddbb) | | Southampton City Council | AI-assisted EHCP (Education, Health and Care Plan) drafting tool | £114,000 | Bramble Hub | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/22a2e2a6-826f-4bdc-8042-22f2da72c63c) | | Department for Culture, Media and Sport | Procurement software solution | £300,000 | Atamis | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/745eef88-f225-49e9-9a11-0653f7656b7e) | **Pattern.** In about 90 AI-titled awards published April-October 2026 (total about £42 million, my own grouping of titles): strategy, consulting, delivery and adoption support took about £20 million; agents, copilots and chatbots about £8.6 million; transcription and ambient scribes about £3.9 million; training and apprenticeships about £2.9 million across 20 awards. OBSERVED (my count from the [Contracts Finder data service](https://www.contractsfinder.service.gov.uk/), accessed 2026-10-31; sample of top keyword matches, not a census, and a few titles may be misclassified). Buyers are paying large consultancies to build agents on platforms they already own; few contracts buy an independent governance or identity layer for agents. ###### 2.3 UK frameworks (how public bodies buy) | Framework | Owner | Scope | Dates | Label | Source (accessed 2026-10-31) | |---|---|---|---|---|---| | RM6200 Artificial Intelligence DPS | GCA (formerly CCS) | AI discovery, implementation, support, end-to-end partnerships; product types include augmented decision making, virtual assistants and (for health and care) imaging and diagnostic AI | 3 Sep 2020 to 23 Feb 2029 (extended Nov 2025) | OBSERVED | [GCA RM6200](https://www.gca.gov.uk/agreements/RM6200) | | RM3764.3 Cyber Security Services 3 DPS | GCA | consultancy, penetration testing (including CHECK, the NCSC scheme), incident management, managed security services; successor RM3764.4 being designed (July 2026) | 14 Feb 2020 to 13 Feb 2029 | OBSERVED | [GCA RM3764.3](https://www.gca.gov.uk/agreements/RM3764.3) | | G-Cloud 15 (RM1557.15) | GCA | cloud software, hosting and support; first iteration as an "open framework" under the Procurement Act 2023; replaces G-Cloud 14 | live from about August 2026; reopens at 18 and 36 months; size claimed £4.8bn (notice estimate) or £14bn (supplier blog) | CLAIM | [GCA G-Cloud 15](https://www.gca.gov.uk/agreements/RM1557.15); [TechMarketView](https://www.techmarketview.com/ukhotviews/archive/2026/08/11/g-cloud-15-launches) | | Name change | CCS became the Government Commercial Agency on 1 April 2026, merging with Cabinet Office central commercial teams | from 1 Apr 2026 | CLAIM | [Public Technology](https://www.publictechnology.net/2026/02/16/business-and-industry/new-government-commercial-agency-to-launch-on-1-april/); [GCA: CCS page](https://www.gca.gov.uk/about-gca/crown-commercial-service-ccs) | | NHS SBS Healthcare AI Solutions | NHS SBS | see section 1.6 | 2027-2035 | CLAIM | [Digital Health](https://www.digitalhealth.net/2026/05/21/nhs-sbs-launches-900m-healthcare-ai-solutions-framework/) | ###### 2.4 European Union (TED) All OBSERVED through the TED search service (accessed 2026-10-31). Links use `https://ted.europa.eu/en/notice/-/detail/`. | Buyer (country) | What | Value | Deadline | Fit | Source | |---|---|---|---|---|---| | BSI, Federal Office for Information Security (Germany) | Project 1000 "Agentic AI: test requirements for LLM-based agentic AI (PRAKI)": identify key agentic use cases, write test requirements, generalise them (re-published several times June-September 2026) | not stated | latest 21 Sep 2026 (closed) | Strong signal for agent assurance | [TED 584449-2026](https://ted.europa.eu/en/notice/-/detail/584449-2026) | | Agency for Innovation in Cybersecurity, "Cyberagentur" (Germany) | [AI-as-OS] Artificial Intelligence as Operating System: prior information for a partnering event on autonomous agents running on computers | not stated | prior notice | Judiverse agent lab | [TED 575139-2026](https://ted.europa.eu/en/notice/-/detail/575139-2026) | | European Commission, DG CNECT | Supporting the AI Observatory (indicators on AI's impact in strategic sectors) | EUR 2,500,000 | 3 Nov 2026 | Low (research service) | [TED 681546-2026](https://ted.europa.eu/en/notice/-/detail/681546-2026) | | EuroHPC Joint Undertaking | Selection of AI Gigafactory consortia | 10,000,000,000 as recorded (currency not checked) | 3 Dec 2026 | Context only (sovereign compute) | [TED 674672-2026](https://ted.europa.eu/en/notice/-/detail/674672-2026) | | CONSIP (Italy's central buyer) | Cybersecurity framework: solutions and services for resilience of public networks (ID 2733) | EUR 456,379,200 | 4 Nov 2026 | Public Consortium model (non-UK) | [TED 674254-2026](https://ted.europa.eu/en/notice/-/detail/674254-2026) | | Bundesbeschaffung (Austria's federal buyer) | Cybersecurity 2027 framework | 149,200,000 as recorded | 19 Oct 2026 (closed) | as above | [TED 700215-2026](https://ted.europa.eu/en/notice/-/detail/700215-2026) | | Department of Contracts (Malta) | Fully compliant European Digital Identity Wallet (EUDIW) and trust ecosystem | EUR 15,659,714 | 3 Sep 2026 (closed) | EDGE ID identity layer | [TED 483367-2026](https://ted.europa.eu/en/notice/-/detail/483367-2026) | | Digital and Population Data Services Agency (Finland) | Market consultation: certification services for the Finnish EU Digital Identity Wallet | not stated | consultation | EDGE ID identity layer | [TED 463748-2026](https://ted.europa.eu/en/notice/-/detail/463748-2026) | | Utdanningsdirektoratet (Norway's education directorate) | Research: generative AI and changes in upper-secondary schools | 10,000,000 as recorded (currency not checked) | not stated | Public Consortium (education research) | [TED 667798-2026](https://ted.europa.eu/en/notice/-/detail/667798-2026) | - EU grant (not tender): Horizon Europe topic HORIZON-CL4-2026-05-DIGITAL-EMERGING-02 "Next generation AI agents for real-world applications", deadline 30 September 2026 (closed). CLAIM (aggregator). [GrantedAI listing](https://grantedai.com/grants/horizon-europe-horizon-cl4-2026-05-digital-emerging-02-next-generation-a-european-commission-horizon-europe-clust-b5b554bc), accessed 2026-10-31. UK bodies can join Horizon Europe as an associated country: CLAIM not checked here. - A TED search for social prescribing or early detection combined with AI returned no notices in 2026. OBSERVED (absence). [TED](https://ted.europa.eu/), accessed 2026-10-31. ###### 2.5 United States (SAM.gov and SBIR/STTR) - SAM.gov could not be queried directly (registered key needed). UNKNOWN: any open US federal solicitation for AI agent identity or governance. [SAM.gov](https://sam.gov/), accessed 2026-10-31. - The Pentagon's Chief Digital and AI Office (CDAO) awarded prototype agreements of up to $200 million each to Anthropic, Google, OpenAI and xAI in July 2025 to develop "agentic AI workflows" across mission areas; OpenAI's estimated completion was July 2026. CLAIM (trade press). [Nextgov](https://www.nextgov.com/acquisition/2025/07/pentagon-awards-multiple-companies-200m-contracts-ai-tools/406698); [GovConWire](https://www.govconwire.com/articles/cdao-frontier-ai-contract-awards-national-security), accessed 2026-10-31. - SBIR and STTR lapsed from 1 October 2025 and were reauthorised through 30 September 2031 by a law signed 13 April 2026; restart across agencies has been uneven. CLAIM (law firm). [Crowell & Moring](https://www.crowell.com/en/insights/client-alerts/sbirsttr-programs-reauthorized-after-six-month-lapse), accessed 2026-10-31. SBIR is open only to US-owned small businesses: CLAIM not checked here. - NIST (the US standards institute) launched an AI Agent Standards Initiative in February 2026 and consulted on identity and authorisation for AI agents (comments closed 2 April 2026). CLAIM (law firm and trade press). [Pillsbury](https://www.pillsburylaw.com/en/news-and-insights/nist-ai-agent-standards.html); [ExecutiveGov](https://executivegov.com/articles/nist-nccoe-concept-paper-ai-agent-access-controls), accessed 2026-10-31. This is a standards signal, not money. --- ##### 3. Venture capital, 2025-2026 ###### 3.1 Trend reports | Measure | Figure | Label | Source (accessed 2026-10-31) | |---|---|---|---| | Global venture, Q3 2026 | $159bn across ~6,000 start-ups; down 25% on Q2 ($212bn), up 53% on Q3 2025; AI $102bn (64%); record 27 rounds of $1bn+ | OBSERVED | [Crunchbase News, 5 Oct 2026](https://news.crunchbase.com/venture/q3-2026-global-startup-funding-ai-billion-dollar-rounds-exits-data/) | | Global venture, first nine months of 2026 | $679bn, the highest on record | CLAIM (search summary of Crunchbase) | [Crunchbase News](https://news.crunchbase.com/venture/q3-2026-global-startup-funding-ai-billion-dollar-rounds-exits-data/) | | Europe, Q3 2026 | $25bn (up 77% year on year); UK $7.5bn, the largest in Europe; AI 75% of European funding; defence, data centres, energy, aerospace and robotics about half | OBSERVED | [Crunchbase News, 8 Oct 2026](https://news.crunchbase.com/venture/q3-2026-europe-strong-quarter-ai-uk-germany-france/) | | Fintech, Q3 2026 | $10.6bn (down 28.4% on Q2, down 21.4% year on year); 687 rounds, fewest since Q1 2018; UK sixth at $337.8m | OBSERVED | [Crunchbase News, 20 Oct 2026](https://news.crunchbase.com/venture/q3-2026-fintech-startup-funding-deal-count-fell-ai/) | | Cybersecurity, Q3 2026 | $6.8bn, up ~50% on Q2, driven by bigger rounds; at least 18 rounds of $100m+; identity security a focus | OBSERVED | [Crunchbase News, 15 Oct 2026](https://news.crunchbase.com/cybersecurity/q3-2026-large-venture-investments-ai-defense-startups/) | | US venture (PitchBook-NVCA) | $515.8bn year to date at Q3 2026; AI 82.7% of the year's deal value, 65.9% in Q3 | CLAIM (SiliconANGLE summary; primary report not opened) | [SiliconANGLE](https://siliconangle.com/2026/07/09/pitchbook-us-venture-funding-hits-412-7b-first-half-ai-deals-dominate/); [NVCA Venture Monitor](https://nvca.org/research/pitchbook-nvca-venture-monitor) | | UK equity (Beauhurst), H1 2026 | £14.4bn across 2,799 deals; three AI companies took 29% of capital | OBSERVED | [Beauhurst: The Deal H1 2026](https://www.beauhurst.com/research/the-deal-h1-2026/) | | UK AI equity, Q2 2026 | record £4.56bn, 57% of all UK equity; eight £100m+ AI raises took ~80% of disclosed AI money | CLAIM (press summary of Beauhurst) | [UKTN](https://www.uktech.news/ai/uks-ai-investment-reaches-record-4-56bn-in-q2-20260716) | | Edtech (HolonIQ), H1 2026 | ~$1bn, down 26% on H1 2025; workforce training took most VC; K-12 AI personalisation gaining | OBSERVED | [HolonIQ, 14 Jul 2026](https://www.holoniq.com/notes/1b-in-edtech-venture-capital-to-date-funding-falls-short-of-last-years-midpoint-asia-mena-buck-the-trend) | | US digital health (Rock Health), H1 2026 | $7.4bn across 244 deals; 20 rounds of $100m+ took 45% | CLAIM (secondary coverage; primary not opened) | [MarketScale](https://www.marketscale.com/industries/healthcare/rock-health-74b-went-into-digital-health-in-h1-2026-and-20-rounds-took-45) | | Dealroom | No standalone 2026 agentic AI report found; Dealroom's European AI-agent list (updated June 2026) is gated | UNKNOWN | [Dealroom list](https://dealroom.co/lists/ai-agent-startups-europe/) | ###### 3.2 Notable rounds by theme | Theme | Company (base) | Round | Investors | Date | Label | Source (accessed 2026-10-31) | |---|---|---|---|---|---|---| | Agent identity | Oak (Israel) | $60m seed: one identity system for people, machines and AI agents | Accel, Greylock, CRV (co-leads) | 16 Jul 2026 | OBSERVED | [The Next Web](https://thenextweb.com/news/oak-60m-seed-ai-native-identity-platform) | | Agent identity | NewCore | $66m seed, security-first identities for AI agents | not checked | Jun 2026 | CLAIM | [Threatcluster summary](https://threatcluster.io/cluster/newcore-has-emerged-from-stealth-with-66-million-in-funding-e446d016) | | Agent identity | Oasis Security | $120m to manage and secure AI agent identities | not checked | Mar 2026 | CLAIM | [Shopifreaks](https://www.shopifreaks.com/oasis-security-raises-120m-to-manage-and-secure-ai-agent-identities-inside-enterprise-systems/) | | Agent identity | Rig Security (Tel Aviv) | $12m seed (agents running under employee credentials) | not checked | 29 Sep 2026 | CLAIM | [SC World](https://www.scworld.com/brief/rig-security-launches-with-12-million-to-manage-ai-agent-identities) | | Identity security | Keyfactor | $1bn growth round | Summit Partners | Jul 2026 | OBSERVED | [Crunchbase News: cyber Q3](https://news.crunchbase.com/cybersecurity/q3-2026-large-venture-investments-ai-defense-startups/) | | Agent and data security | Island | $400m Series F at $6.4bn (governs how people and AI agents use apps and data) | Evolution Equity Partners | Sep 2026 | OBSERVED | [Crunchbase News: cyber Q3](https://news.crunchbase.com/cybersecurity/q3-2026-large-venture-investments-ai-defense-startups/) | | Agent and data security | Cyera | $400m at $12bn (data security for AI agents), three months after $600m | Goldman Sachs Growth Equity | Sep 2026 | OBSERVED | [Crunchbase News: cyber Q3](https://news.crunchbase.com/cybersecurity/q3-2026-large-venture-investments-ai-defense-startups/) | | Agent governance | Geordie AI (London) | $30m Series A: governance for enterprise AI agents (security, monitoring, compliance, policies) | Balderton (lead), Crosspoint, General Catalyst, Ten Eleven | 28 May 2026 | CLAIM (secondary page read) | [Raising.fi](https://raising.fi/news/geordie-ai-series-a-may-2026) | | Agent governance | Onyx Security | $113m Series B to control enterprise AI agents (valuation $640m) | not checked | Jul 2026 | CLAIM | [Crypto Briefing](https://cryptobriefing.com/onyx-security-113m-ai-agents-enterprise/) | | Agent governance | Runlayer (New York) | $30m Series A | Felicis (lead), Khosla | Jun 2026 | CLAIM | [The SaaS News](https://thesaasnews.com/news/runlayer-raises-30m-series-a) | | AI governance | AI Score (London) | $5.4m seed | Fuel Ventures | Sep 2026 | CLAIM | [Global Legal Post](https://www.globallegalpost.com/news/uk-ai-governance-intelligence-platform-ai-score-raises-54m-in-seed-funding-round-2125790039) | | Agentic payments | Natural | $30m Series A: wallets, vaults and agent-initiated pay and request flows (earlier $9.8m seed) | Forerunner (Kirsten Green, lead) | May-Jul 2026 (sources disagree) | CLAIM | [The SaaS News](https://www.thesaasnews.com/news/natural-raises-30m-series-a/) | | Agentic payments, investor view | Forerunner's Kirsten Green: agent payments moved from "if" to "when"; more opportunity in fraud prevention, identity verification and agent permissions; tools for agents to complete purchases are overfunded | — | — | Oct 2026 | OBSERVED (as reported) | [Crunchbase News: fintech Q3](https://news.crunchbase.com/venture/q3-2026-fintech-startup-funding-deal-count-fell-ai/) | | Agentic payments volume | Agentic payment volume reportedly fell in 2026 after a late-2025 spike | — | — | 2026 | CLAIM | [American Banker](https://www.americanbanker.com/payments/news/agentic-payment-rail-shows-volume-decline) | | Fraud and scams | Savi Security | $7m seed against digital fraud | Acrew Capital (lead) | Jul 2026 | CLAIM | [Seedtable](https://seedtable.com/companies/savi-security/funding-rounds/seed-2026-07) | | Fintech (largest Q3) | Ant International (Singapore) | $1.2bn | Alibaba Group, Ant Group | Q3 2026 | OBSERVED | [Crunchbase News: fintech Q3](https://news.crunchbase.com/venture/q3-2026-fintech-startup-funding-deal-count-fell-ai/) | | Defence | Anduril (US) | $5bn Series H at $61bn | Thrive Capital, a16z | 13 May 2026 | OBSERVED | [TechCrunch](https://techcrunch.com/2026/05/13/anduril-raises-5b-doubles-valuation-to-61b/) | | Defence | Helsing (Germany) | $1.8bn Series E at $18bn | Lightspeed, General Catalyst (new); Prima Materia, Accel, Greenoaks | 13 Jul 2026 | OBSERVED | [Defense News](https://www.defensenews.com/global/europe/2026/07/13/helsing-raises-18-billion-in-europes-biggest-defense-startup-round/) | | Defence | Quantum Systems (Germany) $1.2bn Series D at $8bn; Kraken Technology (UK) $175m Series B at $1bn; Stark (Germany) €500m; ICEYE (Finland) €450m Series F; Harmattan AI (France) $200m Series B | — | various | Jan-Jul 2026 | OBSERVED | [Defense News](https://www.defensenews.com/global/europe/2026/07/13/helsing-raises-18-billion-in-europes-biggest-defense-startup-round/) | | Govtech | GovWell (US) | $25m Series A, "AI operating system for government" | Insight Partners (lead) | May 2026 | CLAIM | [Willkie Farr](https://www.willkie.com/news/2026/05/willkie-advises-insight-partners-on-series-a-investment-in-govwell) | | Early detection | Bios Life (Boston) | $25m seed, AI virtual clinic for cancer survivors and high-risk patients | not checked | Aug 2026 | CLAIM | [Dealroom news](https://dealroom.co/news/144359-bios-life-raises-25m-seed-for-ai-cancer-surveillance-startup/) | | Early detection | LungLife AI (UK) | $19m, round type undisclosed | not checked | Jul 2026 | CLAIM | [aifunding.me](https://aifunding.me/companies/lunglife-ai) | | Edtech | Preply $150m; Multiverse $70m (workforce training, VC) | — | not checked | H1 2026 | OBSERVED (HolonIQ list) | [HolonIQ](https://www.holoniq.com/notes/1b-in-edtech-venture-capital-to-date-funding-falls-short-of-last-years-midpoint-asia-mena-buck-the-trend) | | Sovereign AI | Callosum (London) | $100m seed, largest in Europe in Q3 | not named | Q3 2026 | OBSERVED | [Crunchbase News: Europe Q3](https://news.crunchbase.com/venture/q3-2026-europe-strong-quarter-ai-uk-germany-france/) | - **Not found.** A 2026 Series B or later for a UK authorised push payment (APP) fraud start-up: UNKNOWN. A UK social prescribing start-up venture round in 2026: UNKNOWN. A defence-logistics software venture round in 2026: UNKNOWN. [Web searches, 31 October 2026](https://news.crunchbase.com/), accessed 2026-10-31. --- ##### 4. Synthesis: who is paying, for which problem Strength of demand: **Strong** = money is being spent now by several payers, with open routes; **Medium** = real money but narrow, early or one main payer; **Weak** = mostly talk, research or closed calls. | Problem area | Who is paying | Evidence (accessed 2026-10-31) | Fit to app | Strength of demand | |---|---|---|---|---| | Building and governing AI agents inside organisations | Home Office, FCA, Pension Protection Fund, councils, police; VC (Geordie AI, Onyx, Runlayer) | [Home Office £3m x2](https://www.contractsfinder.service.gov.uk/Notice/22eb09e9-74c8-4c8e-90d8-453dfd05838d); [FCA £625k](https://www.contractsfinder.service.gov.uk/Notice/71f13f64-750d-4790-b895-5789f4e2f06f); [Geordie AI](https://raising.fi/news/geordie-ai-series-a-may-2026) | Judiverse (assistants, agent lab); Public Consortium | **Strong** (OBSERVED), though public money mostly goes to consultancies on existing platforms | | Identity for AI agents (machines and agents as accountable identities) | Venture investors; standards bodies; ARIA | [Oak $60m](https://thenextweb.com/news/oak-60m-seed-ai-native-identity-platform); [Crunchbase cyber Q3](https://news.crunchbase.com/cybersecurity/q3-2026-large-venture-investments-ai-defense-startups/); [ARIA Scaling Trust](https://www.aria.org.uk/opportunity-spaces/trust-everything-everywhere/scaling-trust/funding) | EDGE ID core; all four apps | **Strong in VC, Weak in public procurement** (no public tender found) | | AI assurance, testing and compliance evidence | Innovate UK (£5m open), FCA (Project Libra), BSI Germany, AISI, EPSRC | [BridgeAI Assurance](https://www.ukri.org/opportunity/bridgeai-deployment-catalyst-ai-for-assurance-and-compliance/); [FCA Libra £448k](https://www.contractsfinder.service.gov.uk/Notice/42657d5a-24f9-48c8-aebc-648827417d29); [BSI PRAKI](https://ted.europa.eu/en/notice/-/detail/584449-2026); [AI^2 £1.05m](https://gtr.ukri.org/projects?ref=UKRI567) | All four, especially EDGE Finance and Public Consortium | **Strong** (grants open now) | | Identity verification and identity fraud in public services | HMRC, Home Office, DVLA, DWP, Department for Business and Trade | [HMRC £20m](https://www.contractsfinder.service.gov.uk/Notice/5add23e7-2df5-441c-a60e-6894f8325b4a); [Home Office £3.06m](https://www.contractsfinder.service.gov.uk/Notice/4e4a9a81-f998-4c18-acb5-1be72b68ebe9); [DVLA £2.25m](https://www.contractsfinder.service.gov.uk/Notice/0abb34c3-01ac-4fd9-8d67-0b89981b545f) | EDGE Finance; Public Consortium (vetting) | **Strong** (incumbents: Experian, TransUnion, Big Four) | | Payment fraud, scams and agent payments | Fintech investors (shrinking overall); GCA Open Banking DPS; UKRI Trusted AI Adoption | [Crunchbase fintech Q3](https://news.crunchbase.com/venture/q3-2026-fintech-startup-funding-deal-count-fell-ai/); [Open Banking DPS](https://www.contractsfinder.service.gov.uk/Notice/d7d25e48-faf9-4d7f-b4df-e6e1660559bd); [UKRI-6347](https://www.find-tender.service.gov.uk/Notice/094267-2026) | EDGE Finance | **Medium**: investors want fraud, identity and agent permissions, not more agent checkout tools | | Public-sector cyber resilience | MHCLG (£30m), DSIT (up to £14m), NHS England, GCA DPS (£800m), CONSIP Italy (EUR 456m) | [MHCLG](https://www.contractsfinder.service.gov.uk/Notice/e9d30f07-2719-4ed7-ab82-5f5698cd3af5); [DSIT](https://www.contractsfinder.service.gov.uk/Notice/953c29c9-0e3e-4b7c-a609-58b2832f11a9); [DSIT Cyber Accelerator, open](https://www.find-tender.service.gov.uk/Notice/103182-2026) | Public Consortium (cyber governance) | **Strong** (crowded with established cyber firms) | | Early detection and screening | NHS England (£215m screening digital), Innovate UK (£25.4m cancer, closing 4 Nov), NHS SBS (£900m AI framework), SBRI | [NHS screening](https://www.contractsfinder.service.gov.uk/Notice/a7e8665f-95fe-4e36-8c1b-7414b8792aac); [IFS 2527](https://apply-for-innovation-funding.service.gov.uk/competition/2527/overview/d48f61f8-2bd5-4067-9c05-29dd08487623) | EDGE Clinic | **Strong**, but needs clinical validation and medical device status | | Clinical documentation (AI scribes) | NHS trusts, ICBs (integrated care boards), councils | [ULTH £3.37m](https://www.contractsfinder.service.gov.uk/Notice/5f7d6277-d794-45c0-8b24-bb0fa565ec7b); [AVT registry](https://www.find-tender.service.gov.uk/Notice/009501-2026) | EDGE Clinic (adjacent); Judiverse | **Strong** but a crowded, regulated lane | | Social prescribing coordination | ICBs (for example a Leicestershire platform), councils; research councils | [LLR platform](https://www.find-tender.service.gov.uk/Notice/100676-2026); [GtR social prescribing](https://gtr.ukri.org/search/project?term=%22social+prescribing%22) | EDGE Clinic | **Medium-Weak**: local, small contracts; mainly link-worker staffing | | Public procurement tooling with AI | FCA proofs of concept, DCMS, Social Work England, Financial Ombudsman | [FCA PAIaaS](https://www.find-tender.service.gov.uk/Notice/095939-2026); [DCMS £300k](https://www.contractsfinder.service.gov.uk/Notice/745eef88-f225-49e9-9a11-0653f7656b7e) | Public Consortium (procurement) | **Medium** (early pilots) | | Vetting | MoD (ship vetting), combined authorities, King's Trust (background checks) | [MoD ShipVetting](https://www.find-tender.service.gov.uk/Notice/094205-2026); [Core-Vet](https://www.contractsfinder.service.gov.uk/Notice/e7a33ce4-6a27-4af0-8b31-b255a594bfff) | Public Consortium (vetting) | **Weak** (small, specialised) | | Defence and government logistics | MoD estates and support contracts; UKDI open calls (status unclear); defence VC booming but in hardware and autonomy | [Defense News](https://www.defensenews.com/global/europe/2026/07/13/helsing-raises-18-billion-in-europes-biggest-defense-startup-round/); [UKDI Security Open Call](https://iuk-business-connect.org.uk/opportunities/uk-defence-innovation-security-open-call/) | Public Consortium (defence logistics) | **Medium in VC, Weak/UNKNOWN for logistics software** | | Education and university research | Universities (copilots, OpenAI pilot via Jisc), EEF digital service (open), KTP; edtech VC down 26% | [Jisc OpenAI](https://www.find-tender.service.gov.uk/Notice/088540-2026); [EEF £730k](https://www.contractsfinder.service.gov.uk/Notice/612ec727-9bb1-4e65-99f4-1c68f8c05873); [HolonIQ](https://www.holoniq.com/notes/1b-in-edtech-venture-capital-to-date-funding-falls-short-of-last-years-midpoint-asia-mena-buck-the-trend) | Public Consortium (education); Judiverse (office) | **Medium** | | AI skills and training for staff | Many departments and councils (20 awards in my sample) | [Contracts Finder](https://www.contractsfinder.service.gov.uk/) | Judiverse | **Strong but low value per contract** | | Agent-to-agent trust and multi-agent security research | ARIA (~£50m Scaling Trust), AISI Alignment Project (£27m), multi-agent safety joint call ($10m) | [ARIA](https://www.aria.org.uk/opportunity-spaces/trust-everything-everywhere/scaling-trust/funding); [AISI](https://www.aisi.gov.uk/blog/funding-60-projects-to-advance-ai-alignment-research) | Judiverse agent lab; EDGE ID core | **Medium** (grant money, research outputs) | | Digital identity wallets (EU) | Malta, Finland, Germany (SPRIND) | [TED 483367-2026](https://ted.europa.eu/en/notice/-/detail/483367-2026) | EDGE ID identity layer | **Medium** (EU-only buyers) | ##### 5. Dates to act on (from the evidence above) | Date | What | Source (accessed 2026-10-31) | |---|---|---| | 31 Oct 2026 (today) | ARIA Scaling Trust cycle cut-off; later proposals roll to the next cycle | [ARIA](https://www.aria.org.uk/opportunity-spaces/trust-everything-everywhere/scaling-trust/funding) | | 4 Nov 2026 | Innovate UK accelerated cancer diagnosis (£25.4m) closes | [IFS 2527](https://apply-for-innovation-funding.service.gov.uk/competition/2527/overview/d48f61f8-2bd5-4067-9c05-29dd08487623) | | 5 Nov 2026 | UKRI-6347 Trusted AI Adoption tender (£800k) closes | [Find a Tender 094267-2026](https://www.find-tender.service.gov.uk/Notice/094267-2026) | | 12 Nov 2026 | MoD ShipVetting tender closes | [Find a Tender 094205-2026](https://www.find-tender.service.gov.uk/Notice/094205-2026) | | 18 Nov 2026 | Innovate UK Efficient, Structured and Controllable AI Systems closes | [IFS 2582](https://apply-for-innovation-funding.service.gov.uk/competition/2582/overview/9fdb47a5-011c-4c47-ac5e-c426e295fe89) | | 27 Nov 2026 | DSIT Cyber Accelerator Programme Design Partnerships (£1.2m) closes | [Find a Tender 103182-2026](https://www.find-tender.service.gov.uk/Notice/103182-2026) | | 30 Nov 2026 | EEF Digital Public Service product team (£730k) closes | [Contracts Finder](https://www.contractsfinder.service.gov.uk/Notice/612ec727-9bb1-4e65-99f4-1c68f8c05873) | | 8 Dec 2026 | Innovate UK BridgeAI Deployment Catalyst, three strands of £5m each, close | [UKRI](https://www.ukri.org/opportunity/bridgeai-deployment-catalyst-ai-for-assurance-and-compliance/) | | 27 Jan 2027 | KTP 2026-27 Round 4 closes | [IFS list](https://apply-for-innovation-funding.service.gov.uk/competition/search) | ##### 6. What I could not access or confirm - Innovation Funding Service page 3 (about 10 of 29 competitions) did not return its listings: UNKNOWN whether it holds relevant calls. [IFS](https://apply-for-innovation-funding.service.gov.uk/competition/search?page=3), accessed 2026-10-31. - Contracts Finder and Find a Tender public notice pages refuse automated reading; details were checked through their official data services instead. Find a Tender coverage is limited to records updated 4 September to 30 October 2026 because of request-rate limits. [Find a Tender](https://www.find-tender.service.gov.uk/), accessed 2026-10-31. - SAM.gov (needs a registered key), PitchBook, Dealroom and Rock Health primary reports (gated), Fintech Futures and Digital Health article pages (refused automated reading), Helsing's newsroom (rate limited) and ARIA's Q1 2026-27 transparency PDF (would not decode) were not read directly. - Not found: an open SBRI Healthcare AI competition; a 2026 AISI Challenge Fund round; the opening date of the 2026 Alignment Project round; any open UKDI challenge today; any UK public tender for AI agent identity; any US agent-identity solicitation; UK social prescribing or defence-logistics software venture rounds in 2026. All UNKNOWN. ### research/market/HEADLINES-2026-10.md `research/market/HEADLINES-2026-10.md` · sha256 `bff44c43df89bb8a01667554bd85f588195dc0897e8784e5aadb68af9a1f9366` #### Headlines, October 2026 Searched 31 October 2026. Each line is labelled OBSERVED (reported by a primary or reputable source) or CLAIM (secondary or still developing). Figures for the Binance hack were one to two days old when read and may change. ##### Crypto: the largest theft on record - **What happened:** on 29 October 2026 about **$6.36bn** was moved out of Binance wallets across ten blockchains, mostly within about 20 minutes from 20:05 UTC. That is about four times the $1.5bn Bybit hack of February 2025. OBSERVED. Sources: [TRM Labs](https://www.trmlabs.com/resources/blog/north-korea-steals-usd-6-36-billion-from-binance-in-the-largest-hack-in-history), [CNBC](https://www.cnbc.com/2026/10/29/circle-tether-freeze-funds-impacted-by-potential-binance-security-issue.html), [Washington Post](https://www.washingtonpost.com/technology/2026/10/29/suspected-binance-hack-sets-off-scramble-lock-down-billions-crypto/). - **Cause:** Binance's initial investigation points to one engineer's compromised laptop, which gave access to make withdrawals. TRM describes it as a compromise of hot-wallet signing infrastructure. CLAIM, as the investigation is ongoing. Sources: [CNBC](https://www.cnbc.com/2026/10/30/billions-in-crypto-frozen-in-binance-hack-heres-what-happens-next-.html), [TRM Labs](https://www.trmlabs.com/resources/blog/north-korea-steals-usd-6-36-billion-from-binance-in-the-largest-hack-in-history). - **Attribution:** TRM Labs and Chainalysis link it to TraderTraitor, a North Korea-linked group. Binance has not confirmed this. CLAIM. - **Response:** Binance suspended deposits and withdrawals and said it would cover all losses. About $3bn was frozen on-chain, with Circle, Tether and Ripple freezing linked funds. Bitcoin fell below $80,000. OBSERVED. Sources: [Al Jazeera](https://www.aljazeera.com/economy/2026/10/30/binance-says-2-54bn-in-crypto-funds-frozen-amid-potential-security-issue), [Axios](https://www.axios.com/2026/10/29/binance-cyberattack-crypto-heist), [Wikipedia](https://en.wikipedia.org/wiki/2026_Binance_hack). - **Earlier the same autumn:** Bitget was hacked on 24 September 2026, with about $387.5m lost. CLAIM. Source: [Wikipedia](https://en.wikipedia.org/wiki/Bitget). - **What it means for the estate (HYPOTHESIS, to test):** the loss came from one compromised person's access moving billions in minutes. EDGE Finance's crypto wallet capability puts every transfer through four layers: 1. rehearsal before effect; 2. a mandate with per-transfer and per-period limits; 3. approval beyond those limits by a separate accountable owner; 4. evidence and monitoring that alert on a rise in outflows. The test: replay the published outflow pattern (ten chains, about 20 minutes) through the engine, and measure how much would have left before limits, approval or monitoring stopped it. ##### Energy - **Price cap:** Ofgem's cap rose 4% on 1 October 2026 to **£1,723 a year** for a typical dual-fuel household, until 31 December. VAT is removed from domestic electricity until 31 March 2027. The rise is driven by wholesale gas prices since the Middle East conflict. CLAIM: secondary sources, so check against Ofgem. Sources: [Uswitch](https://www.uswitch.com/gas-electricity/guides/price-cap/), [House of Commons Library](https://commonslibrary.parliament.uk/research-briefings/cbp-9714/). - **January 2027 forecasts:** £2,065 to £2,080. These are unconfirmed forecasts. CLAIM. - **Grid:** no grid-specific October news was found. UNKNOWN. - **What it means for the estate (HYPOTHESIS):** the estate's own measures (carbon and compute per verified outcome, and deterministic operations using no model at run time) cut the energy cost of digital work. Whether this addresses the energy crisis needs its own measure, energy per verified outcome against today's way of doing the same work, before it is a claim. ##### Fraud, scams and cyber (UK) - **Account hacking:** Report Fraud ran a passkey campaign from 5 October. Hacking is the most reported UK cyber crime, with reports up 34% on the previous year. OBSERVED. Source: [Report Fraud](https://www.reportfraud.police.uk/social-media-hacking). - **Payment diversion fraud:** about £101m lost in 2025/26 across 3,657 reports, per City of London Police, as cited by a roundup. CLAIM, so verify against the primary source. Source: [ScanMyShadow](https://www.scanmyshadow.com/blog/2026/10/uk-scam-news-october-2026.html). - **ASOS:** unusual customer app notifications on 6 October after unauthorised access to its communications systems. CLAIM. Source: [Join the Claim](https://jointheclaim.com/scam-alert-update-october-2026/). ##### AI agents and regulation - **EU:** first formal requests for information under the AI Act were sent to leading AI providers, after AI agents breached external systems during testing. Announced 14 October. CLAIM: trade press. Sources: [Unite.AI](https://www.unite.ai/eu-announces-first-ai-act-enforcement-actions-after-ai-agent-incidents/), [Lexology](https://www.lexology.com/library/detail.aspx?g=cdc3a626-bf58-48c1-a9b3-d80923381e31). - **UK:** on 19 October the FCA promised a trusted framework for agentic finance, covering agent identity, delegated authority, accountability and execution. It says existing rules already apply and that significant AI failures and near misses must be reported. CLAIM: trade press. Sources: [FinTech Global](https://fintech.global/2026/10/26/fca-builds-agentic-finance-rules-as-ai-reshapes-uk-markets), [Reinvently](https://reinvently.co.uk/blog/fca-agentic-finance-existing-rules/). - **What it means for the estate:** the FCA's framework headings are what the estate already specifies for every operation: agent identity, delegated authority (mandate), accountability (a named owner, never an agent), execution and evidence. ##### NHS - **Waiting list:** 7.37m treatments for 6.24m patients at the end of August, up from 7.33m in July. 113,339 people have waited over a year. OBSERVED. Sources: [LBC](https://www.lbc.co.uk/article/nhs-waiting-list-hits-737million-despite-record-treatment-levels-5HjdjZs_2/), [Medscape](https://www.medscape.com/viewarticle/current-trends-nhs-misses-18-week-target-wide-margin-2026a10011wj). - **GP pressure:** patients waiting for referrals add demand to general practice. CLAIM. Source: [GPonline](https://www.gponline.com/nhs-delays-piling-extra-work-gps/article/1972270). ### research/public-consortium/COMPETITORS-REGULATORS-REUSE.md `research/public-consortium/COMPETITORS-REGULATORS-REUSE.md` · sha256 `f78f97e663b8a5d2407a0c5c68f632aea4a9da2935631f9ce5ae960dc5267750` #### EDGE Public Consortium: Competitors, Regulators and Reusable Systems Research note for **EDGE Public Consortium** (part of JudeAI | EDGE ID). Prepared 31 October 2026. This is research, not legal advice. It does not cover what is funded or tendered (see `research/market/FUNDING-AND-PROCUREMENT.md`) or the costliest problems (see `research/market/COSTLIEST-PROBLEMS.md`). --- ##### 0. How to read this document ###### 0.1 Evidence labels | Label | Meaning | |---|---| | **OBSERVED** | We read it in a primary source: a government or regulator page, legal text, an official contract notice, or a project's own repository (licence file, tags, commits). News reports of official facts are marked **OBSERVED (secondary)**. | | **CLAIM** | Said by a vendor, a vendor's filing or marketing, or a single commentator. It may be true, but nobody independent has checked it. | | **AGREEMENT** | A rule or standard that is in force or formally adopted. We give the date it took effect. | | **PROPOSED** | A bill, consultation, draft or planned rule that is not yet in force. | | **UNKNOWN** | We looked and could not confirm, or we ran out of search budget. The line says what is missing. | | **ANALYSIS** | Our own reasoning from the facts above it. Not a fact, so it has no source. | ###### 0.2 How the facts were checked - Every source was accessed on **2026-10-31**. - **Licences** in section 5 were read from the actual `LICENSE` (or `COPYING` / `LICENCE`) file in each repository on `raw.githubusercontent.com`, not from a badge. - **Versions** are the highest numeric tag returned by `git ls-remote --tags` on 2026-10-31. **Last commit** is from a shallow clone of the default branch on the same day. - Some official pages (parliament.uk, security.gov.uk, bills.parliament.uk) refused automated fetches (HTTP 403). Where that happened we say so and rely on the search-engine extract or a secondary source. - **Search budget ran out** before we could research Drata, IBM public sector, Accenture public sector, Microsoft UK government agreements, JSP 440, and university-industry collaboration platforms (for example IN-PART). Those rows are marked UNKNOWN. A follow-up session can fill them in. ###### 0.3 Acronyms (first use) | Acronym | Meaning | |---|---| | AI | Artificial Intelligence | | AIP | Palantir's Artificial Intelligence Platform | | API | Application Programming Interface (a way for software to talk to software) | | ATRS | Algorithmic Transparency Recording Standard (UK government) | | ATT&CK | MITRE's catalogue of attacker Adversarial Tactics, Techniques and Common Knowledge | | BODS | Beneficial Ownership Data Standard (who really owns a company) | | BPSS | Baseline Personnel Security Standard (UK pre-employment checks) | | CAF | Cyber Assessment Framework (UK National Cyber Security Centre) | | CDP | Central Digital Platform (UK public procurement website, the enhanced Find a Tender service) | | CMMC | Cybersecurity Maturity Model Certification (US Department of Defense) | | CRA | Cyber Resilience Act (EU) | | CSM | Cyber Security Model (UK Ministry of Defence supplier cyber rules) | | CTC / SC / DV | Counter Terrorist Check / Security Check / Developed Vetting (UK national security vetting levels) | | CTI | Cyber Threat Intelligence | | D3FEND | MITRE's catalogue of defensive cyber techniques | | DEFCON | Defence Condition (a standard UK MoD contract clause), not the US alert level | | Def Stan | UK Defence Standard | | DfE | Department for Education (England) | | DIU | Defense Innovation Unit (US) | | DSIT | Department for Science, Innovation and Technology (UK) | | DUAA | Data (Use and Access) Act 2025 (UK) | | FedRAMP | Federal Risk and Authorization Management Program (US cloud security approval) | | GDS | Government Digital Service (UK) | | GRC | Governance, Risk and Compliance | | JSP | Joint Service Publication (UK MoD policy document) | | KCSIE | Keeping Children Safe in Education (statutory guidance, England) | | LLM | Large Language Model | | LTI | Learning Tools Interoperability (plugging tools into a learning platform) | | MCP | Model Context Protocol (a way for AI agents to call tools) | | MoD | Ministry of Defence (UK) | | NCSC | National Cyber Security Centre (UK) | | NIS / NIS2 | Network and Information Systems rules (UK 2018 regulations / EU 2022 directive) | | NSV | National Security Vetting (UK) | | OCDS | Open Contracting Data Standard | | OfS | Office for Students (higher education regulator, England) | | OGL | Open Government Licence (UK) | | OMB | Office of Management and Budget (US) | | PPN | Procurement Policy Note (UK Cabinet Office) | | SBOM | Software Bill of Materials (list of parts inside software) | | SOAR | Security Orchestration, Automation and Response | | SOC | Security Operations Centre | | SPDX | Software Package Data Exchange (SBOM format and licence-id list) | | STIX / TAXII | Structured Threat Information Expression / Trusted Automated Exchange of Intelligence Information | | UKSV | United Kingdom Security Vetting (the vetting agency) | | xAPI | Experience API (records learning activity) | --- ##### 1. Ten things a new engineer should know 1. **OBSERVED.** Big platform vendors now sell AI agents inside procurement, cyber and logistics tools: Coupa, SAP Ariba, JAGGAER, CrowdStrike, Palo Alto Networks, Microsoft, Vanta, OneTrust, Dun & Bradstreet (sources in section 2). 2. **CLAIM.** Several say their agents are governed: role-based access, human approval, audit logs (Palo Alto AgentiX, JAGGAER JAI, Palantir's ontology). These are vendor statements, and the logs stay inside each vendor's product (section 2.6). 3. **ANALYSIS.** The gap EDGE can fill is not "no logs". It is that no one gives a **portable, independent, glass-box record** of *which agent, acting for whom, under what mandate, did what, on what evidence*, that a buyer, a supplier, an auditor and a regulator can all read. 4. **OBSERVED.** UK government's own transparency tools do not yet name agents: the ATRS scope policy and PPN 017 (AI in procurement) do not mention AI agents (section 4). 5. **OBSERVED.** The NCSC's May 2026 guidance says: "If you cannot understand, monitor or contain an agent's actions, it is not ready for deployment." That is the best one-line test for EDGE's design (section 4). 6. **AGREEMENT.** The UK Procurement Act 2023 has been in force since 24 February 2025, and the Central Digital Platform publishes notices in OCDS format under the Open Government Licence (section 3). 7. **PROPOSED.** The UK Cyber Security and Resilience Bill finished Lords report stage on 26 October 2026, with Lords third reading scheduled for 2 November 2026. It is not yet law (section 3). 8. **AGREEMENT.** In the US, CMMC Phase 2 (third-party Level 2 assessments) starts on 10 November 2026, ten days from today (section 3). 9. **AGREEMENT.** EU Cyber Resilience Act reporting (24-hour early warning) has applied since 11 September 2026 (section 3). 10. **OBSERVED.** Almost everything EDGE needs for data formats already exists under permissive licences: OCDS and BODS (Apache-2.0), D3FEND (MIT), CycloneDX (Apache-2.0), GOV.UK Frontend (MIT), OpenCTI Community Edition (Apache-2.0). Watch out for copyleft in Wazuh (GPL-2.0 / AGPL-3.0) and Canvas (AGPL-3.0) (section 5). --- ##### 2. Competitors How to read each table: **What they offer**, **Agents?** (do they sell AI agents that act, not only chat), **Glass-box or black-box?** (can an outsider see why a decision was made), and **Label**. ###### 2.1 Intelligence platforms for government and defence | Company / product | What they offer, with evidence | Agents? | Label | |---|---|---|---| | **Palantir (UK MoD partnership)** | Signed 18 September 2025. Palantir to "invest up to £1.5bn"; partnership "will identify opportunities with UK Defence which could be worth up to £750 million over the next five years"; covers "data analysis, intelligence, decision support and targeting systems" and part of the Digital Targeting Web ([source](https://www.gov.uk/government/news/new-strategic-partnership-to-unlock-billions-and-boost-military-ai-and-innovation), accessed 2026-10-31). | Not stated in the release. | OBSERVED (release); £1.5bn is a CLAIM of future investment | | Palantir (UK MoD partnership): governance | The release does not mention ethics, human oversight or human control ([source](https://www.gov.uk/government/news/new-strategic-partnership-to-unlock-billions-and-boost-military-ai-and-innovation), accessed 2026-10-31). | n/a | OBSERVED | | Palantir (UK MoD contract, December 2025) | A £240 million MoD contract signed December 2025 by direct award, criticised in a parliamentary debate on 10 February 2026 ([source](https://westminsterbrief.co.uk/archive/debate/10-february-2026/ministry-defence-palantir-contracts-08d0), accessed 2026-10-31). | Unknown | CLAIM (one side of a debate; contract notice not checked) | | **Palantir (NHS Federated Data Platform)** | Awarded November 2023, £330 million over seven years, consortium with Accenture, PwC, NECS and Carnall Farrar ([source](https://digitalhealth.net/2023/11/nhs-england-awards-330m-federated-data-platform-contract-to-palantir), accessed 2026-10-31); award notice on Find a Tender ([source](https://www.find-tender.service.gov.uk/Notice/037776-2023), accessed 2026-10-31). In 2026 the Health and Social Care Committee urged ministers to consider ending it ([source](https://www.computing.co.uk/news/2026/government/mps-urge-government-cancel-palantir-nhs-contract), accessed 2026-10-31). | Not stated | OBSERVED (secondary) | | **Palantir AIP** | Palantir's FY2025 annual report says AIP "can turn organizations' actions and processes into tools for humans and LLM-driven agents" ([source](https://www.sec.gov/Archives/edgar/data/0001321655/000132165526000021/fy2025palantirars.pdf), accessed 2026-10-31). An analyst summary says Palantir frames its "ontology" (data model) as the control layer for "who authorized an agent action" ([source](https://futurumgroup.com/insights/palantir-q1-fy-2026-revenue-beats-estimates-us-demand-drives-outlook-raise/), accessed 2026-10-31). | Yes | CLAIM | | **Anduril (Lattice)** | UK MoD awarded Anduril a £17m, 31-month force-protection contract (could grow to £24m) using Lattice software to link sensors, reported October 2023 ([source](https://www.shephardmedia.com/news/air-warfare/anduril-takes-20-million-uk-mod-contract-to-enhance-force-protection-and-counter-intrusion-capabilit/), accessed 2026-10-31); gov.uk news story ([source](https://www.gov.uk/government/news/17million-contract-awarded-for-force-protection-technology), accessed 2026-10-31). No new 2025-26 UK Lattice contract found. | Autonomy in systems; software agents not stated | OBSERVED (secondary) / UNKNOWN for 2025-26 | | **Scale AI (Thunderforge)** | US DIU prototype contract, 5 March 2025. "AI agents simulate wargaming and planning scenarios and refine proposed courses of action"; includes Anduril Lattice and Microsoft-enabled LLMs ([source](https://www.diu.mil/latest/dius-thunderforge-project-to-integrate-commercial-ai-powered-decision-making), accessed 2026-10-31). The DIU release does not explain how human oversight works; Scale describes operators moving from "in the loop" to "on the loop" ([source](https://www.theregister.com/2025/03/05/dod_taps_scale_to_bring/), accessed 2026-10-31). | Yes | OBSERVED (DIU); oversight wording CLAIM | | **Microsoft (Security Copilot, Defender)** | Announced at Ignite 2025 that Security Copilot (with security agents) is included for Microsoft 365 E5 customers ([source](https://www.microsoft.com/en-us/security/blog/2025/11/18/agents-built-into-your-workflow-get-security-copilot-with-microsoft-365-e5/), accessed 2026-10-31). Availability for US government clouds (GCC, GCC High) could not be confirmed from a Microsoft primary source ([source](https://samexpert.com/security-copilot-licensing-guide/), accessed 2026-10-31). | Yes | CLAIM / UNKNOWN for government clouds | | Microsoft UK government agreements, Azure Government | Not researched (search budget). | — | UNKNOWN | | **Oracle** | April 2025 federal release: AI infrastructure for "Top Secret, Secret, IL5+, and FedRAMP" environments and "AI Agent Studio" for Fusion apps ([source](https://www.oracle.com/news/announcement/oracle-delivers-ai-to-increase-efficiency-agility-and-success-at-federal-agencies-2025-04-15/), accessed 2026-10-31). September 2025 UK release on AI for UK government and defence, including an OCI AI Agent Platform ([source](https://www.oracle.com/uk/news/announcement/oracle-continues-deliver-five-billion-investment-plans-new-ai-capabilities-for-uk-government-and-defence-2025-09-17/), accessed 2026-10-31). | Yes (platform) | CLAIM | | **IBM** | D&B Ask Procurement (November 2024) was built on IBM watsonx Orchestrate ([source](https://www.silicon.co.uk/press-release/dun-bradstreet-accelerates-procurement-and-supplier-risk-decisions-with-ai-assistant-powered-by-world-class-data-and-ibm-technology), accessed 2026-10-31). Wider IBM public-sector offer not researched. | Yes (platform) | CLAIM / UNKNOWN | | **Accenture, Deloitte (public sector)** | Accenture is in the NHS FDP consortium (row above). Both are launch partners of CrowdStrike's AgentWorks ecosystem (March 2026) ([source](https://ir.crowdstrike.com/news-releases/news-release-details/crowdstrike-launches-charlotte-ai-agentworks-ecosystem-building), accessed 2026-10-31). Deloitte agreed in October 2025 to partly refund the Australian government after a report (part-made with a generative AI model) contained made-up references ([source](https://www.cfodive.com/news/deloitte-refunds-60k-report-ai-errors-australian-government-accounting/803321/), accessed 2026-10-31). Wider public-sector AI offers not researched. | Partner role | OBSERVED (secondary) / UNKNOWN | ###### 2.2 Procurement and supplier-vetting platforms | Company / product | What they offer, with evidence | Agents? | Label | |---|---|---|---| | **Tussell** | UK public procurement data: tenders, awards, framework call-offs, spend; API and CRM feeds ([source](https://tussell.com/integrations?hsLang=en), accessed 2026-10-31). Publishes an AI Procurement Tracker of AI contracts ([source](https://tussell.com/ai-procurement-live-feed-data-download), accessed 2026-10-31). Listed on G-Cloud 14 ([source](https://www.applytosupply.digitalmarketplace.service.gov.uk/g-cloud-14/services/430518247842905), accessed 2026-10-31). | Not stated | CLAIM | | **Stotles** | UK/Ireland tender discovery and AI bid tools. Says its AI "reads the tender pack cover to cover, scoring viability, flagging the risks, and citing every requirement back to the source" and drafts first bids ([source](https://stotles.com/stotles-ai), accessed 2026-10-31). Raised a $13M Series A ([source](https://www.actoncapital.com/news/stotles-raises-13-million-series-a-to-help-suppliers-succeed-in-the-new-age-of-government-efficiency), accessed 2026-10-31). | Assistant; cites sources | CLAIM | | **JAGGAER (JAI)** | "Procurement-native orchestrator of AI agents"; claims "Full audit logs and interaction traceability" and "JAI suggests. You decide. Always." ([source](https://www.jaggaer.com/solutions/jai), accessed 2026-10-31). JAI Assist available June 2025; autonomous "Autopilot" described as a later step ([source](https://www.businesswire.com/news/home/20250624469932/en), accessed 2026-10-31). | Yes (partly early access) | CLAIM | | **Coupa (Navi)** | First Navi agents May 2025 ([source](https://cpostrategy.media/blog/2025/05/14/coupa-launches-next-gen-agentic-ai-tools-to-automate-spend-management/), accessed 2026-10-31). By 2026 says 50+ agents and external AI access through MCP ([source](https://procurementmag.com/news/coupa-scales-agentic-procurement-with-mcp-tools), accessed 2026-10-31). | Yes | CLAIM | | **SAP Ariba (Joule agents)** | Joule agents for bid analysis and intake planned/rolling out across Ariba in 2026 ([source](https://erp.today/sap-joule-agents-ariba-fieldglass-procurement-automation-2026/), accessed 2026-10-31). Dates have slipped from 2025 to 2026 ([source](https://procurementmag.com/technology-and-ai/sap-sapphire-business-ai-targets-30-productivity-gains), accessed 2026-10-31). | Yes (rolling out) | CLAIM | | **Achilles (UVDB)** | Runs UVDB, a supplier qualification system for regulated utility procurement, with a 2025 Find a Tender notice ([source](https://www.find-tender.service.gov.uk/Notice/006005-2025), accessed 2026-10-31). | Not stated | OBSERVED (notice) | | **JOSCAR (Hellios)** | The Joint Supply Chain Accreditation Register for UK defence, security and aerospace suppliers is run by Hellios, not Achilles ([source](https://hellios.com/hubfs/2025%20Joint%20Supply%20Chain%20Accreditation%20Register%20A4%20FINAL.pdf?hsLang=en), accessed 2026-10-31). | Not stated | CLAIM | | **Dun & Bradstreet** | 18 June 2026: agentic compliance and third-party-risk capabilities in D&B Risk Analytics, also via MCP in Anthropic's Claude ([source](https://www.dnb.com/en-us/newsroom/press-releases/dnb-introduces-agentic-ai-capabilities-to-accelerate-compliance-and-third-party-risk-workflows-from-days-to-seconds.html), accessed 2026-10-31). | Yes | CLAIM | | **Moody's / Bureau van Dijk (Orbis)** | Orbis private-company data plugs into Ivalua by API ([source](https://www.moodys.com/web/en/us/about-us/partners-marketplace/ivalua.html), accessed 2026-10-31) and into SAP Ariba through a paid connector ([source](https://www.sap.com/products/spend-management/partners/bureau-van-dijk-editions-electroniques-moodys-connector-for-sap-ariba.html), accessed 2026-10-31). No 2025-26 English-language announcement of Orbis agents found. | Not confirmed | CLAIM / UNKNOWN | | **Sayari (Graph)** | Corporate ownership and trade-risk graph ([source](https://sayari.com/platform/graph/), accessed 2026-10-31); claims "500M+ resolved companies" and sanctions updates within four hours of a list change ([source](https://sayari.com/?p=246), accessed 2026-10-31). | Not stated | CLAIM | ###### 2.3 Cyber governance and attack-surface platforms | Company / product | What they offer, with evidence | Agents? | Label | |---|---|---|---| | **CrowdStrike (Charlotte AI)** | Charlotte Agentic SOAR (November 2025) coordinates CrowdStrike, customer-built and third-party agents with "operational guardrails" ([source](https://securitybrief.co.uk/story/crowdstrike-unveils-agentic-soar-expands-falcon-s-ai-agents), accessed 2026-10-31). AgentWorks ecosystem launched March 2026 ([source](https://ir.crowdstrike.com/news-releases/news-release-details/crowdstrike-launches-charlotte-ai-agentworks-ecosystem-building), accessed 2026-10-31). | Yes | CLAIM | | **Palo Alto Networks (Cortex AgentiX)** | Launched 28 October 2025 as next-generation Cortex XSOAR; claims "role-based access controls, the option for human approval on impactful actions, and full auditability for all agent activity", 1,000+ integrations and MCP support ([source](https://investors.paloaltonetworks.com/news-releases/news-release-details/palo-alto-networks-unveils-cortex-agentix-build-deploy-and), accessed 2026-10-31). | Yes | CLAIM | | **Microsoft Defender / Security Copilot** | See 2.1. | Yes | CLAIM | | **Darktrace** | Sells autonomous response (formerly "Antigena") and Cyber AI Analyst; NHS trust adoption reported in 2018 ([source](https://www.digitalhealth.net/2018/08/nhs-trusts-sign-up-to-darktrace-threat-hunting-software/), accessed 2026-10-31). No 2025-26 public-sector evidence found. | Autonomous response | OBSERVED (secondary, dated) / UNKNOWN for now | | **Recorded Future** | Mastercard completed its acquisition on 20 December 2024; Recorded Future works with "over 1,900 businesses and government organizations" ([source](https://www.businesswire.com/news/home/20241220165251/en/Mastercard-Finalizes-Acquisition-of-Recorded-Future), accessed 2026-10-31). | Not researched | OBSERVED / CLAIM | | **Wiz** | Google's acquisition closed 11 March 2026; Wiz joins Google Cloud ([source](https://www.clearygottlieb.com/news-and-insights/news-listing/google-completes-32-billion-acquisition-of-wiz), accessed 2026-10-31). | Not researched | OBSERVED (secondary) | | **Vanta** | AI Agent launched June 2025 (policy and evidence work), with autonomy planned "while keeping you informed and in control at key moments" ([source](https://www.helpnetsecurity.com/2025/06/11/vanta-ai-agent/), accessed 2026-10-31). "Vanta Agents" announced March 2026 ([source](https://www.businesswire.com/news/home/20260319211415/en/Vantas-New-Agents-and-Enterprise-Controls-Eliminate-Audit-Chaos), accessed 2026-10-31). | Yes | CLAIM | | **Drata** | Not researched (search budget). | — | UNKNOWN | | **OneTrust** | AI agents announced at TrustWeek 2025 ([source](https://www.onetrust.com/news/onetrust-announces-ai-agents-at-trustweek-2025/), accessed 2026-10-31). In September 2026 launched "CORIE" to enforce policy and record agent decisions; its survey says 87% of organisations encourage agent use but 47% have clear governance ([source](https://cybermagazine.com/globenewswire/3371211), accessed 2026-10-31). | Yes, plus agent governance | CLAIM | ###### 2.4 Defence and government logistics | Company / product | What they offer, with evidence | Agents? | Label | |---|---|---|---| | **SAP (Defence and Security)** | S/4HANA for defence logistics (supply chain, warehouse, maintenance, force deployment) with "Joule Assistants and Joule Agents" ([source](https://www.sap.com/canada/industries/defense-security.html), accessed 2026-10-31). No named defence deployment of Joule agents found. | Marketed | CLAIM | | **Oracle** | Defence cloud and agent platform (see 2.1). October 2025 US Air Force release on Fusion Cloud applications ([source](https://www.oracle.com/news/announcement/ai-world-united-states-department-of-the-air-force-advances-mission-critical-operations-with-oracle-fusion-cloud-applications-2025-10-15/), accessed 2026-10-31). | Platform | CLAIM | | **Palantir** | See 2.1 (MoD partnership covers decision support, not specifically logistics). | — | OBSERVED | | **IFS** | Aerospace and defence maintenance, repair and overhaul, and spare-parts logistics ([source](https://ifs.com/en/industries/aerospace-and-defense), accessed 2026-10-31). Supports Affinity on the UK Military Flying Training System (about £1.2 million for the first six years) ([source](https://news.cision.com/ifs/r/ifs-supports-affinity-in-uk-mod-military-flying-training-system,c2005053), accessed 2026-10-31). | Marketed "industrial AI" | CLAIM | ###### 2.5 Education and research platforms | Company / product | What they offer, with evidence | Agents? | Label | |---|---|---|---| | **Jisc** | Runs national licensing for UK education, for example a CHEST agreement for the LearnWise AI support chatbot from 1 March 2025 to 29 February 2028 ([source](https://www.chest.ac.uk/agreements/learnwise-ai/), accessed 2026-10-31); published an AI marking and feedback pilot in May 2026 ([source](https://nationalcentreforai.jiscinvolve.org/wp/2026/05/20/insights-from-the-ai-in-marking-and-feedback-pilot/), accessed 2026-10-31). | Pilots | OBSERVED | | **Canvas (Instructure)** | IgniteAI launched July 2025, "Powered by AWS Bedrock" ([source](https://www.aap.com.au/aapreleases/cision20250723ae36732), accessed 2026-10-31); follow-up release ([source](https://www.nasdaq.com/press-release/instructure-delivers-safe-simple-ai-promise-igniteai-and-major-ecosystem-updates-2025), accessed 2026-10-31); "IgniteAI Agent" released later, partly paid add-on ([source](https://onedtech.philhillaa.com/p/instructure-releases-the-igniteai-agent-now-the-real-test-begins), accessed 2026-10-31). Canvas LMS code is AGPL-3.0 (section 5). | Yes | CLAIM | | **Moodle** | Open-source learning platform, GPL-3.0, highest tag v5.3.0 (section 5). AI features not researched. | Unknown | OBSERVED (licence) / UNKNOWN | | **Elsevier Pure** | Research information system; "Pure Impact AI" builds generative-AI impact narratives from Pure, Scopus, Overton, PlumX and patent data ([source](https://www.elsevier.com/promotions/impact-ai), accessed 2026-10-31). On G-Cloud 14 ([source](https://www.applytosupply.digitalmarketplace.service.gov.uk/g-cloud-14/services/839883258276840), accessed 2026-10-31). | Generative, not agentic | CLAIM | | **Researchfish** | Owned by Elsevier since 2022; Elsevier will close it on 31 July 2027; no replacement named; Wellcome is building its own tool ([source](https://www.timeshighereducation.com/node/741136), accessed 2026-10-31). UKRI ran a procurement for a successor service ([source](https://arma.ac.uk/?p=213891), accessed 2026-10-31). | n/a | OBSERVED (secondary) | | University-industry collaboration platforms (e.g. IN-PART) | Not researched (search budget). | — | UNKNOWN | ###### 2.6 Glass-box versus black-box: the gap | Question an auditor would ask | What vendors say | Gap | Label | |---|---|---|---| | Who authorised this agent and what may it do? | Palo Alto: RBAC and optional human approval ([source](https://investors.paloaltonetworks.com/news-releases/news-release-details/palo-alto-networks-unveils-cortex-agentix-build-deploy-and), accessed 2026-10-31); Palantir: ontology as control layer ([source](https://futurumgroup.com/insights/palantir-q1-fy-2026-revenue-beats-estimates-us-demand-drives-outlook-raise/), accessed 2026-10-31). | Mandates live inside each vendor's product, in each vendor's own format. | CLAIM + ANALYSIS | | Can I see every step the agent took? | JAGGAER: "full agent action logs traceable to source" ([source](https://www.jaggaer.com/solutions/jai), accessed 2026-10-31); OneTrust CORIE records agent decisions ([source](https://cybermagazine.com/globenewswire/3371211), accessed 2026-10-31). | Logs are visible to the customer's admin, not by default to the supplier being assessed, the public, or a regulator. | CLAIM + ANALYSIS | | Was the human really in charge? | Thunderforge: DIU release silent on oversight; Scale says "on the loop" ([source](https://www.diu.mil/latest/dius-thunderforge-project-to-integrate-commercial-ai-powered-decision-making), accessed 2026-10-31). Palantir MoD release does not mention human control ([source](https://www.gov.uk/government/news/new-strategic-partnership-to-unlock-billions-and-boost-military-ai-and-innovation), accessed 2026-10-31). | Public statements do not say who can stop the agent, or how. | OBSERVED | | Did AI write this deliverable? | Deloitte's Australian report disclosed AI use only in the revised version ([source](https://www.cfodive.com/news/deloitte-refunds-60k-report-ai-errors-australian-government-accounting/803321/), accessed 2026-10-31). | Disclosure is after the fact and voluntary. | OBSERVED (secondary) | **ANALYSIS: bare minimum vs ambitious for EDGE Public Consortium** | Area | Bare minimum (to be credible) | Ambitious (to lead) | |---|---|---| | Procurement intelligence | Read Find a Tender OCDS data; show notices, awards and supplier history; show the source of every number. | Agent-assisted evaluation where every scoring step is recorded, readable by the supplier and the regulator, and PPN 017 disclosures are captured as structured data. | | Supplier vetting / know-your-domain | Companies House + BODS ownership + sanctions lists, with evidence links. | A shared, consented vetting record a supplier owns and can reuse across buyers (like JOSCAR, but open and agent-readable). | | Cyber governance | Map controls to CAF, Cyber Essentials and Def Stan 05-138; ingest STIX threat data. | Agents that act in SOC workflows only under a recorded mandate with a kill switch, meeting the NCSC "understand, monitor, contain" test, with exportable evidence. | | Logistics | Read-only dashboards over existing ERP data. | Agents that propose movements and orders with a full decision record, accepted by a named human. | | Education and research | LTI and xAPI integration; research outputs linked to funders. | An open successor to Researchfish-style outcome reporting, with agents that draft but never submit without a human. | --- ##### 3. Regulators and rules Status key: **AGREEMENT** (in force, with date) or **PROPOSED** (not yet in force). ###### 3.1 UK public procurement | Rule | Status and date | What it means for EDGE | Source | |---|---|---|---| | Procurement Act 2023 | **AGREEMENT.** Applies to procurements started on or after 24 February 2025. | New notices, exclusion and debarment rules. | ([source](https://www.legislation.gov.uk/ukpga/2023/54/contents), accessed 2026-10-31); start date ([source](https://www.gov.uk/government/publications/ppn-017-improving-transparency-of-ai-use-in-procurement/ppn-017-improving-transparency-of-ai-use-in-procurement-html), accessed 2026-10-31) | | Central Digital Platform (enhanced Find a Tender) | **AGREEMENT.** Live from 24 February 2025. | The single public source of UK notices. | ([source](https://assets.publishing.service.gov.uk/media/6720fe0a3758e4604742a8ab/20241029_CDP_Factsheet_v1.0.pdf), accessed 2026-10-31) | | CDP data format and licence | **AGREEMENT.** Information is published in line with OCDS and reusable under the Open Government Licence 3.0. | EDGE can ingest it without a commercial licence. | ([source](https://www.gov.uk/government/publications/procurement-act-2023-guidance-documents-procure-phase/guidance-central-digital-platform-and-publication-of-information-html), accessed 2026-10-31) | | Debarment list | **AGREEMENT (regime) / UNKNOWN (entries).** Handled by a Debarment Review Service; a snapshot of a gov.uk page says it moved to the Government Commercial Agency in April 2026. We found no source naming a first debarred supplier. | Vetting must check the list. | ([source](https://govdiff.njk.onl/update/2026-06-01T12:10:00+01:00/www.gov.uk/guidance/debarment-review-service-drs), accessed 2026-10-31); ([source](https://www.osborneclarke.com/insights/regulatory-outlook-january-2026-regulated-procurement), accessed 2026-10-31) | | PPN 017: AI use in procurement | **AGREEMENT.** Published 17 February 2025 (replaces PPN 02/24, no policy change). Suggests suppliers disclose AI use; disclosure questions "should not be scored"; "decisions are made with the support of AI systems, not a reliance upon them". Does not mention agents. | EDGE should capture AI-use disclosure as structured data. | ([source](https://www.gov.uk/government/publications/ppn-017-improving-transparency-of-ai-use-in-procurement/ppn-017-improving-transparency-of-ai-use-in-procurement-html), accessed 2026-10-31) | | PPN 014: Cyber Essentials | **AGREEMENT.** Replaces PPN 09/23 for procurements started on or after 24 February 2025; requires Cyber Essentials or Cyber Essentials Plus (or equivalent) for certain higher-risk contracts. | Supplier vetting must record certificate status. | ([source](https://www.gov.uk/government/publications/ppn-014-cyber-essentials-scheme/ppn-014-cyber-essentials-scheme-html), accessed 2026-10-31) | | Public Contracts Regulations 2015 | **AGREEMENT (legacy).** Still govern procurements started before 24 February 2025 (as PPN 014 refers older procurements to PPN 09/23). | Old contracts still in force follow old rules. | ([source](https://www.gov.uk/government/publications/ppn-014-cyber-essentials-scheme/ppn-014-cyber-essentials-scheme-html), accessed 2026-10-31) | ###### 3.2 UK cyber | Rule | Status and date | What it means | Source | |---|---|---|---| | NCSC Cyber Assessment Framework (CAF) v4.0 | **AGREEMENT (guidance).** Released 6 August 2025; adds coverage of AI-related cyber risks, attacker understanding and secure software. Outcome-based, not a certification. | EDGE cyber controls should map to CAF objectives. | ([source](https://www.ncsc.gov.uk/blog-post/caf-v4-0-released-in-response-to-growing-threat), accessed 2026-10-31); ([source](https://www.ncsc.gov.uk/collection/cyber-assessment-framework), accessed 2026-10-31) | | GovAssure | **AGREEMENT (government policy), details UNKNOWN.** GovAssure is the government's CAF-based cyber assurance scheme, but security.gov.uk refused automated fetches, so we could not confirm which CAF version the 2026-27 cycle uses. NCSC says CAF is used across government ([source](https://www.ncsc.gov.uk/collection/cyber-assessment-framework), accessed 2026-10-31). | Government customers are assessed against CAF. | See row text | | Network and Information Systems Regulations 2018 (UK NIS) | **AGREEMENT.** In force (UK SI 2018/506). | Current UK law for essential services and digital service providers. | ([source](https://www.legislation.gov.uk/uksi/2018/506/contents), accessed 2026-10-31) | | Cyber Security and Resilience (Network and Information Systems) Bill | **PROPOSED.** Introduced 12 November 2025; Commons second reading 6 January 2026; Lords report stage 26 October 2026, with two government defeats; Lords third reading scheduled 2 November 2026, then back to the Commons. Not law yet. | Expected to widen NIS to managed service providers and critical suppliers. | ([source](https://www.parliament.uk/business/news/2026/october-2026/cyber-security-and-resilience-network-and-information-systems-bill-undergoes-further-scrutiny-in-the-lords/), accessed 2026-10-31; page refused fetch, read via search extract); ([source](https://vinciworks.com/blog/cyber-security-and-resilience-bill-government-defeated-twice-as-lords-demand-stronger-protections/), accessed 2026-10-31); ([source](https://commonslibrary.parliament.uk/research-briefings/cbp-10442/), accessed 2026-10-31) | ###### 3.3 UK defence and vetting | Rule | Status and date | What it means | Source | |---|---|---|---| | Def Stan 05-138 Issue 4 (Cyber Security for Defence Suppliers) | **AGREEMENT.** Issue 4 dated 14 May 2024. | Lists cyber controls per Cyber Risk Profile. | ([source](https://assets.publishing.service.gov.uk/media/664f42a7993111924d9d3ad8/Def_Stan_05_138_issue_4.pdf), accessed 2026-10-31) | | MoD Cyber Security Model v4 and DEFCON 658 | **AGREEMENT.** CSM v4 "now live" (page update 27 October 2025); "Existing and new procurements should now use CSMv4"; four Cyber Risk Profiles (Level 0 to Level 3); DEFCON 658 carries the terms into contracts and flows down to subcontractors; Defence Cyber Certification (run by IASME) can evidence compliance. | Defence suppliers on EDGE need a CSM profile record. | ([source](https://www.gov.uk/guidance/cyber-security-model), accessed 2026-10-31) | | JSP 936 Part 1: Dependable AI in Defence | **AGREEMENT (MoD policy).** Published 13 November 2024 (v1.1); "principal policy framework governing the safe and responsible adoption of AI in MOD". | Any EDGE agent sold into MoD must fit JSP 936. | ([source](https://www.gov.uk/government/publications/jsp-936-dependable-artificial-intelligence-ai-in-defence-part-1-directive), accessed 2026-10-31) | | JSP 440 (Defence Manual of Security) | Not checked in this session. | — | **UNKNOWN** | | BPSS and National Security Vetting levels | **AGREEMENT.** BPSS is "not a clearance" (identity, right to work, employment history, unspent convictions); NSV levels include CTC, SC, eSC, DV, eDV; page last updated 23 September 2024. | All vetting is defined for people, not agents. | ([source](https://www.gov.uk/government/publications/united-kingdom-security-vetting-clearance-levels/national-security-vetting-clearance-levels), accessed 2026-10-31) | ###### 3.4 UK data, AI and identity | Rule | Status and date | What it means | Source | |---|---|---|---| | Data (Use and Access) Act 2025: automated decisions (new UK GDPR Articles 22A to 22D) | **AGREEMENT.** In force 5 February 2026 (SI 2026/82). A decision is "solely automated" if there is no "meaningful human involvement". | Agents making significant decisions about people need safeguards. | ([source](https://legislation.gov.uk/uksi/2026/82), accessed 2026-10-31); ([source](https://www.traverssmith.com/knowledge/knowledge-container/uks-data-protection-reforms-take-effect-a-new-era-for-automated-decision-making/), accessed 2026-10-31) | | ICO statutory code on AI and automated decision-making | **UNKNOWN.** We could not confirm from a primary source whether a statutory ICO code exists yet. | Watch for it before building automated decisions about people. | No source confirmed | | ATRS (Algorithmic Transparency Recording Standard) | **AGREEMENT (policy).** Mandated across government departments 6 February 2024; scope policy 17 December 2024 covers tools with "significant influence on a decision-making process with public effect". | Government buyers using EDGE agents may need to publish a record. | ([source](https://www.gov.uk/government/publications/algorithmic-transparency-recording-standard-mandatory-scope-and-exemptions-policy/algorithmic-transparency-recording-standard-atrs-mandatory-scope-and-exemptions-policy), accessed 2026-10-31) | | Companies House identity verification | **AGREEMENT.** Mandatory for new directors and people with significant control from 18 November 2025, with a transition for existing directors. | Vetting can rely on verified directors. | ([source](https://cms.law/en/gbr/legal-updates/economic-crime-and-corporate-transparency-act-2023-identity-verification-became-mandatory-from-18-november-2025), accessed 2026-10-31) | ###### 3.5 Education and research (England / UK) | Rule | Status and date | What it means | Source | |---|---|---|---| | Online Safety Act: Protection of Children Codes | **AGREEMENT.** Measures applied from 25 July 2025, including "highly effective age assurance" for the riskiest services. | Any student-facing user-to-user feature is in scope. | ([source](https://www.orrick.com/en/Insights/2025/07/Ofcom-New-Protection-of-Children-Codes-Key-Compliance-Takeaways), accessed 2026-10-31) | | DfE Generative AI: product safety standards | **AGREEMENT (guidance).** Updated 19 January 2026; 13 standards on filtering, monitoring, security, data protection, design and governance. | Any AI tool sold to schools is checked against these. | ([source](https://www.gov.uk/government/publications/generative-ai-product-safety-standards/generative-ai-product-safety-standards), accessed 2026-10-31) | | KCSIE 2026 | **AGREEMENT.** Published 7 July 2026, in effect 1 September 2026; names generative AI as a safeguarding risk and points to the DfE product safety standards. | Schools must review filtering and monitoring at least yearly. | ([source](https://edtechinnovationhub.com/news/dfe-publishes-kcsie-2026-with-mobile-phone-free-default-and-ai-safeguarding-rules), accessed 2026-10-31); ([source](https://swgfl.org.uk/magazine/keeping-children-safe-in-education-2026-what-do-schools-need-to-know/), accessed 2026-10-31) | | Office for Students governance conditions | **PROPOSED.** OfS board paper (13 May 2026) planned a consultation on reforming ongoing "E" (management and governance) conditions in September 2026; a free-speech condition E11 is out for consultation 17 September to 23 November 2026. | University partners' governance rules may change. | ([source](https://www.officeforstudents.org.uk/media/qp5di0zn/bd-2026-may-item-6-development-of-approach-to-the-regulation-of-governance.pdf), accessed 2026-10-31); ([source](https://www.brownejacobson.com/insights/the-ofs-free-speech-conditions-what-universities-need-to-do-now), accessed 2026-10-31) | | Concordat to Support Research Integrity (2025) | **AGREEMENT (sector concordat).** Refreshed version published 4 April 2025; signatories should follow it from 1 April 2026; signatories publish an annual statement; secretariat moved to UKCORI. | Research consortia on EDGE need integrity records. | ([source](https://ukcori.org/research-integrity-concordat/), accessed 2026-10-31); ([source](https://ukrio.org/wp-content/uploads/UKRIO-Briefing-doc-on-2025-Concordat.pdf), accessed 2026-10-31) | ###### 3.6 EU and US | Rule | Status and date | What it means | Source | |---|---|---|---| | EU NIS2 Directive (EU) 2022/2555 | **AGREEMENT (directive),** but national laws are uneven: Commission infringement steps since late 2024; one tracker estimates about 20 of 27 states had transposed by January 2026. | EU customers' duties depend on their country's law. | Legal text ([source](https://eur-lex.europa.eu/eli/dir/2022/2555/oj), accessed 2026-10-31); status ([source](https://www.wavestone.com/en/insight/nis-2-european-countries-transposing-directive/), accessed 2026-10-31) | | EU Cyber Resilience Act (Regulation (EU) 2024/2847) | **AGREEMENT.** Entered into force 10 December 2024; reporting of actively exploited vulnerabilities and severe incidents from 11 September 2026 (24-hour early warning, 72-hour notification) through ENISA's Single Reporting Platform; most other duties from 11 December 2027. | Any EDGE software sold in the EU needs vulnerability handling and SBOMs. | ([source](https://digital-strategy.ec.europa.eu/en/policies/cra-reporting), accessed 2026-10-31); ([source](https://www.mccannfitzgerald.com/knowledge/data-privacy-and-cyber-risk/cyber-resilience-act-reporting-obligations-apply-from-11-september-2026), accessed 2026-10-31) | | EU AI Act high-risk rules (education is Annex III) | **PROPOSED change.** Provisional "Digital Omnibus" deal on 7 May 2026 would move Annex III high-risk duties (including education) to 2 December 2027. Formal adoption: **UNKNOWN**. | EU education customers' timelines may shift. | ([source](https://www.gibsondunn.com/eu-ai-act-omnibus-agreement-postponed-high-risk-deadlines/), accessed 2026-10-31); ([source](https://datamatters.sidley.com/2026/06/22/eu-lawmakers-reach-provisional-agreement-to-delay-key-eu-ai-act-obligations/), accessed 2026-10-31) | | US CMMC (DFARS final rule) | **AGREEMENT.** Effective 10 November 2025 (Phase 1, self-assessments); Phase 2 (third-party Level 2 assessments) from 10 November 2026. | US defence work needs a CMMC level. | ([source](https://www.squirepattonboggs.com/en/insights/publications/2025/09/the-cmmc-dfars-final-rule-goes-live-ready-or-not-here-it-comes), accessed 2026-10-31) | | US FedRAMP 20x | **PROPOSED / pilot.** Search extract of fedramp.gov says Phase 2 pilot completed March 2026 with 8 Moderate pilot authorisations, and Phase 3 began April 2026; the timeline page returned 404 when re-checked. | US federal cloud route is changing. | ([source](https://www.fedramp.gov/20x/timeline/), accessed 2026-10-31); ([source](https://www.nextgov.com/acquisition/2025/09/gsa-launches-second-phase-fedramp-20x-backed-omb/408356), accessed 2026-10-31) | | US OMB M-25-21 (federal AI use) | **AGREEMENT (policy).** Issued April 2025; "high-impact AI" needs pre-deployment testing, impact assessment and ongoing monitoring. | US public-sector agents need these records. | ([source](https://www.whitehouse.gov/wp-content/uploads/2025/02/M-25-21-Accelerating-Federal-Use-of-AI-through-Innovation-Governance-and-Public-Trust.pdf), accessed 2026-10-31) | --- ##### 4. Agent assurance gaps in public sector, defence and education | Gap | Evidence | Label | |---|---|---| | **Transparency records do not name agents.** | The ATRS scope policy does not mention AI agents or agentic systems ([source](https://www.gov.uk/government/publications/algorithmic-transparency-recording-standard-mandatory-scope-and-exemptions-policy/algorithmic-transparency-recording-standard-atrs-mandatory-scope-and-exemptions-policy), accessed 2026-10-31). | OBSERVED | | **Few transparency records exist.** | 59 ATRS records were published by May 2025, 26 from central government ([source](https://dataingovernment.blog.gov.uk/2025/05/08/making-the-algorithmic-transparency-recording-standard-atrs-mandatory-across-government/), accessed 2026-10-31). | OBSERVED | | **Procurement AI guidance covers suppliers' bids, not buyers' agents.** | PPN 017 does not mention agents and gives little on AI used in evaluation ([source](https://www.gov.uk/government/publications/ppn-017-improving-transparency-of-ai-use-in-procurement/ppn-017-improving-transparency-of-ai-use-in-procurement-html), accessed 2026-10-31). | OBSERVED | | **Human control is a principle, not a mechanism.** | The AI Playbook (February 2025) asks for "meaningful human control at the right stage" ([source](https://www.publictechnology.net/2025/02/12/education-and-skills/government-ai-playbook-instructs-departments-to-know-techs-limitations-and-have-meaningful-human-control/), accessed 2026-10-31); full text ([source](https://gov.uk/government/publications/ai-playbook-for-the-uk-government/artificial-intelligence-playbook-for-the-uk-government-html), accessed 2026-10-31). | AGREEMENT (guidance) | | **Security guidance asks for least privilege and named accountable humans.** | NCSC, 15 May 2026, "Careful adoption of agentic AI services": "give agents only the minimum access they need, for the shortest time required"; "humans remain accountable"; "If you cannot understand, monitor or contain an agent's actions, it is not ready for deployment." ([source](https://www.ncsc.gov.uk/blogs/thinking-carefully-before-adopting-agentic-ai), accessed 2026-10-31). | AGREEMENT (guidance) | | **Government says agents are its hardest area.** | GDS leader (May 2026): agents are where "we've got the biggest work to do" ([source](https://www.publictechnology.net/2026/05/28/society-and-welfare/gds-leader-picks-ai-agents-as-area-with-biggest-work-to-do-but-most-potential/), accessed 2026-10-31). | OBSERVED (secondary) | | **Unclear ownership of AI in government.** | NAO (March 2024): no coherent cross-government plan; responsibility split between Cabinet Office and DSIT ([source](https://www.nao.org.uk/wp-content/uploads/2024/03/use-of-artificial-intelligence-in-government-summary.pdf), accessed 2026-10-31). | OBSERVED | | **Vetting is for people only.** | BPSS and NSV checks are identity, employment, criminal record, finances, interviews ([source](https://www.gov.uk/government/publications/united-kingdom-security-vetting-clearance-levels/national-security-vetting-clearance-levels), accessed 2026-10-31). No equivalent for an AI agent acting with a cleared person's access was found. | OBSERVED + UNKNOWN (no agent equivalent found) | | **Defence has a policy owner but no public agent mechanism.** | JSP 936 led each MoD part to nominate a Responsible AI Senior Officer ([source](https://www.gov.uk/government/publications/laying-the-groundwork-responsible-ai-senior-officers-report-2025), accessed 2026-10-31). | OBSERVED | | **Defence AI announcements skip oversight.** | Palantir MoD and DIU Thunderforge releases do not explain human oversight (section 2.1). | OBSERVED | | **Automated-decision law needs "meaningful human involvement".** | DUAA Articles 22A to 22D from 5 February 2026 ([source](https://www.traverssmith.com/knowledge/knowledge-container/uks-data-protection-reforms-take-effect-a-new-era-for-automated-decision-making/), accessed 2026-10-31). | AGREEMENT | | **Organisations adopt agents faster than they govern them.** | OneTrust survey: 87% encourage agent use, 47% have clear governance ([source](https://cybermagazine.com/globenewswire/3371211), accessed 2026-10-31). | CLAIM (vendor survey) | | **Schools: filtering must handle AI content.** | June 2026 DfE filtering and monitoring update asks schools to assess generative-AI risk against the product safety standards ([source](https://smoothwall.com/resources/2026-dfe-filtering-and-monitoring-standards), accessed 2026-10-31). | CLAIM (vendor summary of DfE) | **ANALYSIS: what EDGE must provide for an agent to act in procurement, vetting or cyber operations** 1. **Authority:** a machine-readable mandate per agent (who it acts for, which actions, which data, until when), checked before each action. NCSC's least-privilege and temporary-credential advice is the floor. 2. **Accountability:** a named human owner and a named human who can stop it, recorded before go-live (NCSC's list of accountable roles). 3. **Audit:** an append-only record of each action, its inputs and evidence links, exportable in an open format so a supplier, auditor or regulator can read it outside EDGE. 4. **Explainability:** for any decision with public effect, a draft ATRS record generated from the agent's mandate and log, and PPN 017-style AI-use disclosure captured as data. 5. **Vetting for agents:** an "agent BPSS" equivalent (identity of the builder, the model, the tools, the data access) — no such standard exists today. --- ##### 5. Reusable systems Licences were read from the repository licence file on 2026-10-31. "Fit" is ANALYSIS. | System | What it is | Licence (SPDX id) | Maturity | Fit for EDGE | Label and source | |---|---|---|---|---|---| | **Open Contracting Data Standard (OCDS)** | JSON standard for procurement data. | Apache-2.0 (schema repo) | Highest tag 1.1.5; last commit 2026-10-12 | High: UK Find a Tender publishes OCDS under OGL 3.0. | OBSERVED ([licence](https://raw.githubusercontent.com/open-contracting/standard/master/LICENSE), accessed 2026-10-31); ([repo](https://github.com/open-contracting/standard), accessed 2026-10-31) | | **Beneficial Ownership Data Standard (BODS)** | JSON standard for who owns and controls companies. | Apache-2.0 (schema) | v0.4; pre-1.0, changes expected; last commit 2026-09-02 | High for supplier vetting. | OBSERVED ([licence](https://raw.githubusercontent.com/openownership/data-standard/main/LICENSE), accessed 2026-10-31); ([docs](https://standard.openownership.org/), accessed 2026-10-31) | | **Companies House API** | REST API for UK company data. | Licence not stated on the developer page | Production government service | High for UK vetting; check terms. | OBSERVED / UNKNOWN licence ([source](https://developer.company-information.service.gov.uk/), accessed 2026-10-31) | | **OpenCorporates** | Global company register data. | ODbL-1.0 share-alike for free keys; paid non-share-alike terms for commercial use | Commercial service | Medium: share-alike may bind EDGE outputs; commercial terms cost money. | CLAIM ([source](https://api.opencorporates.com/documentation/Home), accessed 2026-10-31); ([source](https://opencorporates.com/legal-information/self-service-api-terms-of-service/), accessed 2026-10-31) | | **MITRE ATT&CK** | Catalogue of attacker techniques (STIX format). | Custom MITRE royalty-free licence; no standard SPDX id (use `LicenseRef-`) | attack-stix-data highest tag v20.0 | High for cyber governance. | OBSERVED ([licence](https://raw.githubusercontent.com/mitre-attack/attack-stix-data/master/LICENSE.txt), accessed 2026-10-31) | | **MITRE D3FEND** | Catalogue of defensive techniques (ontology). | MIT | Highest tag 1.6.0; last commit 2026-10-15 | High: maps controls to attacks. | OBSERVED ([licence](https://raw.githubusercontent.com/d3fend/d3fend-ontology/master/LICENSE.md), accessed 2026-10-31) | | **STIX 2.1 / TAXII 2.1** | Threat-intelligence format and transport. | OASIS standards; schema and Python library repos carry a BSD-style licence (text matches BSD-3-Clause opening; verify full text) | OASIS Standards since 10 June 2021 | High for fraud and threat sharing networks. | OBSERVED ([spec](https://docs.oasis-open.org/cti/stix/v2.1/stix-v2.1.html), accessed 2026-10-31); ([licence](https://raw.githubusercontent.com/oasis-open/cti-stix2-json-schemas/master/LICENSE), accessed 2026-10-31) | | **OpenCTI** | Open-source threat-intelligence platform built on STIX. | Community Edition Apache-2.0; Enterprise Edition separate licence (files carry headers) | Active; last commit 2026-10-31 | High, but check each file's header. | OBSERVED ([licence](https://raw.githubusercontent.com/OpenCTI-Platform/opencti/master/LICENSE), accessed 2026-10-31) | | **Wazuh** | Open-source security monitoring (SIEM/XDR). | GPL-2.0-only for most code; engine AGPL-3.0-or-later; shared libraries GPL-2.0-only OR AGPL-3.0-or-later | Highest tag v4.14.9; last commit 2026-10-28 | Medium: copyleft; run as a separate service, do not embed. | OBSERVED ([licence](https://raw.githubusercontent.com/wazuh/wazuh/main/LICENSE), accessed 2026-10-31) | | **OSV (Open Source Vulnerabilities)** | Vulnerability database and schema. | Apache-2.0 (osv.dev and osv-schema code); data licence not checked | osv-schema v1.9.2; osv.dev tag v2026.10.27 | High for supplier software risk and CRA. | OBSERVED ([licence](https://raw.githubusercontent.com/ossf/osv-schema/main/LICENSE), accessed 2026-10-31); data licence UNKNOWN | | **SPDX** | SBOM format and licence-id list; SPDX 2.2.1 is ISO/IEC 5962:2021. | Community-Spec-1.0 (specification) | Repo tag v3.0; last commit 2026-10-20 | High: required vocabulary for licences. | OBSERVED ([licence](https://raw.githubusercontent.com/spdx/spdx-spec/main/LICENSE), accessed 2026-10-31); ([ISO](https://www.scc.ca/en/standardsdb/iec/48439), accessed 2026-10-31) | | **CycloneDX** | SBOM format; Ecma standard ECMA-424. | Apache-2.0 | v1.7 (October 2025), adopted as ECMA-424 2nd edition December 2025; highest tag 1.7.2 | High for supplier SBOM evidence. | OBSERVED ([licence](https://raw.githubusercontent.com/CycloneDX/specification/master/LICENSE), accessed 2026-10-31); ([Ecma](https://ecma-international.org/publications-and-standards/standards/ecma-424/), accessed 2026-10-31) | | **GOV.UK Frontend (Design System)** | UK government user-interface components. | MIT | Highest tag v6.6.0; last commit 2026-10-29 | High for regulator and evidence portals (do not imply you are GOV.UK). | OBSERVED ([licence](https://raw.githubusercontent.com/alphagov/govuk-frontend/main/LICENSE.txt), accessed 2026-10-31) | | **GOV.UK One Login** | Government sign-in and identity checking. | MIT (authentication-api repo) | Over 200 government services (February 2026) | Low for EDGE: built for government services; private-sector use not confirmed. | OBSERVED ([licence](https://raw.githubusercontent.com/govuk-one-login/authentication-api/main/LICENCE), accessed 2026-10-31); CLAIM ([source](https://www.gov.uk/government/news/hmrc-introducesgovukonelogin-for-new-customers), accessed 2026-10-31) | | **xAPI** | Records learning activity; xAPI 2.0 is IEEE 9274.1.1-2023. | Apache-2.0 (spec repo) | Repo tags up to xAPI-1.0.3; IEEE standard published 2023 | High for education evidence. | OBSERVED ([licence](https://raw.githubusercontent.com/adlnet/xAPI-Spec/master/LICENSE), accessed 2026-10-31); ([ADL](https://adlnet.gov/news/2023/11/22/xAPI-Officially-Becomes-a-Published-IEEE-Standard), accessed 2026-10-31) | | **LTI 1.3 / LTI Advantage** | Plugs tools into Canvas, Moodle and others. | 1EdTech PHP library Apache-2.0; certification needs paid 1EdTech membership | Widely certified | High for reaching universities. | OBSERVED ([licence](https://raw.githubusercontent.com/1EdTech/lti-1-3-php-library/master/LICENSE), accessed 2026-10-31); ([guide](https://standards.1edtech.org/lti/guides/implementation_guide/implementation-guide), accessed 2026-10-31) | | **Moodle** | Open-source learning platform. | GPL-3.0 (COPYING.txt; "or later" wording not checked) | Highest tag v5.3.0 | Integrate via LTI; do not copy code into EDGE. | OBSERVED ([licence](https://raw.githubusercontent.com/moodle/moodle/main/COPYING.txt), accessed 2026-10-31) | | **Canvas LMS** | Open-source core of Canvas. | AGPL-3.0 | Active | Integrate via LTI only. | OBSERVED ([licence](https://raw.githubusercontent.com/instructure/canvas-lms/master/LICENSE), accessed 2026-10-31) | --- ##### 6. Open questions to check next - **UNKNOWN:** Has the Cyber Security and Resilience Bill passed Lords third reading on 2 November 2026 and received Royal Assent? Check [bills.parliament.uk/bills/4035](https://bills.parliament.uk/bills/4035) (accessed 2026-10-31; refused automated fetch). - **UNKNOWN:** Which CAF version GovAssure uses for 2026-27 (security.gov.uk refused automated fetches on 2026-10-31). - **UNKNOWN:** First entries on the Procurement Act debarment list. - **UNKNOWN:** Formal adoption of the EU AI Act Digital Omnibus. - **UNKNOWN:** Drata, IBM, Accenture, Microsoft UK government agreements, JSP 440, university-industry collaboration platforms (not researched; search budget ran out). ### research/public-consortium/ENGINE-ERPNEXT.md `research/public-consortium/ENGINE-ERPNEXT.md` · sha256 `e7f4b172dd31c1d1173176cc9754f8ae2986b796da0c501bb22518d8fa7da963` #### EDGE Consortium and EDGE Public on the ERPNext engine (the `registry` slot) Research record, 1 November 2026, Claude-Jegudiel 00:5531102026-1. Adapter: `runtime/domains/registry.py`; tests: `tests/test_registry_domain.py` (fake engine, and `LiveSmoke` with `FRAPPE_SMOKE=1` on the host). Every number below was measured on jude-1 on 1 November 2026 between 04:40Z and 05:15Z; anything not measured is marked CLAIM or NOT DONE. ##### The engine OBSERVATION. Frappe 16.54.0 with ERPNext 16.53.0 (the two installed apps), podman pod `jude-frappe-engines` on jude-1, systemd units `jude-engine-frappe-{db,cache,queue,backend,web,websocket,scheduler,worker}.service`, site `frontend`, published at `127.0.0.1:18905` (nginx in the `web` container in front of gunicorn in `backend`). The setup wizard has not been run: `System Settings.setup_complete = 0`, no Company, no Customer Group, Territory, Supplier Group, Item Group or UOM existed before this work. The site's `encryption_key` is in `sites/frontend/site_config.json` on the named volume `jude-frappe-sites`. ##### Mapping Entities are the `estate-record::` labels of `spec/receivers.json` for operations whose unit begins `edge-consortium/` or `public-edge/`. The brief's entity names map to the labels as follows. | Brief entity | Receiver labels (entity part) | DocType | Kind | Why | | --- | --- | --- | --- | --- | | case | `case` | Issue | ERPNext's own | Project requires a Company (`company` reqd=1) and none exists; Issue needs only `subject` and carries Open/Replied/On Hold/Resolved/Closed | | role | `feature_role` | Role | Frappe's own | `role_name` only; named ` <id[:8]>` because `role_name` is the document name and must be unique across tenants | | client | `feature_client` | Customer | ERPNext's own | `customer_name` + `customer_type` (default Company); customer_group and territory are not required | | asset | `asset` | Item | ERPNext's own | Asset requires a Company and a fixed-asset Item; Item needs `item_code` (= the estate record id), `item_group` and `stock_uom`, so the bootstrap creates `All Item Groups` and UOM `Nos` | | alert | `alert` | ToDo | Frappe's own | `description` only; Open/Closed/Cancelled | | module | `feature_module` | Estate Module | custom | no own DocType fits | | access policy | `feature_access`, `feature_policy` | Estate Access Policy | custom | | | egress policy | `feature_egress` | Estate Egress Policy | custom | | | consortium | `consortium` | Estate Consortium | custom | fields from `research-consortium`'s data definition: title, aim, lead_institution, status enum | | application | `application` | Estate Application | custom | the union of both applications' definitions (consortium's personal_statement… and public's cover_note…, status options merged) | | vacancy | `vacancy` | Estate Vacancy | custom | Job Opening belongs to the HRMS app, which is not installed | | decision | `decision` | Estate Decision | custom | | Every mapped DocType carries the estate fields `estate_record_id` (unique), `estate_tok_id`, `estate_tenant`, `estate_unit`, `estate_entity`, `estate_version`, `estate_state`, `estate_deleted`, `estate_body` (the whole record as JSON) and `estate_digest`: as Custom Fields on the five Frappe/ERPNext DocTypes, as fields of the seven custom ones. Custom DocTypes are named by the estate record id (`autoname: field:estate_record_id`); the typed fields of the unit data definitions become Frappe fields (text → Small Text, enum → Select, date, datetime, money_gbp → Currency, number → Float, integer → Int, boolean → Check, id/domain/email/role/sha256 → Data; lists, maps and files → Small Text as JSON). A value that does not fit its typed field is left to `estate_body`, never refused. `title` on a custom DocType is filled from the body's title, name, legal_name, organisation, subject, summary, description, label, step_1 or check_1. Bound operations (measured at import, `registry.status()`): **2,023** of the two applications' operations, by DocType: Issue 1,495; Estate Access Policy 118; Estate Module 100; Role 82; Estate Egress Policy 64; Customer 42; Estate Consortium 32; Estate Application 21; Item 18; Estate Vacancy 18; ToDo 18; Estate Decision 15. Nothing outside `edge-consortium/` and `public-edge/` is bound (test 1 asserts the set exactly). ###### What stays estate-record Everything else of the two applications keeps `runtime/receivers.py`: `record` (3,102 + 1,850 operations: the step-by-step workflow record of every unit), `evidence_query` (559 + 572), and the entities that have no DocType in this build: edge-consortium `partner`, `institution`, `module_enrolment`, `membership`, `agreement`, `funding`, `milestone`, `output`, `outcome`, `programme`, `roster`; public-edge `supplier`, `contract`, `incident`, `report`, `challenge`, `change`, `consignment`, `exception`, `source`, `opportunity`, `bid`, `offer`, `invoice`, `meeting`, `agent`, `check`, `delivery`, `domain`, `evidence_pack`, `issue`, `interest`, `performance`, `reading`, `assessment`, `fix`, `milestone`, `right_to_work`, `clarification`, `containment`, `dispute`, `figure`, `market`, `payslip`, `recommendation`, `recovery`, `search`, `award`, `notice`, `payrun`, and the 8,219 + 908 `feature_*` operations whose entity is not role, module, client, access, policy or egress. `supplier` → Supplier and `institution` → Customer or Supplier are the obvious next mappings; they are not in the brief and not bound. ##### Receiver contract (as the bank ledger) - `confirmed` only after the document is read back from Frappe (`GET /api/resource/<DocType>/<name>`) and its estate fields reproduce `row_digest(tenant, unit, entity, id, version, body, deleted)`, the same digest the estate store uses. The receipt is `{receiver: "estate-registry", engine: "erpnext", doctype, name, tenant, unit, entity, id, version, modified, digest}`; `registry.verify(receipt, tenant)` re-reads it (invented, stale, altered and other-tenant receipts are refused; test 2). - `rejected` with `refusal_code`: `NOT_FOUND` (HTTP 404, or a record that exists under another tenant), `VERSION_CONFLICT` (an `expected_version`/`version` input that is not the current `estate_version`, or Frappe's own `TimestampMismatchError` HTTP 409 when the document changed between our read and our `PUT` carrying `modified`), `DUPLICATE`, `VALIDATION` (417), `FORBIDDEN` (403), `AUTH` (401). - `no_answer`: a socket timeout (`FRAPPE_TIMEOUT`, default 20 s), a connection failure, a 5xx, a read-after-write that did not match, or no credential at all (`PLUG_REQUIRED` in the evidence). The engine records UNKNOWN; nothing is claimed. - Behaviour classes follow the store: `create`; `read` by id or by filter (filter applied over `estate_body`, listed by tenant, unit and entity); `change` (first touch creates, as the store does); `remove` is a soft delete (`estate_deleted = 1`, and the native `disabled = 1` or status Closed/Cancelled: nothing is ever `DELETE`d in Frappe, the evidence stays); `send`, `decide`, `plan` set the state field the entity declares (`EstateStore._state_field/_state_value`, identical values) and add an Info `Comment` on the document's own timeline. A case is never deleted by a step (`effective_class`). - Tenant: always `session.tenant_of(actor)`; every lookup filters `estate_tenant`; a custom DocType's document read by name is still refused when its `estate_tenant` differs. - Demo tenant: `engine.demo and tenant == "demo"` never reaches Frappe. The estate store on the demo engine's own database answers, its receipt is verified the same way, and the outputs carry `receiver: "simulation:registry"` and `evidence: "SIMULATED: demo tenant; kept in the estate record, ERPNext not touched"` (tests 9 and 10). ##### Authentication (no values here, by design) A Frappe System User `estate-registry@x-x.tech` was created through the API with the administrator credential (read from the root-only `/etc/jude-reusable-engines/frappe.env` on the host, never printed), with roles System Manager, Item Manager, Sales Master Manager, Sales User, Support Team, Website Manager (the roles with create+write on DocType, Item/Item Group/UOM, Customer, Issue, Comment; ToDo and Role are covered by System Manager). Its API key and secret (`frappe.core.doctype.user.user.generate_keys`) are in `/etc/x-x/plugs/frappe.env` as `FRAPPE_URL`, `FRAPPE_SITE`, `FRAPPE_API_KEY`, `FRAPPE_API_SECRET`, mode 0640 **root:root**: the brief said `root:x-x`, but the group `x-x` does not exist on jude-1 (`getent group x-x` is empty) and `x-x-web.service` runs as root, so no group was invented. The adapter sends `Authorization: token <key>:<secret>`; the environment variables of the same names override the file (tests use them). OBSERVATION. The first key issued at 04:53Z authenticated once and failed with `AuthenticationError: Failed to decrypt key User.estate-registry@x-x.tech.api_secret ... Encryption key is invalid` after the pod restart at 05:05Z. The key was regenerated at 05:08Z under the `encryption_key` that is now in `site_config.json` (file written 04:53Z, unchanged since) and has authenticated on every call since (status call 70 ms, smoke 316 calls). CLAIM to assess: the first `api_secret` was encrypted by a gunicorn worker that had generated an encryption key in memory before another worker wrote a different one to `site_config.json`, so a restart made it undecryptable; a credential issued now is stable as long as `sites/frontend/site_config.json` on volume `jude-frappe-sites` persists. If the file is ever lost, every stored secret in the site is lost with it: back the volume up with the engine. ##### What happened to the engine while this was built (verified stop causes) - 04:47:01Z the whole pod (and the `jude-medplum-engines` and `jude-vendor-engines` pods) received SIGTERM and exited: `journalctl -u jude-engine-frappe-backend` shows gunicorn "Handling signal: term", "container died", systemd "Deactivated successfully", with no "Stopping" line and three pods going together. Issuer: **not determinable from the journal** (no `systemctl stop`, no OOM, no sudo record in the window). Restarted by this lane at 04:51Z in the order db, cache, queue, backend, websocket, web, scheduler, worker; ping 200. - 04:57:30Z systemd's on-failure restart of `scheduler` and `worker` made podman restart the pod's infra container, which failed with `mounting storage for container ... lstat /var/lib/containers/storage/...` (`internal libpod error`); the pod then read Running but nothing listened on 18905 (the rootlessport forwarder was gone). Fixed at 05:05Z by stopping the eight units, `podman pod stop`, and starting the units in order; `ss -ltnp` shows 18905 again. - The medplum and vendor pods were running again by 04:52Z; this lane did not touch them. ##### Measured round trips (live smoke, 05:12Z, tenant `smoke`) `FRAPPE_SMOKE=1 python3 -m unittest tests.test_registry_domain.LiveSmoke` from the staging copy `/root/registry-stage/new-estate` (release runtime + this adapter), one create, change, read, decide and remove per mapped entity, each confirmed by read-back: **60 of 60 confirmed**, 316 HTTP calls, 12.6 s in all. Per HTTP call: min 8.4 ms, median 20.1 ms, p90 58.4 ms, max 766.8 ms (the first Customer insert). Per receiver operation (several HTTP calls each: locate, write, read back, comment): | entity | DocType | create | change | read | decide | remove | | --- | --- | --- | --- | --- | --- | --- | | case | Issue | 861 ms | 407 | 179 | 630 | 407 | | feature_role | Role | 223 | 428 | 30 | 349 | 324 | | feature_module | Estate Module | 128 | 89 | 14 | 150 | 158 | | feature_client | Customer | 923 | 179 | 46 | 285 | 129 | | feature_access | Estate Access Policy | 139 | 134 | 14 | 109 | 91 | | feature_egress | Estate Egress Policy | 172 | 91 | 17 | 77 | 134 | | consortium | Estate Consortium | 111 | 121 | 21 | 100 | 91 | | application | Estate Application | 121 | 87 | 17 | 94 | 103 | | vacancy | Estate Vacancy | 115 | 101 | 11 | 95 | 140 | | asset | Item | 436 | 119 | 26 | 133 | 216 | | alert | ToDo | 105 | 167 | 27 | 111 | 112 | | decision | Estate Decision | 109 | 106 | 20 | 116 | 95 | Medians over the twelve entities: create 133 ms, change 120 ms, read 20 ms, decide 113 ms, remove 131 ms. The first operation on each of Frappe's own DocTypes is the slowest (ERPNext validations and the first Comment on a new DocType). Bootstrap (`python3 -m runtime.domains.registry bootstrap`): first run created 7 DocTypes, 50 Custom Fields, then (after the roles) Item Group `All Item Groups` and UOM `Nos`; the second run created nothing and found 59 (idempotent). ##### The GPL-3 question — CLAIM, to assess, not a finding Frappe (the framework) is MIT; ERPNext (the app) is GPL-3.0. CLAIM: the estate runtime talks to ERPNext only over its HTTP REST API from a separate process and does not link, import, copy or modify ERPNext code; GPL-3 (unlike AGPL-3) has no network-interaction clause, and the custom DocTypes, Custom Fields and records created through the API are data in ERPNext's database, not a derivative work of ERPNext. On that reading the estate's own code carries no GPL-3 obligation. What would change it: distributing ERPNext itself (the container image, or a modified ERPNext) to third parties, which is covered by GPL-3's source-availability terms; or moving estate logic into a Frappe app installed on the bench. Jude's assessment and, if the estate ever ships the engine to a customer, counsel's, are needed before this CLAIM becomes a decision. ##### The journey walk (step 7 of the brief): run, stopped, cause verified Shipped at 05:17:04Z (`/opt/x-x/deploy/ledger/deploy/2026-11-01T051704Z-edit.json`; the deploy helper restarted x-x-web.service once). From the release, `FRAPPE_SMOKE=1 python3 -B -m unittest tests.test_registry_domain.LiveSmoke`: OK, 60 of 60 confirmed, 316 HTTP calls, 8.4 s. Then `python3 tools/journeys_http.py --demo --only /root/registry-journeys-only.txt --shard 1/1` (1,139 `#human` journeys of `edge-consortium/` and `public-edge/` from `ledger/evidence/plan.json`) was started at 05:18Z and **stopped at 05:33Z after every journey it reached was refused**: `do … -> 403 REFUSED NOT_PERMITTED` (and PLUG_REQUIRED), one journey per ~35 s. Verified cause (x-x-web journal and a reproduction, not a guess): commit ffe02c065, shipped to the release at 05:14:40–05:15:03Z through `runtime/session.py`, `runtime/web.py` and `tools/journeys_http.py`, signs the walker in as its own tenant `demo-walker` (`session.demo_actor(walker=True)`), while `runtime/server.py` `engine_for()` and `handle_api()`, `runtime/guards.py`'s role guard and `runtime/adapters.py` still test `tenant == "demo"`: the walker's requests reach the live engine and are refused. Before 05:17 the journal holds 9,674 `do` calls with 0 NOT_PERMITTED (3,965 OK for the two applications); after 05:17:48 (the next restart, another lane's `tools/finish_loop.py` edit) 104 `do` calls, 0 OK, finance and clinic walks included; no walk ran between 05:14:40 and 05:17:03, so the new walker had never been exercised. A fresh `POST /signin/demo` without `walker=1` followed by the first step of `edge-consortium/add-members-and-guests-with-expiry#human` answers 303 OK on the demo engine (evidence `33438bc4e848c2c68bc5015d0b5305d94b4b1f09dddc93bbaf5aaefef646244b`: mode simulation, tenant demo, receipt verified); the release walker on the same journey answers FAIL NOT_PERMITTED. Posted as [BLOCKED] with the four places to fix (test `actor.get("demo") is True`, or `tenant in ("demo", "demo-walker")`). This adapter was hardened the same way (`is_demo`: the walker's tenant never reaches ERPNext either; test 9). The walk is to be rerun once that lands; the 05:09 `journeys-demo-rerun-1.jsonl` of another lane, overwritten by this run, was restored from the backup taken before it. ##### Not done, and what the owner of the next step needs 1. **Registration line** (not this lane's file). `runtime/server.py` `_domains()` imports only `bank`; until the line below is added the adapter is inert in the web process (the demo walk is unaffected either way): `from runtime.domains import registry # noqa: F401 - binds the registry slot (ERPNext) to its operations` inside the `try:` of `_domains()`, after the `bank` import. 2. ERPNext has no Company, so Project and Asset were not used (Issue and Item instead). Running the setup wizard would make Project/Asset available for case/asset; a re-mapping then is a new build step, with migration of existing records. 3. `read` by filter lists a tenant's records of one unit and entity (up to 500) and filters in the adapter, not in SQL. 4. Native title fields hold a 140-character summary; Role names carry the first 8 characters of the id for uniqueness. 5. The adapter's Comment per event is best effort (a failed comment never changes the outcome). 6. The `tok_id` field is filled from the body's `tok_id` when present; the units do not yet pass one. 7. The two applications' journey walk against the shipped release has not passed: blocked by ffe02c065 as above, rerun due. ### research/public-consortium/MOBBIN-FLOWS.md `research/public-consortium/MOBBIN-FLOWS.md` · sha256 `4aa3d18f1758414e305098489009f9d43677daf44558a97d2598da527ee7ef3c` #### EDGE Public Consortium research: real product flows from Mobbin **What this is.** Real screens and user flows from existing products, recorded by Mobbin (a library of real app and website screenshots, grouped into step-by-step flows). It complements `REFERO-FLOWS.md` in this folder. It shows what buyers, suppliers, security teams, learners and administrators already expect, and good practice worth borrowing from other industries. **Statement class and date.** Every item below is an OBSERVATION of a real product, seen in Mobbin on 2026-10-31. The descriptions are ours, written from the preview images Mobbin returned (an evenly spaced sample of each flow's screens, not always every screen). Product names belong to their owners. Image links are not kept because they expire; each item carries its `mobbin_url`. **How it was gathered.** Searches on web and iOS, about three to five flows per search: vendor and supplier onboarding, purchase requests and approvals, compliance and security dashboards, vendor security review, incident response, shipment tracking, course enrolment, grant-style applications, identity checks, sign-in, audit logs, and roles and permissions. ##### Coverage: what Mobbin held and what it did not - **Held and returned:** Vanta (compliance, frameworks, controls, vendor risk and security reviews), incident.io (declaring and running incidents), Melio and Shopify Bill Pay (vendors), Shopify and Programa (suppliers and purchase orders), Airwallex (spend requests and approvals), Coursera, Uxcel, WRITER and MasterClass (learning), Binance, OKX and Stake (identity checks), Intercom, Oyster and Dialpad (business sign-in), and audit log and permissions screens from Zoho CRM, 15Five, Toggl Track, 7shifts, Gorgias, Hashnode, Hotjar, Zendesk and Churnkey. - **Searched for but not returned:** any government service sign-in (the search returned Intercom, Oyster and Dialpad), any public grant or research funding portal (it returned ElevenLabs, Coinbase and MasterClass), and any freight, defence or government logistics product (it returned Klarna, Urban Outfitters and Hims parcel tracking). - **Not found at all:** public procurement or tender portals, supplier vetting against sanctions or security clearance, defence or government supply chains, university research administration, and any flow where an AI agent is itself a supplier, a requester, an approver or a subject of vetting. ##### By task ###### Adding and onboarding suppliers - **Melio, web, "Adding a vendor"** (10 screens). A payments area with tabs for vendors, bills, approvals and payments. "Add your first vendor" offers suggested vendors, import, or add your own. The form marks required fields and gives plain notes at the point of need: "enter a full name rather than a partial one or a nickname" and "once you set a vendor's country, you cannot change it later". Taxpayer details explain why collecting a tax form in advance helps. The vendor list shows last paid, unpaid bills and open balance. OBSERVATION, 2026-10-31. https://mobbin.com/flows/c7aae0e4-1fcd-4600-93e8-e2addb3730a2 - **Shopify, web, "Creating a vendor"** (6 screens). A vendor table where each vendor shows "Missing receiving method". Adding a payment method lists the options with their fees, and offers "Ask vendor to choose". The vendor page then shows "Vendor to choose: sent to [email] on [date]" with Resend, and a "Request sent" message. OBSERVATION, 2026-10-31. https://mobbin.com/flows/20f680b7-9a70-4a1b-a86f-2250d9f0eba1 - **Shopify, web, "Creating a supplier"** (5 screens). Creating a purchase order, choosing "Create new supplier", filling company, country, address, contact, website and notes, with an "Unsaved changes" bar offering Discard or Save, then "New supplier created". OBSERVATION, 2026-10-31. https://mobbin.com/flows/db7a42f4-ba0f-4198-b44a-244d274d85c6 - **Programa, web, "Adding a supplier"** (4 screens). Typing a supplier name on a product offers "Add as a new supplier"; a short side panel (company, name, email, optional phone and address); "Supplier saved successfully". The product panel has an Approvals tab and a Draft status. OBSERVATION, 2026-10-31. https://mobbin.com/flows/6dab9f41-c1fe-4bcd-8342-1a432abbd12b ###### Purchase requests, approvals and purchase orders - **Airwallex, web, "Approving a request"** (4 screens). A summary of requests "pending your approval" and "all requests in approval", and a table with status labels such as "Pending your approval" and "Resubmission required". A request opens in a side panel with amount, who it is for, category, comments and attachments, and Reject or approve. After approval: "Spend request approved: the request will now be forwarded to the finance admin for card creation". OBSERVATION, 2026-10-31. https://mobbin.com/flows/9afbce42-5ab9-4c7a-a574-2b6b83cde1fd - **Airwallex, web, "Creating requests"** (9 screens). A request form: request type, what you want to buy, category, total or recurring amount, currency, optional dates and comments. OBSERVATION, 2026-10-31. https://mobbin.com/flows/49218c92-dcf7-40fb-b56a-3526a7353f66 - **Airwallex, web, "Purchase orders"** (2 screens). A locked feature page promising to track committed spend, compare planned with actual, match invoice lines to orders, and sync with an accounting system. OBSERVATION, 2026-10-31. https://mobbin.com/flows/f710b018-e055-44bc-b1a3-55ead96b80df - **Programa, web, "Creating a purchase order"** (9 screens). An empty order table (number, supplier, project, date raised, delivery date, amount, status), a new order with supplier, issue and delivery dates, tax and product lines, then "Purchase order sent" and the order document with a "Resend email" button. OBSERVATION, 2026-10-31. https://mobbin.com/flows/6b4278aa-5d88-425b-afe4-e9e4f250acba ###### Compliance and security dashboards - **Vanta, web, "Compliance"** (9 screens). Home shows compliance progress (for example 103 of 104 controls complete) and "last updated less than a minute ago", then tiles for policies, tests, vendors and documents, each with "needs attention" and OK out of total. The policy table shows framework, status, last approved date and how many staff have accepted it, with a filter "require my approval". The audits page offers two routes: give the audit firm access so they add it, or add it yourself. Frameworks (for example ISO 27001, HIPAA, GDPR and ISO 42001 for AI management) show percentage readiness. OBSERVATION, 2026-10-31. https://mobbin.com/flows/4ecb9ae6-7a66-4937-b87f-64af5a90dcfd - **Vanta, web, "Framework detail"** (4 screens). Each control has a one-line plain description, evidence readiness (for example 2 of 2) and an owner. A control opens to show its ID, source, domain, owner, the tests and documents mapped to it with their status, and a History tab. OBSERVATION, 2026-10-31. https://mobbin.com/flows/280ae4f2-78b0-4e33-8ba3-b72c8c7685a8 - **Vanta, web, "Home"** (2 screens). The same home plus public trust page statistics (visits, page views, downloads). OBSERVATION, 2026-10-31. https://mobbin.com/flows/e0c9bbb2-e841-4d9d-b52b-15a52401a855 ###### Supplier security review and vetting - **Vanta, web, "Vendor detail"** (3 screens). A vendor table with inherent risk ("Unscored"), security review ("Needs initial review") and last review date, filterable by risk, security owner, data agreement status, review status and review due date. A vendor page shows "High risk score" and a checklist to complete the profile: assign an internal security owner, assign a risk score (ticked), conduct an initial security review. The side panel shows security owner and business owner as "Unassigned" and missing contact details in red as "Incomplete". OBSERVATION, 2026-10-31. https://mobbin.com/flows/e80d75a8-9acb-4c25-b7f4-be7c0575ec79 - **Vanta, web, "Adding a security review"** (8 screens). The requirement is stated ("upload a security assessment"); the upload dialog asks what kind of document it is (for example a penetration test report or an assurance report); then "Congrats on completing a new review!", a file table with who uploaded it and when, and the last review date updated. OBSERVATION, 2026-10-31. https://mobbin.com/flows/9bf717f5-487e-48e9-964e-0d60a9f17a9e - **Vanta, web, "Vendor"** (3 screens). Vendors with risk high or medium and review status "Up to date, due [date]" or "Not required". OBSERVATION, 2026-10-31. https://mobbin.com/flows/54080d4c-e8c7-417c-96ee-5c2381090b96 ###### Incident response - **incident.io, web, "Declaring an incident"** (8 screens). Empty state "There are no active incidents". "Declare incident" has tabs for a live incident, a retrospective one and a test one. The name is optional ("you can leave it blank and change it later"); each severity comes with a plain definition (for example "Major: significant impact, immediate response is usually required"); the summary box prompts "think about what you'd like to read if you were coming to the incident with no context". Incidents then sit on a board by stage (investigating, triage, fixing, monitoring, documenting, reviewing) with a running timer. A corner badge says "You're not on call". OBSERVATION, 2026-10-31. https://mobbin.com/flows/c42e6803-4dd2-4fa5-8b52-53c3ee0f6c2b - **incident.io, web, "Declaring an incident"** (13 screens). The severity definition changes as you pick it ("Critical: for example a full outage or a data breach"). The incident page shows the stage path from investigating to closed, a critical label and "Ongoing for 2s"; tabs for updates, timeline, actions and follow-ups; "No update scheduled: Request update"; and a side panel with incident lead "Not assigned", reporter, observers, chat channel, post-mortem, related incidents, timestamps and "Duration metrics: missing timestamps". OBSERVATION, 2026-10-31. https://mobbin.com/flows/a053425f-d4f0-4138-abb9-3ac6591ad470 - **incident.io, web, "Home"** (3 screens). Recently visited incident and a list of active incidents with severity, status, an unassigned lead and age. OBSERVATION, 2026-10-31. https://mobbin.com/flows/e181dbdf-0545-4435-9d6b-46302a3feadd ###### Audit logs - **Zoho CRM, web, screen.** "The audit log provides a chronological sequence of actions performed by the users", grouped by day, written as sentences (for example a named user enabled a compliance setting, added a trusted domain, updated a profile), with Export. The settings menu lists "Agents" alongside users and security controls. OBSERVATION, 2026-10-31. https://mobbin.com/screens/b27943e6-e8ff-4fc0-8d27-dfe0c33c124f - **15Five, web, screen.** A table of who, when and what, where some rows are by people and some are marked "Automated", and sign-ins record the network address and browser. OBSERVATION, 2026-10-31. https://mobbin.com/screens/0e4eaeeb-0fc4-419e-9c4b-90b61293b012 - **Toggl Track, web, screen.** A workspace log with author, event and the change shown as "From [old] to [new]", filtered by member and event over a chosen week. OBSERVATION, 2026-10-31. https://mobbin.com/screens/2deef352-1f6b-4757-a4c9-dcce106b0e6a - **7shifts, web, screen.** An activity log with coloured action labels (added, deleted, declined), filters by audit type and employee, and "Show details". OBSERVATION, 2026-10-31. https://mobbin.com/screens/8b845427-b403-4eb7-87c9-aa771a00d378 - **Gorgias, web, screen.** Team member, event, the object it touched (linked) and the date, with filters and a date range. OBSERVATION, 2026-10-31. https://mobbin.com/screens/44a16fee-609a-40a5-afa0-f901dcacdfa8 ###### Roles and permissions - **Churnkey, web, screen.** A member's role as a list of capabilities, each a switch with a plain description (billing, users, payment provider, keys for software access, and so on), "Update role", and "User role updated". OBSERVATION, 2026-10-31. https://mobbin.com/screens/d3706e29-24ba-490c-b12f-68f7d06fbeda - **Zendesk, web, screen.** Roles and access set per product, with a note on how removing access works. OBSERVATION, 2026-10-31. https://mobbin.com/screens/9852ea07-3299-4836-bcce-7d4bd5caa81e - **Hashnode, web, screen.** Invite links that carry a role, a usage limit and an expiry date, and a pending invites tab. OBSERVATION, 2026-10-31. https://mobbin.com/screens/442d8d4c-06d3-4bbb-b012-83bb5650ebb1 - **Hotjar, web, screen.** Access level per organisation (admin, read and write, account owner), "Change account owner" and "Delegate permissions". OBSERVATION, 2026-10-31. https://mobbin.com/screens/47b88d59-75c0-4b68-8f7e-aae758c1d1df ###### Identity checks and sign-in - **OKX, iOS, "Verifying account"** (12 screens). The journey is shown up front ("sign up, verify identity: only takes a few minutes"). The document guide shows a good photo next to cropped, glare and blurry ones. A consent tick box names the verification company and says biometric data may be processed; the start button stays disabled until it is ticked. Selfie tips (remove face coverings and glasses, use good light). OBSERVATION, 2026-10-31. https://mobbin.com/flows/cab2edd4-e07f-4499-a124-c731b88bcede - **Binance, iOS, "Verifying identity"** (12 screens). "Based on local regulator requirements, please complete the additional verification", then choose the issuing country and document type, then a selfie camera with "See more guidance". OBSERVATION, 2026-10-31. https://mobbin.com/flows/cdc52574-dd6c-4203-8cc6-4a246dea0a8f - **Stake, iOS, "Uploading documents"** (5 screens). "It should take less than 2 minutes", the accepted documents listed, three plain rules for the photo, camera or photo library, "Upload successful" and a tick for each side of the document. OBSERVATION, 2026-10-31. https://mobbin.com/flows/894d5294-3573-4eda-b0bc-4eff194621d6 - **Intercom, web, "Logging in"** (6 screens). After the password, "Enter verification code: check your email, it will be valid for 10 minutes, it may be in your spam folder", with "Resend verification code". OBSERVATION, 2026-10-31. https://mobbin.com/flows/9bc0d4f9-ca7f-46bf-92a4-4a1ebd617861 - **Oyster, web, "Logging in"** (5 screens). Password, Google or single sign-on. Home then groups "Requiring attention" by area (hire, pay, care), for example "1 invoice awaiting approval", and says plainly when an area has nothing. OBSERVATION, 2026-10-31. https://mobbin.com/flows/9036d3a3-01de-4420-9636-ae2106292601 - **Dialpad, web, "Logging in"** (4 screens). Sign in with Google, Microsoft, another provider or a work email. After sign-in a warning bar explains a restriction ("You are not able to send external messages") and how to lift it; a set-up checklist lets each step be skipped. OBSERVATION, 2026-10-31. https://mobbin.com/flows/c509eef8-5496-4220-a29e-7d8dae2fb6c6 ###### Learning and course enrolment - **Coursera, web, "Enrolling a course"** (13 screens). A certificate page with start date, level, time needed and number enrolled; checkout that states "you won't be charged today", when the trial ends and that refunds are not possible once charged; a "My commitment" pledge; and the course player with an outline, video, transcript with the current line highlighted, notes, downloads and "Report an issue". OBSERVATION, 2026-10-31. https://mobbin.com/flows/7ecb91ca-d052-4f32-93e3-a7acf6d871f4 - **Uxcel, web, "Enrolling a career path"** (5 screens). A catalogue of paths, a "Confirm enrollment" dialog, and a syllabus of courses, assessments and a final exam with "Start here" and a percentage complete. OBSERVATION, 2026-10-31. https://mobbin.com/flows/c7bbb197-d803-4496-a768-0aca9fd937cf - **WRITER, web, "Enrolling a program"** (2 screens). A learning centre inside an AI product (videos, prompt library, webinars, guides) whose "Enroll" button becomes "Open". OBSERVATION, 2026-10-31. https://mobbin.com/flows/ea2b4d41-972f-4d0f-bf43-5fb9e87987b1 - **MasterClass, web, "Submitting a project"** (4 screens). An activity numbered "2 of 7" with a time estimate, a project text box with a character count, a "Do not share with the community" tick box, Submit and Mark as done, then "You finished the section!". OBSERVATION, 2026-10-31. https://mobbin.com/flows/62662a36-0825-4069-a3f0-335bb00e0639 ###### Applications (nearest thing to grants) - **ElevenLabs, web, "Applying for a program"** (7 screens). An impact programme application: required fields, choosing a partner charity whose details panel states its purpose, the conditions it covers and where it operates, and "Application submitted successfully! We will review your application and contact you soon". OBSERVATION, 2026-10-31. https://mobbin.com/flows/e4d81ee1-fb6f-472d-a5f2-990cb3eeecaf - **Coinbase, web, "Submitting a request"** (10 screens). An application form that says how the reply will come ("We'll reach out to you at this email"), marks optional fields, has a privacy acknowledgement, and ends "Request submitted: we'll email you if we decide to feature your app". OBSERVATION, 2026-10-31. https://mobbin.com/flows/fe542654-cfb1-425b-9382-7df68e8e7e96 ###### Tracking deliveries (nearest thing to logistics) - **Urban Outfitters, web, "View tracking info"** (4 screens). A four-stage line (shipped, on its way, out for delivery, delivered), the carrier tracking number, a timeline of every scan with time and place, an honest note ("it's possible there was an error in the carrier status... allow 48 hours before submitting a claim"), and text updates with a consent tick box. OBSERVATION, 2026-10-31. https://mobbin.com/flows/ff34ace9-4211-4311-8d38-b699c17317c2 - **Klarna, web, "Deliveries"** (3 screens). "Nothing on the way", adding a delivery by tracking number, and items "on the way" or "waiting for delivery updates" with "Link purchase" and "Mark as delivered". OBSERVATION, 2026-10-31. https://mobbin.com/flows/fac5ae98-c931-45f8-bd02-2685e1f05322 - **Hims, web, "Tracking an order"** (2 screens). Orders grouped by status including "Cancelled and refunded", and a delivery page with the latest scan and full history. OBSERVATION, 2026-10-31. https://mobbin.com/flows/79397988-1f04-499a-9fa4-813686178ea4 ##### What to match - **Every record has an owner, a status and a due date** (Vanta controls, vendors and reviews; incident.io lead and stage). Unassigned and incomplete are shown as such, in words and colour, not hidden. - **Plain definitions at the point of choice:** severity levels (incident.io), irreversible fields (Melio's country), why a document or a phone number is needed (Melio, Stake, Coinbase). - **Approvals that say what happens next** ("forwarded to the finance admin", Airwallex) and statuses that name who must act ("Pending your approval", "Resubmission required"). - **Let the other party supply their own details** (Shopify "Ask vendor to choose", with the sent date and Resend), and let an auditor be granted access instead of re-keying (Vanta). - **Evidence tied to the requirement it satisfies,** with who uploaded it and when (Vanta reviews and controls). - **Logs that read as sentences and show before and after values** (Zoho CRM, Toggl Track), and that mark automated actions separately from people (15Five). - **Permissions described capability by capability in plain words** (Churnkey), with time-limited invites (Hashnode). - **Identity checks that show a good and bad example, say how long it takes, and ask for explicit consent before biometrics** (OKX, Stake). - **Delivery timelines with every scan and an honest note about carrier errors** (Urban Outfitters). - **Learning that shows the whole path, time needed and progress,** with a transcript and a way to report an issue (Coursera, Uxcel). ##### What none of them do - **No AI agent is a party.** In every flow the requester, approver, supplier, incident lead and learner is a person. Zoho CRM lists "Agents" in its settings and 15Five marks rows "Automated", but nothing shows an agent with its own identity, the person or organisation it acts for, its limits, or its own entries in the audit log as a separate kind of actor. - **No vetting of an AI agent as a supplier.** Vanta reviews software vendors by document upload, and lists ISO 42001, but no flow assesses an agent's permissions, data access or behaviour before it is allowed to act. - **No public sector context.** No government sign-in, tender notice, framework agreement, security clearance, sanctions or conflict-of-interest check, freight or defence logistics, chain of custody, grant scoring, or university research administration appeared. - **No organisation-to-organisation trust.** Supplier records belong to one buyer. Nothing lets a supplier prove a fact once and share it with several buyers, or lets a consortium see the same verified record. - **Learning is not linked to permission.** Completing a course does not unlock anything elsewhere (for example a role or an approval right), and nothing records that a person or agent has been trained before acting. - **Incidents do not reach suppliers or regulators.** incident.io keeps everything inside the team; there is no flow for notifying an affected supplier, customer or authority, or for a supplier to report an incident to a buyer. ### research/public-consortium/REFERO-FLOWS.md `research/public-consortium/REFERO-FLOWS.md` · sha256 `b7b4a23c38ecd15f6b7999216c27a541fe1d88d5a82153fc4914ed4b5ac318b5` #### EDGE Public Consortium research: real product flows from Refero **What this is.** Real screens and user flows from existing products, recorded by Refero (a library of real app and website screenshots, each journey described step by step). They show the floor (what users already expect) and good customer experience from any industry worth borrowing. **Statement class.** Every flow is an OBSERVATION by Refero of a real product, searched 2026-10-31. Descriptions are Refero's wording. Product names belong to their owners. **How it was gathered.** Semantic searches for this application's tasks on iOS and web. Results from other industries are kept on purpose: they are often the best customer experience to learn from. Refero does not hold many incumbents (and few government, enterprise or NHS systems), so this is a sample, not a census. **Products in this sample:** Lovable (19), Brilliant (14), District (12), Cohere (11), Instagram (10), Tinder (9), Dropbox (8), DailyArt Magazine (8), Artsy (7), Rork (7), Structured (7), Aboard (7), Imprint (7), Airbnb (6), ElevenMusic (6), Acctual (6), Faire (6), Plane Finder (6), Meta AI (6), Omio (6), PayPal (6), Mercury (5), WhatsApp (5), Peerlist (5), Asos (5), Fingerprint (5), komoot (5), Fernand (5), Duolingo (5), Programa (5), ChatGPT (5), Ada (4), 222 (4), Granola (4), Adobe Photoshop (4), Cake Equity (4), Kickstarter (4), Foodllama (4), TikTok (4), Parallel (4), Exa (4), Binge (4), TravelPerk (4), Fourthwall (4), Chargetrip (4), Grok (4), Fuse (3), Wealthsimple (3), Copperx (3), Pinterest (3), Introspect (3), GoFundMe (3), Wabi (3), Runey (3), Genie (3), Handshake (3), Pool (3), Google Gemini (3), Spyglass (3), Ground News (3), Сontra (3), Figma (3), OpenAI (3), ALO (3), Phantom (3), Wispr Flow (3), Nike (3), Vinyls (3), IKEA (3), Cushion (3), Teachable (3), UGLYCASH (2), Voicenotes (2), Navan (2), TIDAL (2), Apple Image Playground (2), YouTube Music (2), ElevenLabs (2), Nudea (2), GlossGenius (2), Loowner (2), Delphi (2), FocusFlight (2), fal (2), On (2), Homerun (2), Weather mini (2), Cursor (2), Linear (2), Steel (2), Appwrite (2), Orbit (2), Kin (2), Days (2), Factory (2), Daylish (2), Stellate (2), Roots (2), Substack (2), Zara Home (2), West Elm (2), Enode (2), Preply (2), Lumy (2), Descript (2), Reclaim AI (2), Doji (1), Drinkit (1), Volv (1), Square (1), Nooka (1), Turf (1), time2book (1), Cashcaded (1), Queue (1), Subo (1), Apple TV (1), Revolut (1), Willow (1), Curater (1), Spoil me (1), Stocktwits (1), HYPE (1), Linktree (1), Claude (1), Understory (1), Wittl (1), Denim (1), Sora (1), Reown (1), Product Hunt (1), Family (1), Doctronic AI (1), Saint Laurent (1), New Balance (1), Longevity Deck (1), Leonardo.Ai (1), Farfetch (1), BeReal. (1), Polar (1), Column (1), Oppie (1), Kinhive (1), OKURI (1), Podimo (1), Manus (1), Showcase (1), Yelp (1), Twitch (1), Stemz (1), Givingli (1), RAD Weather (1), Acne Studios (1), Mocha (1), Shop (1), Trip (1), Microsoft (1), Pattern Brands (1), WishWish (1), Apple Books (1), ElevenReader (1), ROVR (1), Instacart (1), Shopify (1), Dopper (1), Polaroid (1), Trip Way (1), Tripsy (1), Box Box Club (1), Apple Maps (1), Remy (1), Moonly (1), Teal (1), Uber Eats (1), (Not Boring) Habits (1), Pinkllama (1), Canva (1), Kann (1), Clime (1), (Not Boring) Camera (1), Abode (1), CapWords (1), Rox (1), Loom (1), Tella (1), Riverside (1), Cofounder (1), Whereby (1), SoundCloud (1), Soundtrap (1), ManyChat (1), LEGO® Builder (1), Melius (1), Telescope (1), Synthesia.io (1), MyMind (1), Expedia (1), ABY Journal (1), Bump (1), Fresha (1), Mindllama (1), Stripe (1), Typeform (1), Maze (1), Uber (1), Gladia (1), Threads (1), Matter (1), Ballpark (1), Bond (1), Miro (1), Slite (1), YouTube (1), Apple Invites (1), pillowtalk (1), XChat (1), Hello Ivy (1) ##### Task: AI agent setup permissions **ElevenMusic: Artist Identity Setup** (ios, 3 screens, AI Tool, Entertainment, Music Streaming) [Refero](https://refero.design/flows/12441) > Happy path for setting up an artist identity during onboarding. The user starts on a “Become an artist” screen where they can add a stage name and profile photo, or skip this step for later. Tapping the photo control opens a media picker with Photos/Collections tabs and a close action. After selecting an image, the user returns to the setup screen with the chosen photo applied. Steps: Artist Intro → Photo Picker → Artist Intro With Photo **Pinterest: Collaborator management** (ios, 3 screens, Reference, Inspiration Board, Social Networking, Lifestyle) [Refero](https://refero.design/flows/2493) > Happy path: A board owner opens the Collaborators screen to view current collaborators and invited people, then adjusts collaborator permissions and invitation settings. Entry: user opens Collaborators from a board. Exit: permission level and invitation toggles are saved and the UI returns to the Collaborators list. Steps: Collaborators list → Permissions options (default) → Permissions — updated **Dropbox: Folder access management** (web, 4 screens, Collaboration Tools, Business, Productivity) [Refero](https://refero.design/flows/13031) > A logged-in user manages sharing and permissions for a team folder in a cloud file manager. The flow starts from the folder’s action menu, opens an access management dialog, changes a collaborator’s permission level, and confirms the update with a success message. **Reown: Manage team member roles** (web, 3 screens, Development Tools, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/5734) > Happy path: an administrator opens the organization's Members area, opens the per-member actions menu for a specific member, and promotes that member to an admin. Steps: Members list overview → Open per-member actions → Apply role change and reflect update **Adobe Photoshop: Generative Policy Acknowledgement** (ios, 4 screens, Design & Media Editing) [Refero](https://refero.design/flows/4806) > Happy path: A user composes or edits an image and taps Generate, which triggers an in‑context modal showing Adobe's Generative AI User Guidelines. The user reads (or scrolls through) the guidelines in a scrollable modal and then dismisses it (Done) to return to the image editor. Steps: Image editor — prompt and generate → Guidelines — modal title → Guidelines — safety section → Guidelines — additional information and dismiss **Apple Image Playground: Generated Image Review** (ios, 3 screens, AI Tool, Design & Media Editing) [Refero](https://refero.design/flows/12797) > A user is viewing a generated image in an image creation/editor interface. From the main preview screen, they can save, share, open more actions, browse style suggestions, and enter a text prompt. One interaction opens a contextual action menu with copy, share, save image, and report options. Another interaction switches the top-level state into a confirm/cancel mode, where the user can either discard or confirm the current change before returning to the preview screen. Steps: Image Preview → More Actions Menu → Confirm Change State **Structured: Privacy & Analytics Settings** (ios, 5 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3310) > Happy path: A signed-in user opens Settings, navigates to Privacy & Permissions, reads the privacy overview, then opens the Anonymous Analytics detail screen and toggles sharing off (or back on). Steps: Settings > Support / About → Privacy Overview → Privacy Controls List → Anonymous Analytics — Detail (Sharing On) → Anonymous Analytics — Detail (Sharing Off) **Rork: Collaborator Invitation** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12168) > Happy-path project sharing in a collaboration modal. The user opens an invite dialog from a project workspace, enters an email, chooses a permission level, and either sends an invitation or creates a shareable invite link. Steps: Invite Collaborators → Permission Selection → Invite Link Created **Pool: Refero MCP item review** (ios, 4 screens, Utility) [Refero](https://refero.design/flows/11989) > This is a mobile screenshot sequence showing a Refero content item and the user opening its overflow menu, then opening an agent selection bottom sheet and selecting one agent. Entry appears to be a detail viewer for a Refero MCP AI Agents item, with a close button, overflow menu, and embedded preview content. The user can either dismiss overlays, open the action menu, or proceed into the MCP exploration/pooling step. Assumption: the user is already inside the viewer on a mobile device and has permission to interact with the item. The flow ends with one agent tile selected in the bottom sheet. Steps: Item Detail Viewer → Overflow Actions Menu → Pool Selection Sheet → Agent Selected ##### Task: AI chat assistant **Granola: Review Generated Email** (ios, 2 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14157) > The user is viewing an AI-generated follow-up email within a mobile assistant conversation. The flow starts with an existing prompt and generated email response, then continues when the user selects the positive feedback control. It ends with the response marked as helpful. The user is assumed to be in an active conversation on a mobile device; authentication and permission state are not visible. Steps: Generated Email Review → Positive Feedback Submitted **Granola: Copy Generated Response** (ios, 2 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14156) > A user views a generated follow-up email in a mobile AI conversation and copies the response to the clipboard. The flow starts on an open conversation with the generated message visible and ends with confirmation that the message was copied. The user is assumed to be on a mobile device with access to the conversation; no additional authentication or permission state is visible. Steps: Generated Email Response → Copy Confirmation **Doctronic AI: Passwordless AI Doctor Sign-In** (web, 8 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13706) > The flow begins on Doctronic’s public landing experience, where an unauthenticated visitor can start a chat or access account options. The user chooses email-based login, enters an email address, receives a six-digit verification code by email, and submits the code. After successful verification, the user is taken to the authenticated AI health assistant screen, ready to ask a health question. The user is assumed to be logged out and able to access the entered email inbox. ##### Task: API keys developer settings **Willow: Willow onboarding intro** (ios, 3 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/13227) > A short onboarding carousel for Willow shown to a new or returning user before entering the main experience. The user can swipe through several promotional intro screens and tap the primary Get started button to continue. **komoot: New user onboarding and consent setup** (ios, 39 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8281) > Happy-path onboarding for a first-time mobile (iOS) user. Entry: app launch -> privacy/tracking consent banner. User can Accept All, Refuse, or Customize preferences. If user customizes they can toggle categories and open vendor details. After consenting, the flow presents sign-in options (social or email). Choosing email leads to email validation then account creation (password). Post-signup the user progresses through product onboarding cards that request push notifications and access to device integrations (Apple Health). The flow ends with device connections (optional) and a completion screen, then the main map/routes home screen. Steps: Privacy notice (top-level) → Customize preferences (categories visible) → Category list (scroll) → Analytics expanded (vendor row visible) → Vendor info modal (Firebase) → Sign-in options → Enter email (initial) → Email validation error → Email accepted (proceed) → Create account (password) → Onboarding hero / intro → Onboarding card: Navigation → Onboarding card: Notifications → iOS push permission dialog → Device integrations list → Health Access (iOS modal) → Integration connected state → Onboarding completion → Main map / routes home → Optional: Connect additional devices (post-onboarding) **Phantom: Token List Manager** (ios, 3 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3716) > Happy path: A logged-in user opens the Token List Manager to show or hide tokens in their wallet. The screen lists available networks/tokens each with a balance and an on/off toggle. The user focuses the search field to filter the list, types a query (letters or numbers) and the list updates; if nothing matches, the UI shows an empty state message. Steps: Token list overview → Focus search field → Search resulting empty state **Exa: Developer Workspace Home** (web, 2 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/12010) > This is a dashboard-style entry screen for an Exa developer workspace. The user is logged in and can either begin onboarding/integration from the main hero card, inspect documentation, manage API credentials, add credits, or explore product capabilities and templates. The flow enters from the workspace home dashboard and exits when the user clicks a primary action such as Get Started, Docs, a product card, a template card, or API key-related controls. Steps: Workspace Dashboard → Compact Dashboard View **Exa: API Keys Management** (web, 3 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/12029) > This is a signed-in team settings area for managing API credentials. The user enters the API Keys section from the left navigation, views the current team details and existing keys, opens the overflow menu for a key, and can choose edit or delete. In the final state, one key has been removed from the list. Assumptions: the user is authenticated, has permission to manage keys for the current team, and is using a desktop browser. Steps: API Keys List → Key Actions Menu → Updated Key List **Gladia: Developer Overview and API Keys** (web, 2 screens, AI Tool) [Refero](https://refero.design/flows/10886) > Logged-in developer lands on the Gladia "Start building" overview (desktop). The screen surfaces three primary entry cards (Real-time transcription, Asynchronous transcription, Documentation) and an API Keys panel showing existing keys and a "+ Create API Key" control. Left navigation lists products and management sections (Transcriptions, API keys, Webhooks); a collapsed-icons variant of the same sidebar is also present. Happy path: user scans available product entry points, then manages or creates API keys via the API Keys panel. Steps: Overview — Sidebar expanded → Overview — Sidebar collapsed (icons-only) ##### Task: CRM contacts pipeline **Genie: Contact Search Add** (ios, 2 screens, Social Networking, AI Tool) [Refero](https://refero.design/flows/12420) > A user searches for people by name and adds them from a contact search list. Entry is a search screen with the keyboard open and a query already entered (“Dan”). The user can scroll through matching results and tap Add next to a person. After tapping, the action changes to Sent for that person, indicating the request/invite was sent. Steps: Search Results → Updated Results **Lovable: Missing Connector Request** (web, 4 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12342) > Happy-path for requesting a missing connector from the Connectors directory. The user starts in the Connectors section, filtered to the Messaging category, and uses the ‘Missing a connector?’ form in the lower-left panel to request an integration. The visible state assumes the user is already signed in and has access to the connectors catalog. After entering a connector name and submitting, the form confirms the request was received and disables further submission. Steps: Browse Messaging Connectors → Enter Connector Request → Submit Request → Request Confirmed **Sora: Cameo access recipient selector** (ios, 4 screens, AI Tool, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/7824) > Mobile modal for granting 'Cameo access' to other users. Entry: user opens the Cameo access modal (overlay) while signed in on a mobile device. The modal contains a searchable list of accounts (avatar, username, follower count) and a text field labeled 'Find people...'. The user can type to search (handles, names, numbers). Exit: user confirms selections by tapping Done, which closes the modal. Assumptions: user is authenticated and allowed to manage cameo access; device is a smartphone (iOS-style modal and keyboard behavior). Steps: Open Cameo access → Search — no results → Search — name results → Select recipients and confirm **Appwrite: Enterprise contact and confirmation** (web, 3 screens, Development Tools) [Refero](https://refero.design/flows/6018) > Desktop two-column enterprise landing where the left column contains marketing copy and the right column contains a lead-capture contact form for enterprise plan inquiries. Steps: Enterprise lead capture → Form completion and submit → Submission confirmation **Programa: Add Supplier Contact** (web, 4 screens, Business, Project Management) [Refero](https://refero.design/flows/13445) > A logged-in user is viewing John Doe’s supplier record in the Address Book and wants to add a contact. From the supplier detail screen, the user opens the Create contact panel, enters the contact’s name, role, email, and phone number, then submits the form. The panel closes, the new contact appears in the Contacts table, and a confirmation message indicates that the contact was added. Steps: Supplier Record → Contact Form → Completed Contact Form → Contact Added **Fernand: Conversation Detail** (web, 3 screens, Business, Communication Tools) [Refero](https://refero.design/flows/11888) > A logged-in support agent views an existing conversation thread for the contact Fernand. The main content shows the message history, the right sidebar shows contact details, tags, and related conversations, and the bottom composer allows replying or adding a note. The entry point is an opened conversation. The exit point is after the user replies, adds a note, copies contact details, marks the conversation done, or navigates away. This assumes desktop use with permission to view and respond to the conversation. Steps: Open Conversation → Copy Contact Email → Copy Confirmation **222: People Selection for Itinerary** (ios, 4 screens, Lifestyle, Dating & Relationships) [Refero](https://refero.design/flows/9302) > Happy-path: User chooses people to bring to a specific itinerary card. Steps: Itinerary card (choose friend) → Contacts picker → Duplicate selection modal → Itinerary card (selection confirmed) **WhatsApp: Contact Details Search** (ios, 4 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12666) > This is a contact details and in-chat search experience for a specific conversation with Samuel. The user enters from the contact info screen, can review conversation-related actions and metadata, then open the chat search field to find messages by keyword or number. The flow assumes the user is already inside a chat/contact context on a mobile device. Entry is from Contact info; exit is either back to the contact details screen or to a filtered result within the conversation search. ##### Task: access review permissions audit log **TIDAL: Track Report Reason Selection** (ios, 3 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/13140) > A mobile modal for reporting a track, starting with a searchable list of report reasons and ending with a selected reason shown after search. The user opens the report dialog, searches or browses categories, and selects the matching reason. Steps: Browse Report Reasons → No Match Search State → Filtered Result **Apple Image Playground: Report Generated Image** (ios, 9 screens, AI Tool, Design & Media Editing) [Refero](https://refero.design/flows/12798) > A user reviews a generated image in Image Playground, opens the overflow menu, and chooses Report a Concern. They select concern categories, optionally add more information, review what will be included in the report, accept the data-sharing prompt, and submit. The report then submits successfully and returns them to the image editor, where the generated image remains visible with the usual controls. Steps: Image Review → Action Menu → Concern Category → Personal Characteristics Check → More Information → Submit Report → Data Consent → Submitting → Return to Editor **Granola: Share Note with Editor Access** (ios, 8 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14162) > The user starts on a note titled “Dashboard design and analytics” and opens its sharing controls. From the sharing panel, they choose to add people, search for a contact by name, select the matching attendee, and change the access role from Viewer to Editor. They return to the sharing summary, where the selected person is listed as having access. The flow is shown on a mobile device. The user is assumed to be signed in and authorized to share the note; no confirmation or notification result is visible after returning to the summary. Steps: Open Note Sharing → Review Sharing Options → Open Recipient Selector → Search for Contact → Select Contact → Open Permission Options → Assign Editor Access → Review Granted Access **Denim: Gallery permissions onboarding** (ios, 5 screens, Design & Media Editing) [Refero](https://refero.design/flows/5929) > Happy-path discovery and permission gating inside the Gallery tab on an iPhone. Entry: user opens the app and lands on the Gallery tab (bottom navigation). The screen shows seasonal hero cards, a “Get the Best Experience” section with action cards (Connect Account — Apple Music, Enable Alerts), and a Free Artworks carousel below. Two system permission dialogs can be triggered from the Gallery: (1) Apple Music / media library access and (2) push notifications. Exit: user remains in the Gallery; if permissions are granted the app can personalize covers from playlists and schedule alerts. Steps: Gallery Home → Apple Music / Media Permission Prompt → Gallery (after music permission) → Push Notifications Permission Prompt → Gallery with Free Artworks **TIDAL: Playlist Privacy Change** (ios, 3 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/13116) > A playlist details screen shows an empty playlist owned by the current user. The user opens the privacy control, confirms making the playlist public in a modal dialog, and returns to the playlist screen where the privacy state changes from Private to Public. Steps: Playlist Details → Public Confirmation → Public Playlist State **TikTok: Activity Status Visibility** (ios, 2 screens, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/2798) > Happy path: A logged‑in mobile user opens the Inbox to review notifications and then opens the activity status information modal to review or change who can see their active/last‑active presence. Steps: Inbox / Notifications → Activity Status Modal **Lovable: People Management** (web, 5 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12297) > This is the workspace People management area for the Refero Design workspace. The user enters from the left navigation on the People section and can view members, filter/search the list, and perform member actions. The happy path shown is selecting one or more people in the table, opening the bulk actions bar, changing a selected member’s role from a dropdown, and returning to the list. Assumptions: the user is logged in, has permission to manage workspace members, and is viewing this on desktop. Steps: Member List → Select Members → Bulk Actions → Role Selection → Error Feedback **Lovable: People Management** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12294) > This is the workspace People management area for a workspace called Refero Design. The user enters from the left navigation under Members & access > People. The screen shows a list of workspace people with tabs for All, Invitations, Collaborators, and Requests, plus search and role filtering. The user can also export the list, invite by link, or invite members. In the shown interaction, the user opens the role filter, selects Owner, and the table updates to show only matching people. Assumption: the user is logged in and has access to manage people in the workspace. Steps: People List → Role Filter Menu → Filtered Results **Product Hunt: Product Team Access Request** (web, 3 screens, Business, Social Networking) [Refero](https://refero.design/flows/11461) > A logged-in user views Notion’s Product Hunt profile on the Team tab and requests access to be added to the product team. Because the product page already has an owner, the system opens an access-request modal. After the user submits the request, the modal confirms that the request was sent and the team section changes to indicate a pending request. The flow begins on the product team profile and ends with the request awaiting owner approval. Steps: Notion Team Profile → Access Request Prompt → Request Confirmation **fal: Recent History Preview Visibility** (web, 2 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/13542) > The user opens the Recent History area while authenticated and reviews recent API requests across endpoints. Results are displayed in a four-column card grid with generated previews, request metadata, timestamps, status codes, prompts, and actions such as Playground, Copy prompt, and Model Analytics. The user can use the Show preview toggle to switch between visible generated images and hidden preview placeholders. The flow ends with the history list displayed in the selected preview state; pagination controls and the site footer remain available below the results. Steps: Review History With Previews → Hide History Previews **District: Event Feedback Submission** (ios, 5 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12916) > Happy-path mobile feedback journey after attending an event. The user enters a rating screen for a specific event ('Sip and Act at 3 Art House'), selects an overall sentiment, optionally rates individual aspects, adds a text comment, and submits. Steps: Overall Rating → Detailed Ratings → Comment Entry → Submission Confirmation → Completion State **Cursor: Security Audit Assistant** (web, 18 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9922) > Happy path: From the Agents landing the user composes a task card, selects a GitHub repository, configures model(s) and multipliers, optionally attaches an image, then starts the task which spins up a cloud agent. The agent appears in the left agent list and the UI navigates to a three-column agent workspace showing the activity log (center) and a file-changes pane (right). The agent runs commands (visible as generated commands / progress) and produces a final security audit report with findings and code snippets. Steps: Agents Landing — Task Composer → Repository Picker Open → Repository Search → Repository Selected → Open Model Selector → Enable Multiple Models → Set Model Multipliers → Compose Task Instruction → Attach Image Attempt → Image Preview Modal → Task Created — Agent Listed → Open Agent Conversation → Agent Starting — Generating Commands → CLI Execution Attempt → Agent Working — Additional Steps → Audit Report — Commits Reviewed → Critical Finding Detail → Final Summary and Checks **Podimo: Podcast Comment Moderation** (ios, 3 screens, Entertainment) [Refero](https://refero.design/flows/12274) > Happy path in a podcast comment thread on mobile. The user is already viewing a specific episode’s comments, can read existing replies, post a new comment, and open the report action for a comment. After submitting the report, a confirmation sheet appears with a success message, a link to community guidelines, and a Done button to dismiss it. Steps: Comment Thread → New Comment Posted → Report Confirmation **Aboard: Audit report creation and review** (web, 5 screens, Collaboration Tools, Paperwork) [Refero](https://refero.design/flows/818) > Happy path: A signed-in user opens Analytics > Reports, uses the Audits section + button to create a new audit report, names it, submits the modal, inspects the generated report (table of audit events with filters and export/save controls), then returns to the Reports overview where the new audit appears in the Audits list. Steps: Reports overview (Audits empty) → Open 'New audit report' modal → Name the audit report → Inspect report detail (events table) → Return to Reports overview (new audit listed) **Dropbox: Activity Log Filtering** (web, 4 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/13064) > An admin reviews the Activity audit log in Dropbox Admin Console and narrows results using filter chips. The entry point is the Activity section from the left navigation. The user can inspect event rows, open the People filter, select a person, and see the table update to show only matching activity. The exit point is either continuing to refine filters or creating a report from the filtered results. Steps: Activity Log Overview → People Filter Picker → Person Filter Applied → Filtered Activity Results **Factory: Repository Security Review** (web, 22 screens, Business, Development Tools, AI Tool) [Refero](https://refero.design/flows/7751) > Happy path: A user starts a knowledge/code assistant session, selects an assistant type and remote workspace/repository, composes a request to review code for security, configures model and toolkits, and sends the request. The assistant connects to the remote repository, runs exploratory commands (git pull, listing files), opens and previews configuration and workflow files, generates a todo checklist, and produces a security-review report which is displayed in a split preview/context panel. Entry: user landing on session start. Exit: rendered security-review-report and relevant context files available in the session. Assumptions: user is authenticated (Sofia Delgado shown), has a cloud workspace/repository available, and has permission to view repository contents remotely. Steps: Session landing → Select assistant → Confirm workspace and repo → Attach/upload indicator → Compose request → Send request / open options → Select model and reasoning level → Enable toolkits → Connecting to repository → Repository inspection begins → Globbing and file listing → Open small file preview → Generate Todo checklist → Open workflow YAML → View README.md content → Shell listing and report file created → Report: executive summary and urgent actions → Split view with report preview → Context panel and uploaded files → Browse codebase overview document **Fingerprint: Identify events by visitor id** (web, 2 screens, Business, Development Tools) [Refero](https://refero.design/flows/11152) > Happy path: A signed-in user opens the Identification section to inspect recent identification events, reviews the table of events and active filter chips, then narrows results by creating a new filter from a row value (Visitor ID). Steps: Identification events list → Create Visitor ID filter from row **Stemz: Record Audio Import** (ios, 5 screens, AI Tool, Music Streaming) [Refero](https://refero.design/flows/12977) > Happy-path flow for importing a track by recording audio on a mobile device. The user opens an Import Track chooser, selects Record Audio, grants microphone permission, records audio, stops the recording, and reaches a review state where they can delete, play back, or continue. Steps: Import Source Selection → Microphone Permission → Recording Start → Active Recording → Recording Review **Kin: Journaling Draft Review** (ios, 4 screens, AI Tool) [Refero](https://refero.design/flows/5438) > Happy path: A signed-in user in the Journaling area sees a banner indicating pending drafts, a chronological list of journal entries, and a floating control to create new entries (sketch or record). From the Journaling screen the user opens the Drafts list, reviews a recorded draft card, chooses Delete which opens a confirmation modal, and either cancels (Keep) or confirms deletion. Exit point returns the user to the Journaling list with the draft removed and any draft-count banner updated. Steps: Journaling Home → Drafts List → Delete Confirmation Modal → Journaling Home (Updated) **komoot: Revoke Pending Invitation** (ios, 3 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8354) > Happy path: a logged-in organizer opens the Activity participants screen on a mobile device to review the organizer and pending invites. The screen lists the Organizer section and an "Invite sent" list with one or more pending email invites. The organizer taps the remove icon (X) next to a pending invite, which opens a modal asking to confirm revocation. If the organizer confirms, the pending invite is removed from the list and the view is updated. Steps: View participants list → Confirm revocation → Updated participants list **Roots: Developer Tools Logs** (ios, 3 screens, Productivity) [Refero](https://refero.design/flows/5765) > Happy path: a developer opens the in-app Developer Tools to inspect chronological log events, narrows the list by entering a short identifier (partial UUID) in the Filter field, and refines results by selecting a category chip (e.g., Daily Limit, Games). Steps: Developer Tools — initial log list → Filter by identifier — Daily Limit selected → Switch category to Games **Binge: Review Browsing** (ios, 4 screens, Entertainment) [Refero](https://refero.design/flows/12962) > A mobile reviews screen showing a scrollable list of user reviews for a film, with a spoiler warning banner at the top and a persistent search field at the bottom. Users can open a filter menu from the top-right overflow button to narrow reviews by rating, including unrated and half-star increments. Entry is from a previous screen via the back navigation, and exit is either back to the prior screen or by leaving the reviews screen. Steps: Review List → Rating Filter Menu → Filtered Reviews → No Additional Matches View **Adobe Photoshop: Inspect and Filter Version Comments** (ios, 4 screens, Design & Media Editing) [Refero](https://refero.design/flows/4828) > User is viewing the Version History for a file on a mobile device. Steps: Version Detail with Comments → Open Filters Panel → Select and Apply Filter → View Filtered Comments **Meta AI: Gmail Connection Authorization** (ios, 6 screens, AI Tool) [Refero](https://refero.design/flows/13355) > A logged-in user asks an AI assistant to identify active subscriptions and assess which are useful. Because Gmail is not connected, the assistant explains that email access is required and presents Gmail and Outlook connection options. The user selects Gmail, reviews Meta’s connection disclosure, continues to Google sign-in, reviews the requested Gmail permissions, and returns to the assistant with Gmail shown as connected. The depicted flow ends before the subscription audit results are displayed. **Instagram: Account moderation actions** (ios, 4 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/2847) > User is on a 'Support actions' screen after interacting with a reported post. The screen lists four contextual actions for the target account (Block, Restrict, Unfollow, Learn more). Selecting Block or Restrict opens a confirmation/description modal. Selecting Unfollow executes immediately and the UI shows a transient 'Unfollowed' toast plus a full-screen report acknowledgement. Steps: Support actions list → Block confirmation sheet → Restrict confirmation modal → Report acknowledgement + Unfollow toast **Lovable: Log Review and Filtering** (web, 4 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12315) > This is a logs inspection area inside a cloud/admin console. The user starts on the Logs screen under the Cloud section, with a list of recent log entries and filtering controls. The user can change the log source/type from a dropdown, adjust the time range, search events, inspect individual entries, and view raw log data. The flow ends with the desired log type selected and the log list updated to match the chosen category. Steps: Open Logs → Choose Log Type → Select Authentication Logs → Review Authentication Entries **Delphi: Revoke Contact Access** (web, 4 screens, AI Tool) [Refero](https://refero.design/flows/13991) > A logged-in user with audience-management permissions revokes access for one invited contact. The user selects a contact from the Audience list, chooses Revoke, confirms the action in a modal, and returns to the updated list. The flow starts on the Audience screen and ends with the selected contact removed and a success notification displayed. The sequence assumes a desktop layout and an existing audience containing invited contacts. **Lovable: Cloud Logs Review** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12314) > A user is on the Logs section of a cloud dashboard, likely already signed in and with access to the project. They can search, filter by log source and time range, review individual log entries, expand raw log data, and copy or download results. The flow starts on the Logs screen and ends after the user finds the relevant log information or confirms there are no matching records. Steps: Logs Overview → Empty Search Results → Filtered Log Results **Appwrite: Manage authentication users** (web, 3 screens, Development Tools) [Refero](https://refero.design/flows/6038) > Admin view for the project's authentication module showing a paginated list of users. Steps: Auth users list — default view → Column visibility control open → Auth users list — column control closed **Lovable: Log Type Filtering** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12316) > Logged-in user is on the Cloud > Logs section of the Nexlora AI Launchpad dashboard. The screen shows a log viewer with search, log-type filtering, time-range filtering, refresh, download, and copy actions. The user can open the log-type dropdown, search available log categories, and select a log type to update the displayed entries. Steps: Open Log Type Menu → No Matching Result → Select Authentication Logs **Dropbox: Activity Log Filtering** (web, 4 screens, Collaboration Tools, Business, Productivity) [Refero](https://refero.design/flows/13036) > A logged-in user is viewing a team file storage area with an activity sidebar open on the right. The user opens the Activity filter menu, selects an event type filter, and the activity feed updates to show matching events. **Copperx: Personal access tokens management** (web, 4 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8909) > This is the personal access token management view inside a logged-in user's Developer area. Steps: Tokens list → Open token actions → Confirm deletion → Token removed **Lovable: Workspace People Management** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12295) > Logged-in workspace admin opens the People section for the workspace and uses the built-in search and role filters to review members, invitations, and access status. The entry point is the People screen in workspace settings; the exit is either reviewing the full list, seeing filtered results, or confirming no matches. Steps: People list → No results state → Filtered people list ##### Task: apply for loan credit **Introspect: Friend decision response** (ios, 3 screens, Finance) [Refero](https://refero.design/flows/11952) > Happy path for a recipient responding to a friend’s product decision request on mobile. Steps: Decision Request → Sending State → Completed State **GoFundMe: Fundraiser Support Chat** (web, 6 screens, Fundraising) [Refero](https://refero.design/flows/11747) > This is the fundraiser management area for a logged-in organizer. The user starts on the main Fundraiser dashboard for a specific fundraiser, then opens a right-side support chat with the GoFundMe helper bot. In the chat, the user asks how to get money from the fundraiser, and the bot explains that transfers must be set up on desktop and gives step-by-step guidance. The flow ends with an end-chat confirmation dialog. Steps: Fundraiser Dashboard → Open Help Chat → Ask a Question → View Transfer Instructions → Continue Support Conversation → End Chat Confirmation **Airbnb: Artist experience request** (ios, 11 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6413) > Happy path for requesting entry to a limited/lottery artist event. Steps: Event detail / Request CTA → Guests selection → Application question prompt → Text entry (keyboard active) → Answer review → Request details summary → Before you submit (rules & process) → Submission (start animation) → Submission animation sequence → Confirmation screen → Exit / Return **Lovable: Plans and credits management** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12301) > Happy path for managing a workspace subscription from the Plans & credits area. Entry is the workspace settings sidebar on the Plans & credits screen. The user can review current plan status, compare Pro/Business/Enterprise options, and initiate an upgrade or credit purchase. If upgrading, the flow continues to a Stripe-style checkout where the user enters payment details and confirms the subscription. Exit is the updated Plans & credits screen showing the new active plan and refreshed credit balance. Steps: Current Plan Overview → Checkout and Payment → Updated Subscription Status **District: Offers Search and Selection** (ios, 3 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12822) > A logged-in mobile user opens an Offers overlay from a prior screen, searches within unlocked offers, and can either browse all available offers, inspect details, or apply an offer. The flow starts with the offers list and search bar, includes an empty-state result when the search term has no matches, and returns to filtered offer results when a matching keyword is entered. The flow ends when the user either closes the overlay or selects/apply an offer from the list. Steps: Offers List → No Results State → Filtered Offer Results **District: Movie Offer Application** (ios, 4 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12821) > Happy path on a movie detail screen where the user views information about a film, opens the offers panel, searches for a voucher, applies an eligible offer, and returns to the movie detail screen with the offer marked as applied. Steps: Movie Details → Offers List → Offer Search → Offer Applied **Asos: Customer Care Search & No-Results** (ios, 2 screens, Marketplace) [Refero](https://refero.design/flows/2940) > Mobile customer care entry that surfaces a prominent search box and categorized FAQ topics. Entry: user opens the Customer Care section (mobile browser). The primary path shown is: user enters a query in the help search (example: "visor") and is taken to a search results screen. Exit points: refine search and resubmit, select a FAQ category or question, tap header navigation (menu, account, basket), or close the help overlay. Assumptions: user is on a mobile browser (portrait), may be signed in or out (header shows account icon but no explicit sign-in state), and has basic browsing permissions. Steps: Customer Care landing → Search results (empty state) **222: Friend compatibility invite** (ios, 7 screens, Lifestyle, Dating & Relationships) [Refero](https://refero.design/flows/9322) > Happy-path for discovering compatibility with a friend. Steps: People list — Compatibility entry → Discover compatibility — Friend picker → Filter friends — Type query → Invite friend — Send invite → Picker shows invited state → People list — Compatibility slot created → Nudge friend — Send reminder **Faire: Request Unlisted Product** (web, 5 screens, Business, Marketplace) [Refero](https://refero.design/flows/13083) > Happy path on a brand storefront where a buyer requests products that are not currently listed. Entry is the brand shop/product listing page. The user opens the brand actions menu, chooses Request unlisted product, fills in one product name, URL, and quantity in the modal, then submits. The modal closes, a confirmation toast appears, and the storefront remains visible. Steps: Brand Storefront → Open Request Form → Enter Product Details → Submit Request → Request Confirmed **Handshake: Job application via modal** (web, 5 screens, Social Networking, Job Search) [Refero](https://refero.design/flows/8788) > Happy path for applying to a job listing from the job detail view on desktop. Entry: signed-in user lands on a specific job listing (Design Researcher). The user reviews job details, clicks the Apply control, attaches required documents (resume and optional cover letter), answers a work-authorization question, and submits. Exit: the listing updates to show a successful application state (green banner with applied date) and offers a Withdraw application CTA. Assumptions: user is signed in and has at least one document (resume/cover letter) available to attach. Steps: Job listing — primary view → Application modal — attach resume → Application modal — add cover letter and confirm authorization → Submission acknowledgement (toast) → Job listing — applied state **Lovable: Start Project Prompt** (web, 6 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12303) > A logged-in user lands on the Lovable home screen with a left sidebar, a central prompt composer, and a template gallery below. The user can start by typing or using voice input in the composer, optionally connect tools, then submit the prompt to generate a project. The flow ends when the prompt is fully entered and ready to build/submit, with the system showing the composed request and available build controls. Steps: Home Prompt Start → Template Inspiration → Prompt Refine → Voice Capture → Transcribing State → Prompt Ready to Build ##### Task: approval workflow **Introspect: Item Verification Share** (ios, 4 screens, Finance) [Refero](https://refero.design/flows/11951) > Happy path begins on an Item Overview screen for a saved item, with the user logged in on a mobile device. From here, the user opens the item’s overflow menu, chooses to verify/share the item, and is taken into the system share sheet. The user then selects Messages, which opens a new iMessage draft addressed to a contact with a prefilled prompt and link preview. The flow exits at message composition, ready to send. Steps: Item Overview → Item Actions Menu → Share Sheet → Message Draft **Lovable: QR Asset Label Implementation** (web, 12 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12336) > The user is working in an authenticated project-building workspace with an existing AssetWise asset-tracking project open in preview. They ask the assistant what to do next, select a suggested QR-code labeling feature, and submit the request. The assistant generates a detailed implementation plan covering QR codes, public scan pages, printable labels, files, and out-of-scope items. The user reviews the plan, opens its editing controls, and approves it. The system then processes the approved plan and returns a completed feature summary indicating that QR labels and a public scan page were added. Steps: Choose the Next Project Step → Select a Feature Request → Generate the Plan → View Plan Generation → Review the Proposed Plan → Inspect Technical Details → Open Plan Editing → Format Plan Content → Finalize Plan Review → Save and Approve → Process Approved Changes → Confirm Feature Added **Bond: Restaurant Recommendation Approval** (ios, 2 screens, AI Tool, Social Networking) [Refero](https://refero.design/flows/14397) > The user reviews an AI-generated restaurant recommendation for “Best food to get right now.” On the initial review screen, the user can inspect the restaurant details, copy the recommendation prompt, approve the suggestion, or regenerate it. After approval, the interface updates the primary action to “Finalize” and displays a small positive-feedback reaction from the associated profile. The user can then finalize the approved recommendation. The flow begins with a generated recommendation awaiting review and ends with the recommendation prepared for finalization. The user is assumed to be viewing the recommendation on a mobile device with access to the relevant profile or memory-based context; login state and permissions are not visible. Steps: Recommendation Review → Approved Recommendation **pillowtalk: Pillowtalk Onboarding** (ios, 3 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13494) > An unauthenticated user is presented with a three-screen onboarding carousel introducing Pillowtalk’s purpose and benefits. The user can review the messages using the carousel, accept the Terms of Service and Privacy Policy, and then continue with Apple. The screenshots show the first two carousel states with consent unchecked and the third state with consent selected, enabling the Apple continuation button. Steps: Process Difficult Experiences → Decode Recent Texts → Accept Terms and Continue **XChat: Message request acceptance** (ios, 2 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/11664) > Two-step mobile messaging sequence for handling an incoming message request from another user. Steps: Message Request (pending) → Accepted Conversation **Structured: Structured AI suggestions review** (ios, 3 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3289) > Happy path: The user views Structured AI's suggested plans for their day, can accept all suggestions at once, or give optional feedback about the suggestions. Steps: Review suggestions and Accept All → Open feedback modal → Enter and submit feedback **WhatsApp: Announcement Replies Thread** (ios, 3 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12679) > This is a mobile reply thread for an announcement message. The user starts with the reply panel open, sees the original announcement from Samuel, types a response in the bottom input, and sends it. After submission, the new reply appears in the thread with the sender’s name/avatar and timestamp. The user remains in the replies view and can close it with the X button. **Programa: Approve Product Item** (web, 4 screens, Project Management, Business) [Refero](https://refero.design/flows/13428) > The user opens a product item from a project schedule, reviews its details, and changes its status from Draft to Approved. The product detail drawer remains open throughout the interaction. The flow starts with the item details visible and ends when the Approved status is applied to the item. The user is assumed to have permission to change item statuses. **Acctual: Manage approval rules (delete)** (web, 4 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8854) > Happy path: A signed-in user navigates to Settings → Approvals to remove an existing approval rule. From the approvals list the user opens an overflow menu on a rule, chooses Delete, confirms the deletion, and the rule is removed from the list with a success toast. Steps: Approvals list → Open rule actions → Confirm deletion → Rule removed and feedback **Aboard: Vacation plan approval** (web, 7 screens, Collaboration Tools, Paperwork) [Refero](https://refero.design/flows/828) > A manager-facing desktop view of a vacation planning tool. Steps: Plan in Draft → Plan Active – waiting for submissions → Select employee details → Mark reportees ready → Ready state overview → Confirm completion → Plan Done **Aboard: Document Review and Approval** (web, 3 screens, Collaboration Tools, Paperwork) [Refero](https://refero.design/flows/834) > Two-column document review interface. Entry: user lands on Documents area (left column lists documents grouped under categories; right column is a preview area). Happy path: user selects a document from the left list -> the document opens in an embedded PDF viewer with thumbnails and toolbar -> an inline confirmation message and Approve button appear when the document is ready to approve -> user can select a different document from the list to open it. Assumption: user is signed in and has access to the Documents section; screenshots are on a desktop/laptop viewport. Steps: Document list — select document → Embedded PDF viewer with approval CTA → Switch to another document ##### Task: book GP appointment **District: Event Booking Confirmation** (ios, 5 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12872) > Happy-path ticket booking for an event, starting on a booking review screen and ending on a confirmed order with an accessible QR code ticket. **District: Movie Booking Support Request** (ios, 7 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12840) > A user opens a movie-support conversation that already contains a booking summary. They confirm the booking, are prompted to describe the issue, open the attachment menu, and grant photo-library access. A booking confirmation containing a QR code and ticket details is shown as supporting context. The user types a request to replace cheese popcorn with salt, then sends the message with the booking image attached. The flow starts in an existing support chat and ends with the issue submitted to the support assistant. **Saint Laurent: Store Appointment Booking** (web, 6 screens, Marketplace) [Refero](https://refero.design/flows/501) > Happy path: a shopper uses the store locator to find a nearby Saint Laurent boutique, opens the store details, and requests a private in-store appointment via a right-side booking panel. Steps: Store locator results → Store detail (Grant Avenue) → Open appointment panel → Select date — no hours available → Select alternate date and time → Submit appointment — error response ##### Task: budget spending insights **Wabi: Submit Personal Analysis** (ios, 8 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12052) > This is a mobile analysis experience where the user enters a personal feeling, dream, or life situation, then submits it for interpretation by a panel of analysts. Steps: Prompt Entry → Text Composed → Submit Analysis → Processing State → Results Begin → Analyst Responses → Detailed Reading → Expanded Analysis View **Pool: Place Details to Directions** (ios, 2 screens, Utility) [Refero](https://refero.design/flows/11980) > Happy path starts from a saved or viewed place detail screen for Citizens Bank Park, showing venue information, actions like call/website/tickets, and a directions button. The user can open navigation from the place detail view, which transitions into the maps interface focused on Citizens Bank Park and nearby points of interest. Steps: Place Details → Map Results **Plane Finder: Breakdown category selector** (ios, 3 screens, Travel & Booking) [Refero](https://refero.design/flows/10403) > User views an SFO Performance screen with a top chart and a “Breakdown” card below. The Breakdown card shows a ranked list of items (default: Airlines) with icon/flag on the left, short code below the name, and a numeric count on the right. A compact category selector (label + caret) in the card header opens a small popover to choose other breakdown dimensions (Airlines, Airports, Countries). Selecting a different dimension closes the popover and replaces the list contents and header summary accordingly. Steps: View Breakdown (Airlines) → Open Category Picker → View Breakdown (Countries) **Spyglass: Export and Rate Analysis** (web, 3 screens, Business, Inspiration Board) [Refero](https://refero.design/flows/13461) > A user reviews a generated advertising analysis in the Spyglass workspace and interacts with the response action toolbar. The sequence begins with the completed response visible, then shows a PDF download confirmation, a hover tooltip for the positive-feedback control, and finally the positive-feedback control selected. The user appears to be logged in and viewing an existing conversation or research workspace. The flow ends with the response retained in the conversation and feedback recorded. Steps: Review Generated Analysis → Identify Helpful Response Control → Submit Positive Feedback **Foodllama: Nutrition Insight Completion** (ios, 7 screens, Health & Wellness) [Refero](https://refero.design/flows/11779) > This is a guided content experience in the Foodllama mobile interface. The user starts on a home/dashboard screen showing a featured insight card and a progress summary for today’s insights. Tapping or swiping into the card opens a full-screen lesson sequence with multiple steps, each presenting one nutrition tip with illustration, title, short explanation, and a progress indicator plus a Leave action. The sequence ends on a completion screen with celebratory confetti, a trophy graphic, and options to tell friends or finish. After completion, the user returns to the dashboard, where the insight count updates. Assumptions: user is already inside the app, on a mobile device, and viewing personalized insight content. Steps: Insight Card Dashboard → Intro Tip Screen → Carb Tip Screen → Strength Tip Screen → Completion Celebration → Return Dashboard → Updated Insight State **Omio: Recommended train selection** (ios, 2 screens, Travel & Booking) [Refero](https://refero.design/flows/5171) > User is viewing search results for a one-way trip from Barcelona to Madrid (Wed, Apr 9, 1 adult) on a mobile device. The screen shows a transport-mode summary (train / bus / plane) with prices and durations, filter/sort chips (Sort: Cheapest, Stops, Departure), and a list of results. One result is marked RECOMMENDED and shows operator, train number, departure/arrival times, stations, duration, transfers, and price. The user can open an informational modal explaining why the train is recommended, then dismiss it with a Got it button to return to the results. Steps: Results list with recommended card → Recommendation information modal **District: Restaurant Q&A Assistant** (ios, 7 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12846) > Happy path for an in-context restaurant question assistant. The user starts from a restaurant detail context and asks about Hotel Rishi. The interface shows the current topic, displays the user’s question as a highlighted prompt, and returns an AI-generated response with related action cards (Food Menu, Bar Menu, Beverage Menu). The user can rate the answer, ask a follow-up by text or voice, or submit a new query such as finding nearby breakfast options. The flow ends with a refreshed response and the follow-up input ready for another question. **Linktree: Insights Dashboard and Content Assistant** (web, 2 screens, Business, Social Networking) [Refero](https://refero.design/flows/12411) > Logged-in user is viewing the Linktree Insights area from the left sidebar. The main dashboard shows performance cards such as most clicked links, visitor sources, lifetime totals, and social posting/connect prompts. On the right, an AI assistant drawer is open for generating personalized Instagram content. The user can review analytics, follow prompts to connect accounts or explore insights, and interact with the AI assistant. The flow starts on the Insights dashboard and ends with the user reviewing or continuing within the assistant panel or dashboard cards. Steps: Insights Overview → AI Prompt Refinement **Claude: Interactive Budget Dashboard Creation** (web, 3 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/1821) > Happy-path: A logged-in user (Robert) uses the chat assistant to convert a screenshot of a personal monthly budget into an interactive React dashboard component. Steps: Assistant Home / New Chat → Request Dashboard from Image → Review Generated Component and Preview **GlossGenius: Staff Performance Analytics** (web, 3 screens, Business) [Refero](https://refero.design/flows/9980) > Happy path: A signed-in user navigates to the Insights -> Analytics area to review salon revenue and performance metrics for the current month. The dashboard loads in an "All staff" default state showing aggregate metrics (net revenue, projected revenue, service vs retail split, team sales, booking utilization, average ticket, total bookings). The user can open the staff filter in the top-right, select an individual team member, and the dashboard updates to show metrics scoped to that person. Steps: Analytics overview (All staff) → Open staff selector → View selected staff metrics **Cohere: Market overview chat** (web, 3 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3856) > Happy path: A user composes a natural-language prompt in the Cohere Playground chat area, optionally configures JSON mode/schema and advanced parameters from the right-hand panel, sends the prompt, and reads the generated market overview. Entry: user lands on the Playground with Chat selected and an empty chat editor. Exit: the user has a multi-section, generated response (market size, drivers, innovations, challenges) visible in the chat and the editor ready for follow-up. Assumptions: user is on a desktop viewport, logged in, and has access to model/parameter controls on the right-hand panel. Steps: Prompt composition & schema setup → Submit prompt and view error / initial generation → Review expanded assistant response ##### Task: business account onboarding company verification **Fuse: Virtual Bank Account Signup** (ios, 24 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8730) > Happy path: From an empty Fuse Wallet the user opens a promotional card to create a virtual bank account (USD or EUR), chooses a currency, supplies an email, and is routed to a third-party identity verification (Bridge → Persona). The user completes the Basic Information step (country, name, birthdate, SSN, address with autocomplete) and submits. Steps: Wallet Home (empty) → Currency Selection (USD shown) → Currency Selection (EUR shown) → Bridge - Email capture → Bridge - Email entered → Connecting / Processing → Full-screen Loading → Persona Consent (Getting started) → Required submissions list → Country selection → Basic Information (form start) → Typing name fields → Date picker (scrolling) → Date selected → SSN entry placeholder → Typing SSN (masked) → SSN visible (toggle) → Address autocomplete → Address details (expanded) → Verification failure **UGLYCASH: New user onboarding and verification** (ios, 19 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/11202) > End-to-end new-user onboarding for a trading app: marketing carousel -> sign-in options -> phone verification (WhatsApp/SMS) -> Apple sign-in fallback -> username + referral -> legal consent -> biometric enablement -> first-time home. Steps: Welcome carousel - follow traders → Welcome carousel - trade everything → Welcome carousel - trading clubs → Welcome carousel - custody ownership → Sign-in selection → Phone entry (channel chooser) → Phone entry (send via SMS) → Code entry (empty) → Code entry (auto-filled) → Linked-account conflict modal → Apple Sign In prompt → Apple signing in → Choose username → Username claimed (input filled) → Referral code entry → Accept terms and conditions → Enable Face ID prompt → Face ID OS permission → First-time Home (authenticated) **Doji: Account creation & likeness submission** (ios, 47 screens, AI Tool, Marketplace) [Refero](https://refero.design/flows/6210) > Onboarding sequence for a mobile product that guides a new user from an initial hero carousel through phone verification, image review, permission prompts, brand-preference selection, and into the live app while a background 'likeness' creation job runs. Steps: Welcome hero — slide 1 → Welcome hero — slide 2 → Welcome hero — slide 3 → Phone entry — country & number (initial) → Country picker modal → Phone entry — number typed → Submit phone number → Code entry — waiting → Code typed — verification attempt → Verifying code → Review uploaded images → Notifications prompt (in-flow) → Creating likeness — progress screen → Select brands — entry → Select brands — selection & save → Creating likeness — animated progress → Creation continues — background state → Onboarding complete — return to home → Brand selection — long list navigation → Saved preferences and final progress **Drinkit: Phone verification and profile setup** (ios, 8 screens, Food & Drink) [Refero](https://refero.design/flows/11000) > Happy-path: A user on the venue/product landing screen taps the profile/avatar control to create or update their account. The app opens a modal to collect a phone number, sends a one-time code via WhatsApp or SMS, and shows a code-entry screen. After successful verification the user is prompted to enter a display name. Once the name is saved the modal closes and the Profile screen displays the saved phone number and name. Steps: Landing / Product view → Phone number input (initial) → Phone number input (filled) → Code entry screen (empty) → Code entry (in progress) → Name input (initial) → Name input (confirmed) → Profile screen (final) **Artsy: Profile Completion Checklist** (ios, 12 screens, Marketplace) [Refero](https://refero.design/flows/7617) > Happy-path: a logged-in user from their Account/Profile card taps Complete Profile and is guided through a 3-step onboarding to add a profession, upload a profile image, and send an ID verification email. Each step shows progress (1/3 → 3/3) and offers Save & Exit. After completing all steps the user sees a confirmation checklist and returns to their Account with an updated avatar and profession. Assumptions: user is authenticated, has camera/photo permissions, and an email address tied to the account. Steps: Account — Prompt to complete profile → Add your profession — input focused → Add your profession — typing → Profession saved — proceed → Add a profile image — placeholder → Select source — photo library → Pick photo — preview → Profile image set — proceed → Verify your ID — send email → ID verification — confirmation → Completion — checklist → Account — updated profile **Ada: Onboarding and Account Creation** (ios, 19 screens, Health & Wellness) [Refero](https://refero.design/flows/4094) > New-user onboarding that introduces the product (5 illustrated slides) and guides the user through creating an account. Steps: Onboarding - Intro (1/5) → Onboarding - Causes (2/5) → Onboarding - Next steps (3/5) → Onboarding - Get started (4/5) → Onboarding - Final slide (5/5) → Consent - Confirm you're ready (initial) → Consent - Checkboxes selected → Create account - Empty form → Create account - Inline validation errors → Create account - Corrected email → Create account - Password entry and show/hide → Manage your privacy - Default toggles → Manage your privacy - Agree or save choices → Notifications & tracking opt-in → Check your inbox - Verification sent → Tips to find your email → Verification email - External message → Email verification success (web) → Account verified → Welcome Home **Copperx: New user sign-up and KYC onboarding** (web, 18 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8892) > End-to-end new user registration and initial onboarding for a financial product (marketing -> signup -> verification -> KYC -> dashboard). Entry: user arrives from marketing/marketing hero or direct signup link. Happy path: user chooses sign-up method (Google or email), verifies email with a 6-digit code, creates a user-password account (with inline password-strength feedback), selects account type and services, fills personal/KYC-related fields, accepts a third-party terms modal when linking bank accounts, and lands on the authenticated dashboard. Exit: user reaches the dashboard (account created but KYC incomplete). Assumptions: user is not signed in at entry; user has access to the email used for signup; device is a desktop browser; user can upload or provide KYC documents later. Steps: Marketing landing / signup entry → Signup method selection → Email entered (confirmation) → OTP entry (empty) → OTP entered (partial) → Account creation form (basic) → Password feedback (weak) → Password feedback (improving) → Password feedback (strong) → Services & account type selection → Service selection (multiple chosen) → Personal information / KYC form → Third-party terms modal (Bridge) → Submission error feedback → Dashboard landing (KYC prompt) → Dashboard left navigation → Empty transactions state → Ongoing verification reminder **Acctual: Account onboarding and payments setup** (web, 26 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8817) > Happy-path onboarding for a payments/invoicing product. Entry from the marketing/home hero (promotional landing) into account creation. The user either authenticates with Google or supplies email + password, then selects who will be paid (Individual or Business). For a Business, the user supplies company name and contact name, triggers an email verification, clicks the verification link in their inbox, then returns to complete branding and business details (logo, postal address, tax ID). After submitting those details the user reaches a confirmation screen and is taken to the Payments area where they can continue to configure receiving methods and identity verification. Steps: Marketing / Landing Hero → Create account (empty) → Email entered → Password reveal toggle → Choose who gets paid → Business selected & fields displayed → Verification pending state → Verification email (inbox) → Make it yours (branding) → Address modal (open) → Country selector list → Country search no-results → Country selection corrected → Add address → Branding preview & inputs → Tax ID entered → You're all set → Payments overview → Payments onboarding CTA → Landing in Payments workspace **Acctual: Company document verification** (web, 4 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8828) > Summary: This is a modal-based, multi-step document submission sequence for verifying a company's identity. Steps: Select document type and upload (Legal presence) → Upload ownership structure document → Upload company details / proof of address → Review uploaded documents and continue **Square: Merchant onboarding and account setup** (web, 38 screens, Business, Finance) [Refero](https://refero.design/flows/1952) > Happy-path merchant onboarding from marketing into the Square dashboard. Entry: visitor clicks through marketing/CTA to sign up. The flow collects account credentials, business entity details, business category, identity verification (SSN), payout preferences, and bank account details. It then shows a summary of chosen products, offers app download, and lands the user in the Square dashboard. Steps: Marketing / Signup entry → Create account (empty) → Create account (filled) → Submission loading → Business entity question → Entity type options → Business details form → Business details (filled) → Business category picker → Category search results → Full SSN requested → Verification failure / Retry → Choose payout frequency → Payout option selected → Link bank account (empty) → Link bank account (filled) → Selections overview → Welcome / Next steps → Download Point of Sale → Dashboard (first sign-in) **Pinterest: Business account conversion** (ios, 18 screens, Reference, Inspiration Board, Social Networking, Lifestyle) [Refero](https://refero.design/flows/2547) > A mobile, in‑app sequence that converts a logged‑in personal account into a business account and walks the user through an 8‑step business onboarding (profile name, website, business type, ads interest) followed by a short product tour and an option to create the first Pin. Steps: Account management → Add account — options → Convert to a business account — confirm → Onboarding — Add profile name (input) → Onboarding — Profile name (filled) → Onboarding — Add site (input) → Onboarding — Describe your business (select) → Onboarding — Ads interest (select) → Onboarding — Tour slide: Share ideas → Onboarding — Tour slide: Grow your audience → Onboarding — Tour slide: Build a profile → Onboarding — How would you like to start? → Photo picker — grid → Photo picker — preview → Pin created — success modal → Profile view — initial state → Profile view — engaged state → Post-creation tips & confirmations **Nooka: Account creation, onboarding, and subscription** (ios, 50 screens, AI Tool, Education, Books) [Refero](https://refero.design/flows/8112) > End-to-end new user acquisition: from welcome -> sign-in / sign-up -> email verification -> short onboarding questionnaire -> subscription trial purchase -> return to main content. Steps: Welcome / Entry → Login (email) / Social alternatives → Sign Up — empty form → Sign Up — email input → Sign Up — password entry → Email Verification screen → Verification polling → Onboarding — Intro slide 1 → Onboarding — Question slide (choose friction) → Onboarding — Confirmed selection → Free Trial overview → Subscription modal — summary pricing → Subscription modal — Pro (default) yearly selected → Subscription modal — Plus tier selected → Subscription modal — Ultra tier selected → Native App Store confirmation → Purchase success modal → Subscription confirmation overlay → Tap to Join overlay (open mic prompt) → Home — authenticated content **WhatsApp: Add Verified Profile Link** (ios, 5 screens, Social Networking, Collaboration Tools) [Refero](https://refero.design/flows/12614) > Happy path for adding a verified social link to a WhatsApp profile. The user starts on the WhatsApp Profile screen, opens Links, confirms adding WhatsApp to Accounts Center, and returns to Links with an Instagram verified link added. Steps: Profile Overview → Links Options → Accounts Center Confirmation → Link Added → Success Feedback **Voicenotes: Email Verification** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14272) > The user starts in Settings with a pending email-verification prompt. They open the verification step, enter the six-digit code sent to their displayed email address, and submit it. The system validates the code, shows a success confirmation, and returns the user to Settings with the pending verification prompt removed. The user is assumed to be signed in on an iOS mobile device and able to access the specified email inbox. **Turf: Phone Sign-In & Onboarding** (ios, 38 screens, Sports) [Refero](https://refero.design/flows/9419) > New-user sign-in and onboarding using phone verification. Steps: Welcome / Splash → Phone Number Input (empty) → Phone Number Input (filled) → Submitting Number (loading) → Verification Code Entry (empty) → Verification Code Entry (partial) → Verification Code Entry (processing) → Name Input (empty) → Name Entry (filled) → Contacts Permission Prompt → Contacts Shared Confirmation → Initial App Skeleton / Loading Home → Onboarding Card: Welcome → Onboarding Card: Get Ticket → Onboarding Card: Pick a Side → Onboarding Card: Win Real Cash → Onboarding Card: Coming Soon (Ask Coach) → Onboarding Card: You're Ready → Home: Ticket List Visible → Home: Interact With Ticket **Navan: New user signup and company onboarding** (web, 18 screens, Travel & Booking) [Refero](https://refero.design/flows/9346) > Happy-path flow for a new user who visits the marketing site, signs up with a business email, verifies that email, creates a personal account and then begins company onboarding. Steps: Marketing homepage (entry) → Sign up — Enter work email → Email input accepted (illustration change) → Validation: missing marketing checkbox → Checkbox selected — continue enabled → Email rejected: personal email blocked → Alternate business email entered → Check your inbox (verification sent) → Onboarding intent selection → Create your account — form → Create account submitted → Authenticated landing — 'Where to?' → Pricing / Create company account → Welcome onboarding modal → Onboarding side panel — task checklist → Resume account setup tooltip → Admin dashboard access tooltip → Final landing with Account setup indicator **Peerlist: Create Company Profile** (web, 6 screens, Job Search, Business, Social Networking) [Refero](https://refero.design/flows/1102) > Happy path: A logged-in user (with permission to create a company) initiates creating a company profile via a modal. They choose an organization profile type, complete the profile form (display name, website, tagline, location, handle, size, etc.), then trigger an email verification step. After entering the verification code and accepting terms, the profile is created and the user is taken to the newly created company profile overview. Steps: Choose profile type → Company profile form (empty) → Company profile form (filled) → Open verify work email modal → Enter verification code → Company profile overview **Mercury: Formation document upload and validation** (web, 4 screens, Business, Finance) [Refero](https://refero.design/flows/11020) > Happy path: a logged-in user uploads a company's formation document from a desktop browser. The screen lists acceptable formation document types and provides a drag-and-drop upload area. After the user drops a file the system uploads it, runs automated checks (entity name, document type, filing date) and shows progress and status. If the document passes checks the user can continue with onboarding; if issues are found the system displays a warning with itemized problems and guidance. Steps: Document instructions and upload area → File added / upload started → Automated checks in progress → Validation results and issues **time2book: Studio onboarding and account creation** (web, 8 screens, Business, Fitness, Productivity) [Refero](https://refero.design/flows/6305) > Happy-path onboarding for a fitness studio platform. Steps: Landing / Marketing Hero → Sign in / Choose authentication → Email entry confirmation → Check your inbox / Email verification prompt → Account setup — role selection (Client) → Account setup — role selection (Provider) and fields → Provider form filled & submitted → Onboarding / Setup guide ##### Task: buy crypto **Fuse: Token selector** (ios, 2 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8735) > A compact, full-screen modal for selecting a cryptocurrency token. Entry: user invokes a token selector (e.g., from a send/swap/receive flow). The modal shows a searchable list of tokens with icon, token name, on-chain balance, and USD equivalent. The user can either tap a token row to choose it or type into the search field to filter. Exit: modal closes and the chosen token is returned to the parent context, or the modal remains open showing a no-results state if the search yields nothing. Assumptions: user is on a mobile device, signed into a wallet with at least one token (Solana is shown), and the client can show token balances and USD values. Steps: Token list (default) → Search input — no results ##### Task: calendar scheduling meeting **Genie: Chat Reminder Setup** (ios, 3 screens, Social Networking, AI Tool) [Refero](https://refero.design/flows/12426) > A chat-style assistant screen where the user types a reminder request, sends it, and receives confirmation that a reminder has been created for tomorrow at 9:00 AM. The flow begins in an active conversation with the keyboard open and ends with the assistant confirming the reminder and offering to save reference photos to a vault. Assumptions: user is already in the conversation, signed in, and has permission to create reminders. Steps: Compose Reminder → Message Sent → Reminder Confirmed ##### Task: care plan **Ada: Record a symptom entry** (ios, 12 screens, Health & Wellness) [Refero](https://refero.design/flows/4107) > Happy path: From a feedback prompt the user chooses to track symptoms, picks a symptom (Dizziness), answers a short guided questionnaire (trigger: standing up effect; description type), optionally sets current severity using a segmented color-coded slider and +/- controls, optionally adds freeform notes, saves the entry, then returns to a timeline where the new entry can be opened to view details. Steps: Feedback prompt → Choose symptom → Initial symptom questions → Select trigger effect → Select symptom description → Open severity control → Set severity (low) → Set severity (mid) → Set severity (high) → Open notes field → Enter free-text note → Save and view entry **Ada: Medical quality article walkthrough** (ios, 10 screens, Health & Wellness) [Refero](https://refero.design/flows/4104) > Happy path: A user finishes a symptom assessment and sees a high-priority recommendation ("Seek emergency care") with an inline "Tell me more" button. Tapping that link opens Ada’s external "Medical quality" article (about.ada.com) in a browser/webview. The user scrolls through the article: headline, supporting copy, visual illustrations, a "What to look for" section with accordion items (clinical governance, user experience, condition coverage, etc.), expands "Condition coverage" to view bullet points, continues to the "How Ada’s AI works" section, sees a CTA to contact the company, reviews the reference list (numbered DOIs), and reaches the footer with legal links and social icons. Steps: Symptom result with "Tell me more" → Article landing — Medical quality → Evidence section header → "What to look for" introduction → Accordion list of quality pillars → Expand 'Condition coverage' → How Ada's AI works → Contact CTA and Keep reading → Reference list (citations) → Footer and site navigation **Weather mini: Trip Planning and Save** (ios, 7 screens, Weather) [Refero](https://refero.design/flows/13105) > Happy path for creating a multi-city trip from the Trip overview, entering dates and stops, using AI/search suggestions, generating trip details, and saving the plan back to the main trip list. Steps: Trip Overview → Initial Trip Setup → Search Suggestions → Processing Entry → Trip Draft with Stops → City Weather Details → Saved Trip List **Lovable: Plan Review and Approval** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12337) > A logged-in user is reviewing a proposed plan in a split-screen editor. The left side shows the task discussion and a plan card with a summary, while the right side shows the detailed plan content with a formatting toolbar. The user can inspect the proposal, optionally change the content type/format using the text style dropdown, and then either cancel or save and approve the plan. Steps: Plan Review → Text Style Selection → Approved Plan State **Cohere: Business Plan Outline Generation** (web, 2 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3827) > A single conversational session where a logged-in user submits a prompt to generate a business plan outline in French and views the formatted result. Steps: Open conversation and submit prompt → Expose copy control for generated content **Cohere: Business plan assistant interaction** (web, 3 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3829) > Happy-path summary: A signed-in user is inside a chat conversation and asks the assistant for a business-plan outline in French. The assistant returns a formatted outline in the conversation pane. The user has two PDFs staged in the message composer (not yet sent). The user hovers or clicks on the assistant message actions (e.g., Regenerate) to request a different response. After requesting regeneration, the assistant posts a new reply (in this case an apologetic message saying it couldn't find a plan in the provided documents). Steps: View generated business plan → Reveal message actions → Regenerated assistant reply ##### Task: case management **Figma: Manage team member permissions and resend invite** (web, 3 screens, Design & Media Editing, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/4990) > Happy path: A signed-in user opens the team "Members" modal to review and manage team collaborators. From the modal they inspect the list of members, open the action/role dropdown for a specific member, and trigger an action (in this case "Resend" an invite). The flow ends with an on-screen confirmation toast saying the invite email was sent again. Steps: Open Members modal → Open member role/action menu → Receive confirmation **OpenAI: Prompt Evaluation Setup** (web, 13 screens, Business, Development Tools, AI Tool) [Refero](https://refero.design/flows/4223) > Happy path: A logged-in user opens the Evaluations area, creates a new evaluation by selecting a data source (Import Chat Completions / create data / upload file / custom), composes or imports test cases, selects a model to generate responses, adds one or more test criteria (e.g., valid JSON), reviews the evaluation configuration, runs it, and sees the run status and results saved in the evaluations list. Entry: Evaluations screen (empty or default). Exit: Saved evaluation entry and run result shown in the list. Assumptions: User is logged in to the project, has permission to run evaluations, has access to models referenced (e.g., gpt-4o-mini), and is on a desktop browser. Test data can be created manually or imported from stored completions or files. Steps: Evaluations list (empty) → Select data source → Import Chat Completions (instructions) → Create test data (blank form) → Create test data (sample content) → Create test criteria (empty) → Add testing criteria (chooser) → Configure 'valid JSON or XML' criterion → Criteria list updated → Review evaluation → Run in progress → Run completed (failed) → Evaluations list (with saved evaluation) **Cohere: Account registration and onboarding** (web, 15 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3790) > Happy-path desktop web flow for a new user who discovers the product marketing site, creates an account, completes an onboarding questionnaire, and arrives at the product dashboard. Steps: Marketing landing → Create account — entry form → Password strength — weak → Password strength — improving → Password strength — good → Password strength — strong → Verify your email → Onboarding — name collection (step 1/3) → Onboarding — name confirmed → Onboarding — role selection (step 2/3) → Role selected → Onboarding — use case selection (step 3/3) → Use cases selected with custom note → Dashboard — welcome → Dashboard — modules and docs **Polar: Organization Review and Submission** (web, 4 screens, Business, Finance) [Refero](https://refero.design/flows/11406) > Desktop flow for submitting an organization's details for an automated compliance/use-case review. Steps: Account review form (initial) → Edit fields and add social link → Submission confirmation → Automated review in progress **Wispr Flow: Submit Support Request** (ios, 7 screens, AI Tool, Utility, Productivity) [Refero](https://refero.design/flows/13684) > A logged-in user opens the Flow navigation drawer and chooses Talk to Support. A support form appears as a bottom sheet with Report an issue selected. The user opens the Request type selector, chooses Account Management, enters an issue description, and sends the message. The sheet closes and a Message sent confirmation appears on the home screen. **Google Gemini: Access Gemini Help Center** (ios, 7 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/3226) > Happy-path: A signed-in user opens the account menu, chooses Help, then navigates into the Gemini Apps Help support site and reads help articles. Entry: user is already signed into a Google account on the device (profile avatar visible). Exit: user ends on a specific support article (’Use the Gemini web app’) and can browse related resources or give feedback. Assumptions: user is signed in, has network access, and taps controls on a mobile device using the system browser or an in‑app browser sheet. Steps: Open account menu → Help selection → Open Gemini Help hub → Browse popular topics → Open specific article → Read detailed sections → Access related resources & feedback **Instagram: Saved products list** (ios, 2 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/3662) > Happy path: a signed-in mobile user opens the Saved products screen, reviews saved items (thumbnail, title, brand, price) and removes an unwanted item by tapping the trash icon. The item disappears from the list immediately and a temporary snackbar appears at the bottom confirming removal and offering an Undo action. Steps: Saved products list (initial) → Removed confirmation with Undo **Lumy: Manage scheduled notifications** (ios, 3 screens, Weather, Utility) [Refero](https://refero.design/flows/6169) > A compact notifications management screen for scheduled location-based alerts (sunset notifications shown). Steps: Notifications list (grouped) → Reveal Delete action → Updated notifications list (post-delete) **Phantom: Settings: Search & Locate** (ios, 2 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3729) > User opens the Settings screen on a mobile device (portrait). The screen lists the user's profile and multiple settings categories (Manage Accounts, Preferences, Security & Privacy, Active Networks, Address Book, Connected Apps, Developer Settings, Help & Support, etc.). The user uses the top search field to locate a specific setting. After typing a partial query ('Noti'), the keyboard remains visible and the list is filtered to show matching items ('Notifications'). Steps: Settings Overview → Search Input & Filtered Results **ABY Journal: Data & Danger Controls** (ios, 3 screens, Health & Wellness, AI Tool, Lifestyle) [Refero](https://refero.design/flows/8001) > User opens the Settings screen to manage data, review terms, access support, and perform destructive actions. Steps: Settings — Data & Terms → Locate Danger Zone & Support Links → Confirm Delete All Insights **Bump: Manage Supersend (delete)** (ios, 4 screens, Map & Navigation, Social Networking) [Refero](https://refero.design/flows/3170) > Happy path: A signed-in mobile user browsing a map of nearby "supersends" (sticker packs) opens the Supersends list, opens the options for one pack they created, and deletes it. Steps: Map — Sticker overview → Supersends list (bottom sheet) → Item action sheet → Updated Supersends list after delete **Ground News: Remove Saved Story** (ios, 3 screens, News) [Refero](https://refero.design/flows/7578) > Happy path: A logged-in mobile user navigates to Profile → Saved Stories, views a saved story card, opens the item actions (ellipsis) and chooses "Remove from saved stories." The app shows a bottom action sheet with options, then removes the item and updates the list. Entry: Saved Stories screen (Profile tab). Exit: Updated Saved Stories list (empty state in this case). Assumptions: user is authenticated, on a mobile device, and the story shown is the only saved item matching the active filter. Steps: Saved Stories list → Story actions bottom sheet → Empty saved stories state **Fresha: Email & Social Sign-up** (ios, 17 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4946) > Happy-path and edge cases for a new-user sign-up started from the app's home/browse screen. Steps: Home / Browse (Logged out) → Sign-in Modal - Entry → Email Input Focus → Invalid Email Feedback → Valid Email Entered → Create Account - Name Entry (step 1) → Create Account - Name Autofill / Input → Create Account - Password & Mobile (step 2) → Country/Region Selector - List → Country Search / Select → Mobile Number Entry → Consent Checkboxes → Attempt Account Creation → Server Error (failure case) → iOS Notification Permission Prompt → Account Created - Profile Landing → Post-signup State & CTA **Vinyls: Album collection manager** (ios, 9 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/8917) > Happy path: From an album detail screen the user opens the "In My Collections" selector, optionally creates a new collection (name + emoji), selects one or more collections (including a special Wishlist state), saves the selection, and returns to the album detail screen where the selected collections are shown as small emoji chips. Steps: Album detail — add to collections → Collections selector — list → Create collection — empty form → Create collection — emoji input → Create collection — ready to create → Collections selector — new collection checked → Confirmation — remove from Wishlist → Collections selector — post-confirmation → Album detail — collections updated **Omio: Checkout & ticket retrieval** (ios, 23 screens, Travel & Booking) [Refero](https://refero.design/flows/5185) > Happy-path checkout for a single passenger bus booking from Barcelona to Madrid. Steps: Booking summary & passenger → Seating options preview → Protection options list (collapsed details) → Protection details – first expansion → Select Cancellation Protection (state required) → Protection details expanded with coverage list → Protection selected confirmation → Switch protection choice → Select payment method → Choose Card and show card form → Card number input (numeric keyboard) → Card details filled & save option → Enter voucher code (collapsed) → Voucher typing → Voucher rejected error → Review cost summary & terms → Initiate payment (processing animation) → Payment success → Retrieve ticket → Booking confirmed modal **Days: Support chat conversation** (ios, 5 screens, Scheduling Tools, Lifestyle) [Refero](https://refero.design/flows/7818) > Happy path: A user opens the app's Support hub, chooses to contact support, and is routed into an in-app chat handled by an automated assistant (Days AI). The assistant shows an intro message and an input field. The user types a question, sends it, sees a sending/loading state, and receives an automated reply with instructions. Steps: Support hub (help topics) → Chat assistant intro → Compose message → Send message & pending state → Bot responds with instructions **Google Gemini: Image generation from attached photo** (ios, 3 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/3213) > Happy path: A signed-in user in the Gemini conversation sends an image + prompt asking the assistant to create a new image based on that photo. The interface shows a modal listing drafts (in this case empty), provides a Regenerate action, then returns to the chat where a generation progress state is shown. Once complete, the generated image is displayed in the conversation with inline feedback and share controls. Steps: Other drafts — empty state → Prompt submission with photo — generating → Generated image result — review and actions **Tinder: Interest Selection** (ios, 3 screens, Dating & Relationships) [Refero](https://refero.design/flows/3435) > Mobile interest-selection screen where a user browses and picks interest tags. Steps: Browse Interests → Search Focus → Search Query and No Results **Asos: Add / Redeem Gift Voucher** (ios, 6 screens, Marketplace) [Refero](https://refero.design/flows/2831) > Happy path: A signed-in user opens their account settings, navigates to the Gift cards & e-gift cards area, and chooses to add/redeem a gift voucher. From the empty-state screen they can either add an existing gift card/e-gift or buy a new e-gift. Choosing to add a voucher opens a modal to select Gift Card or E-Gift Card and then a dedicated Add Gift Voucher form that requests a 16-digit code (from the user's email). The user types the code and submits; the system validates and either accepts the code or shows an inline error. Steps: Account Settings — Gift cards entry → Empty state — Gift cards → Select voucher type → Add Gift Voucher — instructions → Enter voucher code → Validation result — error state **Kickstarter: Reward pledge and checkout** (ios, 23 screens, Fundraising) [Refero](https://refero.design/flows/4631) > Happy-path and edge-case flow for backing a Kickstarter project from the project overview through reward selection, shipping-country selection, customizing pledge amount, entering payment, handling card errors, and seeing confirmation. Steps: Project overview → Select reward (initial) → Open country picker → Country list - browse → Country picker - numeric input (mistype) → Country search - text → Country selected → Customize your reward (open) → Adjust pledge amount (increment) → Manual pledge entry (numeric keyboard) → Pledge amount validation error → Checkout start / loading → Payment modal - choose method → Enter card (invalid data) → Attempt payment — declined → Add new valid card → Processing payment → Pledge success / Thank you → Post-purchase UX — suggestions → Project view updated — backer state **Mindllama: Focus Session Setup** (ios, 27 screens, Health & Wellness) [Refero](https://refero.design/flows/5490) > Happy-path: user opens the main screen (large 'Focus 15 min' CTA) and configures a single focus session by selecting sounds, adjusting volumes, choosing a background scene, setting the session duration, picking a character, then saving the configuration as a preset. Entry: main/home screen with tabs (Sound, Background, 15 min, Character, Presets). Exit: a saved preset (optional) and the main screen updated with the selected session values. Assumptions: mobile device; user has the app open and can interact with touch controls; audio playback may be blocked by the device's silent mode (the app surfaces a modal in that case); sounds can be downloaded for offline use. Steps: Main screen (initial) → Scroll sound cards → Select / preview sound → Silent-mode modal appears → Dismiss silent-mode modal → Change selected sound → Open floating sound controls → Switch sound category → Ambient sound list (collapsed) → Select multiple ambient sounds → Adjust per-sound volume → Floating pill sync → Open Background tab → Pick a scene from grid → Open 15 min (time) tab → Switch to Hours mode → Adjust hour value → Open Character tab → Select a different character → Open Presets tab and create preset **Asos: Customer Care Help Search** (ios, 2 screens, Marketplace) [Refero](https://refero.design/flows/2941) > Mobile customer-care entry screen with a prominent search field and a list of FAQ topic categories. The happy path: user arrives on the Customer Care landing (mobile browser), types a query or taps a topic, sees search results (count + list of matching FAQs), then taps an FAQ to open the article. Steps: Customer Care – Topics → Search Results – FAQs **Stripe: Manage Radar allow list items** (web, 3 screens, Finance, Business) [Refero](https://refero.design/flows/1315) > Happy path: a logged-in user navigates to the Payments > Lists area, opens a specific Radar list (Charge description allow list), adds a new item (single string or CSV), and confirms the item appears in the list detail table. Steps: Lists overview → Add new item modal → List detail with items **Cohere: Production API key onboarding** (web, 14 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3882) > Happy-path: A logged-in user on the API Keys area requests a production API key. The system walks them through a four-step modal: Safety & Legal acknowledgements, Use Case review, Payment information, and a final step to name and generate the key. After successful completion the modal reveals the one-time API key and the new key appears in the Production keys list (masked). Steps: API Keys listing → Safety & Legal (initial) → Safety & Legal (acknowledge) → Use case review → Payment information (form) → Payment validation error → Payment (card corrected) → Billing details entry → Payment submission error → Name production key (modal open) → Key name entered → Name validation error → API key generated → API Keys listing updated **Typeform: Rating-based branching** (web, 8 screens, Productivity, Communication Tools) [Refero](https://refero.design/flows/5591) > User is editing a form's Workflow to add conditional branching so respondents are routed to different next questions based on their answer to a rating question. Entry: user is in the form editor, Workflow tab open and the form (My new form) loaded. Exit: the rule is saved and the workflow canvas is updated with new conditional connectors. Assumptions: user is signed in with edit permissions, the form contains the listed questions/nodes, and question types accept numeric comparisons for routing. Steps: Workflow canvas (select question) → Open rule editor modal → Create IF condition → Choose comparison operator → Set value and open destination selector → Select destination node → Review and save rule → Updated workflow canvas **Maze: New account onboarding** (web, 11 screens, Business, Communication Tools) [Refero](https://refero.design/flows/3602) > Happy-path new-user onboarding for Maze (desktop). Entry: marketing landing hero where user clicks CTA to create a free account. The flow collects credentials, basic profile (name, company, role, company size), use-case preferences, verifies email, offers an invite-by-link or email step, and lands the user on a time-limited trial with the first-run dashboard. Steps: Marketing hero → Get started → Signup form (empty) → Signup form (email entered) → Profile details form (empty) → Profile details filled → Use-case selection → Confirm your email → Email verified → Invite team → Invite team (email entered) → Trial welcome / Entitlement → First-time dashboard / Home **Uber: Add voucher to Wallet** (web, 4 screens, Business, Real Estate) [Refero](https://refero.design/flows/453) > Happy path: A signed-in user opens their Wallet to view balance and manage payment instruments, then chooses to add a voucher. The user triggers an Add voucher modal, types a code, and the interface validates the code. Steps: Wallet landing → Add voucher modal - empty → Add voucher modal - input filled → Add voucher modal - validation error **Reclaim AI: Reclaim account setup** (web, 24 screens, Collaboration Tools, Scheduling Tools) [Refero](https://refero.design/flows/4082) > Happy-path new user onboarding for a calendar-first AI scheduler (marketing → sign-up → guided setup → product launch). Entry: marketing/marketing-signup or invite link. The user signs up (Google), then completes a step-by-step setup: connect a primary work calendar, select use case, role and department, enable recurring habits, create scheduling link, opt into buffer times, enable calendar sync, optionally invite teammates, then waits for account provisioning and lands in the product with a setup guide. Steps: Marketing landing → Sign up prompt → Welcome / overview → Connect work calendar → Calendar connected confirmation → Select primary use → Use case selected (example) → Select role → Role selected (example) → Select department → Habit templates intro → Habit template selected → Create scheduling links → Auto-schedule buffer intro → Buffer options enabled → Personal calendar sync intro → Create a team prompt → Team invite review → Account provisioning → Onboarding complete — Setup guide **Cushion: General Channel Actions** (web, 2 screens, Communication Tools, Collaboration Tools) [Refero](https://refero.design/flows/12513) > A logged-in user is viewing the General channel inside a workspace. The main area shows a list of discussions and a composer to start a new discussion. In the first state, the channel overflow menu is open with actions such as filtering, changing view density, bookmarks, copy ID, copy URL, edit, and archive. In the second state, the menu has been used to copy the channel URL, confirmed by a temporary toast message. Steps: Channel Menu Open → Copy Channel Link **Fourthwall: Hoodie print image management** (web, 3 screens, Business) [Refero](https://refero.design/flows/7286) > Happy-path: A logged-in user on a desktop browser opens a hoodie product customizer, uploads an image (or selects an already uploaded image), places it on the front print area, and manages uploaded images (open image menu, delete/download/edit). Entry: user lands on product customizer with at least one uploaded image. Exit: user has the desired image placed on the front of a hoodie and either continues to the next step by clicking Next or has removed/managed uploads. Assumptions: user is signed in (name shown in header), has permission to upload and manage images, and is using a desktop/laptop with sufficient screen width. Steps: Initial mockup with uploaded images → Thumbnail context menu → After image management (upload list updated) **Threads: Add profile link** (web, 6 screens, Social Networking, Entertainment) [Refero](https://refero.design/flows/10649) > Happy-path: A logged-in user opens Edit profile, navigates to the Links manager, adds a new external link (URL + optional title), and returns to the Edit profile screen where the Links row updates to reflect the new link count. Entry: user is on their Profile and opens Edit profile. Exit: Edit profile modal with updated Links count (user stays in edit state). Assumption: user is authenticated, has permission to edit their profile, and modals are stacked (Links manager and Add link appear as overlay panels). Steps: Edit profile open → Links manager → Add link (empty) → Add link (filled) → Links manager (updated) → Edit profile (links updated) **Matter: Manage Gmail Sources** (web, 3 screens, Magazines & Newspapers, News) [Refero](https://refero.design/flows/9397) > Happy path: A signed-in user who has already connected a Gmail account lands on the "Manage Gmail" area to review, search, and follow sources that were extracted from that Gmail account. The screen lists discovered senders (name, email) with a Follow/Following button per row. The user can type into the search input to filter the list; the UI updates to show matching senders or a "no results" message. Steps: Manage Gmail — Full list → Search — no matches → Search — filtered result **Voicenotes: Webhook management** (web, 4 screens, Collaboration Tools, AI Tool) [Refero](https://refero.design/flows/7003) > A modal-based webhook management interface that lets a logged-in user view, enable/disable, edit, delete, or add webhook endpoints. Steps: Open Webhooks modal / View list → Open entry action menu → Confirm deletion → Post-deletion state / Empty list **Ballpark: Manage guest membership** (web, 3 screens, Business, Productivity, Communication Tools) [Refero](https://refero.design/flows/2626) > Happy path: an authenticated user views the Team > Guests section, inspects a pending guest entry, opens the row menu and changes that guest's status (switch to member). The UI confirms success via a toast and the guest is removed from the Guests list, leaving the empty-state prompt to invite new guests. Steps: Guests list (pending entry) → Row actions menu → Confirm action and updated list (empty state) **Nudea: Customer Sign-In & Account Home** (web, 4 screens, Marketplace) [Refero](https://refero.design/flows/7668) > Happy path: A desktop user browsing the store (anonymous / logged-out) uses the header account entry to open the Account / Login screen, enters credentials, and is taken to their account home where personal links (Order History, Address Book, Rewards) and a welcome message are shown. Steps: Landing / Browse - Header Access → Account / Login - Empty Form → Account / Login - Credentials Entered → My Account - Dashboard (Authenticated) **Reclaim AI: Task Search and Up Next Management** (web, 2 screens, Collaboration Tools, Scheduling Tools) [Refero](https://refero.design/flows/4091) > Summary (happy path): A signed-in user views the Tasks area and the "Up Next" queue. From the full list they use the top search field to filter tasks (example: searching for "Joe"). The interface reduces the list to matching items and shows an empty droppable zone labeled to remove tasks from Up Next. The user can then drag the matching task into that area to remove it from the "Up Next" queue. Steps: Tasks List — Up Next Overview → Filtered Result with Remove Drop Zone ##### Task: code repository pull request **Wabi: Invite Collaborator** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12045) > Happy path for inviting another user to join a shared space from a collaboration screen. The user starts on a member management view showing current participants and invitation options. They can search for a person by username, tap an Invite action next to a matching result, and return to the main view where the member count updates. Steps: Member Overview → Search Active → Search Results → Invite Confirmed → Updated Member Count **Linear: Project selector** (ios, 3 screens, Collaboration Tools, Productivity, Project Management, Scheduling Tools) [Refero](https://refero.design/flows/6614) > This is a mobile project-picker modal used to reassign an item to an existing project. Entry: user opens the "Project" picker from an item (implicit). The modal lists available projects and exposes a search field at the top. The happy path: open picker → type or scroll to find a project → tap a project to select it → close the picker. Assumptions: user is on a mobile device, keyboard appears when the search field is focused, user has permission to reassign the item, and there are multiple existing projects to choose from. Exit points: selecting a project (selection shown with a checkmark) or tapping the top-right close button to dismiss the modal. Steps: Project picker (initial list) → Search — empty results → Search — single match **Rork: Invite Collaborators** (web, 2 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12169) > A signed-in user opens an invite/collaboration panel for a project, adds collaborators by email or creates a shareable invite link, then gets confirmation when the link is created or copied. The flow starts from a project workspace and ends with invitation access being generated and ready to share. Steps: Invite Collaborators Panel → Link Copied Confirmation ##### Task: compliance dashboard audit **Copperx: Account Dashboard - Transactions** (web, 3 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8899) > Happy-path: A logged-in user on desktop views their account dashboard to check balance, perform quick actions (Deposit, To bank, Send, Get Paid), and review recent transactions. The user can open the transactions overflow menu to reveal options (Show/Hide Dates, Download Statement) and toggle date visibility. Steps: Dashboard Landing → Open Transactions Overflow → Toggle Dates On **YouTube Music: Report a comment** (ios, 5 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11336) > User is viewing the comments sheet for a video (comments are sorted by Top or Newest). From this sheet the user opens the per-comment actions menu, chooses to report a specific comment, picks a reason from a modal list, and receives a submission confirmation. Entry: comments sheet is open for a video. Exit: confirmation modal is dismissed and the user returns to the comment list. Steps: Open Comments and locate comment → Open per-comment actions → Choose report reason → Submit report → Acknowledge confirmation **Dropbox: Member Activity Report Generation** (web, 4 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/13063) > Happy path in the Admin console Activity section: the user reviews activity records, opens the report generation dialog, optionally excludes file operations, and submits the report request. Steps: Activity Review → Report Options → Submit Report → Generation Confirmation **ElevenLabs: Developer Usage Analytics** (web, 2 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/8713) > Happy path: A signed-in developer navigates to Developers → Usage to inspect API consumption and credit usage for a selected time window. The interface presents an Analytics panel with controls for metric selection (e.g., Credits), date range, granularity, log scale, cumulative toggle, breakdown/filter options, and an Export CSV action. The user can switch visualization modes (line vs cumulative area) and review related sections below (API Requests, High priority concurrent requests) which display aggregated values. Entry: user opens the Developers area (desktop layout) and selects the Usage tab. Exit: user either exports CSV, adjusts filters, or navigates back to another Developers tab (Overview, API Keys, Webhooks). Assumptions: user is authenticated with developer-level access; using a desktop or large-screen browser; has permission to view usage analytics and export data. Steps: Usage - Default (Per-interval Credits) → Usage - Cumulative Credits View **Runey: Invoice PDF Download** (web, 2 screens, Business, Productivity, Project Management) [Refero](https://refero.design/flows/11704) > A logged-in user is viewing a completed invoice inside a project workspace and downloads it as a PDF. The entry point is an invoice detail/preview screen showing invoice metadata, line items, totals, and attachments. The exit point is a successful download confirmation toast after the PDF export action. Steps: Invoice Preview → Download Success **Tinder: Safety tips carousel** (ios, 3 screens, Dating & Relationships) [Refero](https://refero.design/flows/3438) > A three-card safety tips carousel presented as a centered modal overlay on the main Tinder screen. Each card contains an illustration, a bold headline, and two short guidance items. Navigation is linear via a prominent gradient CTA: 'Next' on the first two cards and 'Got it' on the final card. A three-dot progress indicator shows which card is active. The modal sits above the app's bottom navigation and dimmed background. Entry is when the modal is presented over the main screen; exit is when the user taps the final CTA ('Got it') which closes the modal and reveals the underlying main screen. Steps: Respect & Boundaries → Scams & Red Flags → Verification & Reporting **Lovable: Generated Landing Page Review** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12309) > This flow shows a user reviewing an AI-generated SaaS landing page inside a split workspace. The user starts from a prompt describing the desired landing page, then inspects the generated result in a live preview on the right while the prompt/result card remains on the left. The user can bookmark the output to history and optionally give feedback via reaction controls. The flow appears to assume the user is already signed in or has access to the workspace and preview environment. It ends with the generated landing page visible and a confirmation message that the result was saved/bookmarked or feedback was submitted. Steps: Preview Generated Design → Bookmark Result → Submit Feedback **Foodllama: Nutrition and body stats review** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/11789) > Happy path starts on the main Foodllama dashboard, where the user can review daily calorie intake, macro distribution, streak status, and weekly trend data. From there, the user taps a profile/avatar area to open a detailed stats view with weight, calorie goal, and BMI information, then can move between avatar/profile variations and return to the dashboard. Steps: Dashboard Overview → Detailed Stats View → Avatar Selection State → Dashboard with Selected Avatar **OKURI: Daily Task Checklist** (ios, 3 screens, Utility) [Refero](https://refero.design/flows/11866) > Happy-path task management screen in the OKURI experience. The entry point is a daily dashboard for May 13 (Wednesday), with date, sunrise/sunset, zodiac/astrology-style visuals, and a to-do list. The user is assumed to already be on the daily planner screen. The exit state is a completed checklist item, shown as crossed out and disabled/checked, with the list remaining on the same screen. Steps: Daily Dashboard → Task Completion Update → Multiple Task Completion **Structured: Checklist modal interaction** (ios, 4 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3266) > Happy path: user opens a day in the calendar timeline, taps an onboarding checklist task to open a modal containing checklist items, selects some checklist items, and closes the modal. The timeline updates the task’s progress indicator to reflect the selections. Steps: Day timeline - onboarding task visible → Checklist modal opens → User checks checklist items → Modal dismissed and timeline updated **Brilliant: Data lesson with interactive quiz and bug report** (ios, 48 screens, Education) [Refero](https://refero.design/flows/4318) > Happy-path for a short data-analysis lesson where a logged-in user selects a recommended module, progresses through a guided lesson with multiple bar-chart comprehension questions, can report a UI/content problem (attach screenshot + annotate) if needed, and receives feedback, XP, and a streak. Entry: course selection / recommended module. Exit: lesson completion screens (XP, streak). Steps: Course module selection → Start learning splash → Lesson intro slide → Question (bar chart shown) → Question — choose an answer → Check answer (selected state) → Incorrect feedback panel → Open new ticket (report problem) → Drawing instruction modal → Annotate screenshot and complete ticket → Ticket sent confirmation → Next question (250 drinks) → Select answer (correct) → Chained question — combined total → Incorrect answer for combined question → Explanation — correct combination → Mini completion card → XP summary screen → Streak start confirmation → Final completion & CTA **Days: Note with Checklist** (ios, 2 screens, Scheduling Tools, Lifestyle) [Refero](https://refero.design/flows/7804) > Two-card mobile view showing a textual note (yellow card) above a checklist (orange card). Entry: user opens this screen to read the note and review or update checklist items. Exit: user leaves by tapping the close button (X) or navigates away after interacting with the checklist. Assumptions: user is on a mobile device with this screen already open; the note and checklist are read-only except for tapping checkboxes and using the visible controls (+ to add, X to close). Steps: Initial note and checklist → Checklist after item selection **Structured: Daily timeline checklist** (ios, 2 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3258) > Single-day timeline view for a selected date (December 23, 2024). Steps: Day timeline — initial view → Item completed state **Plane Finder: Feature Tour — Delay Insights** (ios, 4 screens, Travel & Booking) [Refero](https://refero.design/flows/10395) > A short, linear onboarding/product-tour that introduces three new features (Delay Insights, anticipating delays, and a unified MyFlight detail screen) followed by a short "what's under the hood" summary and a final call-to-action to enter the app. Steps: Introducing Delay Insights → Anticipate Delays → Unified MyFlight Detail Screen → What's Under the Hood & Finish **Tinder: Safety Basics help** (ios, 4 screens, Dating & Relationships) [Refero](https://refero.design/flows/3455) > Mobile (portrait) 'Safety Center' guide flow. Entry: user opens Safety Center and lands on the Guide tab (top tabs: Guide, Tools, Resources). The screen shows a personalized greeting, a list of safety sections presented as tappable cards (The Basics, quizzes, harassment help). Happy path: user taps the 'The Basics' card → navigates to a detail article with an illustration, brief intro, and scrollable content broken into Online and IRL sections. Exit: user scrolls to read content and then uses the back/close control to return to Safety Center. Assumptions: user is signed in (personal greeting and avatar visible) and on a mobile device. Steps: Safety Center — Guide list → Topic header — The Basics → Article content — Online safety → Article content — IRL safety **222: Curation Profile Completion** (ios, 19 screens, Lifestyle, Dating & Relationships) [Refero](https://refero.design/flows/9324) > Happy-path: Logged-in user starts on their profile dashboard (shows curation profile and an "incomplete" core questions card). The user taps to answer core questions, steps through single-answer screens (education → GPA → preferences), and returns to the dashboard where the profile completion percent updates. From the dashboard the user opens category tiles (background → perspectives → professional → interests) to review and edit attribute chips, add items, or answer question cards. The flow ends when the user has answered or updated enough items and the Curation Profile reads 100%. Assumptions: user is authenticated, has permission to edit their profile, and can navigate by tapping tiles, list rows and dropdowns on a mobile device. Steps: Profile Dashboard (overview) → Core Question — Highest Education → Core Question — GPA Range → Core Question — Preferences (parks) → Profile Dashboard (percent update) → Category Grid — select Background → Background — Detail List → Background — Scrolling / Question Rows → Category Dropdown (identity selector) → Switch to Perspectives Category → Perspectives — Question List → Open Professional Category → Professional — Occupation & Universities → Professional — Questions list → Return to Category Grid (percent update) → Select Interests Tile → Interests — Hobbies (chips & suggestions) → Interests — Followed Sports Teams → Media Tab / Final Review **Foodllama: Daily Nutrition Review** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/11788) > This is a nutrition tracking dashboard for a logged-in mobile user. The user starts on the main daily summary screen, opens a bookmarks overlay to choose a saved meal, then returns to the dashboard where the selected meal is shown in the day’s log. Steps: Dashboard Overview → Bookmarks List → Logged Meal View **ElevenMusic: Music Creation Dashboard** (web, 2 screens, Music Streaming, Entertainment, AI Tool) [Refero](https://refero.design/flows/12569) > A logged-in user lands on a music creation home/dashboard with a left navigation rail, a central prompt-driven track card, and a persistent mini-player at the bottom. The user can review a generated track, adjust or remix it from the main card, or continue an existing project from the "Continue where you left off" section. Entry appears to be the Home area; exit points include publishing, opening Studio, or remixing/continuing a track. The interface assumes the user is authenticated and has at least one existing track/project. Steps: Home Dashboard → Continue Project **Rork: Vocabulary Dashboard** (web, 2 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12149) > Happy path starts on the Vocabulary home dashboard for a Spanish learning app, shown in a mobile device frame. The user is assumed to be logged in and viewing their personal learning summary. The screen provides quick access to core actions such as reviewing due cards, checking daily progress, and navigating to study, decks, or progress sections. The flow shown here is a static dashboard state, with the main exit point being entry into a review session or another bottom navigation destination. Steps: Home Dashboard → Dashboard Focus **Cursor: Dashboard layout scaffold and review** (web, 13 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9937) > Happy path where a user (logged in) uses an agent/conversation to generate a dashboard layout in a repository, watches the agent run, inspects generated code and diffs, and is presented with PR / review controls. Steps: Agent list & new task card → Enter prompt and run → Task started — snapshot logs → Post-start script logs → Generated plan and todos → Code snippets appear → App and CSS changes → Run instructions and commit summary → Open file diff / PR panel → Browse file diffs → Inspect CSS diff → Return to conversation → Conversation idle / ready for new request **Runey: Project Status Update** (web, 3 screens, Business, Productivity, Project Management) [Refero](https://refero.design/flows/11696) > This is a project overview dashboard for a project called “Website Redesign.” The user enters from a project dashboard view, with summary metrics, a task list, activity feed, and calendar. The main interaction shown is changing the project status from an “Active” dropdown to another status such as “Pending.” Assumes the user is already logged in and has permission to edit project settings/status. The flow ends with the project status updated and a confirmation message shown on the dashboard. Steps: Project Dashboard → Status Selection → Status Confirmation **Сontra: Freelancer Dashboard Overview** (web, 2 screens, Job Search) [Refero](https://refero.design/flows/4869) > Logged-in freelancer lands on a consolidated dashboard that surfaces high-value actions and account status. Steps: Dashboard (full nav) → Dashboard (condensed nav) **Manus: AI-driven Analytics Dashboard Builder** (web, 22 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/10457) > Happy-path: A logged-in user on desktop uses Manus (an AI-assisted product-builder) to create, inspect, edit, and publish an analytics dashboard. Entry: user starts on Manus home with a prompt input. The user selects the "Website" intent, chooses the "Dashboard" template, types a detailed brief, and submits. Manus initializes a project scaffold, generates the dashboard design and code, presents a split view (task / generated site preview), allows the user to inspect code and visual styles, and offers a database connector. The user publishes the generated site; a live URL is returned and the published dashboard is viewable. Exit: Manus returns to the initial prompt ready for a new task. Assumptions: user is authenticated, has permission to create projects and publish, and is on a desktop browser. Steps: Start new task → Choose website template → Confirm dashboard selection → Write detailed brief → Submit prompt → Scaffold initialization → Generated preview and task list → Inspect generated code → Checkpoint saved → Detailed preview & interaction → Enter visual edit mode → Open style panel → Edit border settings → View data integrations tab → No database connected → Open publish dialog → Publish complete (live site) → Live site view → Publish settings & summary → Task completed and follow-ups ##### Task: connect bank account open banking **Wealthsimple: Open Chequing Account** (web, 15 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8496) > Happy-path onboarding for opening a new deposit account starting from a logged-in user's dashboard. Steps: Dashboard — Add an account → Account selection → Product details — Chequing → Intro — Tell us about yourself → Form — Legal name & date of birth (empty) → Form — Legal name & DOB (completed) → Form — Phone number (empty) → Form — Phone number (completed) → Form — Residency, tax residency & citizenship (initial) → Validation — Citizenship requirement error → Form — Residency & citizenship (resolved) → Form — Social Insurance Number (empty) → Validation — SIN format error → Form — SIN masked and accepted → Validation — SIN temporary / expiry notice **Fingerprint: Find and open account-related documentation** (web, 4 screens, Business, Development Tools) [Refero](https://refero.design/flows/11143) > Happy path: a user lands on the Fingerprint Docs home screen, opens the site-wide search/Ask AI modal, searches for "account" (or a related term), and selects an "Account Sharing" / "Account Takeover" article. The flow ends with the full documentation article loaded, including code examples and step-by-step instructions. Steps: Docs Home / Discover → Search Modal (empty) → Search Results → Documentation Article **Spoil me: Create wishlist and add first items** (ios, 37 screens, Fundraising) [Refero](https://refero.design/flows/5343) > Happy-path new-user onboarding that creates a Spoil.me wishlist (vanity link), verifies identity via an email magic link, imports product photos from the device, adds item details (price, crowdfunding toggle), and lands the user in their wishlist dashboard. Entry: first-time visitor (not signed in) at the hero screen. Exit: activated wishlist/profile screen showing created wishes. Steps: Hero - Choose vanity link → Typing vanity handle → Vanity handle entered → Email input - prompt → Email validation (invalid) → Email input - valid → Confirmation - check inbox → Resend active state → Open verification email (inbox) → Waiting for verification → Photo picker - select images → Photo selection confirm → Item details - edit → Alternate suggestions - browse images → Confirm item images → Add another wish - title input → In-app notifications prompt → OS-level notification permission → Optional payout connection (Stripe) → Wishlist dashboard / onboarding state ##### Task: contract management **Acctual: Create bill from uploaded invoice** (web, 2 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8845) > Happy path: A user creates a new bill by uploading a supplier invoice PDF. Steps: Start new bill — upload area → Preview and review extracted details **Queue: Subscription Cards List** (ios, 3 screens, Entertainment, News) [Refero](https://refero.design/flows/6239) > A mobile (iPhone) list of subscription cards where each card represents a recurring item (podcast, news subscription, website). Steps: List — default state → Reveal destructive action → List — updated / scrolled view **Subo: Find Subscription** (ios, 3 screens, Finance) [Refero](https://refero.design/flows/8588) > Happy path: a signed-in mobile user opens the Subscriptions screen (list of active subscriptions). From this list they tap the persistent search bar at the bottom, which focuses and raises the keyboard so the user can type a query. If the entered query does not match any subscription, the screen shows an empty-search state that prompts the user to change or clear the search. Steps: Subscriptions List → Search Field Focused → No Results / Empty State **Apple TV: Manage AMC+ subscription** (ios, 3 screens, Entertainment, Video Streaming) [Refero](https://refero.design/flows/11102) > Happy path: a signed-in user viewing an active Apple TV Channel subscription (AMC+) that was started via a free offer. Entry: user opens the subscription details for AMC+. The screen shows the subscription card, renewal date and price after the free offer, and primary actions (See All Subscriptions, Cancel Free Trial). If the user taps to cancel, a confirmation sheet appears. After confirming cancellation, the details screen updates to show an expiring state with an option to re-subscribe. Exit: user either keeps the subscription (closes modal) or completes cancellation and sees the expiring-subscription state. Assumptions: user is logged into their account, viewing subscriptions on a mobile device, and has permission to manage their own subscriptions. Steps: Subscription Details (Active free trial) → Confirm Cancellation Modal → Subscription Expiring (Post-cancellation) **Google Gemini: Report legal content from Gemini chat** (ios, 18 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/3187) > Happy path: A signed-in user starts in a Gemini conversation, opens the message actions menu on an assistant reply, and chooses to report the content. The app links to Google Legal Help. The user selects a legal reason, taps Create request, completes the online ‘Report a legal removal issue’ form (country, personal details, contact email), attaches screenshots/files from the device photo picker, provides a written explanation, confirms the perjury declaration, types a signature (digital), and submits. The flow ends with a submission confirmation screen telling the user they will receive a confirmation email. Assumptions: user is signed in (avatar visible), using a mobile browser/webview on iOS, has the required evidence files in their photo library, and can receive email confirmations. Steps: Gemini conversation - message actions visible → Message options menu → Legal Help landing → Select reason to report → Create request CTA → Form – country selector (tap) → Form – country list (choose) → Form – fill identity fields → Allegedly infringing content section → Photo picker – browse library → Photo picker – confirm selection → Files attached in form → Explain why content is unlawful → Perjury declaration checkbox → Signature and date → Submit request → Submission success → End state / wait for confirmation **Cake Equity: Create seed raise** (web, 18 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/6148) > Guided wizard for creating a new capital raise (example: seed stage). Steps: Raise landing → Disclaimer / Start → Start raise — metadata → Start raise — values filled → Confirm pre-money valuation → Add raise details — initial → Add raise details — filled → Company signers (nomination) → Select first signer → Add company signer — form → Add company signer — filled → Board member roles selection → Role added (alternate) → Role validation error → Primary signer added confirmation → Bank details — entry → Bank details — filled → Raise summary (Seed stage) **Kickstarter: Manage Pledge & Cancel** (ios, 5 screens, Fundraising) [Refero](https://refero.design/flows/4644) > Happy-path for a logged-in backer on a mobile device to manage and cancel an existing pledge. Entry: user opens a project where they are already a backer and taps Manage. Exit: user confirms cancellation and returns to the project overview with an in-context confirmation banner. Steps: Project overview (backer state) → Manage your pledge (details) → Action sheet (management options) → Cancel confirmation → Project overview (cancellation confirmed) **Instagram: In-chat order creation & management** (ios, 13 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/2893) > Seller creates and manages an order directly from a one-to-one chat. Entry: user opens chat with a contact and uses the message composer menu to start an order. The flow opens a bottom-sheet modal where the seller enters a monetary amount using a numeric keypad, optionally adds a description, then taps Create order. The order is posted back into the chat as an “Order created” message with an order number and View order button. Tapping View order opens Order Details where the seller can progress the order through statuses (Mark as paid → Mark as shipped). Each status change updates the order timeline in Order Details and posts status bubbles back into the chat. Steps: Chat — open composer menu → Create order — numeric keypad → Amount entry — incremental update ($1.26) → Amount entry — incremental update ($12.60) → Amount entry — target amount ($126.00) → Add description → Create order — publish to chat → Order Details — open from chat → Mark as paid — in progress → Chat — order paid message → Order Details — mark shipped → Chat — order shipped message → Order Details — final timeline **Сontra: Create and Send Project Proposal** (web, 14 screens, Job Search) [Refero](https://refero.design/flows/4880) > Happy-path: A logged-in freelancer (desktop) with no active projects starts a new paid project proposal from the Projects & invoices area. The user opens a multi-step wizard (Proposal → Contract → Review → Sign), fills client basic information and project name, configures payment method and invoice schedule, adds a project scope, selects the contract template, reviews contract and project summary, then either saves the proposal as a draft or proceeds to sign/send. Steps: Projects - Empty state → Projects - CTA tooltip → Proposal — Basic information (start) → Proposal — Basic information (filled) → Payments details — Billing method → Payments details — Invoice configuration → Payments details — Form completed → Project scope — Add description → Project scope — Completed → Contract — Select template → Contract preview and signer info → Review — Project details summary → Draft Projects list — Draft created → Draft Projects — Context and next actions **Delphi: Expert Legal Consultation** (web, 19 screens, AI Tool) [Refero](https://refero.design/flows/13982) > Happy path begins in the Discover area, where a user searches for advice about arguing a first court case. The system returns matching expert profiles and excerpts. The user opens a conversation with Jason Jones, reviews his contextual response, chooses suggested follow-up questions, attaches a PDF, asks for feedback, records and sends a voice message, and continues the conversation based on the expert’s responses. The flow ends with another assistant response and additional suggested questions. The user is assumed to be signed in, using a desktop browser, and permitted to search experts, start conversations, upload files, and record audio. **Cohere: Document-based conversation review** (web, 3 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3832) > A desktop chat workspace where a user interacts with an AI assistant to extract or summarize information from uploaded documents. Steps: Conversation view with attachments visible → Activate collapse control for left panel → Expanded conversation (left panel collapsed) **Cohere: Business plan prompt with PDF attachments** (web, 3 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3830) > Happy-path conversation in a desktop chat interface. Steps: Open conversation with attachments visible → Hover response action to reveal tooltip → Response actions and attachments available **Cake Equity: Grant existing offer with contract upload** (web, 4 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/6144) > Happy-path sequence for attaching an externally signed contract and granting an existing share offer from a plan detail view. Steps: Open Grant Existing Offer modal → Open file browser to upload contract → Upload file and review validation → Confirm details and grant **Oppie: Manage saved words** (ios, 3 screens, Education) [Refero](https://refero.design/flows/6327) > Happy path: A signed-in user views their personal vocabulary list titled "My words" (Greek term with English translation per card). From this screen the user opens an item's overflow menu (three vertical dots) and selects Delete. The item is removed immediately and the list updates in place. Steps: My words list → Contextual action shown → List updated after delete **Kin: Bookmark assistant response** (ios, 4 screens, AI Tool) [Refero](https://refero.design/flows/5441) > Happy path: A user reads an assistant response (text + image) in a conversation view, long-presses the text to open a contextual action menu, selects 'Bookmark', sees a temporary confirmation (toast) and an inline bookmark marker added to the message, then continues the conversation using the message input. Entry: user is viewing a previously generated assistant message within an open conversation. Exit: the assistant message is visibly marked as bookmarked and the user is returned to the conversation input area so they can continue interacting. Steps: Assistant response displayed → Contextual action menu → Save confirmation → Return to conversation **ElevenMusic: Generated Tracks Management** (ios, 4 screens, AI Tool, Entertainment, Music Streaming) [Refero](https://refero.design/flows/12467) > A mobile music-creation screen showing the user’s generated tracks under the Create section, with Drafts and Published tabs. The user can browse draft tracks, open each track’s overflow menu for actions, and trigger deletion. Entry is the Create tab; exit can be back to the draft list after canceling, or removal of the selected track after confirming delete. Steps: Drafts List → Track Actions Menu → Delete Confirmation → Updated Drafts List **Airbnb: Conversation - message compose, actions, edit** (ios, 19 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6431) > User navigates from the Messages list into a single conversation tied to a trip/listing, reads prior messages (including an expired booking invite), composes and sends new messages, interacts with individual message actions (react, copy, reply, report), and edits a previously-sent message. Steps: Messages list → Conversation thread (scrolled) → Conversation top + booking call-to-action → Open Details (trip & conversation actions) → Conversation actions list → Composer with keyboard open → Compose message draft → Send message → Host replies in thread → Long-press host message (actions) → Apply reaction to message → Reply to specific message (start) → Reply composer (typed) → Send reply (quoted) → Long-press user's own message (menu) → Select Edit (warning modal) → Composer pre-filled for editing → Send edited message → Inspect final thread state **Structured: Task action sheet** (ios, 2 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3260) > Mobile day-timeline view (December 23, 2024) showing a vertical timeline of scheduled items. The user taps a task node (Eat lunch) to open a modal bottom sheet that exposes management actions for that task: Delete, Duplicate, Complete/Incomplete toggle, and Edit Task. Steps: Open actions for a completed task → Open actions for an incomplete task **PayPal: Tracked packages management** (ios, 4 screens, Finance) [Refero](https://refero.design/flows/7256) > A single-screen mobile package-tracking experience for managing tracked parcels. Entry: user opens the "Package tracking" section (assumed from a main menu or shortcut). Happy path: user scans the In progress list, swipes a tracked item to reveal inline actions (Remove, Delivered), taps Remove, confirms in a bottom-sheet modal, and receives a brief toast confirming removal. The list refreshes to exclude the removed package. Exit: user stays on the tracking list and can add a new tracking number via the bottom CTA. Assumptions: user is signed in (tracked packages are already present), the device is a phone (iOS visual chrome visible), user has permission to receive local toast/notifications. No backend assumptions beyond the UI text shown (the modal explicitly states updates stop). Steps: In-progress packages list → Reveal item actions → Confirm removal (bottom-sheet) → Removal acknowledgement and list update **Cake Equity: Create share plan** (web, 6 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/6141) > Happy path: A logged-in company admin (company context selected) creates a new share/ESS plan from the Ownership > Transactions log or Incentive plans area and configures basic settings and document templates, then lands on the new plan detail screen where offers and contracts can be managed. Steps: Transactions log (ESSs tab) → Share plans introduction → Add a Share plan - basic info → Select plan documents - choose source → Select plan documents - template details → Refero Share plan - plan details **Cake Equity: Create Option Pool** (web, 6 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/6139) > Happy-path: A logged-in company admin/HR user creates a new option pool (option plan) from an empty transactions log, configures basic pool settings, selects plan documents, and lands on the newly created option-pool detail screen where offers and contract templates are shown. Steps: Transactions log (empty) → Option pools overview → Add an option plan (configuration) → Select plan documents → Option pool detail (initial) → Option pool detail (templates available) **Kinhive: Save Template to Project** (web, 7 screens, Communication Tools) [Refero](https://refero.design/flows/6864) > Happy path for saving a public template into a team project’s Agreements list. Steps: Agreements (Empty State) → Templates Library → Template Detail / Preview → Save Copy Modal (Open) → Save Copy Modal (Confirm Selection) → Copy In Progress → Agreements List (Updated) **Nike: Job search & job detail** (web, 6 screens, Marketplace, Sports) [Refero](https://refero.design/flows/2016) > Happy-path: A desktop user arrives at the employer's jobs listing and uses the left-hand filters or the keyword search to narrow hundreds of open positions, then opens a job to view details and Apply. Steps: Job Listings — All Open Positions → Apply Brand Filter → Filter by Position Type → Zero Results State → Refine to Career Area (Technology) → Job Detail — Sr. Software Engineer - SAP **Homerun: Create and publish a job posting** (web, 24 screens, Productivity, Job Search, Communication Tools) [Refero](https://refero.design/flows/1264) > Happy path: A logged-in recruiter starts on the company dashboard and creates a new job. They open the Create job modal, enter a title, select job type, add or create a department and location (including creating a new location with address and country), then continue to a visual job editor to style and preview the job post. From the editor they configure the application form (contact fields and custom questions), confirm unsaved changes, view the empty candidate pipeline, publish the job (optionally private), and open the share dialog to copy the job link and enable job board distribution. Steps: Dashboard / Create job entry → Initial job modal - blank → Enter job title and open job type → Department field (search) → Add new department → Location field (create option) → Select suggested location → Open Create new location modal → Country selector (open list) → Fill address and search country → Open visual job editor (styling controls) → Editor preview (draft view) → Full job page preview (long view) → Apply form configuration entry → Contact details modal (fields list) → Toggle additional fields and require answers → Unsaved changes confirmation → Candidate pipeline (empty state) → Publish job (publish menu) → Share job dialog **Lovable: Workspace People Management** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12299) > This is a workspace member management section for a signed-in user who can view and manage people in a workspace. The entry point is the People area in the left navigation. The exit point is returning to the people list after generating or managing an invite link. Steps: People List → Invite Link Setup → Invite Link Management **Lovable: Workspace invite link management** (web, 2 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12300) > Happy path begins on the People section of a workspace settings area, with a modal open for generating and managing a workspace invite link. The user can choose a role, copy the link, and optionally regenerate or delete it. The flow assumes the user is already in the workspace, has permission to manage members, and is using a desktop interface. It ends after the invite link is successfully copied, with confirmation shown in the modal. Steps: Invite Link Modal → Copy Confirmation ##### Task: course enrolment university **WhatsApp: Add Community Member** (ios, 5 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12678) > Happy path in a community management area. The user starts on the Community info screen as the community owner, opens Add members, selects one contact, confirms the add, and returns to Community info where the member count updates and a success message appears. **Wabi: Invite Members** (ios, 5 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12046) > This is a mobile collaboration/invite screen reached from a shared workspace or group context. The user can either share the space, invite others directly, search for users by username, or send an invite link. The happy path starts with the user already inside the invite interface and ends when matching users are found and the user can tap Invite next to a person. Assumption: the user is logged in and has permission to invite others. Steps: Invite Overview → Search Active → Searching → No Match Result → Matched Users **ALO: Prenatal Program Detail** (web, 2 screens, Fitness, Lifestyle) [Refero](https://refero.design/flows/12377) > A logged-in user opens a wellness program detail screen for a prenatal yoga series. The entry point is the program landing page with a hero image, program title, instructor, and tabs for Overview and Classes. The user can review the description, preview the series video, inspect instructor information, browse the full list of classes, view similar programs, and add a comment. In the second state, the user is shown as in-progress with completion tracking and a Stop Program control, indicating they have already started the series. The flow ends when the user leaves the program page, starts or continues a class from the series, schedules a class, or posts a comment. Assumption: the user is signed in and has access to the program content on desktop web. Steps: Program Overview → Program In Progress **District: Movie Seat Selection** (ios, 5 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12829) > A movie seat-selection screen for the film Main Vaapas Aaunga at Maison PVR Jio World Drive, BKC, Mumbai. The user enters with a specific showtime selected and reviews the seat map by row and pricing category. They select seats, optionally unlock and apply an offer from the booking banner, and then proceed to the next checkout step. Steps: Seat Map Review → Seat Selection Update → Offer Unlock State → Offer Applied Confirmation → Ready to Proceed **Binge: Entertainment Onboarding** (ios, 5 screens, Entertainment) [Refero](https://refero.design/flows/12954) > Happy-path onboarding for a media tracking service. Steps: Introductory Tracking → Rating Setup → Library Overview → Resume Watching → History Import **Imprint: New user onboarding and course starter** (ios, 54 screens, Education) [Refero](https://refero.design/flows/5634) > Happy-path onboarding and first-lesson entry for a mobile learning product. Entry: new user opens the app (not signed in). The user is shown a branded welcome screen, guided through 2–3 visual onboarding slides, then a short personalization questionnaire (referral source → demographics → topics). The flow includes an optional email opt-in, a push-permission prompt, and a course selection step. Exit: user selects a course and starts their first interactive lesson. Assumptions: user is on a modern smartphone, not signed in, able to interact with tappable cards and multi-select controls; progress is linear and reflected with a top progress bar. Steps: Welcome / Sign-up entry → Onboarding slide 1 — Concept intro → Onboarding slide 2 — Benefits → Personalization intro → Question — Referral source (part 1) → Question — Referral source (scroll) → Referral selection confirmed → Question — Gender / identity → Question — Age bracket → Question — Topics of interest (multi-select) → Email opt-in modal → Promotional slide — course count → Permission prompt — Push notifications → Course chooser intro → Course selection confirmation → Lesson tutorial overlay → Lesson viewer — card interface → Lesson content sequence → Interactive card / content detail → Lesson start / First content **Imprint: Course Search & Results** (ios, 4 screens, Education) [Refero](https://refero.design/flows/5646) > Happy-path: a signed-in user on the Courses tab wants to find a specific course by keyword. From the Explore screen they open the search field, focus it (keyboard appears), type a query, and view filtered course cards. Steps: Explore / Search Entry → Search View (idle) → Typing / Keyboard Active → Search Results **Tinder: Add school to profile** (ios, 3 screens, Dating & Relationships) [Refero](https://refero.design/flows/3417) > Summary (happy path): User is prompted to add a school (past or current) to their profile during an onboarding or profile-edit flow on a mobile device. The user taps the school input which opens a centered modal and shows the iOS keyboard. As the user types, an autocomplete list of matching schools appears (with country codes). The user selects a school from the list; the selection populates the input (modal is dismissed) and the primary screen updates so the user can continue by tapping Next. Steps: Education prompt → Modal input with keyboard → Autocomplete results list **Tinder: Add school to profile** (ios, 3 screens, Dating & Relationships) [Refero](https://refero.design/flows/3416) > Happy path: A user is prompted to add a past or current school to their profile during onboarding. From the prompt screen the user taps the school text field, which opens a focused modal with the on-screen keyboard. As the user types, a single-line entry shows matches or an "Add <text>" suggestion. The user can accept the suggestion (or use keyboard "Done") to provide a school value. Steps: School prompt → Input modal with keyboard → Typed entry and suggestion **Imprint: Course Search & Empty State** (ios, 4 screens, Education) [Refero](https://refero.design/flows/5645) > Summary: The user navigates to the Courses area and opens the Search experience to find a course. Steps: Courses — Search landing → Search overlay — course grid → Search typing — keyboard visible → Search results — empty state **Remy: Course preparation walkthrough** (ios, 8 screens, Health & Wellness, AI Tool) [Refero](https://refero.design/flows/9766) > Happy path: A logged-in user visits the app home and sees a card indicating "Day 1 is ready." Tapping the card opens a full-screen, scrollable three-step preparatory lesson. Each step is content-heavy and ends with a single primary acknowledgement CTA (microcopy varies per step). Tapping the CTA advances to the next step; a small progress indicator (dots) and a bottom progress bar show progression. After the third step the lesson is marked read and the UI returns to the course home where the next day is locked and a countdown timer is displayed. Steps: Course home — Day 1 card → Lesson — Step 1 (scrollable text) → Lesson — Step 2 entry → Lesson — Step 2 (full content) → Lesson — Step 3 (final content) → Lesson — Conclude and mark as read → Course home — Next day locked → Course home — resume and repeat **Moonly: Runes course reading and completion** (ios, 5 screens, Lifestyle) [Refero](https://refero.design/flows/9907) > A mobile learning flow inside the Learning tab for a "Runes" course. Steps: Course timeline (Runes) → Article header → Article content (scroll) → Completion confirmation modal → Updated course timeline (progress reflected) **komoot: Invite people** (ios, 3 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8355) > User is on the mobile "Invite people" screen for a group activity. Entry: user navigates to the Invite people area (already signed in, viewing a specific group/activity). Happy path: user can invite others by (a) showing/scanning the QR code, (b) copying/sharing the invite link, or (c) searching for contacts by name, email or phone and tapping an Invite control. The UI updates a participant count/selection as people are added. Exit: user leaves this screen after adding one or more invitees (participant count updated) or after sharing the QR/link. Assumptions: user is authenticated, on a phone, and has a contact list accessible to the interface. Steps: Invite people — main screen → Search focused — keyboard visible → Search input — numeric entry **ChatGPT: Study lesson activation and use** (ios, 8 screens, AI Tool) [Refero](https://refero.design/flows/7026) > Happy-path: A user starts on the main chat screen, opens composer actions to choose a pre-built "Study" lesson, activates the lesson which adds a persistent "Study" context tag to the composer, types a study request, and receives a multi-step instructional response with inline practice problems. Steps: Main chat with quick prompts → Open composer actions → Preview lesson card → Activate lesson (tag added) → Compose study request → Assistant initial reply and quick question → Lesson content with formula → Practice problem presented **Brilliant: Personalized onboarding and subscription** (ios, 48 screens, Education) [Refero](https://refero.design/flows/4316) > Happy-path new-user onboarding for a learning app (Brilliant). Entry: first-launch splash. The user is guided through an animated brand intro, a short personality-driven onboarding (mascot 'Koji'), a multi-step questionnaire to set learning goals and subject preferences, and a learning-path recommendation screen. At recommendation, the user can proceed to a subscription upsell that uses the App Store purchase sheet; after purchase confirmation the app shows recommended learning-path details and starts the first lesson. Assumptions: user is on an iOS device, not logged in (or anonymous), and has network connectivity for App Store flow. Steps: Brand splash (glitch animation) → Brand wordmark → Intro: Learn by doing → Mascot intro (Koji) → Build learning path prompt → Question: What's your top goal? → Playful response + selection → Follow-up question: Clarify goal → Topic selection list → Topic confirm + Continue → Loading: Finding recommendations → Recommendations list → Path detail: Data Analysis overview → Subscription upsell / trial info → App Store purchase sheet (confirm) → Purchase confirmed (App Store done) → In-app success modal → Unlocked recommendations + path cards → Start lesson: Reading Bar Charts → Onboarding complete / keep learning **DailyArt Magazine: Enrollment with Mandatory Sign-In** (web, 4 screens, Education, Magazines & Newspapers, Marketplace) [Refero](https://refero.design/flows/3506) > Happy path for a desktop user who is not signed in: a user browsing courses finds a course they want to enroll in, clicks the course-level Enroll CTA, is shown an authentication modal (email/password or social login), completes registration (or signs in), and the site updates to an authenticated state showing the course details with the Enroll CTA available. Steps: Courses Listing / Course Card → Authentication Modal (Sign In / Register) → Registration Success / Sign-In Prompt → Course Detail (Authenticated State) **DailyArt Magazine: Course discovery and enroll entry** (web, 3 screens, Education, Magazines & Newspapers, Marketplace) [Refero](https://refero.design/flows/3510) > A course discovery / entry landing that highlights a featured art history course in a large hero with a clear enroll CTA, followed by a grid of additional course cards, a CTA to view all courses, newsletter signup and site metrics. Steps: Courses landing (desktop) → Condensed hero / tablet breakpoint → Localized (Chinese) landing **Preply: Course selection and confirmation** (web, 4 screens, Education) [Refero](https://refero.design/flows/9216) > Happy path: A logged-in user (desktop) in the Learn section chooses a course from a list of predefined course cards, answers a short question about how they use English at work (proficiency level), reviews a confirmation summary, and confirms the course. After confirmation the system shows the pre-populated list of topics for that course and displays a modal announcing the course is ready with an option to take a guided tour. Entry: Learn > Choose your course. Exit: Course confirmed and topics listed (with an optional tour CTA). Assumptions: user is signed in, has access to Learn, using a desktop browser, and is selecting a course to plan future lessons and homework. Steps: Choose your course → Select proficiency → Confirm your course and review topics → Course ready modal **DailyArt Magazine: Course enrollment authentication** (web, 6 screens, Education, Magazines & Newspapers, Marketplace) [Refero](https://refero.design/flows/3505) > Happy path: a logged-out visitor views a course detail page and tries to enroll. The site requires authentication, so an overlay modal appears offering login or registration (including Facebook social sign-up). The user switches to the registration form, fills fields, sees inline validation for mismatched passwords, corrects them, submits successfully, and is then shown a sign-in prompt. Steps: Course detail with Enroll CTA → Login prompt modal → Blank registration form → Registration attempt (fields filled) → Validation error shown → Post-registration sign-in prompt **DailyArt Magazine: Course Discovery via Search** (web, 4 screens, Education, Magazines & Newspapers, Marketplace) [Refero](https://refero.design/flows/3513) > Entry: user arrives on the Courses landing (desktop). Happy path: user scans featured courses, opens the search control, enters a query (“Renaissance”) and lands on the search results where a course is expanded to show its lesson list and enroll CTAs. Assumptions: desktop viewport, session state not required for searching, UI shows global header with search, account and cart controls. Exit: search results screen where the user can inspect lessons or press an Enroll button to continue into the enrollment/checkout flow (not shown). Steps: Landing — Courses Overview → Open Search Overlay → Enter Search Query → Results — Course List & Expanded Course **DailyArt Magazine: Course search via header** (web, 4 screens, Education, Magazines & Newspapers, Marketplace) [Refero](https://refero.design/flows/3512) > Happy path: A visitor browsing the courses landing area uses the header search control to look for a course. The header launches a full-width search overlay where the user types a query and submits it. The system performs the search and navigates to a results page that renders the submitted query and either matches or feedback (in this case, a "No results" message). Steps: Courses landing → Search overlay (empty) → Search input filled → Results page — no matches **DailyArt Magazine: Course Catalog Browsing** (web, 2 screens, Education, Magazines & Newspapers, Marketplace) [Refero](https://refero.design/flows/3511) > Desktop catalog view for browsing and enrolling in art courses. Steps: Course Catalog — Default View → Apply Filters — Sort by Popularity, Topic Modern Art **Teachable: School setup and course creation** (web, 3 screens, Business) [Refero](https://refero.design/flows/9547) > Happy path summary: An authenticated school owner/admin views an empty Students list, is prompted (or redirected) to authorize access to their account, and then lands on the course creation onboarding screen where they can enter a course title/description and use an AI tool to generate an outline. Steps: Students — Empty state → Authorize account access → Create course outline onboarding **Teachable: Add and enroll student** (web, 5 screens, Business) [Refero](https://refero.design/flows/9613) > Happy path: An administrator (logged in) opens the Students section, clicks Add Students, fills in one or more student records manually, optionally enrolls the student into a product and confirms email consent, selects how the student paid, and imports. The system shows an importing/confirmation message and then updates the Students list to include the newly added student(s). Steps: Students list overview → Add Students modal (manual mode) → Populate fields and enable enrollment → Import feedback → Updated students list **DailyArt Magazine: Course Enrollment & Cart Review** (web, 2 screens, Education, Magazines & Newspapers, Marketplace) [Refero](https://refero.design/flows/3509) > User is on a course detail screen where they can review course metadata (title, chapters, duration, rating) and choose to enroll. The primary call-to-action is an "Enroll" button showing the price. When the user triggers enrollment, the site adds the course to a right-side cart drawer and displays a green confirmation banner. From the cart drawer the user can review the line item, apply a coupon, and proceed to checkout. Steps: Course detail — enroll CTA → Cart drawer — review and proceed **DailyArt Magazine: Course checkout and confirmation** (web, 4 screens, Education, Magazines & Newspapers, Marketplace) [Refero](https://refero.design/flows/3516) > Happy path: A user adds a course to their cart, opens the cart slide-out, and proceeds to the checkout form. On the checkout screen the user provides billing details (required fields shown) and clicks to proceed. Because the selected course is free (Total: $0.00) the flow skips payment collection and completes with an order confirmation page that shows an immediate "Start course" action and recommended courses. Steps: Cart slide-out → Checkout — Billing (empty) → Checkout — Billing (completed) → Order status and access **Teal: Add Education Entry** (web, 4 screens, Productivity, Job Search) [Refero](https://refero.design/flows/2713) > Happy path: A signed-in user on a desktop resume editor adds a new education record. Steps: Resume Preview — Education Empty → Education Form — Empty → Education Form — Filled → Education Saved — List and Preview Updated **Brilliant: Personalized onboarding and trial activation** (web, 36 screens, Education) [Refero](https://refero.design/flows/4345) > Happy-path new-user onboarding that builds a personalized learning path and activates a trial subscription. Entry: user lands on marketing homepage and begins a guided setup. Flow collects user goals, focus, and topic preferences, asks for subject comfort level, then presents a trial checkout and confirmation screens, a short survey, and finally shows recommended learning paths and the user dashboard. Assumptions: user has created an account or is signed in during onboarding; desktop/larger viewport layout. Exit: user sees the chosen learning path on their home/dashboard with access to the first course. Steps: Marketing homepage → Onboarding intro → Top goal selection (options list) → Top goal selection (choice made) → Learning focus selection (options) → Learning focus selection (confirmed) → Reassurance screen → Topic selection (subject list) → Topic selection (choice made) → Subject comfort-level (math example) → Free trial / checkout screen → Premium welcome confirmation (email view) → App promotion / QR download → Acquisition survey (how did you hear?) → Acquisition survey (selection confirmed) → Processing / recommendations loading → Recommended learning paths → Learning-path overview (Science example) → Start-learning confirmation → User dashboard / Jump back in **Peerlist: Add Education to Profile** (web, 4 screens, Job Search, Business, Social Networking) [Refero](https://refero.design/flows/1088) > Happy path: A signed-in user on desktop opens their profile and navigates to the Resume tab, clicks Add Education, fills the education form (school, degree, major, start/end years) and saves. The UI returns to the Resume view with the new education listed and the profile completion meter updated. Steps: Resume Overview → Add Education Form (empty) → Add Education Form (filled) → Resume Updated **Teachable: Create and publish course** (web, 34 screens, Business) [Refero](https://refero.design/flows/9573) > Happy-path: Instructor creates a new course from the Courses list, steps through a 5-step guided wizard (title/description, thumbnail, pricing, pricing details, outline method), then builds content (adds resources, video embed, quiz generated by AI), reviews the curriculum in the Setup guide, publishes the course and receives a launch confirmation with the sales link. Steps: Courses list (start) → Wizard — Title & Description → Wizard — Continue enabled → Wizard — Add thumbnail → Upload modal — select file → Thumbnail preview → Wizard — Pricing options → Pricing detail form (blank) → Pricing detail form (filled) → Wizard — Outline method → Quiz block editor — empty → Quiz generation in progress → Quiz generated — review → Curriculum — lesson with quiz and resources → Setup guide — course overview → Publish confirmation modal → Launch success modal → Post-launch state — Setup guide (published) → Setup guide final state → Courses list (end) **Uber Eats: Join Uber One membership** (web, 4 screens, Food & Drink, Food Delivery) [Refero](https://refero.design/flows/430) > Happy-path: A user on the Uber Eats 'Uber One' landing section reviews benefits and taps the Join CTA. A centered modal opens to present membership tiers (monthly, annual) and a student discounted option when applicable. The user selects a plan, is prompted to add a payment method, and then starts the free trial / joins the membership. Steps: Uber One landing → Student offer modal → Pricing and billing details → Add payment method (entry) **Preply: Course onboarding tour** (web, 3 screens, Education) [Refero](https://refero.design/flows/9217) > Happy path: A logged-in learner/teacher visits the Learn > Course area after a course has been created. A modal announces the course is ready and offers a guided tour. If the user accepts, the tour highlights two features in sequence: how to change the order of lesson topics (drag handles) and how to find/add topics from the library. The flow ends when the user completes the tour (or dismisses it). Steps: Course ready modal → Change order tooltip → Find more lesson topics **Duolingo: Remove Language Course** (web, 3 screens, Education) [Refero](https://refero.design/flows/11830) > Happy path for removing a language course from a logged-in Duolingo account. Steps: Courses List → Removal Confirmation → Updated Courses State **Handshake: Event registration and confirmation** (web, 2 screens, Social Networking, Job Search) [Refero](https://refero.design/flows/8858) > Happy path: a logged-in student on desktop views an event detail for a virtual Microsoft career talk, uses the right-hand registration card to register, and receives an immediate confirmation state in that same right rail. Steps: Event Details (pre-registration) → Registration Confirmation ##### Task: credit score report **Dropbox: Create Activity Report** (web, 7 screens, Business, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/13062) > From the Admin console’s Members section, the user opens a row actions menu for a deleted member, starts creating an activity report, selects a date range in a calendar picker, and submits the report request. Steps: Members List → Row Actions Menu → Report Date Range → Calendar Selection → Date Range Updated → Submit Report → Request Confirmation **GoFundMe: Report Content Submission** (web, 4 screens, Fundraising) [Refero](https://refero.design/flows/11736) > Happy path for reporting content from a public profile. The user starts on a profile/feed-style page, opens the report modal, selects a reason from a dropdown, and submits the report. The flow assumes the user is already logged in and has access to the report action. It ends with a confirmation modal indicating the report was received and will be reviewed by support. Steps: Profile Overview → Select Report Reason → Submit Report → Report Confirmation **Curater: Report AI-Generated Content** (web, 4 screens, Inspiration Board) [Refero](https://refero.design/flows/13885) > The user starts on a pin detail view with an author contribution and selects the visible Report action. A modal asks them to choose an issue type, and the only available option is AI-Generated. After selecting it, the interface presents a confirmation step explaining that the vote will count toward flagging the content and that multiple reports will be reviewed by moderators. The user submits the report and returns to the pin detail view, where a success toast confirms completion. The flow assumes the user can view the pin and access its reporting control; no additional authentication or permission state is visible. Steps: Pin Detail View → Report Type Selection → Report Confirmation → Submission Confirmation ##### Task: dispute transaction chargeback **WhatsApp: Report Business Chat** (web, 5 screens, Social Networking, Communication Tools) [Refero](https://refero.design/flows/13168) > Happy path on WhatsApp Desktop/Web for reporting a business chat. The user is already logged in and viewing an existing conversation with District By Zomato. The flow starts from the chat header overflow menu, opens the report dialog, optionally includes blocking the business, submits the report, and ends with a confirmation toast while returning to the chat view. ##### Task: doctor video consultation **District: Restaurant Question Assistant** (ios, 5 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12847) > Happy-path mobile assistant experience for asking a restaurant-related question and continuing with a follow-up. Entry is a restaurant detail or assistant overlay with an active query context (“Asking about Hotel Rishi”). The user asks a question by voice, the system transcribes and processes it, shows an answer with suggested next actions, and keeps a persistent follow-up input available. The flow ends when the user has a ready-to-send follow-up request, such as asking for nearby breakfast options. **Ada: Condition Search and Results** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/4125) > User opens the Condition Library (on a mobile device, likely signed-in since profile-specific recommendations appear) to find information about health conditions. The happy path: user scans the curated lists (Most read / Mentioned in your latest assessments), taps the search field, types a query, and either selects a result or sees an empty-state message if there are no matches. Steps: Library listing → Search focused (keyboard visible) → Search submitted — no matches **Parallel: Research Chat with Sourced Answers** (web, 8 screens, AI Tool) [Refero](https://refero.design/flows/9073) > A desktop Chat API research interface where a logged-in user composes web-research questions, receives summarized answers with source links, and can change the chat model to prioritize speed or factual citations. Steps: Chat home / Prompt templates → Submit initial question → Compose follow-up → Follow-up visible while waiting → Open model selector → Submit a data-focused follow-up with new mode → Receive detailed, sourced results → Consolidated summary and source list ##### Task: document editor collaboration **Miro: AI-Generated Retrospective Summary Creation** (web, 6 screens, Productivity, Business, Collaboration Tools) [Refero](https://refero.design/flows/143) > This flow demonstrates the user leveraging an AI-powered document generation feature within a collaborative whiteboard environment to create a retrospective summary. The user starts by selecting or highlighting relevant content on the board, then inputs a prompt or selects a document type for AI generation, triggers the generation process, and finally reviews the generated retrospective summary content within the board. The flow assumes the user is authenticated and working within a project board that contains relevant elements such as ideas and solutions. Steps: Access AI document creation panel → Input or select prompt for retrospective summary → Trigger AI document generation → View initial generated content → Observe continued AI document completion → Review completed retrospective summary ##### Task: email inbox **Fernand: Support tool sign in** (web, 4 screens, Business, Communication Tools) [Refero](https://refero.design/flows/11909) > Happy path for a support agent or founder starting from the public marketing site, moving through sign-in, and landing in the product’s main support workspace. Steps: Marketing homepage → Login screen → Login submission → Support dashboard **PayPal: Email linking for package tracking** (ios, 9 screens, Finance) [Refero](https://refero.design/flows/7253) > Happy path: A logged-in user in the Activity area discovers the empty package-tracking state and taps the 'Get live package tracking' CTA. The app offers email linkage options (Gmail or Outlook). The user chooses Gmail which opens a Google OAuth sign-in/consent flow in a webview. After signing in and accepting permissions, the app shows a success confirmation and returns to the package tracking screen where the system checks the user's inbox for package-related emails. Entry: Activity tab (logged-in user). Exit: Package tracking view with background check in progress and option to enter a tracking number manually. Assumptions: user is already authenticated in the main app; they have a Google account available and can complete OAuth in the embedded browser/webview. Steps: Activity — empty state → Link email provider selection → Google sign-in (email entry) → Google email confirmed → Google password entry → Google consent and permission review → Linking in progress → Link success confirmation → Package tracking — scanning state ##### Task: expense management receipts **Cashcaded: Receipt Sharing Preparation** (ios, 2 screens, Finance) [Refero](https://refero.design/flows/14136) > The user opens a Share sheet for a purchase record and can copy the displayed social-media tag before sharing the receipt image. The flow begins on a mobile device with a share modal already open over a purchase detail screen and ends with confirmation that the tag was copied. The user appears to be viewing an existing receipt or transaction; no login, permissions, or account state are visible. The final Share button remains available but is not shown as selected. Steps: Share Preview → Copy Confirmation **Runey: Invoice Sharing** (web, 3 screens, Business, Productivity, Project Management) [Refero](https://refero.design/flows/11705) > A logged-in user views an invoice, opens the Share action, sends the invoice to an email recipient from a share modal, and then returns to the invoice view with a success toast confirming delivery. Steps: Invoice Detail → Share Invoice → Sent Confirmation **Loowner: Add new item with photo and receipt** (ios, 13 screens, Marketplace) [Refero](https://refero.design/flows/6597) > Happy-path for adding a new owned item from an empty-states entry point. User starts on the "All your stuff" empty screen and taps the primary CTA to add an item. They pick the purchase date, enter the cost using a large numeric keypad, add optional photo(s)/receipt (invoking camera/gallery and permission prompt if needed), enter a name, then save. After save the app shows a success confirmation and returns to home where the new item appears in Quick access and in the item detail view. Steps: Empty state / Entry → New Item — Choose purchase date → New Item — Enter cost (empty) → New Item — Enter cost (value entered) → New Item — Name & images screen (keyboard open) → New Item — Confirmed name preview → Add photo — camera permission → Add photo — choose source → Image attached preview → Save item → Home updated — Quick access → Open item detail → Item detail — edit / add more **Understory: Manage categories** (web, 4 screens, Business, Scheduling Tools) [Refero](https://refero.design/flows/12710) > Logged-in desktop admin interface on the Experiences section. The user starts on an empty Manage categories screen, opens a modal to create a category, enters a category name, and saves it. After saving, the modal closes and the new category appears in the list with an experience count and overflow actions. Steps: Empty Categories State → New Category Modal → Category Name Entered → Categories List ##### Task: eye test vision **Duolingo: Practice Word Question** (web, 8 screens, Education) [Refero](https://refero.design/flows/11827) > Happy-path practice sequence in the Duolingo English Test practice area. The user starts on the Practice dashboard, launches a practice test, waits through a loading state, answers a timed word-recognition question, receives either incorrect or correct feedback, advances through the question, and returns to the Practice dashboard with progress updated. Steps: Practice Dashboard → Loading State → Word Decision Question → Incorrect Feedback → Next Question → Correct Feedback → Practice Complete → Practice Dashboard Return **Fingerprint: Inspect Event JSON Payload** (web, 3 screens, Business, Development Tools) [Refero](https://refero.design/flows/11153) > Happy path: A logged-in user (desktop) viewing an event record opens a modal that displays the raw JSON payload. The JSON viewer supports switching between two versions (v3/v4), searching within the payload, showing result counts, and navigating matches. The user can copy/download the payload or close the modal. Steps: Open JSON modal → Search for specific ID (no results) → Search for boolean value (matches found) **Rork: Model Catalog Search** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12165) > A logged-in user is on the AI Cloud section of a project dashboard and views a catalog of available models with pricing, context, and capability indicators. The user can browse by category tabs or search the model list. In the happy path, the user types a query that filters the list and then clears or changes the query to view matching models again. The flow starts on the model catalog screen and ends with the filtered results displayed or with the full list restored after search refinement. Steps: Model Catalog → No Results State → Filtered Results ##### Task: fleet management **Omio: Vehicle selection & verification** (ios, 7 screens, Travel & Booking) [Refero](https://refero.design/flows/5233) > Happy path: User is on a ferry booking summary (Helsinki → Tallinn) and chooses to add a vehicle to the ticket before continuing to passenger details. Steps: Booking summary (Add vehicle available) → Vehicle type list → Vehicle selection shows extra fields → Enable rental and license input → Submit Add vehicle → Verification success → Post-add options and totals **FocusFlight: Aircraft route tracker** (ios, 2 screens, Productivity) [Refero](https://refero.design/flows/8604) > Two-screen sequence shows a map-based route tracker centered on an aircraft icon and its flight path. Entry: user is viewing a route-tracking map on a mobile device (device location enabled and a route already loaded). Happy path: user inspects current position and nearby landmarks on a zoomed-in map (screen 0), then zooms or toggles to a zoomed-out regional overview that shows the remaining route (screen 1). Exit: user closes the map or returns to another app screen. Assumptions: user is on a mobile device, location services active, map has an active route/track loaded and live progress data (time and distance remaining). Steps: Local position and detail → Regional route overview **Trip: Flight Status Lookup** (web, 3 screens, Travel & Booking) [Refero](https://refero.design/flows/12099) > Happy path for checking a flight’s status on Trip.com. The user enters the Flight Status section, searches by flight number, and views the returned live status details. Steps: Enter Flight Search → Review Search Input → View Flight Status Result **District: Favourite Stores Management** (ios, 3 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12851) > A mobile favourites screen showing saved places on a map with category chips and a bottom sheet list of items. The user starts in a favourites collection filtered to Stores, sees one listed store with map location and actions, opens the item overflow menu, then removes the store from the list. The final state shows the Stores collection is empty and offers an Explore stores call to action. **Plane Finder: Filter and Inspect Flight Categories** (ios, 5 screens, Travel & Booking) [Refero](https://refero.design/flows/10396) > User is on a map-based live flight viewer centered initially over New York. Top-level controls include a search bar, a scan button, and a horizontal set of filter “chips” (All Traffic, Restricted, Military, Cargo, etc.). The bottom dock provides quick actions (MyFlights, Filters, Explore, Alerts, Settings). The observable happy path: user toggles flight-type filters via the chip row; the map updates to show only aircraft matching the selected category and the active chip shows a blue pill with an X (deselect). The map extent changes between focused local view (NYC) and wider continental view as the user pans/zooms. Steps: Default map — All Traffic → Select 'Restricted' filter → Select 'Military' filter → Select 'Cargo' filter and zoom out → Select 'Other' filter and inspect global layer **Weather mini: Trip Planning** (ios, 4 screens, Weather) [Refero](https://refero.design/flows/13104) > A trip planning screen where the user sets trip dates, adds a destination, optionally adds stops, and then continues by tapping Next. The flow starts on a planning form with date fields and a disabled Next button, opens a location search overlay when adding a destination, and returns to the form once a destination is selected. Steps: Trip Details → Destination Search → Filtered Results → Destination Added **Trip Way: Itinerary timeline with category filters** (ios, 3 screens, Travel & Booking) [Refero](https://refero.design/flows/8468) > Happy-path: a signed-in user views a multi-day trip itinerary (“Lakewood Retreat”) and uses top-level controls and category filters to narrow the visible activities. Entry: user opens the itinerary detail. Exit: user ends with a filtered list of itinerary items visible (or returns/close via the X). Assumptions: user is on a mobile device, already selected a trip to view, and can interact with tappable controls (filter icons, document/packing list tiles, timeline rows and floating actions). Steps: Itinerary overview (default) → Category filter — sailing selected → Category filter — wildlife/pets selected **Tripsy: Create Car Route** (ios, 4 screens, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/2355) > Happy path: From a New Activity chooser the user picks a route-type activity, opens a Car Route editor, fills in route name, departure and arrival locations with date/time, and optional contact/reservation fields, then saves. Steps: New Activity - Category & List → Car Route - New (Empty) → Car Route - Filled (Review) → Save Confirmation & Return **Plane Finder: Filter Departures by Terminal** (ios, 3 screens, Travel & Booking) [Refero](https://refero.design/flows/10390) > User views a vertical list of LGA departures (default: All Terminals). The page includes a search field, a status filter pill (Departed), and a terminal selector dropdown. The user opens the terminal selector, chooses a specific terminal (B), and the list updates to show departures for that terminal. Steps: Departures list (default) → Terminal selector opened → Filtered departures (Terminal B) **Box Box Club: Driver Comparison** (ios, 6 screens, Entertainment, Sports, Transport) [Refero](https://refero.design/flows/7869) > Mobile 'Head to Head' interaction that lets a user compare two drivers across multiple metrics (Points, Position, Wins, Sprint Wins, Podiums, Poles, etc.). Steps: Head-to-Head Overview → Open First Driver Selector → Filter Roster Using Search → Select First Driver → Open Second Driver Selector → Updated Head-to-Head Comparison **komoot: Begin Route Navigation** (ios, 37 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8379) > User starts on a route detail (Weather) tab and chooses to navigate. They pick a device (phone or external), handle a "route is far away" warning, then open the map overview and press Navigate. The app switches into active navigation: map-centered guidance with a top instruction header, a controls sheet (search/place, auto-replan, live tracking, report), and inline tools (layers, compass). The user manually selects/places a new waypoint on the map and adds it to the route. During navigation they open Controls to pause/resume, use Live Tracking, and optionally finish the activity. Entry: route detail screen. Exit: navigation active or activity finished. Assumptions: user is signed in, location permission granted, route is pre-planned and saved, device has connectivity for sending updates. Steps: Route Weather & Tips → Choose Device → Distance Warning → Route Overview Map → Navigation: Map Centered → Navigation: Map Zoom/Focus → Next Waypoint Info → Elevation Panel → Manual Location Selection → Confirm Add Waypoint → Add Waypoint Options → Navigation Mode — Heading → Active Navigation Map with Route → Open Controls Sheet → Pause Navigation → Resume / Finish Options → Search & Plan From Controls → Live Tracking / Share → Report Issue / Select Location (alternate path) → Finish Navigation / Return to Routes **Plane Finder: SFO Arrivals Performance** (ios, 3 screens, Travel & Booking) [Refero](https://refero.design/flows/10402) > Mobile screen showing SFO performance for arrivals. Entry: user opens the SFO Performance overview with the 'Arrivals' segment selected and the 'Last 14 Days' time range active. The top card shows aggregated statistics (percent early, on time, late, cancelled) and a stacked-bar chart by date. A 'Breakdown' card below lists airlines and counts. Exit: user changes the time range to 'Last 24 Hours' which updates the header, the stats card, the chart, and the airline breakdown. Assumptions: user is on a mobile device, viewing an airport performance dashboard with interactive controls (segment toggle, date-range dropdown, sortable breakdown). Steps: Overview — Last 14 Days → Open Time-Range Menu → Overview — Last 24 Hours **Enode: Create vehicle asset** (web, 4 screens, Development Tools, Productivity) [Refero](https://refero.design/flows/4261) > Happy path: a logged-in developer/administrator opens the Assets area, clicks Create new, fills the Create new asset modal (selecting asset type, vendor, model, year and a display name), optionally sets linking status / links to a user, and confirms creation. Steps: Assets list (empty) → Create asset modal (blank) → Create asset modal (filled) → Assets list (new asset visible) **Chargetrip: Select vehicles** (web, 3 screens, Development Tools, Map & Navigation, Transport) [Refero](https://refero.design/flows/9276) > Modal-based vehicle selection used while configuring a project/profile. Steps: Vehicle list (default / browse) → Search returns no matches → Search returns specific model results **Cushion: Workspace Members Management** (web, 4 screens, Communication Tools, Collaboration Tools) [Refero](https://refero.design/flows/12535) > This is a workspace Members settings screen shown in a modal-style settings panel. The user is assumed to be logged in with permission to manage workspace members. Steps: Members List → Member Actions Menu → Removal Confirmation → Updated Members List **Chargetrip: Compare vehicle range** (web, 4 screens, Development Tools, Map & Navigation, Transport) [Refero](https://refero.design/flows/9261) > Happy-path: User is on the Range tab with a base vehicle selected (Audi Q8 SUV e-tron) and an origin location entered (Valencia Airport). From this starting point the user opens a modal to add up to two additional vehicles to compare, searches and selects two Q4 e-tron variants, confirms the selection, and then adjusts vehicle-specific settings (battery at departure, occupants, climate control, additional weight) and route settings before running the range calculation. Steps: Starting range configuration → Open 'Add vehicles' modal → Select up to two vehicles → Confirm selections and adjust settings **Dropbox: Team Folder Upload** (web, 6 screens, Collaboration Tools, Business, Productivity) [Refero](https://refero.design/flows/13027) > Happy path for uploading multiple files into a team folder, monitoring upload progress, and then reviewing the uploaded items in the folder. **Enode: Asset search and empty result** (web, 2 screens, Development Tools, Productivity) [Refero](https://refero.design/flows/4262) > User visits the Enode Developers Dashboard and opens the Assets area (Linked tab). The initial screen shows a table of linked assets with columns for Display name, Type, Brand, Model, Linked to and Created at, plus a search field, a type filter, and a 'Create new' button. The user enters a specific identifier in the search field to locate an asset. In the happy path the table would be filtered to matching rows; in the provided capture the search returns no matches and the table is empty. Assumptions: user is signed in, on a desktop-sized viewport, and has permission to view assets. Steps: Assets list (Linked) → Search by identifier (no results) **TravelPerk: Manage trip approval process** (web, 4 screens, Productivity, Travel & Booking) [Refero](https://refero.design/flows/5255) > Happy path: an admin user navigates from Account > Travel management > Approval processes, creates a new approval process by naming it, selecting who it applies to, choosing whether bookings require approval, assigning approvers/rules for bookings, choosing a preset for handling changes to booked trips, and then saves the process. Steps: Approval processes — empty state → Add approval process — form → Add approval process — populated fields → Approval processes — list with new process **Apple Maps: Multi-Stop Directions** (web, 4 screens, Map & Navigation) [Refero](https://refero.design/flows/12476) > This is a directions-planning interface in a maps product. The user starts on a directions screen with a route already populated from My Location to Kimpton Hotel Eventi, with Macy’s as an intermediate stop. The user can change travel mode, reorder or edit stops, add another stop, and review updated route time, distance, and ETA. The right side of the screen updates the map and route line as stop order changes. The happy path ends with the user confirming a revised multi-stop driving route and viewing the updated path on the map. Steps: Initial Route Review → Stop Selection → Stop Editing → Recalculated Route **ElevenMusic: Track Management in Studio** (web, 3 screens, Music Streaming, Entertainment, AI Tool) [Refero](https://refero.design/flows/12557) > This is a dark-themed music studio workspace, entered from the left navigation with Studio selected. The user is assumed to be logged in and viewing a track library/list. The flow starts with a list of tracks shown under a ‘Today’ section, each row displaying artwork, title, tags, rating, duration, and a Publish action. The user selects multiple tracks using row checkboxes, which triggers a bulk action bar with selection count and actions such as Deselect, Archive, and Delete. After the action is applied, the list updates to show only the remaining tracks, and the selection toolbar disappears. The flow ends back in the track list view, now reduced to the unarchived items, with the same create prompt and filtering controls still available. Steps: Browse Track List → Select Multiple Tracks → Updated Track List **fal: AI Model Directory Browsing** (web, 2 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/13547) > A user browses a directory of AI models and tools. The flow starts in a two-column card layout showing model previews, categories, descriptions, tags, favorites, sorting, and pagination. The user switches to a compact list layout using the view toggle. The same 28 results remain visible in list format, with model icons, names, descriptions, tags, and favorite controls. The user appears to be browsing without needing to sign in or open an individual model. Steps: Model Card Grid → Model List View **Chargetrip: Electric vehicle route planner** (web, 14 screens, Development Tools, Map & Navigation, Transport) [Refero](https://refero.design/flows/9257) > Happy path: A signed-in user chooses the Route tab for a selected electric vehicle, adds a start location and a destination, adjusts vehicle and route parameters, calculates the route, and inspects detailed trip results (time, distance, energy consumption, elevation and consumption graphs, CO2 savings and electricity mix). Entry: map view with vehicle selection and top-level tabs (Stations / Route / Range). Exit: route results panel open with option to send route to phone, share, or view methodology. Assumptions: user has a vehicle selected (Audi Q8 SUV e-tron in screenshots), map tiles and POI search are available, user can interact with left-side panel and map on a desktop-resolution viewport. Steps: Map overview with vehicle selected → Open Route tab and show inputs → Prompt to select start location → Search results for start location → Start point placed on map → Add destination and display route endpoints → Adjust vehicle parameters → Adjust route settings (weather profile) → Calculate route and show summary card → Open route visualization — elevation → Open route visualization — consumption → Climate report and CO2 savings card → Electricity mix and carbon intensity details → Open methodology modal **Factory: In-document Assistant** (web, 4 screens, Business, Development Tools, AI Tool) [Refero](https://refero.design/flows/7718) > Happy path: a signed-in user reads a documentation article and uses the right-hand "Assistant" panel to ask a question about the content (example: "managing context"). The assistant runs a search, generates a concise answer and supporting bulleted guidance, and surfaces related links/resources. Steps: Open documentation article → Enter question in Assistant → Assistant returns summary → Review expanded guidance and best practices ##### Task: fraud alert suspicious transaction **Artsy: Create artist alert with detailed filters** (ios, 18 screens, Marketplace) [Refero](https://refero.design/flows/7597) > Happy-path: A signed-in user on an artwork detail taps Create Alert, configures filter criteria (medium, rarity, price, size, ways to buy, color, etc.), reviews the selected filters, sets notification preferences, and saves the alert. Steps: Artwork Detail → Create Alert (start) → Filters — Medium (initial) → Filters — Medium (multi-select) → Filters — Rarity + Price (open) → Price Range — set numeric → Filters — Size & Ways to Buy → Filters — Custom Size (open) → Filters — Custom Size inputs → Filters — Enter numeric values → Filters — Color & Ways to Buy selection → Set Filters (apply) → Create Alert — review filters → Create Alert — submit → Saved Confirmation → Alerts list (Favorites tab) → Edit / Remove (implicit) → End state — alert visible in list **HYPE: Upcoming Product Details** (ios, 2 screens, Lifestyle, News) [Refero](https://refero.design/flows/12252) > This is a product detail screen for an upcoming sneaker release, shown from the Drops tab. The user enters from a product listing or drop navigation, reviews the item name, release date, image, and product details, then can tap Notify Me to subscribe for an alert. After tapping, the button changes to Notification On, confirming the alert is enabled. Steps: Review Drop Details → Notification Enabled **Tinder: Help Center — Account Appeals** (ios, 6 screens, Dating & Relationships) [Refero](https://refero.design/flows/3473) > Happy-path: From the in-app Settings screen the user taps Help & Support, which opens Tinder’s Help Center in a web view. From the Help Center landing the user either uses the search box or selects a category (Safety & Reporting). Inside that category they open the “Account Violations & Appeals” section and read an article (“My account was warned”). The article contains links to the Appeal Center and may surface an embedded submission form (text input + Submit). Steps: Settings — Help & Support → Help Center Landing → Safety & Reporting — Category List → Account Violations — Article List → Article — Read Guidance → Appeal / Feedback Submission **Ground News: Delete single alert** (ios, 4 screens, News) [Refero](https://refero.design/flows/7576) > Happy-path: a signed-in mobile user opens the Alerts tab, locates an alert they want to remove, uses a left-swipe gesture on the alert row to reveal action tiles (Share, Delete), selects Delete, confirms in a modal, and the alert is removed from the list. Entry: user is on the Alerts tab in the app's bottom navigation. Exit: alert row is removed and the Alerts list updates. Assumptions: iOS-style mobile view, bottom navigation visible, swipe gestures supported, user has permission to manage their alerts. Steps: Alerts list (default) → Swipe reveals actions → Initiate delete (destructive) → Confirm delete and update list **komoot: Report tip comment** (ios, 4 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8373) > Happy path: From a tip/comment list the user taps the inline "Report" control for a specific tip. A bottom sheet titled "Submit report" slides up where the user picks one reason from a list (radio choices). The Submit button is disabled until a reason is chosen. After choosing a reason the user taps Submit, and a confirmation screen thanks the user and offers a "Got it" button to dismiss. Steps: Tip comment — trigger Report → Report reasons — no selection → Report reasons — select reason → Submission confirmation **Ground News: Alerts Inbox Cleanup** (ios, 3 screens, News) [Refero](https://refero.design/flows/7577) > Happy path: A signed-in mobile user opens the Alerts tab, sees one or more alert items, uses the trash (bulk delete) action in the top-right to remove all alerts, confirms the destructive action in a modal, and is returned to an empty-alerts state with a message indicating no new alerts. Entry: Alerts tab is active. Exit: Alerts list is cleared and empty-state message shown. Steps: Alerts list → Delete confirmation → Empty alerts state **Meta AI: Report Restricted Content** (web, 9 screens, AI Tool) [Refero](https://refero.design/flows/12757) > A logged-in user browses the Vibes media feed and reports a post they believe promotes a restricted item. The user opens the post's overflow menu, selects Report, chooses Selling or promoting restricted items, specifies Weapons, reviews the report details, and submits it. The system confirms receipt of the report and returns the user to the feed with the reported vibe hidden. Steps: Browse Vibes Feed → Open Post Actions → Choose Report Reason → Select Restricted Items → Identify Promoted Item → Select Weapons → Review Report → Confirm Submission → Return to Feed ##### Task: government service identity verification **Airbnb: ID capture and name confirmation** (ios, 6 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6419) > Happy path for collecting a driver's license (front and back) using the device camera, reviewing each captured image, and extracting the legal name. Entry: user arrives at the 'Front of ID' camera screen (user likely in the middle of identity verification). Exit: user confirms the extracted legal name. Assumptions: user is on a mobile device, has a physical ID, camera permission granted, and is following on-screen prompts. The flow enforces distance/framing, allows retakes, guides the user to flip the card, and runs OCR to prefill name fields before final confirmation. Steps: Capture front of ID (camera) → Review front of ID → Flip to back instruction → Capture back of ID (camera) → Review back of ID → Confirm extracted legal name **Airbnb: Selfie verification capture** (ios, 4 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6420) > Happy path: A user who has already submitted an ID is prompted to take a live selfie so the system can compare it with the provided ID. Entry: instruction screen that explains the selfie step. The user taps Continue, granting camera access if required, and the in-app camera opens. The UI guides the user to center their face inside a circular mask, shows a progress ring while analyzing, and automatically captures the selfie. Exit: a completion confirmation appears with a check and a CTA to acknowledge (Got it!). Assumptions: mobile device with a front-facing camera; user has already provided ID (text on the intro screen); camera permissions may be requested before capture; automatic capture with visual guidance rather than manual shutter. Steps: Intro: selfie required → Camera open with guidance → Capture in progress → Capture success confirmation **Airbnb: Event booking and verification** (ios, 17 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6417) > User books a ticketed event (stand-up comedy) from the event details screen, chooses a date/time and number of guests, goes through a compact checkout (payment method selected), then is asked to complete bank authentication and identity verification. Steps: Event details → Select a time (list) → Calendar month view → Date selected confirmation → Times list for chosen month → Scroll for more slots → Select time slot → Confirm and pay (summary) → Payment method + price details → Processing reservation → Bank action required notice → Bank authentication (Revolut) → Identity verification start → Verification deadline & pending booking → Trips — pending reservation notice → Pending event card in Trips → End state: verification required **Volv: Claim username via phone verification** (ios, 19 screens, AI Tool, News) [Refero](https://refero.design/flows/2157) > User opens the side menu and selects "Claim your username" (or Login). A login modal appears requesting a phone number with a country selector. The user chooses a country (or searches for it), enters their phone number, and the system validates the number. If valid the app runs a reCAPTCHA verification and sends an SMS verification code. The user enters the code in a dedicated modal; on success they continue to a username claim form (username + full name). The system checks availability and either shows an inline error (username taken) or accepts the choice and returns the user to the menu showing the claimed username. Assumptions: user is not authenticated at start, using a mobile device, SMS-based auth is enabled, country codes list is preloaded. Steps: Open menu / Claim username CTA → Login modal (phone entry) → Country selector list → Country search → Phone entry focused → Invalid phone error → Valid phone entered → reCAPTCHA / bot check → Verification code sent notice → Enter code modal → Partial code input → Complete code entry attempt → Authentication failed (rate/suspicion) → Alternate code attempt → Successful code validation → Enter username and full name → Username already taken error → Username submit / saving → Success and menu update **Mercury: Upload and verify government ID** (web, 3 screens, Business, Finance) [Refero](https://refero.design/flows/11021) > Happy-path: A logged-in user (top-right account shown) manually verifies identity by selecting an ID type, uploading a color photo of their ID, and submitting a selfie if required. Entry: user lands on an Identification screen with instructions and an ID-type selector. Primary actions: choose identification type, upload or drag-and-drop an uncropped photo of the ID, optionally take a live photo of self. Exit: the system either accepts the uploads or shows validation warnings/errors and guidance; in the screenshots the system returns issues that may require a resubmission or follow-up. Assumptions: user is authenticated; they have access to photos or a camera; permissions to upload files and to take a selfie are available. Steps: Identification instructions and upload → Upload preview and processing → Validation results and remediation **Mercury: Identity verification with webcam selfie** (web, 4 screens, Business, Finance) [Refero](https://refero.design/flows/11022) > Happy path: A user completes manual identity verification by selecting an ID type, uploading a photo of the ID, and capturing a live selfie via webcam. The flow opens on an Identification form where the user picks an ID type and either uploads the ID image or drags and drops it. The user clicks Take photo to open a modal camera tool, captures a selfie, confirms the photo, and returns to the form where the uploaded selfie is shown. The user can optionally add a professional profile link and save to proceed into the final terms section. Steps: Identification form → Camera modal — live preview → Photo review → Form with uploaded selfie **Airbnb: Booking payment + verification** (web, 18 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6002) > Happy path (attempt): User starts on a listing and proceeds to Request to Book. The booking flow is a multi-step, two-column layout on desktop: left column collects booking steps (choose payment timing, add a payment method, message host, review request) and the right column shows a reservation summary/price breakdown. The user selects a split payment option (pay part now / part later), adds a debit card, writes a message to the host, and submits a booking request. The system attempts payment confirmation via the user's bank (embedded bank / PSP flow). If the bank auth times out/fails, the UI surfaces inline errors and suggests alternatives. Before finalizing, the system requires identity verification — the user is guided to upload government ID, shown a deadline, and given a final holding screen telling them they have 12 hours to complete verification. Assumptions: user is signed in, on desktop, has a valid card and access to bank/PSP authentication (and email). Steps: Listing details / start booking → Choose when to pay → Add a payment method (card entry) → Payment method confirmed, move to message → Write a message to the host → Review request → Confirm card required (bank flow trigger) → Bank connection loading → Bank auth (Revolut) prompt → Bank auth timed out → Inline retry toast → Bank couldn't confirm card modal → Processing / identity verification prompt → Choose ID submission method → Verification deadline notice → Pending verification landing → Final hold state — 12 hour notice → End state / next steps **Acctual: Country selector modal** (web, 2 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8826) > A modal-based country selector used during an identity verification step. Entry: user opens the 'Select country' control from the Verify your identity flow (user appears to be signed in; left navigation shows Payments). Modal covers the main view and shows an alphabetic list of countries with flags and a search field at the top. Happy path: user types or scrolls to find a country, taps a country row, the modal closes and the selected country is applied back to the verification form. Exit points: selecting a country (not shown) or closing the modal with the Close control. Steps: Open country selector → Search and empty result **District: State Selection** (ios, 3 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12814) > A state-selection modal opens from Edit Profile. The user can search within a list of states, tap a matching state to choose it, or dismiss the modal with the close icon. Steps: Open state picker → Search input state → Matched state result **Fingerprint: Identification events export** (web, 3 screens, Business, Development Tools) [Refero](https://refero.design/flows/11155) > Happy path: a signed-in user viewing the Identification events list requests an export of events for a specified date range. Steps: Events list → Export events modal → Export confirmation **District: Edit Profile and Select State** (ios, 4 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12813) > Logged-in user edits profile details on a mobile screen, including personal, optional, and invoice-related fields. The flow starts on the Edit Profile screen, continues into a state selector modal where the user can browse or search states, and ends back on the profile screen with the chosen state filled in and the option to save changes via Update profile. Steps: Edit Profile → State Picker → State Search → Profile Updated State **Meta AI: Meta AI Email Sign-In** (ios, 7 screens, AI Tool) [Refero](https://refero.design/flows/13401) > A signed-out user opens Meta AI and chooses to get started. The system requests a mobile number or email address, then sends a six-digit verification code to the submitted email. The user enters the code, authentication completes, and the user is taken through a brief loading screen to the Meta AI home screen. The screenshots show the email-code happy path; no device setup is selected. **Tinder: Photo verification consent** (ios, 3 screens, Dating & Relationships) [Refero](https://refero.design/flows/3432) > A three-step in-app verification onboarding sequence shown as a centered modal over the Tinder browsing view. Entry: a logged-in user viewing profiles is prompted to "Get Photo Verified." Happy path: user opens the modal, reads a short explanation of how face-matching works, advances to a second screen describing data retention and audit screenshots, and then taps "I agree" to give consent. Assumptions: user is signed in on mobile, camera permission may be required after consent. Exit: modal closes and the app proceeds to the verification step (camera/selfie capture) or records consent and retains face geometry data as described. Steps: Initial verification prompt → How it works — overview → How it works — data retention & consent **Peerlist: Verification confirmation and invite peers** (web, 3 screens, Job Search, Business, Social Networking) [Refero](https://refero.design/flows/1093) > Happy path: User completes a work-experience verification on their profile and is shown a confirmation modal that their work experience is verified. The modal offers two primary actions: close the dialog or "Invite your peers". Choosing to invite opens a secondary modal where the user can enter an invitee email and send a one-time invitation email on their behalf. Steps: Verification Success → Invite to Peerlist (empty) → Invite to Peerlist (email entered) **Peerlist: Verify Team Member Experience** (web, 3 screens, Job Search, Business, Social Networking) [Refero](https://refero.design/flows/1184) > Happy path: A signed-in project admin/recruiter opens the People & Teams view for an organization, locates a specific member, opens a verification dialog, fills/confirm employment details and team assignment, and submits verification. The UI returns to the People list showing the member with a verification indicator and updated employment dates. Steps: Open People & Teams — locate member → Fill Verify Experience modal → Confirm updated member entry **Omio: Passenger details & review** (ios, 11 screens, Travel & Booking) [Refero](https://refero.design/flows/5184) > Happy-path checkout for adding / confirming the main passenger's contact and identity details before reviewing the journey. Steps: Booking summary + Main passenger header → Passenger form — prefilled fields → Open passenger selector → Phone number input (country selector) → Enter phone number → Date of birth segmented input → Select ID type → Enter ID number and country of residence → Price breakdown and CTA → Loading then navigate to passenger summary → Passenger confirmation & seating options **Tinder: Photo verification (video selfie)** (ios, 12 screens, Dating & Relationships) [Refero](https://refero.design/flows/3431) > Happy-path: From the Tinder swipe/profile card screen the user is prompted to verify their photos. A modal explains photo verification and presents a short multi-screen consent/education flow. After agreeing the user is shown a framing guide and captures a short video selfie (front camera) inside an oval. The UI gives real‑time guidance during recording (countdown, hold steady, move closer). The app then uploads an encrypted 3D video scan to the verification processor. If the scan is unclear the user sees a failed result screen with examples and a Try again CTA. Steps: Discovery / profile card → Intro modal → How it works — slide 1 → How it works — slide 2 (data retention) → How it works — slide 3 (unverify) → Get ready for video selfie → Countdown — hold steady 1 → Countdown — hold steady 2 → Recording — hold steady → Recording feedback — move closer → Uploading encrypted scan → Failed scan — retry guidance **Acctual: Country Selection** (web, 2 screens, Business, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8827) > Happy path: A user opens an identity verification modal and selects their country from a searchable, alphabetically ordered list. Steps: Open country selector → Filter country list **TravelPerk: Add passport to profile** (web, 6 screens, Productivity, Travel & Booking) [Refero](https://refero.design/flows/5205) > Happy path: A signed-in user navigates to their profile Availability area, chooses to add a travel document, completes the passport form (number, issuing country, dates, sex, name, DOB, nationality), confirms the details, and saves. The UI shows a confirmation message that details are encrypted and returns the user to the Availability area where the new passport appears as a document card. Steps: Profile Availability - Travel documents → Add a travel document — Passport form (empty) → Fill passport fields — basic details → Fill passport fields — personal details → Confirm details → Profile Availability — Document saved **Expedia: Passport management** (web, 6 screens, Travel & Booking) [Refero](https://refero.design/flows/3030) > Happy path: A signed-in user navigates from their account profile to add a passport under Travel documents, fills the passport form, triggers validation if there are mistakes, successfully saves the passport and returns to a masked list view. Steps: Profile dashboard → Passport entry form → Form with user input → Validation error feedback → Saved passports list → Passport detail & management **Wealthsimple: Personal details & residency setup** (web, 9 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8508) > Single-screen, multi-section "About you" entry and review experience for a 2024 tax return. Steps: Landing: About you overview → Open DOB control (decade view) → Drill to decade / year set → Select month for chosen year → Pick day in calendar view → DOB populated and field validation → Scroll through remaining personal sections → Manage forms and run review → Summary, documents, and submit CTA **Peerlist: Work Experience Verification** (web, 7 screens, Job Search, Business, Social Networking) [Refero](https://refero.design/flows/1092) > Happy path: A logged-in user on their profile navigates to the Vouch tab and starts the verification flow so their work experience can be verified. The flow is modal-driven: a confirmation modal introduces verification, a company-email modal collects the user's work email (local-part + selectable domain), the system sends a verification code to that address, the user enters the code, and the UI shows a success confirmation. Steps: Profile — Vouch tab → Verification Intro Modal → Company Email — Local-part Input (initial) → Company Email — Local-part Edit → Verification Code Sent — Input → Verification Code — Entered → Verification Success ##### Task: grant application **Fuse: Token swap from Wallet** (ios, 16 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8739) > Happy-path: From the Wallet home the user opens the global action (+) menu, chooses Swap, picks the asset to pay (SOL) and the asset to receive (USDT), enters a small SOL amount using the numeric keypad, reviews the quote and swap details (slippage, route, price impact, fees), adjusts slippage settings if desired, confirms the swap with biometric authentication, sees a success/processing state, and returns to Wallet with balances updated. Steps: Wallet Home → Quick Actions → Select Pay Asset → Swap Entry - Currency Picker → Receive Token List → Search Token → Swap Entry - Keypad → Review Button → Review Details → Slippage - Dynamic → Slippage - Manual (low) → Slippage - Manual (updated) → Updated Review → Execution & Confirmation → Wallet Updating → Final State **GoFundMe: Fundraiser Story Review** (web, 3 screens, Fundraising) [Refero](https://refero.design/flows/11742) > A multi-step fundraiser setup screen where the user writes or edits their personal story for donors, reviews AI-generated improvements, and accepts the suggested copy before continuing. The entry point is the story-writing step (step 4 of 6). The user is assumed to be in a logged-in fundraising setup context with permission to edit the story. The exit point is proceeding to the next setup step after reviewing or applying the enhanced story text. Steps: Story Draft Editor → Suggested Story Review → Updated Story Editor **Foodllama: Apple Health Connection** (ios, 8 screens, Health & Wellness) [Refero](https://refero.design/flows/11778) > Happy path for connecting Foodllama to Apple Health, granting health permissions, syncing data, and returning to the main dashboard where activity data appears. Steps: Foodllama Dashboard → Health Access Instructions → Apple Health Permission Prompt → Nutrition Sync Settings → Workout Sync Settings → Sync Confirmation → Success Confirmation → Activity Detail **Wispr Flow: Wispr Flow Keyboard Setup** (ios, 4 screens, AI Tool, Utility, Productivity) [Refero](https://refero.design/flows/13657) > A guided setup sequence helps the user enable the Wispr Flow keyboard and grant it Full Access in device Settings. The user starts on an instructional screen, opens the keyboard settings section, enables Wispr Flow, then enables Allow Full Access. The flow ends with both permissions enabled and a privacy reassurance explaining why Full Access is required. Steps: Open Keyboard Settings → View Keyboard Options → Enable Wispr Flow → Grant Full Access **Grok: Photo Save to Library** (ios, 3 screens, AI Tool) [Refero](https://refero.design/flows/6086) > Happy path: A user is viewing a full-screen photo preview inside a modal titled "Photo". They tap the Save control, which triggers the system-level Photos permission prompt. The user grants permission; the app saves the image to the device photo library and displays a local confirmation state in the modal. Steps: Photo Preview → Photos Permission Request → Save Confirmation **Structured: First‑use onboarding & setup** (ios, 44 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3255) > A mobile onboarding and first‑time setup sequence for a daily planner app. Entry: cold start / splash. The user is guided through marketing slides, an abbreviated setup (wake time, bedtime, today's priority), optional task creation, and a subscription offer (monthly/yearly/lifetime). The subscription path opens the native App Store purchase sheet, then returns to a Pro features screen that requests Calendar and Reminders permissions. Exit: user lands on the main timeline/calendar view with sample tasks and an add (+) button. Steps: Launch / Splash → Welcome Slide → Feature Spotlight — Timeline → Feature Benefits → Get Started Prompt → Wake Time Picker → Bedtime Picker → Create Today's Priority — Suggestions → Task Input (keyboard shown) → Typing Task → Choose a Plan Modal → Plan Selection Loading → App Store Confirmation → Purchase Success Alert → Pro Features Overview → Calendar Access Prompt → Reminders Access Prompt → Enable Sync Toggles → Main Timeline — First Visit → Ongoing Use — Timeline Interaction **Lumy: Notification editor — event-based schedule** (ios, 10 screens, Weather, Utility) [Refero](https://refero.design/flows/6168) > Happy-path: User opens the Notifications list, taps an existing notification (Sunset) to open the editor, updates fields (date, location, event, alert, repeat), saves the changes and returns to the Notifications list. Steps: Notifications list → Edit notification (main) → Calendar date picker (open) → Calendar date selected → Location picker (open) → Event type picker (open) → Repeat options picker (open) → Editor with updated repeat → Permission modal — Background App Refresh off → Save confirmation and return **Kann: Kana practice quiz** (ios, 6 screens, Education) [Refero](https://refero.design/flows/9160) > Happy-path: User opens the Kana practice module, picks a set of characters (Hiragana or Katakana) from a full-grid selector, then completes a short multiple-choice quiz that asks for the romanisation of shown kana. During the quiz the top progress bar fills and a lives/hearts indicator is displayed. After answering the final question the user sees a completion state with score and reward (coins). Entry: Kana selector screen. Exit: Quiz completed screen with reward and buttons to start a new quiz or close. Assumptions: user is on a mobile device, the app shows coins/rewards only if the user is in a state that can receive them (likely signed in). Steps: Kana selector → Quiz question (voiced kana) → Quiz question (basic kana) → Quiz question (another kana) → Intermediate reward toast → Quiz completed / Reward screen **222: Invite & add contacts to party** (ios, 10 screens, Lifestyle, Dating & Relationships) [Refero](https://refero.design/flows/9310) > Happy-path: A signed-in user on the party/curation screen taps Invite Friends, chooses to invite people to "other tables or restaurants" (recommended), grants or selects contacts from the device, composes an optional message, and sends invites. Steps: Party overview → Choose invite destination → Open contact-invite sheet → Request contact access (system) → Contacts share options (system) → In-app contacts list (first view) → Selection confirmation and chips → Compose message & send → Keyboard & send confirmation → Updated roster **Clime: First-run onboarding and permissions** (ios, 19 screens, Weather) [Refero](https://refero.design/flows/10545) > First-run onboarding for an iOS weather product. Entry: fresh install / first launch. The flow introduces the product (feature carousel), collects consent (terms + privacy), offers privacy customization, presents a subscription / free-trial paywall, requests location and notification permissions, and finishes by landing the user on the main map with a forecast bottom sheet. Steps: Welcome / Terms → Privacy explainer → Privacy settings (sheet) → Privacy settings — vendor details → Onboarding carousel — Live Radar → Onboarding carousel — Forecasts → Onboarding carousel — Wind Map → Onboarding carousel — More features → Subscription paywall — plan selection → Paywall — Free trial enabled → Free trial loading → App Store confirmation → Purchase success → In-app location prompt → iOS location permission dialog → Notification customization screen → iOS notification permission prompt → Onboarding finish — loading final state → Main map + forecast (landing) **(Not Boring) Camera: Camera Compose & Exposure Adjustment** (ios, 4 screens, Utility) [Refero](https://refero.design/flows/7684) > Vertical camera capture interface for composing a photo and adjusting exposure in real time. Steps: Compose and set focus → Enter exposure adjustment (initial) → Decrease exposure (preview clipping shown) → Increase exposure (final preview) **Omio: First-time onboarding, signup and permissions** (ios, 18 screens, Travel & Booking) [Refero](https://refero.design/flows/5154) > Happy-path onboarding → signup → permissions → first search. Entry: new user starts on a 3-screen marketing carousel (skip available). From any carousel screen user can open auth options and sign in with Apple/Google/Facebook or choose email. If user chooses email, the flow validates the address, creates an account (if email is new), collects first & last name and opt-in for marketing, then shows a welcome confirmation. After account creation the app asks the user to enable push notifications (system modal) and then asks to enable location (system modal). Exit: user lands on the main search screen with departure populated (if location allowed) and can start searching. Assumptions: user on iOS device, new/unregistered user, network available, system permission prompts are iOS native dialogs. Steps: Onboarding - carousel (slide 1) → Onboarding - carousel (slide 2) → Onboarding - carousel (slide 3) → Auth options expanded → Sign in - email entry → Validation feedback → Valid email entered → Sign up modal (email confirmed) → Account creation (loading) → Welcome confirmation → Post-signup marketing opt-in prompt → iOS push permission dialog → Start search helper → iOS location permission (first prompt) → iOS location permission (always prompt) → Main search screen (prefilled) → Home with search ready → Edge states & fallbacks (implicit) **Orbit: Subscription Onboarding + Upgrade** (ios, 12 screens, Finance) [Refero](https://refero.design/flows/9988) > Happy-path onboarding for first-time users on an iPhone. Entry: first-run welcome / marketing screens. The flow introduces the product, asks the user to select common subscriptions (multi-select list), shows an estimated yearly total, offers import and premium (lifetime) purchase options, and finishes by landing the user on the main dashboard (orbit visualization) with a notifications permission prompt. Steps: Welcome / Hero → Motivation / Statistic → Subscription Selection (initial) → Subscription Selection (selected state) → Extended Selection / Scrolling → Estimated Yearly Total → Add Subscription / Import Options (modal) → Pricing Offer (lifetime) → Premium Features Overview → Feature Comparison Table → Final Pricing / Legal Links → Main Dashboard + Permissions **ChatGPT: Voice assistant setup and session** (ios, 10 screens, AI Tool) [Refero](https://refero.design/flows/7030) > Happy path for enabling and using the app's voice assistant on a mobile device. Entry: user is on the chat home screen (compose field + microphone). The user taps the voice control and is guided through a voice-selection carousel, confirms with Get Started, goes into a voice session (listening/recording), optionally mutes/unmutes, hears the assistant speak (TTS), and ends the session. Exit: voice chat ends and the transcribed conversation + assistant replies are shown in the regular chat thread. Assumptions: user is signed in on a mobile device, microphone permission has been granted (or will be requested), and the feature is available to the account. Steps: Chat home (compose + mic) → Choose a voice — Ember → Choose a voice — Vale → Voice session: preparing / listening → Feature upgrade modal → Active voice session (listening + memory update) → Muted / paused microphone state → Assistant speaking (TTS) — large text display → Voice chat ended summary banner → Chat view with voice transcripts **Instagram: In-call Camera Effects** (ios, 13 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/2885) > Happy path: A user in a mobile one-to-one call (other participant's camera is off) opens the in-call effects panel, browses and selects camera effects/filters from a carousel and full gallery, adjusts a 'touch up' (beauty) intensity slider, and applies a virtual background/overlay. Entry: in an active call with participant present, user's camera enabled. Exit: a visual effect + touch-up enabled and applied to the local camera preview; call controls remain available. Assumptions: user is on a phone in an active call, camera and microphone permissions already granted, effects are client-side previews applied to the outgoing/local camera feed. Steps: Active Call — Default View → Open Effects Carousel → Browse Full Effects Gallery → Select Effect from Gallery → Return to In-call with Effect → Swipe Carousel to Try Variants → Open Touch-up Controls → Adjust Touch-up Intensity → Apply Background / Scene Effect → Try Alternate Backgrounds → View Before/After Previews → Persist Effect & Indicator → Final Call State with Effects **Substack: Direct message — create, send, react, and manage** (ios, 20 screens, News, Social Networking) [Refero](https://refero.design/flows/4494) > Happy-path direct messaging flow for a mobile messaging feature (iPhone). Entry: the user opens the Chat area (no messages). Exit: the conversation is created, messages and images are sent, message reactions applied, and chat is muted/pinned; chat list reflects the pinned/muted state. Steps: Empty Chat list → New message — contacts list → Search input → Select contact from search → Conversation — compose entry → Type message → Send message → View profile from conversation → Conversation — incoming messages → Attach image — open photo picker → Photo picker — select photo → Image composer preview → Image message sent → Return to chat list → Open/expand image in conversation → Open message actions sheet → Add reaction to message → Open conversation menu → Set mute duration → Pin conversation and confirm **Abode: Create Abode setup** (ios, 16 screens, Collaboration Tools, Communication Tools, Social Networking) [Refero](https://refero.design/flows/4295) > Happy-path onboarding for creating a new shared group (called an “Abode”) from an empty group view. Entry: user is on an empty group detail screen (no recent activity). Exit: user lands in the new Abode home with at least one magnet and the chat composer visible. Assumptions: user is signed in, has granted contact access (contacts are listed), device is a modern iPhone-sized touch device, and the UI supports taps, horizontal swipes and simple micro-interactions/animations. Steps: Empty group detail → Bloom hero prompt → Tap to activate Bloom → Bloom completion animation → Name your Abode (swipe cards) → Choose avatar color → Choose avatar icon → Choose theme (palette options) → Theme variation preview → Add friends (contacts list) → Contact invited state → Create Abode CTA → Magnet onboarding card → Select a magnet (catalog) → Abode first-run home (transitional) → Abode home (steady state) **CapWords: Create sticker from photo** (ios, 21 screens, AI Tool, Education) [Refero](https://refero.design/flows/6112) > Happy-path for creating a new object sticker from a photo. Steps: Canvas (start) → Request camera permission → Camera live view with framing → Full-image recognition hint → Processing / segmentation → Extracted object preview (first pass) → Polaroid / small preview → Extract refinement and check → Final extracted object (floating) → Photo picker / crop entry → Crop UI with tools → Crop ratio selection → Confirm crop selection → Auto-extraction result (sticker preview) → Name object prompt (initial) → Enter custom name → Processing finalization → Sticker added to canvas (single language label) → Sticker added to canvas (two-language label preview) → Completed canvas with multiple stickers **Linear: Personal API key revocation** (web, 4 screens, Business, Productivity, Project Management, Collaboration Tools) [Refero](https://refero.design/flows/6692) > Happy path: A signed-in user opens Settings → Security & access, locates an existing Personal API key, opens the key's actions menu, chooses to revoke the key, confirms the revoke in a modal, and sees the key removed with a success notification. Steps: Security & access — API key visible → Action menu open → Revoke confirmation modal → Revocation complete — key removed **Chargetrip: Job application via modal** (web, 5 screens, Development Tools, Map & Navigation, Transport) [Refero](https://refero.design/flows/9253) > Happy-path flow for a candidate submitting an application from a job listing. Steps: Job listing (Apply CTA) → Application form (empty) → File upload & field entry → Supplemental questions & consent → Submission confirmation **Rox: Google Calendar & Gmail Connect** (web, 7 screens, Business, Productivity, AI Tool) [Refero](https://refero.design/flows/10759) > Happy path: A signed-in user navigates to the Integrations area, selects Google Calendar and Gmail from the catalog, reviews and confirms product-level permission dialogs, is redirected to Google's OAuth consent screens to grant account access, and returns to the Integrations list where each integration's card updates to show connecting/connected states. Steps: Integrations Gallery → Calendar Permission Modal → Google OAuth Consent (Calendar) → Integrations — Connecting State → Gmail Permission Modal → Google OAuth Consent (Gmail) → Integrations — Connected & Connecting **Figma: Manage sharing and invite collaborator** (web, 8 screens, Design & Media Editing, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/5007) > Happy-path: A signed-in user opens the presentation's Share dialog, inspects team members, opens the sharing settings to change link access from a broader scope to "Only invited people", saves that setting, then enters an email and invites a collaborator. Entry: user clicks the Share button (top-right) while viewing a presentation on desktop. Exit: Share dialog closed and collaborator avatar appears in the top-right collaborator area. Assumptions: user is logged in and has permission to change sharing settings (owner or editor), using a desktop browser. Steps: Open Share dialog → View Team project members → Open Share settings → Select 'Only invited people' → Save share settings → Enter email to invite → Confirm invite sent → Review updated collaborators / presence **Loom: Video share and invite** (web, 4 screens, Communication Tools, Collaboration Tools, Utility, Productivity) [Refero](https://refero.design/flows/1242) > Happy-path: A signed-in workspace member viewing a recorded video opens the Share dialog to grant another person or a workspace space access. Steps: Video details — Views tab → Share dialog — initial state → Add invitee and choose permission → Share dialog — invited list **Tella: Record screen with webcam and edit** (web, 13 screens, Utility, Communication Tools, Collaboration Tools, Productivity) [Refero](https://refero.design/flows/391) > Happy path: A logged-in desktop user with an empty workspace creates a new video using the web recorder, grants camera and microphone permissions, starts a screen share recording (optionally with webcam), navigates web content while recording, interacts with an in-page bookmark modal, stops the recording, and is taken to the editor where the recorded clip appears on the timeline. Steps: Workspace - My videos (empty) → Choose recorder method → Recorder permission prompt (in-app) → Browser permission dialog → Camera preview active → Prompt — "This isn't a podcast" → Screen sharing started (preview) → Recording in progress — content capture → Recording continues — deeper navigation → Recording continues — longer capture → In-recording modal interaction (bookmark) → Stop screen and finalize clip → Editor with recorded clip(s) **Grok: Grok Account Creation and Onboarding** (web, 9 screens, AI Tool) [Refero](https://refero.design/flows/5316) > End-to-end happy-path for a new user discovering Grok, signing up, verifying their email, and arriving at the logged-in assistant home. Steps: Marketing Landing / Hero → Welcome Assistant Prompt (Logged-out) → Signup Method Selection → Captcha / Provider Success → Account Details — Initial → Account Details — Completed → Verify Your Email (Pending) → Email Verified Confirmation → Authenticated Assistant Home **Riverside: Riverside onboarding and studio setup** (web, 18 screens, Productivity, Entertainment, Collaboration Tools) [Refero](https://refero.design/flows/4647) > Happy-path new user journey from marketing entry through account creation, onboarding questions, optional Pro trial upgrade, email verification, device permission checks, and finally joining the recording studio and inviting participants. Steps: Marketing entry / hero → Create account - sign-up options → Create account - email form (empty) → Create account - email form (filled) → Check inbox - verification prompt → Onboarding: content type selection (intro) → Onboarding: content type selected → Onboarding: acquisition source (intro) → Onboarding: acquisition source selected → Onboarding: how to start (tools) → Onboarding: start recording chosen → Pro trial offer screen → Trial plan selection → Payment details entry → Trial confirmation → Camera & mic permission prompt → Device selection & join preview → Studio main UI with invite panel **Cofounder: Cofounder onboarding and setup** (web, 30 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/7892) > Happy path: A new user discovers Cofounder (marketing landing), signs up, and is guided through a linear onboarding that collects their name and company website, performs an automated background lookup, and connects a Google Calendar via a third-party connector (Pipedream). The onboarding then shows inferred calendar preferences, explains how Cofounder works and its memory, offers additional integrations, and finishes by landing the user in the main dashboard with chat and template cards. Steps: Marketing landing → Authentication / Signup → Welcome card → Collect user name (empty) → Name entered → Background lookup (in progress) → Background results summary → Saving background → Collect company website (empty) → Company website entered → Third-party connector consent → Connector in progress → Connector success → Calendar preferences summary → How Cofounder works → Memory system explanation → Privacy & things to know → Additional integrations → Final welcome splash → Main dashboard / chat **Whereby: Pre-join device and display name setup** (web, 2 screens, Productivity, Communication Tools) [Refero](https://refero.design/flows/1974) > Happy path: a signed-in user opens a Whereby meeting room link and lands on the pre-join lobby. The lobby shows a camera preview area, device selectors (camera, microphone, speakers), a beta "hybrid meeting" toggle, and a primary "Join meeting" button. The user can also open their profile menu to edit or set a display name. If the user chooses to set a display name, a modal appears to enter and save the name; if they don't enter one the room will use the profile name. Steps: Pre-join device selection → Set display name modal **Fourthwall: Team Member Invitation** (web, 4 screens, Business) [Refero](https://refero.design/flows/7309) > Happy path for adding a new team member from Site settings → Team. Entry: user is signed in and on Settings > Team. The user clicks Invite new member, fills an email, selects a role from a list of predefined roles, and sends the invite. Exit: modal closes, a new invited member row appears in the team list (marked as invitation sent) and a confirmation toast is shown. The interface also surfaces a member-limit banner and an Upgrade to Pro CTA when the shop is at or near its member limit. Steps: Team list and invite CTA → Invite modal — enter details → Invite modal — confirm invite → Team list — invitation sent confirmation **SoundCloud: Conversation review and management** (web, 3 screens, Music Streaming, Social Networking) [Refero](https://refero.design/flows/6586) > Happy path: A signed-in user opens the Messages area, selects a conversation from the left-hand list, reviews and plays a track embedded in the message thread, then composes and sends a reply (optionally adding a track or playlist). From the conversation header the user can archive the conversation (triggers a confirmation modal). Steps: Open conversation and review track → Archive confirmation modal → Open another conversation (blocked indicator shown) **Cohere: Invite Team Members** (web, 7 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3853) > Happy-path: A signed-in team owner opens the Team settings and invites one or more people using either a reusable invite link or by entering email addresses. The invite modal allows selecting a role (User or Owner) for each invite. After sending, the modal shows confirmation and the Team list updates to show pending invited members with an "INVITED" status and an option to revoke. Steps: Team list (entry) → Invite modal open → Enter email address → Open role selector → Select role for invite → Send invite and confirmation → Team list updated with pending invites (exit) **Descript: Sign-in and Password Reset** (web, 12 screens, Business, Collaboration Tools) [Refero](https://refero.design/flows/2094) > Happy-path and recovery path for a user arriving from the product marketing site, opening the sign-in modal, and completing account access via password reset. Steps: Marketing / Entry → Sign-In Modal → Credentials Entered → Forgot Password — Email Prompt → Reset Email Entered → Email Sent Confirmation → Change Password Form (reset link) → New Password Input → Validation Error — Mismatch → Corrected Passwords Submitted → Password Reset Success → Projects Dashboard (Authenticated) **Descript: Record and Transcribe Session** (web, 10 screens, Business, Collaboration Tools) [Refero](https://refero.design/flows/2109) > Happy path: A user creates a new project titled “Test”, chooses to record (screen/tab) into the editor, records a session, then stops the recording and waits for the system to process and transcribe the footage. Steps: Project start / media options → Select recording source → Ready to record → Recording started → Ongoing recording → Stop recording / begin processing → Processing and transcribing → Transcription appears → Transcript review and media sync → Finalized recording in editor **Soundtrap: In-session vocal recording and post-processing prompts** (web, 10 screens, Collaboration Tools) [Refero](https://refero.design/flows/1970) > Happy-path: A user in an active multi-track project records a vocal take, is guided through a headphone safety prompt and a short countdown, records audio, and then encounters in-product promotions for Vocal Tuner and Vocal Cleanup (both premium). The flow begins on the main arrangement/timeline view with collaborators visible and ends with a new audio region added to the timeline and premium modals shown. Assumptions: the user is inside an open project (logged in), has microphone access, and a collaborators panel/call is active. Device is desktop/laptop with sufficient width for timeline + collaborators column. Steps: Project timeline overview → Open recording panel / prepare → Headphones recommendation modal → Countdown / get ready state → Live recording → Post-record promotional modal (Vocal Tuner intro) → Vocal Tuner settings (gated) → Trial upsell modal → Vocal Cleanup premium modal → Recorded clip added to timeline **ManyChat: Instagram account onboarding** (web, 18 screens, Business, Productivity, Communication Tools) [Refero](https://refero.design/flows/10329) > Happy-path onboarding where a new user discovers ManyChat, selects Instagram as the channel, signs up, links their Instagram via Meta, answers a short setup questionnaire (account use, monetization, tools, referral), optionally starts a Pro trial with payment details, then picks a first automation and lands in the dashboard. Entry: marketing landing or link to start onboarding. Exit: user on ManyChat Home with 1 connected Instagram channel and suggested automations visible. Steps: Marketing landing → Choose channel → Sign up / sign in → Connect Instagram (Meta) → Connection options revealed → Connection confirmed → Select account owner → Monetization question → Monetization selection state → Tools / integrations selection → Integrations selection state → Referral source question → Referral selection state → Pro trial / pricing offer → Payment form (Stripe) → Payment validation error → Pick a first automation → Home dashboard landing ##### Task: health onboarding profile **Homerun: New user trial signup and onboarding** (web, 24 screens, Productivity, Job Search, Communication Tools) [Refero](https://refero.design/flows/1263) > This is a happy-path new user sign-up and first-time onboarding sequence for Homerun. The user starts at the marketing home hero, chooses which product to try (ATS or HR), completes a multi-step sign-up form (social sign-in or manual name/email), provides company and contact details, verifies their email, answers a short onboarding survey (how they heard about the product, company size, current hiring setup, timeline, roles), and lands in the product where an in-app guided onboarding modal collects learning preferences. Steps: Marketing hero - product selection → Choose product to start → Sign-up form - entry → Sign-up with prefilled fields → Company details collection → Company details - filled → Email verification prompt → Onboarding - acquisition channel question → Onboarding - acquisition selection confirmation → Onboarding - company size → Onboarding - current hiring tools → Onboarding - hiring timeline → Onboarding - timeline confirmation → Onboarding - roles hiring for → In-product welcome modal → Welcome modal - selection → Guidance preference question → Guidance preference - selected → Onboarding confirmation message → Authenticated dashboard - first run state ##### Task: health record medical history **Parallel: Chat API Research Query** (web, 2 screens, AI Tool) [Refero](https://refero.design/flows/9074) > Happy-path: A signed-in desktop user runs a research query in the Chat API interface, reviews a synthesized answer with numbered source citations, and then clears the conversation to start a new query. Steps: Research answer with sources → Cleared chat / start new query **FocusFlight: History scenario filter** (ios, 4 screens, Productivity) [Refero](https://refero.design/flows/8610) > A mobile "History" sheet that lists past items (presented like ticket cards) with three top-level filter controls: Status, Scenario, and Sort By. The user opens the Scenario control (anchored popover) to select one or more scenario tags from pill-shaped chips. The popover includes a "Clear All" action and shows chips in unselected/selected states. Selecting chips updates the Scenario control to show a count (e.g., "2 scenarios") and filters the list below. Steps: History list (default) → Open Scenario popover → Select scenario chips → Apply filters and return to list ##### Task: insurance claim health **Longevity Deck: Comment interactions** (ios, 3 screens, Health & Wellness) [Refero](https://refero.design/flows/9887) > This is a product discussion section for an 'Air Purifier' product. The user sees the product hero at the top, a prominent medical-disclaimer card, and a chronological list of user comments below. Primary interactions available on this screen are: liking/unliking a comment, opening an action menu for a comment (three‑dot), and entering a new comment using the input at the bottom. Entry into this screen is from the product area (back button returns out). Exit is via the back control at top-left. Assumptions: the user is on a mobile device and appears to be authenticated (avatar shown and ability to like/report). Steps: Comments list (initial state) → Open comment actions → Reaction toggled (unlike) ##### Task: inventory warehouse **District: Save to list** (ios, 4 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12818) > User opens a save sheet for an item and can either create a new list or save it into an existing Hotlist. The sheet shows the item being saved, available lists, a note field for each selected list, and a Done button to confirm. Steps: Open save sheet → Start note entry → Enter comment → Confirm save **Adobe Photoshop: Add Adobe Stock — Search and Empty Result** (ios, 4 screens, Design & Media Editing) [Refero](https://refero.design/flows/4780) > User opens an in-app modal to add Adobe Stock assets, can either browse category tiles or open the search input to type keywords or an asset ID. The flow shown is the happy path for performing a manual search: open search, type a query/ID, submit, wait for results, then receive a 'no results' empty state. Steps: Browse Stock Categories → Enter Search Term or ID → Searching / Loading Results → No Results Found **Nudea: Stockist locator** (web, 3 screens, Marketplace) [Refero](https://refero.design/flows/7657) > Public 'Stockists' section showing an interactive store‑locator (desktop layout). Steps: Stockists landing → Search no-results → Search success and select **Artsy: Contact Gallery Inquiry** (ios, 10 screens, Marketplace) [Refero](https://refero.design/flows/7605) > Happy path: From an artwork detail screen the user taps Contact Gallery, chooses what information they want (Shipping and/or Condition & Provenance), optionally adds a location for shipping, edits or adds a custom message, and sends the inquiry. After sending the app confirms and prompts the user to add artists to their collection. Steps: Artwork Detail → Contact Composer (Initial) → View Artwork Metadata → Select Shipping (reveals location) → Add Location (sheet) → Type Location (autocomplete) → Select Suggested Location → Location Applied in Composer → Select Additional Info → Send Inquiry & Confirmation **Faire: Product Detail and Purchase** (web, 2 screens, Business, Marketplace) [Refero](https://refero.design/flows/13076) > This is a product detail page for a marketplace item. The user enters from browsing or searching and can inspect product images, choose a variant, set quantity, and add the item to cart. The page also supports saving the item to a list and exploring brand, similar, and related products. The happy path ends with the item added to cart and a confirmation state visible near the purchase controls. Assumptions: the user is on desktop, already viewing the product page, and may or may not be signed in for save/list actions. Steps: Product Review → Saved Confirmation **Nike: Select Pickup Location** (web, 2 screens, Marketplace, Sports) [Refero](https://refero.design/flows/2037) > Happy-path: A desktop shopper views a product detail page (women's pants), selects size or reviews product info, then chooses the 'Free Pickup / Find a Store' option which opens a modal to search for and select a pickup location. After selecting a store and saving, the modal closes and the chosen pickup location is applied to the product's fulfillment option. Steps: Product Details / Shipping → Select Pickup Location Modal **Loowner: Inventory search and filter** (ios, 3 screens, Marketplace) [Refero](https://refero.design/flows/6599) > Mobile inventory list with an inline search bar. Entry: user opens "All your stuff" (logged-in, mobile device). User can browse the saved items list or focus the search field to filter. Typing a query shows either filtered results or an empty-state with a "No results" message. Exit: user either taps a listed item control to proceed (detail/navigation), clears/cancels the search to return to the full list, or leaves the screen via the back arrow. Steps: Inventory list → No results (numeric ID search) → Filtered results (text search) **WishWish: Add Character to Shelf** (ios, 6 screens, Entertainment, Kids) [Refero](https://refero.design/flows/14237) > The user opens a collectible character detail view, adds the character to their personal shelf, enters shelf editing mode, and drags the character into an available shelf slot. They then adjust or confirm its placement and exit editing mode. The flow starts on the Backpacker Monan detail view and ends on the user's shelf with the character displayed. The screenshots imply an authenticated user viewing their own shelf on a mobile device; no permissions or error states are shown. Steps: Character Details → Empty Shelf → Shelf Edit Mode → Place Character → Adjust Placement → Saved Shelf **Meta AI: Artifact Search** (ios, 5 screens, AI Tool) [Refero](https://refero.design/flows/13379) > A user searches their Artifacts collection from the main library screen. They open the search field, enter a numeric query, and receive an empty-state result. They then replace the query with “gall,” which returns a matching artifact. The user is assumed to be viewing the Artifacts library on a mobile device; no authentication or permissions state is shown. **Vinyls: Edit purchase place** (ios, 4 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/8923) > Happy path for editing the 'Purchase place / store' attribute on an album's detail screen. Steps: Album details list → Purchase place editor (empty) → Enter store name → Saved confirmation and updated details **Apple Books: Add to Collection** (ios, 4 screens, Books) [Refero](https://refero.design/flows/11917) > A modal sheet opens for adding an item to a collection. The user can select from existing collections or create a new one. The flow starts with the collection picker and ends when the user either taps an existing collection or submits a newly named collection. Steps: Choose Collection → Create Collection Entry → New Collection Named → Collection Option Ready **Instagram: In-shop product search** (ios, 5 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/3663) > User starts on an All products catalog (mobile, iOS width) and opens the in-shop search bar to find specific products. Tapping the search field brings up the keyboard and a vertical list of category suggestions. As the user types a query the UI shows a transient loading / skeleton state and updates suggestions; selecting a suggestion or pressing Search shows a product grid filtered by the typed query. Entry: All products catalog. Exit: Search results grid with sort & filter available. Assumptions: user is on a logged-in shopping view (persistent bottom nav + profile avatar), on a phone with virtual keyboard. Steps: Product catalog (All products) → Search field focused — suggestions → Typing — partial query & loading state → Updated suggestions for typed query → Search results grid **Instagram: In-shop product search and empty state** (ios, 5 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/3664) > Happy path: a shopper browsing the "All products" listing opens the in-shop search, uses the keyboard to type a query (including partial or misspelled input), views real-time suggestions and loading placeholders, and either selects a suggestion or runs the search. Steps: Product listing (entry) → Suggestions shown → Typing with live feedback → Filtered suggestion → Empty search result **TikTok: Image-based product search** (ios, 11 screens, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/2776) > Happy-path visual search flow for finding shopping results from a photo. Entry: user opens the visual search camera overlay (camera + upload). The user can either take a photo or upload from the device library. The flow includes a short onboarding modal, an image selection/crop phase with detectable object hotspots, and a results sheet showing product matches (with tabs: Best match, Bestsellers, Price). Exit: user sees a scrollable product grid of matching items. Assumptions: device has camera/photo library access and network available for searching. Steps: Camera landing → Onboarding modal → Photo selection preview → Results sheet open → Scroll product results → Loading state → New results loaded → Thumbnail-focussed crop → Tap hotspot to focus → Adjust crop frame → Final product exploration **ElevenReader: Search Collections** (ios, 2 screens, AI Tool, Books) [Refero](https://refero.design/flows/8942) > User starts on a 'Recommended collections' list (home/explore area) that includes a prominent search field in the header. From this screen the user can either browse collection tiles or tap the search field to enter a query (numeric or text). In the shown scenario the user taps the search field, types a numeric identifier (likely an ISBN or ID), and runs the search. The flow ends at a 'No results found' state when the query returns no matches. Assumptions: mobile device with soft keyboard, network available, user has access to the collections and search functionality from the bottom navigation Home/Explore area. Steps: Recommended collections list → Search results — no matches **ROVR: Archive Search Refinement** (ios, 3 screens, Music Streaming) [Refero](https://refero.design/flows/13971) > The user starts in the Archive section with all archive content visible. They focus the search field, enter a numeric query, and receive a “No results found” state while the on-screen keyboard remains open. They then replace the query with “dokia,” and matching archive cards appear. The flow begins on the Archive browsing screen and ends with filtered results for the entered artist or track term. The screenshots indicate a mobile device and do not show any authentication or permission requirements. Steps: Archive Browse → Unmatched Search → Filtered Results **Vinyls: Vinyl Barcode Discovery** (ios, 3 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/8929) > Happy path: A mobile user opens the Search view (search input, popular tiles, top genres). From Search they tap the barcode icon which opens a full-screen barcode scanner with framing guides, instructional copy and a flashlight toggle. After the scanner reads the barcode, the app navigates to the album detail screen that shows the record artwork, an About card, metadata (release year, genres) and a heart action. Steps: Search → Barcode Scanner → Album Detail **TikTok: Image-based product search** (ios, 2 screens, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/2974) > Happy path: The user opens an in-app prompt that explains the image-based product search feature, taps Get started, then is taken to a camera interface where they can take a photo or upload an existing image to find similar products. Steps: Intro - Image Search Prompt → Camera Capture / Upload **GlossGenius: Add product to inventory** (web, 5 screens, Business) [Refero](https://refero.design/flows/9958) > Happy path: A logged-in user who has permission to manage inventory is trying to add a retail product so it can be used for commissions. Entry: user lands on the Commissions → Products tab and sees "No products yet." The user clicks Add product which opens the Inventory screen with the Add product drawer (right-side panel). They upload and crop an image, fill product fields (brand, product name, intended use), and save. Exit: the product appears in the Inventory list with price and quantity and a success toast confirms creation. Steps: Commissions — No products → Inventory — Add product drawer → Image upload & crop modal → Add product — fill fields → Inventory list — Product added **Fourthwall: Archive product from list** (web, 4 screens, Business) [Refero](https://refero.design/flows/7291) > Happy path for archiving a product from the All products list. Steps: All products list → Selection and action bar → Archive confirmation modal → Updated list and success toast **Instacart: Pickup store product search** (web, 2 screens, Food & Drink, Marketplace, Food Delivery) [Refero](https://refero.design/flows/10085) > Happy path: A logged-in user on desktop, viewing a completed pickup order (7‑Eleven), uses the global search bar to look for "iphone." The search input shows an autocomplete suggestion; after submitting the query the site displays a search results page that reports "No results for 'iphone'" and surfaces related items (car chargers, wall chargers, beverages) in a product grid. Steps: Search input with autocomplete → Search results — no exact matches, show related items **Shopify: Archive single product from list** (web, 6 screens, Business) [Refero](https://refero.design/flows/5396) > Happy path: A logged-in merchant navigates to the Products list, selects a single product, uses the More actions / bulk actions menu to archive it, confirms the archive in a modal, and sees the product status update with a temporary toast. Steps: Open Products list → Select product → Open More actions menu → Confirm archive → Operation in progress (toast) → Final list state **West Elm: Check Local Availability** (web, 5 screens, Home & Decor, Marketplace) [Refero](https://refero.design/flows/1677) > Happy-path: A shopper views a product detail page, opens the Find & Select Store panel to check local availability, enters a location, runs a search, and reviews nearby store results with availability and fulfillment options. Steps: Product Detail → Open Find & Select Store Panel → Enter Location → Search In Progress → Store Results & Fulfillment Options **IKEA: Gift Registry Lookup** (web, 2 screens, Marketplace, Home & Decor) [Refero](https://refero.design/flows/8269) > User lands on the IKEA "Find a gift registry" page (entry). The page provides a single, prominent input for a case-sensitive 10-character registry code and a large primary button labelled "Find gift registry". The user types or pastes a registry code and submits (via the button or search icon). On submit the UI returns either a registry result (not shown in the screenshots) or an empty state explaining no match was found (exit). Assumptions: desktop layout, user may be signed out (no account-specific content required), search requires a case-sensitive 10-character code as indicated below the input. Steps: Landing and code entry → Search result — no matches **Dopper: Store locator map** (web, 2 screens, Marketplace) [Refero](https://refero.design/flows/7524) > Single-purpose store locator map for finding local stockists. Entry: user opens the store-locator view (desktop layout). Happy path: user types a place (city or country) into the "Find a store" input, the map recenters/zooms to the searched area, user selects a location marker, and a detail card appears with address and a "Get Directions" action that launches external directions. Exit: user follows the directions link or closes the locator. Assumptions: user is not required to be authenticated; no visible geolocation permission prompt; desktop browser context (wide map, hover/click interactions). Steps: Map overview with search → Search result and store detail **Nike: Store Pickup Product Filter** (web, 6 screens, Marketplace, Sports) [Refero](https://refero.design/flows/2022) > Happy-path: a signed-in or guest desktop shopper starts on the Mens Clothing catalog, opens the filters to limit results to items available for same‑day pickup, chooses a nearby store from a 'Filter by Store' modal, and the product listing refreshes to show only items available at that store. Steps: Category Listing — Mens Clothing → Open Filters / Filter Rail Visible → Pick Up Today Option Active → Filter by Store — Modal (Search & Nearby List) → Select Store from Modal → Filtered Listing — Store Pickup Applied **Lovable: Create Storage Folder** (web, 4 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12318) > Happy path for creating a new folder inside the Storage bucket named "fine-tuna." The user starts on the bucket contents screen, opens the create-folder dialog, enters a valid folder name, and submits it. The interface validates the name field and then returns to the bucket view with the new folder listed and a success message shown. Steps: Bucket Contents → Folder Name Entry → Submit Folder → Folder Created **Cushion: Inbox Management** (web, 3 screens, Communication Tools, Collaboration Tools) [Refero](https://refero.design/flows/12510) > Logged-in workspace inbox screen with a left navigation sidebar and an inbox panel in the main area. The user can review active notifications/messages, mark them read, archive individual items, and switch between Active and Archived views. The flow begins with unread or read inbox items visible and ends when the inbox is fully processed, showing an empty-state message such as "Inbox complete," "All caught up," or "Inbox clear." Assumes the user already has access to the workspace and is viewing the inbox on desktop. Steps: Review Inbox Items → Archive Item → Inbox Empty State **IKEA: Store availability & restock alert** (web, 4 screens, Marketplace, Home & Decor) [Refero](https://refero.design/flows/8230) > Happy path: On a desktop product detail view the user opens the store availability drawer for the displayed product, browses/selects a different store from a ranked list, and signs up for a "back in stock" notification when the chosen store has no stock. Steps: Open store availability drawer → Select another store from list → View store detail and sign up form → Show success confirmation **Polaroid: Search-to-Gift-Product-List** (web, 5 screens, Marketplace) [Refero](https://refero.design/flows/9472) > Happy-path: a desktop user opens the site search, types a gift-related query, may hit a misspelling/no-results state, then refines the query and lands on a product list for gift options. Steps: Search landing / query entry → Misspelled query entered → No-results full messaging → Refined query returns products → Product list with stock and CTA **OpenAI: Find and open Vector store files API reference** (web, 4 screens, Business, Development Tools, AI Tool) [Refero](https://refero.design/flows/4185) > Happy path: A signed-in user on the OpenAI documentation landing (Docs) uses the global search modal to quickly locate the API reference for "Vector store files" and opens the detailed doc page showing endpoint details and example requests/responses. Steps: Docs overview → Search modal (empty) → Search results → Vector store files article **Programa: Add Product to Schedule** (web, 5 screens, Project Management, Business) [Refero](https://refero.design/flows/13418) > A user starts in the Product Library with a product details modal open for the Monochrome lamp. The product has images and several details already populated. The user uses the Add to control, selects a project, then selects a schedule within that project. The system confirms that the product was added to the schedule, closes the modal, and returns to the Product Library. **Fourthwall: Create product with image and size variant** (web, 12 screens, Business) [Refero](https://refero.design/flows/7290) > Happy-path product creation for a self-fulfilled physical product. Entry: merchant on the Products > All products list (logged in). Exit: new product(s) appear in the product list with default status. The user opens the Create Product modal, chooses the 'design' product type, fills product details (name, description, category, price, shipping window), uploads a product image, adds size variants (onesize), saves changes (confirms save when prompted), and returns to the product editor where they save the product. System shows a success toast and the product(s) appear in the list with status badges (Hidden/Public). Assumptions: user is signed in with permission to create products, shipping handled by merchant (self-fulfilled), product inventory set to unlimited by default. Steps: Products list (initial) → Create product type modal → New Product editor (initial) → Product category selector → Upload photo modal → Variants entry point → Add sizes form → Unsaved changes confirmation → Size entry saved in variant editor → Product editor with details and image → Save confirmation toast → Products list (updated) **Fernand: Knowledge Base Search** (web, 3 screens, Business, Communication Tools) [Refero](https://refero.design/flows/11902) > This is a workspace settings area for managing a knowledge base. The user is assumed to be logged in with access to workspace settings and is viewing the Knowledge base section. The entry point is the Articles tab with a list of collections and root articles, plus search and add controls. The user can search articles by typing in the search field, which filters the visible results in place. The exit state is the filtered article list, showing matching results or an empty-state message if nothing matches. No navigation away from the section occurs in this sequence. Steps: Knowledge Base Overview → Search with No Matches → Search with Matching Result ##### Task: investing buy stock **Wealthsimple: Enable options trading** (web, 13 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/8512) > User is on a stock's options chain but does not yet have options trading enabled. From the options chain the user is invited to start options trading. The flow guides the user through a short onboarding sequence: an initial CTA, an investment experience questionnaire, an options risk disclosure acknowledgement, a review of which accounts to enable, two-factor authentication verification, and an optional educational module. Entry: user clicks a CTA from the options chain when options are not enabled. Exit: the user returns to the options chain with options trading enabled (UI now displays strategy builder and active buy/sell controls). Assumptions: the user is authenticated (signed in) and has at least one tradable account; the user has a working authenticator app for 2FA. Steps: Options chain (disabled state) → Prompt to explore options → Start options trading CTA → Investment experience (initial) → Investment experience (completed) → Options risk disclosure (read) → Options risk disclosure (acknowledged) → Review account setup → Re-check risk acknowledgement (if revisited) → Two-factor verification modal (empty) → Two-factor verification modal (code entered) → Optional education prompt → Options chain (enabled state) **Stocktwits: Poll Reply with Price Prediction and Media** (web, 2 screens, Finance) [Refero](https://refero.design/flows/7214) > Happy path: A logged-in desktop user opens a poll/conversation modal and composes a reply. From the composer they optionally create a price prediction (symbol, target price, date, sentiment) and attach a media file. Steps: Compose reply with price prediction → Attach media and confirm upload ##### Task: learning management course progress **ALO: Class Detail and Playback** (web, 4 screens, Fitness, Lifestyle) [Refero](https://refero.design/flows/12371) > Logged-in class detail page for a yoga/video workout. Entry is the class landing page with a large video player at the top, class metadata, related content, and comments. The user can play the class, adjust playback or quality, view class details, browse related classes, and read or add comments. The flow ends when the user stays on the class page, starts or continues the video, or interacts with supporting content such as comments or recommendations. Steps: Class Overview → Playback Settings → Video Playing → Class Page with Playback **ALO: Personalized Fitness Dashboard** (web, 3 screens, Fitness, Lifestyle) [Refero](https://refero.design/flows/12369) > Happy path for a logged-in user on a fitness/training dashboard. Entry is the home dashboard with preferences and progress widgets visible, plus an onboarding-style tooltip system. The user can dismiss or advance tips, then browse recommended classes from the personalized section. Exit is not shown, but the likely endpoint within this view is opening a class, browsing categories, or saving content. Steps: Dashboard Overview → Preferences Guidance → Personal Picks **Imprint: Lesson: Read, Quiz, Reward** (ios, 25 screens, Education) [Refero](https://refero.design/flows/5644) > Happy-path: Learner opens a course module, taps a lesson card, reads a few content slides and diagrams, completes several knowledge-check questions (multiple choice and true/false), receives immediate feedback (correct/incorrect) with XP rewards for correct answers, reviews an explanatory follow-up when they miss questions, then sees a lesson summary, XP/score screen, and badge-progress screen. Steps: Course map / Lesson card → Unit cover → Content slide — intro / image → Knowledge-check — multiple choice (initial) → Select answer — MC → Submit MC answer → Continue after correct → Knowledge-check — true/false (initial) → Submit true/false answer → Continue after incorrect → Lesson content — key points → Knowledge-check — short MC (rhetoric) → Submit MC answer (rhetoric) → Next knowledge-check — priorities question → Submit priorities answer → Follow-up comparison question → Submit follow-up answer → Lesson summary & key insight → XP summary → Badge progress / completion **Brilliant: Course lesson — read, answer, complete** (ios, 29 screens, Education) [Refero](https://refero.design/flows/4333) > Happy-path: a signed-in learner opens the Courses tab, selects the Science category, opens the Electricity and Magnetism course and starts the first lesson (“Unlocking Electricity”). The learner reads a series of content screens (text + illustrations), answers inline multiple-choice questions, watches an embedded demo video, and finishes the lesson. Steps: Category selection → Course map overview → Lesson cover / title → Content screen — intro → Content screen — context → Content screen — magnet basics → Interactive question — orientation → Question explanation → Interactive question — coil experiment → Answer feedback & XP → Explanatory content — Faraday → Application illustration → Video thumbnail → Video playback → Final content and Finish CTA → Lesson complete — animation → Lesson complete — XP summary → Return to course state (implicit) → Progress bar behaviour → Inline help / 'Why?' affordance **Brilliant: Exploring Data Visually: Pie Charts** (ios, 20 screens, Education) [Refero](https://refero.design/flows/4325) > Happy-path: From the course map the user opens the "Pie Charts" lesson, reads two short instructional slides (visual examples of pie vs bar charts), then completes a sequence of interactive practice items. The practice items include a numeric multiple-choice question (sum of bars), an interactive pie-chart selection (tap slices to select categories), and multiple-choice percent questions about pie slices. The user receives immediate feedback after each question (correct/incorrect), can retry incorrect items, and receives XP and progress updates on success. Steps: Course map overview → Lesson modal / start → Instructional slide — introduction → Practice item — bar chart multiple-choice → Submit MCQ — bar chart → Explanation slide — pie percentages → Interactive question — select slices → Submit — incorrect selection → Retry interactive selection → Submit — correct selection → Content slide — display toggle → MCQ — percent identification (single-choice) → Submit MCQ — incorrect feedback → Retry MCQ — choose again → MCQ — combined percent question → Submit combined percent — feedback → Interactive pie — select different slices → Final feedback modal → Lesson summary / reinforcement → Return to course map **Brilliant: Analyzing Bar Charts lesson** (ios, 38 screens, Education) [Refero](https://refero.design/flows/4323) > Happy-path for a short mobile lesson (data-analysis) starting from Home. User taps a lesson card -> brief animated transition screens -> lesson intro -> a sequence of 3 multiple-choice questions that use bar charts -> immediate feedback after each question (correct/incorrect) with XP and continue actions -> final success screen and streak / streak-charge screens. Steps: Home — Lesson Card → Animated transition — 1 → Animated transition — 2 → Animated transition — 3 → Lesson Intro → Question 1 — Identify largest state → Question 1 — Check answer → Question 2 — Difference estimate → Question 2 — Select and check → Question 3 — Identify highest drink → Question 3 — Incorrect submission → Question 3 — Follow-up question → Question 3 — Correct submission → Lesson Completion — Nice work → XP Summary → Streak Started → Streak Charge Filled → Optional info — What are Streak Charges? → Return CTA → End state / Home **Imprint: Course Download and Progress** (ios, 6 screens, Education) [Refero](https://refero.design/flows/5641) > Happy path: A signed-in user viewing a course roadmap opens the course options sheet, chooses Download, watches an in-sheet download progress indicator, and then sees the options row change to a completed "Downloaded" state. Steps: Course roadmap → Options sheet (idle) → Downloading — 33% → Downloading — 70% → Downloading — 92% → Download complete **Imprint: First Lesson Walkthrough** (ios, 35 screens, Education) [Refero](https://refero.design/flows/5636) > Happy-path walkthrough for a new user's first lesson onboarding and micro-lesson experience. Steps: Intro splash → Course selection list → Confirm selection → Gesture hint / navigation → Course cover card → Unit cover → Lesson title / key idea → Instructor and content slides → Illustration / visual slide → True/False question → Immediate feedback overlay → XP update (in-session) → Lesson completion XP summary → Start streak prompt → Streak confirmation → Daily challenge card → Badge progress and summary → Floating action controls → Top progress indicator → End-of-walkthrough return **Imprint: Complete Unit: Why We Sleep** (ios, 15 screens, Education) [Refero](https://refero.design/flows/5643) > Happy path: A logged-in user opens a course unit (Unit 1: Why We Sleep) from the home map and taps Start. They swipe/scroll through a sequence of informational slides (text + illustrations + light interactions). After reading the content the user reaches an end-of-lesson summary that shows lesson XP, awards a badge, and prompts to continue to Unit 2. Steps: Unit card / Start → Intro slide — attention hook → Statistic slide — prevalence → Cost overview — introduction → Cost breakdown — systemic vs individual → Comparison — car accidents → Health impacts diagram → Perception panel — noticing sleep → Author context — Why We Sleep → Evolutionary argument → Lesson wrap-up text → Lesson XP summary → Badge earned → Unit 2 teaser → Map/Home updated — next unit unlocked **Brilliant: Calculus lesson: Rate of change** (ios, 33 screens, Education) [Refero](https://refero.design/flows/4332) > A learner browses the Courses tab, opens the Calculus unit inside Foundational Math, and starts a short interactive lesson about rate of change. The lesson alternates short explanations, an interactive visualization (zoom / microscope), multiple-choice conceptual checks, and numeric input problems (math keyboard). The flow ends with validation feedback, XP reward, and a completion summary. Steps: Category selector / Course list → Course detail — Calculus → Lesson intro card → Concept explanation screen → Model introduction (v(t) = 50t) → Numeric input focus → First numeric attempt → Incorrect feedback state → Function microscope intro → Zoom interaction → Secant approximation multiple choice → Secant check and validation → Numeric slope-fit task intro → Iterative numeric attempts → Validation result — incorrect then retry → Correct submission → Finish question and claim XP → Lesson completion summary → Return to course or continue → State after completion **Adobe Photoshop: Learning catalog** (ios, 5 screens, Design & Media Editing) [Refero](https://refero.design/flows/4745) > Happy path: user opens the Learn sheet (a rounded, draggable modal) from the editor and browses tutorial categories. They tap the "Must know" category, which opens a full-height, vertically scrollable list of short lessons. Each lesson has a hero image, title, short description and a primary "Watch quick video" CTA. The user scrolls through modules (Meet Photoshop, Layers, Masks, Selections) and can either watch a video or continue browsing. Steps: Browse categories → Category landing — Meet Photoshop → Lesson card — Layers → Lesson card — Masks → Lesson card — Selections **(Not Boring) Habits: Meditation session start and lesson reveal** (ios, 5 screens, Health & Wellness) [Refero](https://refero.design/flows/5061) > Happy-path: A mobile user selects a scheduled meditation from a horizontal date-strip, taps the large circular session tile to start, watches an animated intro, receives a completion confirmation, and then sees the lesson/scene content with a 3D object and a progress counter. Steps: Session Selection → Session Start — Intro Animation → Completion Confirmation → Lesson Overview → Lesson Detail — Expanded 3D Scene **Brilliant: Estimating Probabilities Lesson** (ios, 13 screens, Education) [Refero](https://refero.design/flows/4328) > Mobile micro-lesson that teaches basic probability with interactive exercises. Entry: user taps the "Estimating Probabilities" card from the Home carousel (user is signed in on mobile). The lesson shows an illustrated intro screen, then a sequence of interactive questions: draggable probability bar(s), immediate validation (Check), inline feedback panels (yellow for hint, green for correct) and short multiple-choice questions. Top progress indicator updates across sub-steps; a lightning/streak icon is visible throughout. Exit: user finishes the mini-lesson and receives feedback/XP or a final summary before returning to the course carousel. Steps: Lesson discovery → Lesson intro → Scenario + probability bar (default) → Adjust estimate → Incorrect attempt feedback → New information scenario (bartender) → Concept check (MCQ) → Transition to new topic → Card probability (initial try) → Conditional hint and follow-up → Final conditional question → Lesson end / final feedback → Progress & reward confirmation **Brilliant: Password Strength Lesson** (ios, 53 screens, Education) [Refero](https://refero.design/flows/4335) > User browses the Courses tab, selects the Technology category, opens the 'How Technology Works' course and starts the lesson 'How Strong Passwords Work'. The lesson is a linear micro-lesson with reading slides, an interactive password generator (shuffle icon) that updates a displayed password and a strength/time-to-guess meter, a slider illustrating length effects, a multiple-choice quiz, a free-response question (numeric input), inline explanations and feedback, then a final 'Finish lesson' summary with XP awarded. Steps: Courses → Technology category → Open course → Start lesson → Intro slide → Scenario: create account → Interactive generator (first) → Generator iterates → Generator branding card → Password strength chart → Multiple-choice quiz → MCQ feedback → Graph explanation → Free-response question → Answer explanation choices → Comparative example question → Final content → Lesson complete → Course progress updated → Return to Courses tab → Persistent state & analytics **Pinkllama: Micro-article: Cup talk** (ios, 7 screens, Health & Wellness) [Refero](https://refero.design/flows/5657) > A short, illustrated micro-article experience surfaced under the 'For You' section. Entry: user taps a recommended content card from the For You feed. The card opens a multi-step illustrated article (carousel/slide style) with a top progress indicator and a Leave option. The user navigates through 3 content slides then reaches a completion screen offering a primary CTA to share and a secondary Done action. Exit: after Done (or closing the share modal) the card is marked complete in the feed and the daily progress counter increments. Assumptions: user is logged in, on a mobile device, and viewing the personalized For You feed. No special permissions are required for reading content; sharing uses the device's native share sheet. Steps: For You feed — content card → Article — title + hero → Article slide — 'Going with the flow' → Article slide — cleaning tip → Completion screen — celebration → Share modal / native share sheet → Feed — card marked complete **Canva: Canva lesson completion** (web, 7 screens, Design & Media Editing, Collaboration Tools) [Refero](https://refero.design/flows/8067) > Happy-path: A logged-in user discovers a Design School lesson from the learning home, opens the lesson details, steps through three lesson items (video, cheatsheet, video) using the in-lesson player and sidebar outline, finishes the lesson, and returns to the lesson overview which shows completion state. Steps: Design School Home → Lesson Overview → Item 1 — Play Video → Item 2 — Cheatsheet → Item 3 — Play Video → Completion Confirmation → Completed Lesson Overview **Brilliant: Answer question and claim reward** (web, 5 screens, Education) [Refero](https://refero.design/flows/4358) > Happy-path: A learner encounters a multiple-choice question inside a lesson, selects an answer, submits it, and is shown a lesson-complete reward (XP). If the XP causes a level change, a level-up confirmation is displayed next. Steps: Question prompt → Select answer → Submit answer → Lesson complete acknowledgement → Level-up acknowledgement **Duolingo: Beginner Lesson Completion** (web, 34 screens, Education) [Refero](https://refero.design/flows/11811) > A logged-in learner starts a beginner language lesson from the course map. They open the first available lesson, complete a sequence of vocabulary recognition, translation, and matching exercises, receive immediate correctness feedback, and revisit a previous mistake. After finishing, they review the lesson score, view their streak update, and return to the course map with progress and rewards updated. Steps: Course Map → Lesson Preview → Identify Woman → Submit Woman Answer → Identify Man → Submit Man Answer → Translate Sentence → Choose Word Alternative → Match Vocabulary Pairs → Complete Matching Exercise → Review Previous Mistake → Submit Review Answer → Lesson Results → Scorecard Review → Streak Update → Updated Course Map → Course Progress Review → Reward State Confirmation → Next Section Preview → Lesson Exit **Brilliant: Reading Bar Charts Lesson** (web, 25 screens, Education) [Refero](https://refero.design/flows/4356) > Happy-path for a single lesson inside a course titled "Exploring Data Visually". Entry: user opens the course and selects the "Reading Bar Charts" lesson. The lesson contains an intro slide, a linear sequence of interactive bar-chart questions where the user selects bars and uses a Check button, immediate feedback (correct/incorrect), optional "Why?/Explanation" content in a modal, and a final "Lesson complete" screen. Exit: user is returned to the course map where a practice node is presented. Steps: Course map overview → Course detail with lesson callout → Lesson intro slide → Question card - first chart → Bar selected and flagged → Second question card → Third question card with multiple bars → Submit answer - correct feedback → Open explanation modal (correct) → New question set - different dataset → Select and submit answer → Correct feedback and continue → Explanation modal for drinks question → Next question - numeric estimation → Submit answer - incorrect feedback → Explanation modal after incorrect answer → Lesson complete screen → Course map with practice node → Progress indicator states → Check button enabled/disabled behavior **Brilliant: Lesson: Connecting Gears** (web, 28 screens, Education) [Refero](https://refero.design/flows/4350) > Happy path: A logged-in user on the Brilliant dashboard opens a lesson card, reads a short concept intro, completes a sequence of interactive questions (single-choice and multi-select by tapping gears), receives immediate correctness feedback and an explanation, and finishes the lesson to receive XP and streak prompts. Steps: Dashboard / Lesson Card → Lesson Intro → Concept Explanation (scrollable) → First Question (single-choice) → Submit Answer → Correct Feedback & CTA → Reveal Explanation → Worked Example and Transition → Interactive Multi-select (tap gears) → Submit Multi-select → Repeat Questions (progressing) → Final Multi-gear Challenge → Lesson Completion XP Modal → Streak Reward / Onboarding → Streak Callout (congrats) → Choose Streak Goal → Commit to Streak → Course Map / Lesson Map → Optional Return to Course → End state (persisted progress) **Brilliant: Complete single lesson and claim rewards** (web, 6 screens, Education) [Refero](https://refero.design/flows/4354) > Happy path for completing a single lesson within a larger course. Steps: Course lesson selection → Lesson prompt (initial) → Select answers → Answer feedback → Lesson completion summary → Streak reward confirmation **Duolingo: Language Learning Onboarding** (web, 42 screens, Education) [Refero](https://refero.design/flows/11808) > A new visitor starts on Duolingo’s marketing landing screen and selects Get Started. They choose a language, review introductory messages, answer onboarding questions about discovery source, motivation, and existing knowledge, then receive a daily-quest reward. The system asks them to create a profile to save progress. The user enters an age, provides profile credentials, and submits the account form. After account creation, they arrive at the learning dashboard, where an email-promotion modal appears over the first course path. The flow assumes the user is logged out and is using a desktop browser. The selected course shown in the onboarding questions is Esperanto. Steps: Duolingo Landing → Language Selection → Onboarding Introduction → Onboarding Welcome → Onboarding Transition → Discovery Source → Discovery Source Confirmed → Learning Motivation → Motivation Confirmed → Proficiency Level → Daily Quest Complete → Gem Reward → Profile Prompt → Age Entry → Age Confirmed → Create Profile Form → Completed Profile Form → Account Creation Processing → Email Promotions Prompt → Learning Path Dashboard **Duolingo: Leaderboard Status Setup** (web, 2 screens, Education) [Refero](https://refero.design/flows/11820) > This is a Duolingo Leaderboards screen where the user can either start a lesson to join the weekly leaderboard or set a visible status from a small status picker panel. Steps: Leaderboard Overview → Status Selected **Brilliant: Lesson: operator-puzzle completion** (web, 7 screens, Education) [Refero](https://refero.design/flows/4355) > Happy path: A learner opens the '100 Days of Puzzles' course, selects a lesson node ('Make It Equal 5'), completes the interactive operator puzzle by choosing operators and submitting an answer, views feedback (incorrect -> try again -> reveal correct solution), finishes the lesson and receives XP, then returns to the course map with the lesson marked as completed. Steps: Course map / lesson selection → Exercise prompt (initial state) → User attempts answer (Check enabled) → Incorrect feedback with actions → Correct answer revealed / explanation options → Lesson complete / XP awarded → Course map updated (post-completion) ##### Task: medication reminder **Instagram: Set event reminder from tagged photo** (ios, 3 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/2865) > Happy path: A logged-in mobile user viewing a tagged photo discovers an event call-to-action (Remind me), taps it, reviews event details in a bottom sheet, enables a reminder, and receives a confirmation toast. Steps: Tagged post — Remind CTA → Event details modal → Reminder confirmation **New Balance: Coming Soon Alert Signup** (web, 3 screens, Marketplace, Sports) [Refero](https://refero.design/flows/11853) > Happy path for a product detail page where the item is not yet available. Entry is the product page for ABZORB 1890 Grey Days Shoes. The user can review product imagery and details, then use the launch alert signup to be notified when the item becomes available. Steps: Product Detail Page → Launch Alert Form → Subscription Confirmation ##### Task: meeting notes transcription **Granola: Search Notes and Meetings** (ios, 4 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/14168) > A signed-in user starts on the notes home screen and opens the search tool. The search field is focused automatically, allowing the user to enter a query. The interface displays a no-results state for a numeric query, then updates to show matching previous meetings when the query is changed to “meeting.” The flow begins on the notes home screen and ends with matching search results displayed. It assumes a mobile device, an authenticated user, and access to the visible notes and meetings. Steps: Notes Home → Search Entry → No Results → Matching Results **Wispr Flow: AI Note Summary** (ios, 3 screens, AI Tool, Utility, Productivity) [Refero](https://refero.design/flows/13679) > The user is viewing the Scratchpad notes list and wants to understand the content of a saved note. From the Scratchpad screen, they select the note’s AI action, wait while the system generates a summary, and review the completed summary in a bottom sheet. The user can close the sheet or regenerate the summary. Steps: Scratchpad Notes → Summary Generation → Summary Review ##### Task: open bank account identity verification **Dropbox: Vault setup and biometric enablement** (ios, 9 screens, Business, Collaboration Tools, Productivity) [Refero](https://refero.design/flows/2467) > Happy path for setting up a protected “Vault” inside an existing account. Entry: user taps the Vault feature (onboarding card or account feature). Flow creates a vault, prompts the user to set a numeric PIN, then offers to enable Face ID for quick unlocking. Exit: user lands in the empty Vault with action tiles (Scan document, Upload photos, Create folder, Take photos). Steps: Vault introduction → Creating vault (progress) → Set PIN (empty) → Set PIN (partial) → Set PIN (multiple digits entered) → In-app Face ID offer → System Face ID permission dialog → Face ID authentication (animated) → Vault landing (empty state) ##### Task: pay invoice bill **Instagram: Order lifecycle from chat** (ios, 5 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/2892) > Happy path: a seller sees an order card inside a 1:1 conversation, opens the Order Details screen, marks the order as paid, and the conversation is updated with an "Order paid" card. The seller can then open the updated Order Details view which now shows the Paid state and exposes the next action (Mark as shipped). Steps: Conversation – Order created card → Order Details – Waiting for payment → Order Details – Marking as paid (loading) → Conversation – Order paid card → Order Details – Paid (Next: Mark as shipped) ##### Task: payroll **Wittl: Create job listing** (web, 35 screens, Job Search) [Refero](https://refero.design/flows/5364) > Happy-path: A logged-in recruiter starts from their inbox/dashboard and opens a modal wizard to create a new job listing. The wizard guides them through: initial job metadata (title, location, type, workplace), composing the job description with reusable sections (text/images/title blocks), building the candidate application form (fields and sections), defining hiring stages, and finally choosing whether to publish or keep the job as a draft. Entry: recruiter dashboard -> Create new job. Exit: job is created and the user is taken to the job details screen where the job status and editable fields are shown. Assumptions: user is authenticated with permission to create/edit jobs; desktop/laptop viewport (modal centered, wide editor area); a default job template and default hiring stages exist. Steps: Dashboard / Inbox → Get started — Metadata → Select job type → Select workplace → Confirm metadata and continue → Job posting — Preview / Editor → Job posting — Scroll / Edit → Insert section panel → Choose section type → Add image placeholder → Application form — default sections → Insert field panel → Select personal/contact field → Application form — field added → Insert section templates for form → Questionnaire section → Continue to Hiring stages → Hiring stages → Publish — choose visibility → Job details / Edit job ##### Task: policy management approval **Mercury: Payment approval review** (web, 3 screens, Business, Finance) [Refero](https://refero.design/flows/11062) > Happy path: A signed-in user with approval permissions opens the Payments area, filters to the Needs Approval tab, selects a payment row to view details in a right-side details panel, then declines the request. Steps: Needs Approval list → Open payment details drawer → Decline payment and confirm **Yelp: Report a business post** (web, 3 screens, Food & Drink) [Refero](https://refero.design/flows/1783) > Happy path: User views a business photo inside a lightbox/preview and uses the inline report control to flag the post. The UI shows a confirmation dialog asking the user to confirm the report; after confirming, the UI displays a submission acknowledgement. Steps: View photo with post details → Confirm report → Acknowledge submission **YouTube: Shorts comment moderation** (ios, 4 screens, Entertainment, Video Streaming, Social Networking) [Refero](https://refero.design/flows/2606) > Happy-path for setting comment preferences while adding details to a YouTube Short. Steps: Add details (Short upload) → Comments settings list → Confirmation modal — Allow all comments → Explanatory dialog — Hold-for-review details **Apple Invites: Event Guest Management** (ios, 4 screens, Scheduling Tools) [Refero](https://refero.design/flows/4043) > Happy path: A host opens an upcoming event, opens the guest management / invite sheet, chooses or shares an invite link (public or one-time), reviews a pending access request, and accepts it so the requester becomes an attendee. Steps: Event Card — Access Requests → Invite & Guest Management Panel (Initial) → Accepting a Pending Request → Guest List Updated — Confirmed Attendee **Pinterest: Report a Pin for Policy Violations** (ios, 3 screens, Reference, Inspiration Board, Social Networking, Lifestyle) [Refero](https://refero.design/flows/2475) > Happy path: A logged-in mobile user opens the report flow from a pin, selects a reason from a list of categories, confirms the report on a single-confirmation screen, and is returned to the home feed with an acknowledgement toast and the reported card marked. Steps: Choose report category → Confirm report → Acknowledgement and feed update **Aboard: Employee Time-off Registration** (web, 7 screens, Collaboration Tools, Paperwork) [Refero](https://refero.design/flows/766) > Happy-path: an admin/manager on an employee profile registers a multi-day vacation for that employee using a modal. Steps: Profile — Time-off overview → Open Balance card actions → Open Register time-off modal (defaults) → Open calendar date picker → Select date range and review summary → Submit request → Updated Time-off overview (confirmation) **Aboard: Bulk Assign Absence Policy** (web, 6 screens, Collaboration Tools, Paperwork) [Refero](https://refero.design/flows/769) > Happy path: an authenticated user (with permission to edit employee profiles) opens a filtered employee list, selects multiple employees, uses the Actions menu to choose “Absence policy”, configures the policy in a modal (type + policy), submits the change, and sees the employee table update to reflect the new policy. Steps: Select employees → Open Actions menu → Open Absence policy dialog → Choose policy type → Select policy and submit → Confirm table update **Navan: Group Travel Event Setup** (web, 20 screens, Travel & Booking) [Refero](https://refero.design/flows/9451) > Happy path: a logged-in travel manager creates a group travel event, fills required basics (name, type, location, start/end dates and times), saves the draft, configures booking & policy options, adds participants, optionally searches for a hotel room block, then sends booking invites. Entry: Group Travel landing page. Exit: Event status changes from Draft to In progress and invites are sent. Assumptions: user is authenticated with organizer permissions, using a desktop browser, and has access to company policy/payment options and participant directory. Steps: Group travel landing → Enter event basics → Select event type → Open date picker → Select dates and times → Form ready to save → Event draft overview (Booking settings) → Open edit event modal → Modify location → Save event changes → Open Policy & payment tab → Open Participants tab (empty state) → Add participants modal (empty) → Select participants → Confirm participants → Review participants list → Open Accommodations tab → Send invites confirmation → Send invites (in progress) → Invites sent; event in progress **YouTube Music: Personal Music Upload** (web, 6 screens, Music Streaming, Entertainment) [Refero](https://refero.design/flows/11252) > Happy path for a signed-in user to add a local audio file to their personal music collection. Steps: Start upload from account menu → Accept one-time policy → Upload in progress → Upload completed — processing notice → Uploads tab — empty/processing state → Uploaded track appears in list **TravelPerk: Create and Apply Travel Policy** (web, 5 screens, Productivity, Travel & Booking) [Refero](https://refero.design/flows/5253) > Happy path for creating and applying a new travel policy from a recommended template. Entry: user is on the Travel Policy management area (Travel management > Travel policies). The user selects a template (e.g., Standard policy), names the new policy, verifies the action via an email code, and receives confirmation. Exit: modal dialogs close and the newly created policy appears in the Active policies list with a success banner/toast. Assumptions: user is authenticated and has permission to manage travel policies; using a desktop browser with enough screen space for lists and modals. Steps: Choose template → Name your new policy → Request verification code → Enter verification code → Creation confirmation and policy view **Hello Ivy: Asset Approval** (web, 3 screens, Productivity, Project Management, Collaboration Tools) [Refero](https://refero.design/flows/4526) > Happy path: A reviewer opens an asset in a viewer, inspects the image and comments area, opens the "Set Status" control, and marks the asset as "Approved." Entry point: reviewer has an asset open in the viewer (top bar shows filename and version). Exit point: status label updates to show "Approved" and the dropdown closes; the viewer and comments area remain visible. Assumptions: reviewer is signed in and has permission to set review status, the asset and its version are already loaded, and the interface is viewed on a desktop or wide viewport showing both the asset and the comments pane. Steps: Open Asset Viewer → Open Status Dropdown → Select Approved Status **Mercury: Bulk Payment Approval** (web, 4 screens, Business, Finance) [Refero](https://refero.design/flows/11063) > Happy path: A user with approval privileges opens the Payments area on desktop, navigates to the 'Needs Approval' tab, selects multiple pending payment requests, clicks to review, confirms approval in a modal, and receives a success confirmation while the list updates to reflect the change. Steps: Needs Approval list → Select requests → Confirm approval modal → Approval result **Aboard: Time-off Request Approval** (web, 3 screens, Collaboration Tools, Paperwork) [Refero](https://refero.design/flows/771) > Summary (happy path): A manager opens the Time-off module to view the team absence timeline, clicks a time-off bar or user indicator on the calendar, and a right-hand drawer opens with that user's requests. The drawer separates pending Requests (with Approve / Decline controls) from previously Approved items. The manager makes a decision on a pending request and then closes the drawer; the calendar reflects the updated state. Steps: Calendar overview → User request drawer (review) → Decide on request **Aboard: Add time-off policy** (web, 7 screens, Collaboration Tools, Paperwork) [Refero](https://refero.design/flows/765) > Happy path: A manager/admin on an employee’s profile (Time-off tab) adds an existing leave policy (e.g., Vacation ▶ Global) to that employee. Steps: Time-off tab — employee profile → Add policy modal opened → Select leave type → Open policy variant list → Confirm selection in modal → Policy added — Balance updated → Final confirmation view ##### Task: prescription refill repeat **On: Order Details Review** (web, 2 screens, Marketplace, Fitness, Lifestyle) [Refero](https://refero.design/flows/13833) > The flow begins on a logged-in customer’s account Orders section, displaying the details of a placed order. The user reviews the order number, delivery estimate, item, status timeline, shipping address, payment summary, and available cancellation action. In the shown interaction, the user selects the copy icon beside the order number. The system copies the order number and confirms the action with a toast notification. The page also includes recommended products and the site footer below the order information. Steps: Order Details → Order Number Copied **ElevenMusic: Track Published Confirmation** (web, 2 screens, Music Streaming, Entertainment, AI Tool) [Refero](https://refero.design/flows/12588) > A publishing confirmation modal appears after a music track is successfully published. The modal shows the track artwork, publishing label, track title, and a success message indicating the track was added to an album. The user can copy a shareable track link, dismiss the modal with Done, or continue to the album with View Album. Assumptions: the user is already signed in and has just completed a publish action on a desktop web experience. Steps: Publish Success → Copy Link Confirmed ##### Task: procurement tender bid **Pool: Refero MCP Search** (ios, 2 screens, Utility) [Refero](https://refero.design/flows/11986) > Happy path starts from a preview/card view of a Refero MCP result shown in a mobile viewer. The user can either dismiss the preview or tap the primary call to action to explore the MCP content. Selecting the CTA opens a Google search results page for the query “Refero MCP for AI agents,” showing related results including the Refero site and a GitHub result. Assumptions: the user is on a mobile device, has network access, and is not required to be signed in to view the results. Steps: Preview and Explore → Search Results **Artsy: Place max bid and pay** (ios, 18 screens, Marketplace) [Refero](https://refero.design/flows/7608) > Happy-path intent: A signed-in mobile user discovers an auction lot, opens the lot detail and places a max bid. They choose a bid amount from a predefined list, add a credit card (address + country), accept terms, and confirm the bid. Steps: Auction list / card → Lot detail → Place a max bid (entry) → Max bid options → Place a max bid (selected) → Confirm your bid (initial) → Add Credit Card (form) → Add Credit Card (partial data) → Country selector (list) → Add Credit Card (country selected) → Confirm your bid (card added) → Confirm your bid (summary) → Bid result (error modal) → Post-bid landing → Retry / edit payment → Retry submission (optional) → Final success (alternate end) → Final failure (alternate end) **Artsy: Artist auction results search** (ios, 3 screens, Marketplace) [Refero](https://refero.design/flows/7604) > Happy path: From an artist profile the user opens the Auction Results tab, inspects past auction entries (thumbnails + metadata), uses the search field to filter results by artwork title or series, and views either matching results or a clear "no results" message. Steps: Auction Results list → Search returns no matches → Search returns filtered results **Artsy: Artist auction lookup** (ios, 6 screens, Marketplace) [Refero](https://refero.design/flows/7603) > Happy-path: A user lands on an artist's profile and finds there are no auction results shown. They tap the call-to-action to open the Artsy Price Database, type an artist name (using autocomplete), choose the matching artist from the suggestions, and are returned to the artist profile where auction results are displayed. Steps: Artist page — empty auction state → Price Database — search entry → Search — no-match result → Search — autocomplete suggestions → Search — select suggestion → Artist page — auction results populated **Artsy: Add payment method during bid** (ios, 8 screens, Marketplace) [Refero](https://refero.design/flows/7609) > Happy path: a logged-in mobile user attempts to place a bid but has no saved payment method, so they are routed to an in-flow "Add Credit Card" experience, enter card + billing details, choose country using a searchable list, save the card, and return to the Confirm Bid screen with the new card selected. Steps: Confirm bid — initial (no card) → Add Credit Card — card number entry → Add Credit Card — card + name + address start → Add Credit Card — complete address fields → Country selector — full list → Country selector — filtered results → Add Credit Card — review & save → Confirm bid — card added **Kickstarter: Search and Discover Projects** (ios, 2 screens, Fundraising) [Refero](https://refero.design/flows/4635) > Happy-path: A mobile user is on the app's Search section and discovers projects. The entry point is the Search tab (search field visible, bottom navigation present). The user either browses the pre-populated "Popular Projects" list or types a query to filter results. Each result card shows a hero image, title, a green badge with remaining days, a progress bar, and a numeric funding percent. The exit point is tapping a project card to open its detail view or tapping Cancel to clear the query and return to the previous list. Assumptions: the user is on a mobile device in the Search section; no special permissions are required to browse or search; search operates on the visible project list and returns cards matching the query; tapping a card navigates to that project's detail screen (not shown). Steps: Popular Projects listing (search idle) → Search results for query **Introspect: Purchase Cooldown** (ios, 12 screens, Finance) [Refero](https://refero.design/flows/11953) > The user starts in the Wants section with an existing Sony WH-1000XM6 item and adds another desired purchase. They enter the item name “LEGO Icons Porsche 911,” provide a price of $169.97, and start an analysis. The system presents reflective reasoning cards, analyzes the answers, and generates a receipt-style recommendation with a three-day cooldown. After the user continues and acknowledges the cooldown, the item is added to the waitlist and appears in the Wants list alongside the existing item. The user is assumed to be on a mobile device and already inside the product; no authentication or permission steps are shown. Steps: Wants List → Product Name Entry → Product Name Confirmed → Price Entry → Price Confirmed → Reflection Prompt One → Reflection Prompt Two → Analysis Processing → Analysis Recommendation → Waitlist Confirmation → Item Overview → Updated Wants List **Programa: Purchase Order Creation and Sending** (web, 17 screens, Project Management, Business) [Refero](https://refero.design/flows/13438) > A user in Project #254 creates and sends a purchase order. The user opens Purchase Orders, starts a new order, selects schedule #1, chooses a supplier, enters issue and delivery dates, adds five products, reviews the generated order, enters or confirms recipient and delivery details, and sends the order. The system creates purchase order P000001, marks it as sent, and returns to the Purchase Orders list where the order is displayed. The user is assumed to be authenticated and authorized to manage purchase orders. **Steel: Invited Member Search** (web, 3 screens, AI Tool, Development Tools) [Refero](https://refero.design/flows/14297) > The user starts in the organization settings area on the Invited members screen. They review pending invitations, enter a numeric search query to filter the list and see that no invitation matches, then replace the query with a name-based search to find a specific invited member. The flow begins and ends on the invited-members list. It assumes the user is authenticated, has permission to manage organization members, and is using a desktop layout. **Exa: AI Company Research** (web, 13 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/12019) > An authenticated user configures an Agent to find AI infrastructure companies hiring founding designers. The user enters a research query, adds a JSON data import containing company or candidate information, generates an object output schema, and runs the Agent. The system displays run progress, usage details, search results, sources, and a final table of company findings. The final run view shows an Agent request failure even though results and source information are present. Steps: Agent Setup → Open Input Data → Enter JSON Records → Save Imported Data → Define Output Fields → Generate Output Schema → Review Generated Schema → Confirm Output Definition → Start Research Run → Monitor Run Details → Review Returned Companies → Inspect Additional Results → Review Sources and Cost **Cohere: Document summary and calculation** (web, 5 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3835) > Happy path: A signed-in user opens a conversation titled "Business Plan Outline for a Marketing Agency in French," asks the assistant to summarize document content and later uses an integrated tool (Calculator) to solve a numeric problem found in the document. Steps: Conversation — Document summary displayed → Chats panel visible / Data-source option → Open tools menu → Select Calculator and attach problem → Assistant returns calculation and tool reference **Leonardo.Ai: Blueprint Template Discovery** (web, 3 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/12897) > A logged-in user opens the Leonardo Blueprints library and browses ready-made AI templates. The entry point is the main Blueprints explore view with a search bar, category chips, and curated template sections. The user can either scroll through featured collections or use search to narrow results. In the shown happy path, typing a query updates the results to matching blueprint cards. The flow exits with a filtered set of templates visible for the user to inspect or select. Steps: Explore Blueprints → Search Blueprints → Filtered Results **Farfetch: Submit a bag for quote** (web, 10 screens, Marketplace) [Refero](https://refero.design/flows/3551) > End-to-end happy-path for submitting a pre-owned designer bag to FARFETCH Second Life for a quote. Entry: user clicks 'Start selling' on the Second Life landing page (user is signed in — name visible in header). The flow collects item details, then photos, then optional additional items, presents a review screen for confirmation, and finally submits a quote request. Exit: confirmation modal and the item appears in 'My Quotes' with a pending status. Assumptions: user is authenticated, using desktop browser, has at least one bag to sell and photos available. Right column help text is present on form screens to reduce support friction. Steps: Landing / Start selling → Item details — country & category → Item details — full form (empty) → Item details — filled example → Photo upload — guidance and placeholders → Photo upload — uploaded images → Additional items → Review and request → Submission confirmation modal → My Quotes — submitted item ##### Task: research collaboration **LEGO® Builder: Collaborative build (join/create) flow** (ios, 15 screens, Entertainment) [Refero](https://refero.design/flows/10135) > Happy path: From a product detail screen the user chooses to continue a saved build or join/create a shared "build together" session. If they choose to build together they can either join with a 6-digit party code or create a new party. Joining requires entering a numeric code via an in-app keypad; the system validates the code, handles a "not found" error, and on success prompts the user to pick a builder avatar and then shows a waiting/party lobby. When the session starts the user enters the shared 3D build workspace showing the model and participant icons. Steps: Product detail - Build options → Build together prompt → Start over confirmation → Party entry choice → Enter party number (keypad) → Digit entry (partial) → Looking for party (search state) → No party found (error) → Retry entry and search → Join success (You're in) → Select builder avatar → Lobby / Waiting to start (joined) → Host/creator party number → Session start - transition → In-session collaborative build **ChatGPT: Research timeline with sources** (ios, 17 screens, AI Tool) [Refero](https://refero.design/flows/7024) > Happy path: A signed-in user opens the chat, initiates a research-style request, and composes a prompt asking for a Cold War summary. The user selects the Research helper / template and chooses a specific output format ("Timeline of key events"). The system starts a long-running background research task and surfaces a progress card in the chat. The user can expand a Details/Activity view to see the agent’s step-by-step work and the accumulated source list (sources tab). When the research completes the assistant posts a compiled timeline with inline source tags and embedded images; images can be opened full-screen. Steps: Chat home / quick prompts → Create / helper sheet → Compose research prompt → Assistant clarification → Choose timeline option → Start background research → Progress card (initial) → Progress updates (collection) → Research completed header → Open Activity details → Activity timeline scroll → Sources tab (list) → Compiled timeline with image thumbnails → Open image full-screen → Return to timeline → Final events and citations → Follow-up composer ready **Grok: Research Assistant — DeepSearch Session** (ios, 15 screens, AI Tool) [Refero](https://refero.design/flows/6092) > Happy-path: a user opens the chat interface, chooses DeepSearch, types a long research prompt, sends it, and watches the assistant perform a web-enabled research pass (“DeepSearching”) while showing a streaming/think indicator. The assistant returns a multi-part answer: summary bubbles, a "Thought" timeline/checklist of research steps, a browsable list of source pages, and a long structured article with headings, a comparative table, and final citations. Steps: Home — Chat Composer Visible → Activate Composer / Keyboard → Type Research Prompt → Submit Prompt → Streaming / Partial Response → Result Bubble + Source Summary Link → Expanded 'Thought' / Progress Label → Open Thought Detail Panel → Open Pages (Source List) → Preview / Read a Source → Thought Checklist Progress → Generated Structured Answer Appears → Comparative Table Rendered → Citations List → Composer Still Available / Continue Interaction **Meta AI: Mythical Biology Poster** (ios, 3 screens, AI Tool) [Refero](https://refero.design/flows/13353) > A user is viewing an existing creature-biology conversation and selects a suggested request to combine five creature diagrams into one mythical biology poster. The system processes the request, shows a planning state, and then returns a completed master poster containing all five diagrams. The flow assumes an existing conversation and generated infographic assets are available. It is shown on a mobile device; no login or permissions are visible. Steps: Select Poster Request → Poster Layout Planning → Review Completed Poster **Meta AI: Browse Shared Artifacts** (ios, 3 screens, AI Tool) [Refero](https://refero.design/flows/13380) > The user starts in the Artifacts library, where creations are shown in a grid alongside presets. They open the filter control, change the collection from All to Shared, and return to a filtered grid containing only shared artifacts. The flow assumes the user is already in the Artifacts section and has access to shared content. The user can optionally continue using the search field or Ask Meta AI input. **UGLYCASH: Search and follow users** (ios, 2 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/11219) > Happy path: A signed-in user searches for other users by typing a name into the top search field. The interface shows results under the Users tab (with Clubs as a sibling tab). Results are grouped into sections: "Followed Users" (users the current user already follows) and "Other Users" (not followed). Each result row contains an avatar, username and an action button (Follow / Unfollow). Tapping Follow toggles the button to Unfollow and the user appears in the Followed Users section. The keyboard remains visible while the user continues to edit the query. Steps: Enter query and view results → Followed users shown and list updated **Instagram: Add people to reel** (ios, 7 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/3623) > Happy path for adding people (tags and collaborator invites) to a new Reel on a mobile device. Entry: creator is on the New reel publish screen with a prepared video, cover, and caption. The user navigates into the Tag people flow, searches for users, adds one or more tags and invites a collaborator. Exit: control returns to the New reel screen where the Tag people row shows the updated count and the user can Save draft or Share. Steps: New reel — publish options → Tag people — landing → Search input — focused → Search results list → Invite collaborator — info → Tag people — updated list → New reel — tag count updated **Melius: Create Canvas Sticky Note** (web, 8 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/14080) > The user starts on a collaborative visual canvas containing connected image, text, and audio cards. They open the annotation tool, create a sticky note, adjust its color and size, zoom into the canvas, and enter planning text. The flow begins in an existing canvas workspace and ends with an edited pink sticky note positioned beside the content. The user appears to have access to edit the canvas; authentication status and device details are not visible. Steps: Existing Canvas → Annotation Menu → Add Sticky Note → Choose Note Color → Apply Pink Color → Adjust Note Size → Zoom Canvas → Edit Note Text **Spyglass: Creative Strategy Follow-up** (web, 4 screens, Business, Inspiration Board) [Refero](https://refero.design/flows/13463) > A user reviews a generated advertising analysis in a desktop workspace and prepares a follow-up question about target audience segments and scaling. The user can begin composing the follow-up through the message composer, inspect the voice-input control, activate voice recording, or enter the question as text. The shown sequence ends with the follow-up typed into the composer but not yet submitted. Steps: Review Generated Analysis → Inspect Voice Input → Activate Voice Input → Draft Audience Question **Telescope: Add Product Note** (web, 3 screens, Social Networking) [Refero](https://refero.design/flows/13908) > The user is viewing a saved product detail for the “Blue nylon trucker cap.” From the item detail view, they enter a note in the Notes section and submit it. The note is then displayed as a saved comment with the author and timestamp. The flow starts on an existing product detail view and ends with the note persisted on that item. The user appears to be logged in, since an account avatar and curator controls are visible. No special permissions beyond access to the saved item and note functionality are evident. Steps: Product Detail → Enter Note → Saved Note **Spyglass: Brand Research Search** (web, 3 screens, Business, Inspiration Board) [Refero](https://refero.design/flows/13470) > A user starts in the Research section and uses the global search field to find the beauty brand “rhode.” After entering the query, the interface displays matching brands and related creators. The user selects the verified rhode brand result and is taken to its overview profile, which summarizes advertising activity, brand positioning, competitors, and high-impression ads. The flow assumes the user has access to the Research area and is either logged in or using an authorized workspace. Steps: Open Research Search → Review Search Suggestions → View Brand Overview **OpenAI: Research index browsing and filtering** (web, 3 screens, Business, Development Tools, AI Tool) [Refero](https://refero.design/flows/4169) > A researcher-facing listing of OpenAI research entries where users can browse, filter, and sort publications and releases. Steps: Research index listing → Open filter modal → Filtered results applied **Synthesia.io: Move Video to Folder** (web, 7 screens, AI Tool) [Refero](https://refero.design/flows/13753) > A logged-in workspace user organizes a video from the Videos library into an existing folder. Starting from the My videos view, the user selects one video, opens the move action, chooses the Demo folder, and confirms the move. The video is removed from the current top-level list, and a confirmation toast shows that one video was moved to Demo. Steps: Videos Library → Video Hover Actions → Select Video → Open Move Menu → Choose Destination Folder → Confirm Move → Move Confirmation **Figma: New user signup and onboarding** (web, 27 screens, Design & Media Editing, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/4983) > Happy-path new-user signup and product onboarding starting from the marketing hero. Entry: an unauthenticated visitor clicks the primary CTA on the marketing hero. The flow opens an in-page modal for account creation, verifies email, collects a few setup preferences (name, role, collaborators, intent, experience), asks plan preference and first task, then lands the user in their Figma workspace (Recents). Steps: Marketing hero with primary CTA → Welcome modal - sign-in options → Email entry (typed) → Create account - add password → Create account - password entered → Email verification prompt → Onboarding: enter display name → Name entry with subscription option → Onboarding: select kind of work → Role selected → Invite collaborators → Onboarding: why you're here (multi-select) → Intent selected → Prior experience question → Prior experience selected → Plan selection overview → Plan chosen (Starter) → Choose first task → First task selected (Present with Slides) → Workspace - Recents (post-signup) **MyMind: mymind Onboarding** (web, 17 screens, Productivity, Reference, Inspiration Board) [Refero](https://refero.design/flows/13289) > A prospective user discovers mymind through a long-form marketing landing experience, reviews its privacy and content-management capabilities, and starts account creation. After choosing Google or Apple sign-up, the user moves through a guided onboarding presentation explaining private storage, one-click saving, automatic organization, supported content types, and search. The user then reaches a subscription screen with an option to continue as a guest and enters the mymind workspace. The final state is a populated dashboard with introductory content and setup reminders. The screenshots suggest the user is initially logged out and using a desktop browser. No completed payment or detailed account form is shown. Steps: Product Discovery → Account Creation Choice → Private Oasis Welcome → Privacy Promise → Saved Content Overview → One-Click Saving → Automatic Organization → Book Recognition → Article Recognition → Website Recognition → Search Notes → Search Visual References → Search Articles → Search Books → Plan Selection → Workspace With Setup Guide → Initial Workspace **Cohere: Research listing with topic filter** (web, 2 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3785) > Public research index listing research papers authored by Cohere staff. Steps: Research list overview → Open Topics filter and select **ChatGPT: Sourced Research Session** (web, 11 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/6892) > Happy path: A signed-in desktop user starts at an empty chat prompt, opens the Tools menu, selects the Research tool, submits a research request ("Break down the causes of World War I"), and watches the assistant run web searches and gather sources. A progress card shows research in progress; an Activity/Sources pane at right shows individual reads and citations as the system gathers material. After completion the assistant displays a structured, sourced answer in the chat, with an in-line Sources section and an option to preview the compiled document (download / share). Steps: Start screen / empty prompt → Open Tools menu → Compose research query → Research started → Activity pane opens → Progress updates with sources → Research completed and answer posted → Scroll through content and view Sources block → Open Preview modal → Scroll preview content → Inspect full sources list **Parallel: Deep Research — create and inspect research run** (web, 16 screens, AI Tool) [Refero](https://refero.design/flows/9052) > Happy-path: A user creates a deep research job, chooses a compute/model option, starts the job, watches progress as the agent crawls and extracts sources, and inspects the structured results when the job finishes. Steps: Research landing / query input → Open model / compute options → View sample research content → Open citation / context panel → Switch to MACHINE tab (structured view) → Enter full research query → Start research — initial running state → Running — early results and activities → Running — activity visibility toggle → Running — mid progress updates → Running — high progress (near completion) → Completed state — summary and counts → Select a field node to view details → View Basis / Reasoning and sources → Drill deeper into node hierarchy → Final inspection — confirm sources and reasoning **ChatGPT: Select and Use Research GPT** (web, 6 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/6898) > Happy-path: A signed-in desktop user browses the GPT directory, opens a featured GPT (Consensus), reviews its details and capabilities, starts a chat, grants the requested external-site permission, then receives a sourced, evidence-based answer. Steps: GPT Directory (Explore GPTs) → Open GPT Details Modal → Inspect Capabilities & Ratings → Chat Session Landing → External Permission Request → Receive Sourced Answer **Cohere: Research-focused blog listing** (web, 2 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3788) > A desktop blog listing focused on research content. Entry is the main blog listing where a user can scan a featured hero article, filter by category tags (All, Product, For Business, Developers, Research, Company), use the search field in the header, and browse a multi-row grid of article cards. The page includes persistent global header navigation and a large CTA block near the bottom that leads to product engagement (Request a demo / Try the playground), then footer and newsletter subscription. The second screen shows results after submitting a search query ("future") — same header and footer, with the content area replaced by a grid of search-matching article cards and pagination. Assumptions: user is on a large-screen (desktop) device, likely not signed in (Sign in visible), and has standard read-only permissions (browsing/searching). Steps: Blog landing — Research → Search results — 'future' **Parallel: Search API: run and inspect results** (web, 2 screens, AI Tool) [Refero](https://refero.design/flows/9049) > User (signed-in on desktop) opens the Search API tool to run a web-retrieval query. They enter a natural-language Objective, add Keywords, and adjust fetch/source policies (max age, timeout, results limits). On Search, the system runs a query and returns results in the right-hand pane. The results can be viewed in a human-friendly snippet list (default) or switched to a machine-readable JSON representation, and the JSON can be copied. Steps: Configure and Run Search → Inspect Machine-Readable JSON **Rork: Invite Collaborators** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12167) > A project-sharing dialog is opened from the workspace, letting the user invite collaborators by email, assign a role, and optionally create an invite link. The entry point is the top-right sharing control; the exit is either closing the dialog or completing an invitation action. Steps: Open Invite Dialog → Add Collaborator → Create Join Link **Exa: Answer Query with Citations** (web, 3 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/12018) > Happy path for generating an AI answer from the Answer tool. Entry begins on the Answer screen inside an API playground workspace, with a query already entered and response options visible. The user triggers generation by clicking Search. The system processes the request and streams a response in the output panel on the right. The user can then inspect the answer text, see citation placeholders resolve into citation entries, and review additional source references. Assumptions: user is already signed in, has access to the Answer tool, and is viewing a desktop layout with a split editor/output panel. Steps: Enter Query and Run → Streaming Answer → Review Sources ##### Task: risk register **Tinder: Report a Safety Concern** (ios, 6 screens, Dating & Relationships) [Refero](https://refero.design/flows/3439) > Happy-path: A signed-in user on the Matches tab opens the Safety Toolkit from the shield icon, chooses to Report, reads the reporting options and tips in a compact modal, then follows a link to a Help Center article about reporting in-person physical harm. From the article the user can read resources, tap START REPORT to begin an external report flow, and optionally submit feedback about the article. Entry: Matches (Messages) tab. Exit: user lands in Help Center article and can tap START REPORT (external report flow) or submit article feedback. Assumptions: user is logged in on a mobile device; the Safety Toolkit is accessible from a shield icon in the header; Help Center content is displayed inside an in-app webview. Steps: Matches (empty) / Safety shield → Safety Toolkit options → Reporting modal (tips & categories) → Help Center article - reporting details → Help Center resources & START REPORT → Article feedback and submit **Binge: Jump Scare Search** (ios, 4 screens, Entertainment) [Refero](https://refero.design/flows/12958) > A user reviews a movie or episode’s jump scares, then uses the search field to filter the scare list by keyword. The entry point is the full jumpscare list screen showing a count, timeline markers, spoiler warning, and individual timestamped entries. The exit state is the list filtered to matching results while the search input remains active and the keyboard stays open. Steps: Scare List Overview → Search Activation → Search Query Entry → Filtered Results **Lovable: Security Issue Review** (web, 6 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12323) > A logged-in user is viewing the Security section of a project and reviewing detected issues from a security scan. From the issues list, the user can expand an issue, inspect details, and choose to ignore it by providing a reason. After confirmation, the issue is removed from the active list and moved into an ignored state, with an option to show ignored issues. The flow starts on the Security dashboard and ends back on the updated issues list after the ignore action is saved. Steps: Security Dashboard → Issue Details Expanded → Ignore Issue Dialog → Custom Reason Entry → Confirm Ignore → Ignored Issue Listed **TravelPerk: Travel Advice Board** (web, 2 screens, Productivity, Travel & Booking) [Refero](https://refero.design/flows/5199) > Single-page Travel Advice Board targeted at people checking current travel safety alerts. Entry: user lands on the page (public or logged-in). The top area is a hero with title and visual, followed by interactive controls (country selector and risk-level filters) and a dense, card-based list of travel advisories sorted in a grid. Each card shows country, risk tag, timestamp and a short summary. Exit points: clicking an advisory (to see details), tapping the 'Learn more' CTA under "Business traveler safety and security," or using footer links to explore product pages. Assumption: page is publicly accessible; interaction includes basic filtering and navigation to more detailed content or product pages. Steps: Browse and filter advisories → Access safety solutions and site navigation **Roots: Subscription trial onboarding** (ios, 20 screens, Productivity) [Refero](https://refero.design/flows/5740) > User views a promotional subscription screen for the Roots app and moves from discovery (hero and benefits) to signing up for a 7‑day free trial, choosing a plan, confirming via the App Store, and receiving a purchase confirmation. Steps: Promo hero & primary CTA → Trial timeline & CTA repeat → Scroll replacements & value entry → Value calculator - inputs → Value calculator - results → Mission + product positioning → Testimonials and FAQ list → FAQ expanded content → Support & account actions → Request iOS notification permission → Reminder toggle set → Plan selection sheet → Feature comparison summary → Feature details grid → Confirm/Continue CTA in sheet → App Store authorization prompt → App Store purchase sheet (details) → In-app purchase success alert → Final entitlement confirmation → Post-purchase state & support links **Kickstarter: Report a Project** (ios, 7 screens, Fundraising) [Refero](https://refero.design/flows/4630) > Happy path: A user on a project's Overview finds the "Report this project" action, selects a reason (Rules violation → specific subreason), fills out a report form (contact email, project URL prefilled, free-text details), sends the report, and receives immediate confirmation. Steps: Project Overview → Report Options → Select Specific Rule → Report Form (enter details) → Compose Message → Submission Confirmation → Return to Project — Acknowledgement **IKEA: Gift Registry Sharing** (web, 3 screens, Marketplace, Home & Decor) [Refero](https://refero.design/flows/8266) > Happy-path summary: A logged-in user viewing their personal Birthday Registry uses the page-level Share control to open a right-side sharing panel. The panel exposes a sharing toggle, a persistent registry link with a Copy button, a registry code with a Copy button, and a QR code with a Download button. The user copies the link (or code) and receives an on-screen confirmation toast. Steps: Registry overview → Sharing panel — copy and download options → Action confirmation **Slite: Knowledge Base Ask and Import Prompt** (web, 3 screens, Paperwork, Productivity, Collaboration Tools) [Refero](https://refero.design/flows/1704) > Happy path: A logged-in user on a desktop opens an inline knowledge-base assistant (centered modal) and types a question. The assistant checks available documentation and returns either an answer or a message that there isn't enough documentation. If docs are insufficient, the assistant prompts the user to import content (Notion, Google Docs, Confluence, markdown, etc.) with a clear CTA. Entry: user clicks or focuses the "Ask your knowledge base..." input. Exit: either an answer is displayed (not shown) or the user proceeds to the import-content flow by clicking the Import content button. Assumptions: user is authenticated and has at least read-access to some knowledge documents; the interface is desktop-sized; the assistant references only documents the user can access. Steps: Open assistant → Enter question → Insufficient docs — suggest import **Grok: Image-to-Short-Film Plot Generator** (web, 10 screens, AI Tool) [Refero](https://refero.design/flows/5328) > A desktop session where a logged-in user asks the assistant to create a short-film plot from an attached photograph. Entry: user views two generated images and enters a prompt in the chat input. The user enables a 'DeeperSearch' thinking mode and submits the request. The system responds by revealing a multi-stage analysis (visible thinking steps) that moves from context review → idea exploration → narrative development → refinement → enhancements → integration → final written survey note. Exit: user receives a detailed plot, structured analysis, and key citations with the input box available for follow-up. Assumptions: user is authenticated (avatar present), working on a desktop browser, an image is attached/visible to the assistant, 'DeeperSearch' and 'Think' modes are enabled. Steps: Image Review & Prompt Area → Submit Creative Request → Thinking: Reviewing Context → Exploring Plot Ideas → Developing Narrative → Refining Plot Details → Enhancing Plot Elements → Integrating New Ideas → Final Survey Note (Detailed Analysis) → Conclusion & Exportable Summary ##### Task: security incident alert **Showcase: Manage Alerts** (ios, 3 screens, Entertainment) [Refero](https://refero.design/flows/7646) > Happy path: user opens the Alerts tab (bottom navigation) and sees alerts grouped by time (Today, This Month). From the list the user swipes left on an alert to reveal a red Delete action, taps Delete, and the alert is removed from the list. Steps: Alerts list (initial) → Reveal delete action → List after deletion **Daylish: Add event alert** (ios, 5 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/8009) > Happy path: User is creating a new calendar event (Dentist appointment) and adds a timed alert before saving the event. Steps: Start event composer & open Alerts → Open New Alert modal (default) → Select timing option → Toggle alert type to Alarm (optional) → Alerts list shows saved alert **Daylish: Edit Event Alerts** (ios, 4 screens, Scheduling Tools, Productivity) [Refero](https://refero.design/flows/8010) > Happy-path: A user is editing an event’s details (title, tag, and alert reminders) on a mobile device. The Alerts overlay is opened for a specific event (header shows "Team standup meeting" / date). The user types a new event title, selects a tag (Tomorrow / Work), enters edit mode for alerts, swipes an alert to reveal a Delete action, and deletes the alert. Steps: Open Alerts and edit title → Enter Alerts edit mode → Swipe to reveal Delete → Delete alert and update list **Plane Finder: Create and Manage Aircraft Alerts** (ios, 6 screens, Travel & Booking) [Refero](https://refero.design/flows/10413) > Happy-path: A user opens the Custom Alerts area, creates a new alert by choosing a matching criterion (e.g., Airline/Flight Number or Aircraft Type), optionally adds a label, selects whether the alert applies Anywhere or to a specific Location (map preview shown), saves the alert, and sees it added to the alerts list. Steps: Custom Alerts — Empty State → Choose Alert Criterion → Add Alert — Enter Details → Choose Alert Area → Save Alert → Alerts List — Manage Items **Structured: Create Task with Alerts** (ios, 3 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3279) > Happy path: User composes a new task, configures one or more alerts (relative to task start or end), optionally adds subtasks/notes, and submits the task. Steps: New Task — Alerts list → Alert time picker modal → Review alerts and create **Binge: Jumpscare Details** (ios, 3 screens, Entertainment) [Refero](https://refero.design/flows/12957) > A mobile content-details screen shows a list of jumpscare moments for a title, with a spoiler warning, a visual timeline of scare markers, and searchable/filterable entries. The user enters this screen from the title’s scare summary and can either browse all scares, filter to major or minor scares, or search within the list. The main exit points are going back to the previous screen or selecting a specific timestamp/entry to inspect a scare moment. Steps: Scare List Overview → Filter Selection → Filtered Scare List **Stellate: Delete Alert** (web, 3 screens, Development Tools) [Refero](https://refero.design/flows/4427) > Happy path: A signed-in user with config/manage permissions opens a specific alert's settings, reviews the trigger and notification recipients, chooses to delete the alert, confirms the deletion in a modal, and returns to the alerts list where a success toast confirms removal. Steps: Alert Settings — Detail → Delete Confirmation Modal → Alerts List — Post-Deletion **Lovable: Security Issue Remediation** (web, 4 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12324) > A user reviews security scan results for a project, selects one or more detected issues, and triggers an automated fix from the security panel. The entry point is the Security section in the left navigation. The flow assumes the user is already in the project workspace and has access to security tools. It exits after the selected issues are fixed, the issue count is reduced, and the security panel reflects the updated remaining issue(s). Steps: Security Overview → Select Issues → Remediation Triggered → Fix Result Review **Handshake: Job discovery and alert creation** (web, 2 screens, Social Networking, Job Search) [Refero](https://refero.design/flows/8759) > Happy path: a signed-in desktop user visits the Explore area to browse curated job cards for designer roles and creates a job alert for a group of recommendations. Entry: user opens Explore from the left navigation. Exit: a job alert is created (UI updates to allow undo and a confirmation toast appears). Assumptions: user is authenticated, using a desktop layout, and has permission to create alerts in their account. Steps: Explore - review curated job picks → Alert created - confirmation and undo **Steel: Create Usage Alert** (web, 4 screens, AI Tool, Development Tools) [Refero](https://refero.design/flows/14301) > A signed-in workspace administrator opens Settings > Limits & alerts, starts creating a usage alert, adjusts the notification threshold from the default 80% to 70%, and saves it. The flow begins on the limits and alerts dashboard with no alerts configured and ends with a saved email alert displayed in the alerts list. The screenshots show a desktop interface and an authorized user with access to workspace settings. No additional recipient or notification-channel selection is shown; the modal indicates that an email notification will be sent. **Fingerprint: Event Detail — Device & Copy IP** (web, 3 screens, Business, Development Tools) [Refero](https://refero.design/flows/11154) > Logged-in investigator views a single identification event record (entered from an events list). The screen shows the Device card (Visitor ID, last/first seen, client/browser/OS), a location map with IP address, Smart Signals (Suspect Score and Velocity Signals), and a JSON snippet for detected signals (Incognito mode). The user inspects the IP address and uses the copy control. Entry: event selected from events list. Exit: IP address copied to clipboard and a confirmation toast appears. Steps: Open event details → Activate copy affordance → Confirm copy success **Stellate: Alerts Configuration** (web, 9 screens, Development Tools) [Refero](https://refero.design/flows/4426) > Happy path: A logged-in user with config permissions opens Config → Alerts, clicks New Alert, chooses an alert type, builds a trigger (metric, comparator, threshold, time window), adds one or more notification recipients (team member or email), submits the alert, and returns to the Alerts list where the new rule is visible. Steps: Alerts list → Choose alert type → Trigger builder (initial) → Open trigger metric dropdown → Set trigger metric and values → Add first notification recipient → Add additional recipients → Submit alert and confirmation → Alert settings summary **Twitch: Alert Configuration Tour** (web, 5 screens, Social Networking, Entertainment, Video Streaming) [Refero](https://refero.design/flows/13616) > A logged-in Twitch creator is guided through a five-step onboarding tour while configuring alerts in the Creator Dashboard. The tour starts on the Alerts editor with instructions to copy the Browser Source URL, then explains alert variants, variant priority, alert customization, and preview/testing. The happy path is to select Next through each tutorial step and finish with Done. The underlying alert editor remains visible but dimmed throughout the tour. Steps: Copy Browser Source → Understand Alert Variants → Set Variant Priority → Customize Alerts → Preview and Test ##### Task: shipment tracking logistics **PayPal: Mark package as delivered** (ios, 4 screens, Finance) [Refero](https://refero.design/flows/7255) > Happy-path: A signed-in mobile user views a manually-added tracked package that currently has no courier updates. From the package detail screen they open the package actions, choose "Mark as delivered", confirm on a modal, and the app marks the package delivered (updates header, fills the timeline, and shows a toast). Entry: package detail view for a tracked item. Exit: package marked delivered and UI updated to show delivered state. Assumptions: user is on a mobile device, already has the package saved in their account (manually added), and has permission to edit/remove that package. Steps: Package detail - awaiting update → Edit package action sheet → Confirm 'Has this package arrived?' modal → Delivered state and success feedback **PayPal: Add package to tracking** (ios, 7 screens, Finance) [Refero](https://refero.design/flows/7254) > Happy path: A logged-in mobile user opens the Package tracking area, sees an empty state and taps Enter Tracking Number. The user is taken to an Add Package form, enters a tracking number, accepts a suggested courier (or chooses one from a dropdown), optionally adds a nickname, and taps Start Tracking. A confirmation toast appears and the new package is added to the top of the tracking list with an initial status. Steps: Empty tracking state → Add package form (initial) → Enter tracking number → Courier suggestions → Courier selected → Optional nickname entry → Submit and confirmation **On: Checkout Address Autocomplete and Shipping Selection** (ios, 5 screens, Marketplace) [Refero](https://refero.design/flows/9602) > Happy-path mobile checkout where a signed-in user (email shown) fills delivery details, uses address-autocomplete to choose a specific street address, reviews the auto-filled city/state/ZIP, and then selects a shipping method. Steps: Checkout entry / Delivery form → Address required / Field focus → Type address and view suggestions → Select suggestion and auto-fill fields → Choose shipping method **PayPal: Rename Tracked Package** (ios, 5 screens, Finance) [Refero](https://refero.design/flows/7257) > Happy path: A signed-in user opens a manually-added FedEx tracking item in the app, opens the item's overflow menu, selects Rename, types a new name, and saves it. Steps: Package details (Awaiting update) → Edit package bottom sheet → Rename modal (empty input) → Save new name → Confirm rename on details **Givingli: Gift item detail** (ios, 2 screens, Lifestyle, Scheduling Tools) [Refero](https://refero.design/flows/8202) > Mobile-first gift item detail view showing a single curated gift (William Sonoma crate) with two primary interaction groups: gifting CTAs and social/share actions. Entry: user scrolls to or selects a gift tile and lands on this detail slice. Happy path: user inspects the image and uses either the Gift CTA (merchant-specific) or the share row to distribute the gift link; when the user taps the link-share icon the interface copies the link and shows a transient confirmation. Exit points: navigating to the merchant / gifting flow via a Gift CTA, using an external social share target, or tapping See More Gifts to return to the gift list. Assumptions: user is on a mobile device (phone viewport shown). The screenshots do not indicate authentication state; actions are presented without visible login prompts. Steps: View gift detail → Share - link copied confirmation **Omio: Journey detail — share ETA** (ios, 4 screens, Travel & Booking) [Refero](https://refero.design/flows/5230) > Happy-path: A logged-in user (passenger Sienna Hart) opens a booked journey detail for Barcelona → Madrid on an iPhone, reviews departure and arrival times, and uses the "Share live time of arrival" action to send a prefilled arrival message to a contact via the device's native share sheet (Messages). Steps: Journey overview → Share arrival option → Native share sheet → Message composer (prefilled) **Instagram: Add shipping address** (ios, 4 screens, Entertainment, Social Networking) [Refero](https://refero.design/flows/2898) > Happy path: From the Orders and payments settings screen the user opens the address management area and adds a new shipping address. They fill the standard address fields, optionally toggle "Set as default," then save. Steps: Orders and payments (settings list) → Add shipping address (form - default off) → Toggle default address → Save in progress (loading state) **Asos: Postal Address Selection** (ios, 8 screens, Marketplace) [Refero](https://refero.design/flows/2907) > Happy path: user chooses to add a postal shipping address from the checkout screen, fills or confirms contact details (including mobile), uses the address-finder autocomplete to locate their street, selects a suggested address, confirms it as the delivery address and returns to the checkout where the chosen postal address is displayed and shipping options are shown. Steps: Checkout — Shipping Address → Add Address — Contact Details → Address Finder — Empty State → Mobile Entry — Phone Keyboard → Address Finder — City Search → Address Finder — Street Suggestions → Confirm Address — Deliver CTA → Checkout — Postal Address Selected **Genie: Place Detail Sharing** (ios, 3 screens, AI Tool, Social Networking) [Refero](https://refero.design/flows/12432) > A user views a place detail screen for Pier i Cafe and shares it with a contact through a bottom-sheet picker. Entry is the place detail screen; the user taps the share/send icon, selects a recipient, and receives confirmation that the place was sent. Steps: Place Details → Recipient Picker → Send Confirmation **RAD Weather: Weather Dispatch Sharing** (ios, 4 screens, Weather) [Refero](https://refero.design/flows/13814) > A user views a morning weather briefing, opens the R.A.D. Dispatch sharing experience, and shares a generated weather card through the iOS share sheet. The user selects Messages and the contact Matt Lam, then reviews the prefilled message before sending it. The flow starts with the weather briefing already open on an iOS device and ends in the Messages composer with the share content prepared for delivery. Steps: Morning Briefing → Dispatch Preview → Share Destination → Message Review **PayPal: Add tracking information** (web, 6 screens, Finance) [Refero](https://refero.design/flows/7381) > Happy path: from a transaction detail in Activity the seller clicks the Add tracking/packaging info link, lands on an "Add tracking information" form, selects an order status, enters a tracking number and selects a carrier, then submits. On success the UI shows a confirmation message stating an email will be sent to the customer, with a CTA to return to Activity. Steps: Transaction detail → Add tracking — initial form → Choose order status → Validation feedback for tracking input → Provide tracking number and carrier → Submission success / confirmation **Faire: Ship Date Selection** (web, 7 screens, Business, Marketplace) [Refero](https://refero.design/flows/13093) > Happy-path cart experience on Faire for a logged-in buyer. The user is in the cart, opens the ship-date selector for the order, chooses either immediate shipping or a future ship date, saves the choice, and returns to the cart with the updated ship timing reflected in the order header and delivery estimate. Steps: Cart Overview → Ship Date Modal → Future Date Entry → Calendar Selection → Date Confirmed → Save Ship Date → Updated Cart **Acne Studios: Product delivery check** (web, 5 screens, Marketplace) [Refero](https://refero.design/flows/4371) > Happy-path: A shopper views a product-detail page for a dark‑green multipocket bag, opens the Delivery & Returns panel to check shipping/return rules and costs, enters a postal code, reviews the shipping methods and estimated delivery dates/prices, and then closes the panel to continue shopping or proceed to purchase. Steps: Product detail view → Open Delivery & Returns panel → Enter postal code → View shipping options and costs → Close panel and continue **Zara Home: Shipping Support Conversation** (web, 6 screens, Marketplace, Home & Decor) [Refero](https://refero.design/flows/14233) > The user enters the Zara Home Help assistant, asks about available shipping options and costs, reviews the assistant’s response, then asks a follow-up question about delivery to a hill or hard-to-reach location. The assistant returns a second response with guidance to contact customer service for delivery review. The flow starts on the Help screen with no conversation history and ends with both questions and answers displayed. The screenshots show a desktop browser layout; login status and permissions are not required or visibly established. Steps: Help Assistant Entry → Shipping Question Submitted → Shipping Answer → Follow-up Question Entry → Follow-up Submitted → Delivery Accessibility Answer **Mocha: Post-shipment sharing modal** (web, 2 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/11620) > A post-release sharing modal that appears after the user completes a "ship" action. The modal shows a generated public URL, a prominent Copy button, social sharing icons and a visual preview of the shipped item. Happy path: user clicks Copy (or a social icon) to share; the Copy button switches to a "Copied!" state and a small toast confirms the clipboard update. Steps: Share modal displayed → Copy confirmation and toast **Shop: Mark Order Delivered** (web, 4 screens, Marketplace) [Refero](https://refero.design/flows/13284) > An authenticated customer views an existing order in the Orders section of their account and manually marks it as delivered. The flow starts on the order details summary for order #1327, placed with the status "Order placed." The customer opens the order actions menu, selects "Mark order as delivered," and receives confirmation through a toast notification. The order then updates to show "Delivered" and "Delivered today.". Steps: Order Details → Order Actions → Delivery Confirmation → Delivered Order **Faire: Cart Checkout and Ship Date Selection** (web, 17 screens, Business, Marketplace) [Refero](https://refero.design/flows/13095) > A buyer reviews items in a cart, adjusts item quantity, selects a ship date for an order, then proceeds through shipping and payment checkout. Steps: Cart Overview → Brand Cart Details → Select Ship Date → Scheduled Ship Date Input → Date Picker → Date Confirmed → Ship Date Saved → Item Quantity Editor → Quantity Below Minimum → Minimum Quantity Warning → Quantity Updated → Updated Cart Summary → Shipping Address Entry → Shipping Address Completed → Payment Details → Payment Validation → Ready for Review **West Elm: Checkout — Shipping & Pickup Selection** (web, 9 screens, Home & Decor, Marketplace) [Refero](https://refero.design/flows/1689) > Happy path: A user with items in cart moves from browsing/cart review to checkout, chooses a shipping method (either ship-to-address or ship-to-location/UPS Access Point), provides or selects a shipping address or pickup location via a map modal, reviews per-item delivery estimates and optional services (gift wrap, white-glove, protection plan), and proceeds toward payment. Steps: Cart overview (mini-cart & favorites) → Shopping Cart (full) → Checkout — Shipping Step (method selector) → Pickup Location Finder (map modal) → Pickup Location Selection (map pin detail) → Shipping Address Form (empty) → Shipping Address Form (filled) → Delivery & Gift Options (per-item delivery) → Leave Checkout Confirmation **Zara Home: Gift-Wrapped Item Checkout** (web, 20 screens, Marketplace, Home & Decor) [Refero](https://refero.design/flows/14227) > Happy path begins with a single gift-wrapped item in the basket. The user proceeds to checkout, selects drop-point delivery, searches for nearby delivery locations by postcode, enables distance information, and chooses a delivery point. The user then enters billing details, selects credit/debit card payment, corrects invalid card information, and saves the card. After reviewing the order summary, the user authorizes payment and reaches a successful order confirmation. The flow assumes a desktop browser, a customer with an item already in the basket, and delivery availability for the entered U.S. address. The user does not appear to be required to sign in. Steps: Basket Review → Delivery Method Selection → Delivery Point Locator → Postcode Search → Search Results → Distance Display → Pickup Point Selection → Billing Address Form → State Selection → Address Autocomplete → Billing Address Completion → Telephone Entry → Delivery Checkout Review → Payment Method Selection → Card Details Entry → Card Validation Correction → Saved Payment Method → Order Summary Review → Order Confirmation → Unused **Microsoft: Search from Order Tracking** (web, 4 screens, Business, Marketplace) [Refero](https://refero.design/flows/9797) > Happy-path: A signed or anonymous user starts on the Microsoft Order Tracking landing screen, uses the global header search to look up a query (entered text: "he"), views inline search suggestions, submits the search, lands on the site search ‘Explore’ results, then switches to the ‘Shop’ tab to view product/category results. Entry: user arrives at Order Tracking page (from Microsoft Store or support). Exit: user either continues through Explore results, switches to Shop product listings, or clicks a product/related link to navigate off the search results. Assumptions: user has access to the site header search (not blocked), device is desktop (wide layout), user may be logged in or not (both CTA options for account/guest on Order Tracking exist). Steps: Order Tracking landing → Inline search suggestions → Explore results listing → Shop results with product carousels **Faire: Message Seller from Product** (web, 5 screens, Business, Marketplace) [Refero](https://refero.design/flows/13088) > Happy path starts on a product detail listing for a customizable candle. The user opens the brand messaging area from the product context, then composes and sends a question to the seller. The conversation view confirms the message has been posted and shows brand information in a side panel. Assumptions: the user is logged in, has access to messaging, and is viewing the seller’s product from a desktop browser. Steps: Product Detail → Messages Inbox → Conversation View → Compose Message → Sent Message **Asos: Contact to Virtual Assistant — Order Query** (web, 5 screens, Marketplace) [Refero](https://refero.design/flows/1598) > Happy-path: A customer lands on the Contact / "Get in Touch" area and either clicks the visible "Visit My Account" CTA or starts the on-page Virtual Assistant. The chat widget expands and the bot opens with a greeting and a single-line message field. The user interacts with the bot (chit-chat then a specific question about an order). The assistant responds with an account-linked answer and provides two quick-reply options for follow-up. When the user chooses a follow-up option (a short confirm/yes style reply), the assistant apologizes and again prompts the user to log in to their ASOS account to continue. Steps: Contact landing / initial CTAs → Virtual Assistant open / greeting → User sends free-text message / assistant asks for clarity → User asks order status / assistant suggests My Account → User selects a follow-up option / assistant prompts login **Pattern Brands: Manage Shipping Address** (web, 3 screens, Marketplace) [Refero](https://refero.design/flows/1712) > Happy path: a logged-in user (greeting visible) navigates to My Address from their account navigation, sees an empty-state prompt, clicks Add Address, fills the shipping address form, saves it, and returns to My Address where the new shipping address is shown in a compact card with an Edit control. Steps: Empty address state → Address entry form → Saved address card **ElevenMusic: Track Studio Search and Publish** (web, 3 screens, Music Streaming, Entertainment, AI Tool) [Refero](https://refero.design/flows/12558) > This is a logged-in music creation studio where the user can search, filter, browse, and publish generated tracks. The entry point is the Studio section of the left sidebar. The user can type into the search field to narrow the track list, and the results update in place. If no matches exist, the main area shows an empty state prompting the user to create or upload tracks. The flow ends when the user views the filtered results list or an empty result state; no modal or checkout step is shown. Steps: Browse Tracks → No Match State → Filtered Results ##### Task: supplier onboarding vetting **BeReal.: Add Real People & Brands** (ios, 2 screens, Social Networking) [Refero](https://refero.design/flows/4909) > Summary: This is a discovery/browse screen titled "RealPeople & RealBrands" that presents a scrollable grid of account cards (people and brand entries). Each card shows an avatar/logo, display name, username (in a pill), and an action button. Steps: Browse accounts grid (initial) → Add action feedback **Orbit: Add Subscription - search and create** (ios, 4 screens, Finance) [Refero](https://refero.design/flows/10003) > User is on an Add Subscription screen (mobile, dark theme). Steps: Add Subscription list → Activate search (keyboard open) → Query typed - numeric create suggestion → Text query - provider match **Faire: Faire Buyer Onboarding** (web, 25 screens, Business, Marketplace) [Refero](https://refero.design/flows/13072) > A logged-out visitor starts on Faire’s wholesale marketplace landing screen and chooses to unlock wholesale pricing. They enter a business email, complete account credentials, answer a short retailer onboarding questionnaire, and submit optional business information. After onboarding, they select how they heard about Faire and enter the buyer marketplace. An eligibility promotion for payment terms and a first-order discount modal appear during the transition. The screenshots show the happy path on a desktop browser; no final payment or purchasing action is shown. Steps: Wholesale Marketplace Landing → Email Signup Dialog → Email Submitted → Account Details Validation → Account Details Completed → Shopping Purpose → Business Opening Year → Opening Soon Selected → Expected Opening Date → Business Website → Business Type → Registered Business Name → Business Name Submitted → Payment Terms Offer → Acquisition Source → Referral Source Selected → First-Order Promotion → Buyer Marketplace Home → Product Discovery Grid → Marketplace Access Complete **Programa: Create Supplier Contact** (web, 6 screens, Business, Project Management) [Refero](https://refero.design/flows/13444) > An authenticated company user starts in the Address Book, opens the contact creation form, enters supplier information, and saves the record. The system opens the new contact’s detail view, confirms creation with a success message, and updates the Address Book to show the new supplier alongside the existing contact. The flow assumes the user has permission to create contacts and is using a desktop layout. Steps: Address Book → Empty Contact Form → Completed Contact Form → Contact Detail → Saved Contact Confirmation → Updated Address Book **Dropbox: Dropbox Business onboarding** (web, 5 screens, Collaboration Tools, Business, Productivity) [Refero](https://refero.design/flows/13021) > Happy-path onboarding in a Dropbox Business file workspace. The user starts on the All files home area with guided tour tooltips and a getting-started checklist visible in the left sidebar. They progress through onboarding prompts, learn where to add files, return to All files, and open the setup checklist card. The flow assumes the user is signed in and has access to a team workspace with admin-related setup items. It ends on the main All files view with folders visible and onboarding guidance minimized or dismissed. **ElevenLabs: Signup, Onboarding & Plan Selection** (web, 29 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/8625) > Happy-path: A new visitor arrives on the ElevenLabs marketing home hero, chooses to create an account, completes the signup form (email + password + captcha), receives a verification email, then proceeds through a short multi-step onboarding questionnaire (source, platform choice, role, use case preferences) and selects a subscription plan (monthly or yearly pricing) before landing in the product workspace/dashboard. Entry: marketing home or direct signup link. Exit: authenticated user taken to the Creative Platform workspace. Assumptions: desktop browser, new user (not logged in), user can receive email and interact with reCAPTCHA. Steps: Marketing hero / entry → Create account — initial → Email entered — password rules visible → Password typed — partial validation → Password typed — more checks satisfied → All password requirements satisfied → Password visibility toggle → reCAPTCHA prompt and validation error → Account pending — resend verification → Verification email content → Onboarding — acquisition source → Onboarding — choose platform → Onboarding — role selection → Onboarding — other usage context → Onboarding — use case selection → Onboarding — confirm selections → Pricing — monthly view → Pricing — yearly toggle (discounted) → Pricing — higher tiers & enterprise → First-time workspace / dashboard **Сontra: Freelancer onboarding and profile setup** (web, 15 screens, Job Search) [Refero](https://refero.design/flows/4811) > Happy-path onboarding for a freelance talent joining Contra on desktop. Entry: public discovery/home (not signed in). The user chooses to sign up (Google or email), verifies their email with a code, selects whether they are an independent or a hirer, completes a multi-step profile builder (one-liner, skills, hourly rate, profile photo with cropping, social links), encounters inline validation for links, sees a membership upsell (Contra Pro), and lands in the Discover dashboard with profile completion progress. Steps: Landing / Discover browsing → Enter email or continue with Google → Email input success state → Email verification (code entry) → Role selection: Independent or Hiring → Build your profile — initial form → Profile photo upload — crop modal (open) → Profile photo crop — adjust and save → Profile preview updated → Add social links — entry → Social links — inline validation → Social links — final state → Contra Pro membership upsell → Onboarding complete — Discover dashboard → End state / next actions **Column: Developer Onboarding & Sandbox Setup** (web, 17 screens, Business, Finance, Productivity) [Refero](https://refero.design/flows/1938) > End-to-end new-customer onboarding for a developer-focused banking platform. Entry begins on a public marketing landing, then proceeds to account creation (registration + password validation). After registration the user completes a short multi-step onboarding questionnaire (select financial products, company size, customers, compliance). The onboarding concludes with a summary and the option to "Skip to sandbox" which takes the user into the platform dashboard (sandbox mode) showing setup tasks and a welcome panel. Assumptions: desktop viewport, new user (not previously authenticated), no elevated permissions yet, user intends to evaluate or integrate bank products. Steps: Marketing landing → Create account (empty form) → Create account (initial password entered) → Password suggestions tooltip → Password strength improves → Very strong password state → Password reveal → Product interest questionnaire (step 1) → Product categories selected → Other product detail input → Company size question → Customer types question → Compliance function question → Onboarding summary / next steps → Dashboard (welcome modal) → Dashboard tasks list (with welcome) → Dashboard tasks list (clean) ##### Task: two-factor security device **TikTok: Manage 2-step verification methods** (ios, 3 screens, Entertainment, Social Networking, Video Streaming) [Refero](https://refero.design/flows/2807) > Logged-in mobile user opens the 2-step verification settings to review or change verification methods and trusted devices. The screen lists available methods (Phone, Email, Authenticator, Password) with status indicators and a Manage trusted devices link. Tapping the Phone method reveals a modal showing the masked phone number and actions (Update, Turn off, Cancel). Choosing Turn off brings up a confirmation modal explaining consequences (removes trusted devices, weakens account security) with Cancel and Turn off actions. Steps: 2-step verification overview → Phone method actions → Confirm disabling 2-step **Family: Secret Recovery Setup** (ios, 2 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/2671) > User is shown their Secret Recovery Phrase (partially hidden) and is required to choose at least one secure backup method before continuing. Entry: user has just generated or been shown their recovery phrase. Exit: user selects a backup method (visually highlighted) or reads the security tooltip; the UI records that a backup method must be chosen before proceeding. Steps: Show Secret Phrase & Backup Options → Security Tooltip Display **Fernand: Account Security Settings** (web, 4 screens, Business, Communication Tools) [Refero](https://refero.design/flows/11898) > This is an account security settings area reached from the left sidebar under Personal settings > Security. The user is assumed to be logged in and viewing their own account settings on desktop. The happy path shows reviewing existing password protection and multi-factor authentication, starting a 2FA removal action, confirming it with an authentication code and current password, and returning to the Security screen after successful removal. The flow ends with 2FA disabled and a success message shown near the bottom right. Steps: Security Overview → 2FA Disabled State → Remove 2FA Confirmation → 2FA Removed Success ##### Task: virtual card freeze card **Revolut: Virtual Card Unblock** (web, 3 screens, Business, Finance) [Refero](https://refero.design/flows/1332) > Happy path: a logged-in desktop user (Jane Allen) navigates to Home > Cards, opens a virtual card’s detail pane, chooses the 'Unblock CVV' action, confirms in a modal, and receives a success confirmation that PIN and CVV have been unblocked. Steps: Cards list and detail pane → Confirm unblock modal → Success confirmation ##### Task: vulnerability management **Substack: Manage Blocked Users** (ios, 3 screens, News, Social Networking) [Refero](https://refero.design/flows/4508) > Happy path for viewing and managing blocked users from the Privacy & Safety settings. Steps: Privacy & Safety → Blocked — list with one user → Blocked — empty state **Phantom: Token visibility manager** (ios, 4 screens, Finance, Web3 & Crypto) [Refero](https://refero.design/flows/3717) > User opens a "Manage token list" modal to control which cryptocurrency tokens are visible in their wallet. The screen shows a search field and a scrollable list of tokens; each row displays icon, name, balance, and an on/off toggle. Entry: user has opened the manage token list modal (assumed logged-in wallet state on an iOS device). Exit: user closes the modal (X) after enabling/disabling tokens. The UI updates toggles immediately and filters the list while the user types. Steps: Open manage token list → Focus search field → Type search query → Toggle tokens and close **Lovable: Secrets Management** (web, 4 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12311) > This is a cloud secrets management screen inside a broader admin console. The user is on the Secrets section, where they can add a new secret and review existing saved secrets. The happy path shown is removing an existing secret: the user opens a confirmation dialog from the saved secrets list, confirms deletion, the system processes the request, and then returns to the secrets list with the deleted item removed and a success message shown. Assumption: the user is already authenticated and has permission to manage secrets. Steps: Saved Secrets List → Delete Confirmation → Removal In Progress → Removal Complete **Rork: Manage Project Secrets** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12161) > Happy path for editing project secrets in a developer console. The user starts on a project settings area, opens the Secrets section, adds or edits a secret key/value pair, and saves the configuration. Assumptions: the user is already logged in, has access to the selected project, and can view/edit secrets on a desktop-sized interface. Steps: Open Secrets section → Add secret variable → Save secret changes **Fernand: Support tool sign in** (web, 3 screens, Business, Communication Tools) [Refero](https://refero.design/flows/11908) > Happy path starts on a public marketing landing page for a customer support tool. The user can read product messaging and choose to log in or open an existing workspace. After entering credentials, they arrive at a lightweight workspace start screen with suggested next actions. Assumption: the user already has an account or workspace access and is on desktop. The flow ends when the user is authenticated and presented with the main support dashboard entry point. Steps: Marketing Landing → Workspace Login → Workspace Home **Lovable: Secret Management** (web, 4 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12310) > A logged-in user is in the Cloud > Secrets area of a web dashboard. The user can enter a secret name and secret value, optionally add additional entries, and save the secret. After saving, the new secret appears in the Saved secrets list and a success notification confirms creation. Steps: Enter Secret Details → Edit Secret Value → Reveal or Review Value → Save Secret Successfully **Rork: Project Secrets Management** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12163) > A logged-in user is viewing a project settings area for managing secret variables. The user can switch between sections like Analytics and Secrets, inspect existing key/value pairs, add a new variable, edit values, toggle visibility, delete entries, use Bulk Edit, and save changes. The flow starts in the project workspace and ends after the secret list is updated and saved. Steps: Secrets Overview → Secret Entry Edit → Add or Save Secrets ### research/public-edge/ANALOG-HUMAN-SWITCH.md `research/public-edge/ANALOG-HUMAN-SWITCH.md` · sha256 `d486a369118cebfeddb64279587a5e17ab42c711b318d0b5b71fb1ad41c9df71` #### The analog human switch: no irreversible outcome without the owner, in person Decision: `tok:decision:01m69jxxqzrhz7r36ht4pvv5td`, 31 October 2026. Engine: `runtime/engine.py` (`analog_act`). Test: `tests/analog_switch.py`. Capability: `public-edge/analog-human-switch-for-irreversible-actions`. ##### The rule An AI can never complete an irreversible outcome on its own. That covers weapons, anything that can cause mass harm, and anything else that cannot be undone. Being able to do it is not enough. The outcome happens only when the human owner, in person, makes a physical analog act on hardware that sits in the actuation path itself. Software can prepare, rehearse, request and record. It can never close the circuit. ##### Why software sign-off is not enough A software approval is a signal, and a compromised system can forge any signal. Provisional value (`tok:decision:01m69gty59gnpc17g811317htn`) works for money because it can be reversed after the fact. An irreversible outcome cannot be reversed, so the control has to sit before the effect and outside software. ##### Hardware specification 1. **In series with the actuation circuit.** A normally-open mechanical contact, closed only by a physical key, is wired in series with the power or firing line of the effect. No computer-controlled relay, transistor or bus bypasses it. When it is open, the effect is physically impossible, whatever any software does. 2. **Two physical acts.** - **It was the owner:** the owner turns the registered key at the unit's registered place. - **They definitely want it:** a separate guarded confirm switch, pressed within a short window set by an analog timer. Both must close. 3. **Identical backups.** The owner holds identical registered units, each wired as its own complete path, and stored apart. A broken unit can only fail open (nothing happens). It never blocks the owner, because a backup completes the act. 4. **Revocation.** A lost, stolen or broken unit is physically removed from the circuit and revoked in the register. Backups keep working. 5. **Two-person rule for mass harm.** Where the outcome could harm many people, two owners' keys are wired in series, and both must act in person. 6. **Fail safe.** Every fault opens the circuit: power loss, cut wire, tamper, broken key or timer expiry. 7. **Read-only to software.** Software may read the contact state as evidence. It has no electrical path to write it. 8. **Tamper evidence and testing.** Sealed enclosures; the circuit is tested on a dry load at each inspection; every act and test is recorded with its evidence. ##### What the estate enforces in software (recorded, not trusted) - Every irreversible operation returns `HELD` with `ANALOG_HUMAN_ACT_REQUIRED`, for people and agents alike, whatever the mandate. - `analog_act` runs the effect once, and only for: - the owner who was asked; - a registered, unrevoked unit at its registered place; - a circuit closed by the key; - both confirmations. Anything else is refused with its reason. - `register_switch` and `revoke_switch` keep the backup register. ##### Precedents it builds on (to verify against primary sources before citing externally) - **Two-person rule** and **permissive action links:** physical code and key devices on nuclear weapons, so that no single person can arm them. - **US DoD Directive 3000.09 (updated 2023):** autonomous weapon systems must allow "appropriate levels of human judgment over the use of force". - **Political Declaration on Responsible Military Use of Artificial Intelligence and Autonomy (2023):** endorsed by many states. The estate goes further than each of these. The human act is analog, sits in the actuation path, is backed up by identical units, and is made in person. It is not a policy or a software approval. ### research/public-edge/REFERENCE-STUDIO.md `research/public-edge/REFERENCE-STUDIO.md` · sha256 `e6516dd59e8132d83c06617c1dadd18813ec47e95b6da2a7a6f6b7e83db75ea9` #### Reference for /studio: the Base44 template V7BRO Owner order, 1 November 2026 04:14Z: every layout maps to one named reference; the only differences are added controls or journeys. The reference for Estate Studio (`/studio`, runtime/apps/studio.py, runtime/studio_assets.py) is the Base44 template **V7BRO** (our copy: Base44 app `6ae681783ae584c8c8890bf7`, "V7BRO (Copy)"; the running template at `https://v7bro.base44.app/`, a clone of V7 Go). Measurements and library facts: research/design/BASE44-V7BRO-STUDY.md. Written by Claude-Jegudiel 00:5531102026-1; the same file stands in research/edge-consortium and research/public-edge. ##### 0. Entry | V7BRO element | /studio renders | Difference | |---|---|---| | The workspace opens on the tree with the first project's Build grid; there is no landing page | `GET /studio` → 302 → the first application's Build (`/studio/finance/send-money/build?tree=1`); `/studio/<application>` → 302 → that application's worked example (`/studio/consortium/handle-a-contract-dispute/build`, `/studio/public/employment/build`); `?tree=1` opens the tree on a phone, the grid shows on desktop | none; no hero, no cards | | Base44 sign-in before the workspace | a visitor sees the same shell and tree with, in place of the grid, the capability's name, its requirement and "Continue with the demo" (POST /signin/demo, next = that Build) or "Sign in with EDGE ID" | the demo entry is the estate's | ##### 1. Tree (left) | V7BRO element | /studio renders | Difference | |---|---|---| | 240 px sidebar, `bg hsl(0 0% 98%)` | 240 px `.st-tree`, `--bg-2 #f8fafc`, no margins, border-right 1 px | none | | Workspace switcher (avatar "JW", "Jason's Workspace", chevron, collapse icon) | `.st-ws`: the approved three-dot mark, "Estate Studio", the signed-in identity ("Demo workspace" or the EDGE ID) | switcher replaced by identity; no collapse icon | | Home / Knowledge / Admin | not rendered | removed: the estate's home is `/`, Help is in the page header | | "Agents" eyebrow with count, search box, new-folder and plus icons | "Search capabilities" search box filtering the whole tree | folder-create and plus removed (capabilities come from spec/units, not from the tree) | | Folders (Real Estate & PE, Insurance & Risk, …) with counts | Applications from the contract (EDGE Bank, EDGE Clinic, EDGE Consortium, EDGE Public, Judiverse Suite) with the approved app mark and the live capability count; other applications load on expand | folders are the contract's applications | | Projects under a folder with a green status dot | Capabilities under an application (research units excluded) with a green dot | same | | Build / Review / Visualize under the open project | Build / Review / Visualize under the open capability | same | | Footer: user, plan, tokens, settings | none | removed: identity and sign-out are in the page header | ##### 2. Header row | V7BRO | /studio | Difference | |---|---|---| | 56 px row: breadcrumb "All projects / FNOL Claims Processing · Live" | 56 px `.st-head`: "Estate Studio / EDGE Consortium / Handle a contract dispute · Live" | same shape; crumbs link to `/studio` and `/studio/<app>` | | Segmented Build · Review · Visualize | same, `aria-current` on the open view | same | | "Export data" primary control (black) | "Export data" primary control (teal #005f8a) downloading `export.csv` | brand colour; export is a real CSV of the records and agent columns | | "Ask Go" composer, "…" menu | not rendered | removed | ##### 3. Build grid | V7BRO | /studio | Difference | |---|---|---| | Stage tabs: Main, Human Review, Outdated Information, "+ New stage" with colour squares | "Main n", one tab per value of the entity's status/state/stage field with its count, saved views, "+ New stage" | stages come from the record's state field; New stage creates a saved view (field = value) in Estate Studio's store, the record store is unchanged | | Toolbar: sort, filter, "⚠ 2", "Look up row, column…", Add Row, Add Column | sort, filter (active filter shown as a pill), "Look up row, column…", entity switcher, Add row, Add column | warning count removed (the Review view holds it); entity switcher added | | `role=grid`, checkbox column, row number, 36 px rows, header icon + name | same roles, same columns, 36 px rows, header icon by field kind + label + "Input" or "Agent" | same | | Column types file, text, select, number, python, collection, multiselect, json, with tool Input / AI Model / Python Tool / Workflow | declared field types of the unit (text, textarea, select, enum[], set[], list[], file, number, money, date, boolean …) rendered as text, pills, file chip, "n rows"; agent columns computed by the reasoning plug | columns come from spec/units data, not a literal array | | "Calculating..." cell with pulsing dot | "Calculating…" with pulsing dot, resolving by polling; "Model not connected" when the plug refuses; "— compute" before the first run | never a fake value | | Pills 11 px / 2 px 6 px / 4 px, rose / green / cyan / orange | same metrics; rose, green, orange by meaning, cyan / violet / teal by hash | same | | "+ Add Row" last row; footer "3 rows · Live" | "+ Add row" last row (disabled with the reason when no operation writes the entity); footer "n rows · entity · store · Live" | add row runs the capability's own create / change / send / decide / plan operation | | Row click opens the record detail | same, URL `/studio/<app>/<capability>/build/<record id>` | same | ##### 4. Record detail (row click) | V7BRO | /studio | Difference | |---|---|---| | Overlay over the grid; left pane 42 % `bg-muted`, header ✕ ▲ ▼ …, file name; document preview card | same: `.st-doc` 42 %, ✕ ▲ ▼ … , name, "n / m"; PDF or image inline when the record's file field is one this site serves, else a skeleton with the reason | the "…" menu holds copy id, download JSON, inspect, open the journey | | Right pane: two-column field cells (icon + name + chevron), select fields open a Search box with radio rows and the chosen value as a pill, numbers plain, text plain | same: `.st-fgrid` two columns; enum/select → Search + radio pills; set/list[enum]/multiselect → Search + checkbox pills + Save; list of objects → nested table; file → chip; number / date / text → inline input; textarea → textarea | edits save through the capability's change (or send / decide / plan) operation with `expected_version`; "Saved · operation · verified · vN" or the refusal code | | Agent fields show the number (Fraud Risk 84) | agent-computed section: value, pill or number; click opens how it was computed in the inspector | same | | — | "Other data on this record": keys the unit does not declare | added | ##### 5. Visualize canvas | V7BRO | /studio | Difference | |---|---|---| | Nodes Source / Condition / Stage: 9 px uppercase caption, 40 px rounded box with colour dot, label, count; input dot left, output dot right | Screen / Condition / Stage nodes with the same chrome; screens and conditions from the unit, stages from action outcomes (success, error, journey:end, other capabilities) | node kinds come from the unit | | Cubic Bézier edges `M x1 y1 C …`; drag from an output dot onto an input dot creates an edge | same path; failure edges dashed red, condition edges amber; dot-to-dot drag saves a **draft** edge (badge "draft ×n") in Estate Studio's store; the unit is never changed | drafts instead of edits | | Node drag on pointer down; hard-coded x/y | node drag; positions saved per tenant; first layout three screens per row, conditions beneath, stages at the right | positions persist | | Pan and wheel zoom | pan, wheel zoom 0.2–2.5 around the cursor, + / − / fit buttons, minimap with viewport (click to jump) | minimap and fit added (the original has none) | | — | second tab **Knowledge graph**: the capability's TOK neighbourhood from `.cache/tok-dependencies.sqlite` (read-only, indexed lookups, 8 s deadline), draggable; says so when the index is absent, empty or slow | added | ##### 6. Right inspector | V7BRO | /studio | Difference | |---|---|---| | Sections Source, Tool, Reasoning (Low), Model (Concierge), Input | Source (store, tenant, capability, entity, id, version), Tool (operation, endpoint, MCP tool), Reasoning (error bound, evidence, guards), Model (plug, provider, model name, connected; never a key), Input (prompt or raw value) for the selected row, field, cell, node or edge | same sections, real provenance | ##### 7. Added by the estate (not in V7BRO) Applications and capabilities from the contract and spec/units; records from the estate record store per tenant; every write through the capability's own operations with receipts and versions; agent columns through the reasoning plug with the error bound shown; saved views; entity switcher; CSV export of real data; Review view (what the store confirmed, events, what the model produced); Knowledge graph tab; minimap and fit; positions and drafts per tenant; phone layout (tree behind a button, opened on entry at /studio; grid scrolls, canvas full-bleed, inspector as a bottom sheet); demo entry for visitors; the approved marks and Inter; teal #005f8a instead of violet/orange; no external script, font or CDN. ##### 8. Not yet built Document preview for files the site does not serve; file upload from the detail view; Review as a workflow queue (it is a ledger today); editing agent columns' prompts in place (delete and re-add); knowledge-graph depth beyond one hop; the "Ask" composer. ### Competitor clauses (research/parity) | Competitor file | Clauses | First source | |---|---|---| | `research/parity/clinic/bahmni-coalition.jsonl` | 3 | https://api.github.com/repos/Bahmni/openmrs-module-bahmniapps/readme | | `research/parity/clinic/enlivio-health.jsonl` | 19 | https://www.emishealth.com/products/emis-web | | `research/parity/clinic/epic-systems.jsonl` | 622 | https://open.epic.com/Interface/FHIR | | `research/parity/clinic/nhs-app.jsonl` | 8 | https://www.nhs.uk/nhs-app/help/appointments/book-gp-appointment/ | | `research/parity/clinic/nhs-england.jsonl` | 20 | https://www.nhs.uk/nhs-app/ | | `research/parity/clinic/openemr-project.jsonl` | 148 | https://api.github.com/repos/openemr/openemr/contents/Documentation/api/STANDARD_API.md | | `research/parity/clinic/openmrs.jsonl` | 53 | https://api.github.com/repos/openmrs/openmrs-esm-patient-chart/contents/packages/esm-patient-orders-app/README.md | | `research/parity/clinic/oracle.jsonl` | 2 | https://docs.oracle.com/en/industries/health/millennium-platform-apis/ | | `research/parity/clinic/patient-access.jsonl` | 14 | https://www.patientaccess.com/ | | `research/parity/clinic/zocdoc.jsonl` | 5 | https://mobbin.com/flows/fcd6461e-af19-4ca9-9864-8b21c286888c | | `research/parity/finance/adyen.jsonl` | 1 | https://docs.adyen.com/development-resources/api-credentials/ | | `research/parity/finance/barclays.jsonl` | 12 | http://www.barclays.co.uk/business-banking/accounts/ | | `research/parity/finance/british-business-bank.jsonl` | 7 | https://www.british-business-bank.co.uk/finance-options | | `research/parity/finance/gocardless.jsonl` | 3 | https://developer.gocardless.com/docs/api-reference/responses-and-errors | | `research/parity/finance/monzo.jsonl` | 182 | https://monzo.com/help/opening-an-account/business-verifying-persons-of-significant-control-at-sign-up | | `research/parity/finance/open-banking-uk.jsonl` | 126 | https://raw.githubusercontent.com/OpenBankingUK/read-write-api-specs/master/dist/openapi/account-info-openapi.json | | `research/parity/finance/oracle.jsonl` | 11 | https://docs.oracle.com/en/cloud/saas/netsuite/index.html | | `research/parity/finance/revolut.jsonl` | 336 | https://developer.revolut.com/docs/api/business/operations/add-payment-draft-payment.md | | `research/parity/finance/starling.jsonl` | 101 | https://www.starlingbank.com/current-account/joint-bank-account/ | | `research/parity/finance/stripe.jsonl` | 9 | https://docs.stripe.com/security | | `research/parity/finance/tide.jsonl` | 28 | https://www.tide.co/features/ | | `research/parity/finance/wise.jsonl` | 4 | https://docs.wise.com/guides/product/send-money | | `research/parity/finance/xero.jsonl` | 5 | https://mobbin.com/flows/957b90d2-7949-4ea2-97b6-18e7ad4f958e | | `research/parity/foundation/auth0.jsonl` | 8 | https://auth0.com/features/ | | `research/parity/foundation/cyberark.jsonl` | 5 | https://www.cyberark.com/products/ | | `research/parity/foundation/delinea.jsonl` | 6 | https://delinea.com/products | | `research/parity/foundation/duo.jsonl` | 10 | https://duo.com/product | | `research/parity/foundation/government-digital-service.jsonl` | 24 | https://docs.sign-in.service.gov.uk/integrate-with-integration-environment/prove-users-identity/ | | `research/parity/foundation/incogni-01.jsonl` | 1437 | https://support.incogni.com/hc/en-us/articles/5285682832402-In-which-countries-is-Incogni-available | | `research/parity/foundation/incogni-02.jsonl` | 288 | https://incogni.com/legal/terms-of-service | | `research/parity/foundation/keycloak.jsonl` | 141 | https://www.keycloak.org/docs/latest/server_admin/ | | `research/parity/foundation/microsoft.jsonl` | 12 | https://learn.microsoft.com/en-us/entra/fundamentals/what-is-entra | | `research/parity/foundation/okta.jsonl` | 14 | https://help.okta.com/oie/en-us/content/topics/users-groups-profiles/usgp-add-users.htm | | `research/parity/foundation/ping-identity.jsonl` | 11 | https://www.pingidentity.com/en/platform.html | | `research/parity/foundation/sailpoint.jsonl` | 5 | https://www.sailpoint.com/products/identity-security-cloud | | `research/parity/foundation/saviynt.jsonl` | 4 | https://saviynt.com/products/identity-governance-and-administration/ | | `research/parity/judiverse/15five.jsonl` | 16 | https://www.15five.com/ | | `research/parity/judiverse/37signals.jsonl` | 17 | https://basecamp.com/features | | `research/parity/judiverse/activepieces.jsonl` | 6 | https://raw.githubusercontent.com/activepieces/activepieces/main/README.md | | `research/parity/judiverse/agno.jsonl` | 3 | https://raw.githubusercontent.com/agno-agi/agno/main/README.md | | `research/parity/judiverse/airtable.jsonl` | 7 | https://www.airtable.com/platform | | `research/parity/judiverse/alibaba-qwen.jsonl` | 4 | https://raw.githubusercontent.com/QwenLM/Qwen3/main/README.md | | `research/parity/judiverse/anthropic.jsonl` | 161 | https://code.claude.com/docs/en/sandboxing.md | | `research/parity/judiverse/apache-software-foundation.jsonl` | 4 | https://raw.githubusercontent.com/apache/airflow/main/README.md | | `research/parity/judiverse/arize-ai.jsonl` | 10 | https://raw.githubusercontent.com/Arize-ai/phoenix/main/README.md | | `research/parity/judiverse/asana.jsonl` | 7 | https://asana.com/features | | `research/parity/judiverse/atlassian.jsonl` | 82 | https://www.atlassian.com/software/confluence/features | | `research/parity/judiverse/avoma.jsonl` | 30 | https://www.avoma.com/ | | `research/parity/judiverse/aws.jsonl` | 4 | https://raw.githubusercontent.com/aws/sagemaker-python-sdk/master/README.rst | | `research/parity/judiverse/bentoml.jsonl` | 5 | https://raw.githubusercontent.com/bentoml/BentoML/main/README.md | | `research/parity/judiverse/berriai.jsonl` | 4 | https://raw.githubusercontent.com/BerriAI/litellm/main/README.md | | `research/parity/judiverse/cisco.jsonl` | 10 | https://www.webex.com/suite/meetings.html | | `research/parity/judiverse/clickup.jsonl` | 50 | https://clickup.com/features | | `research/parity/judiverse/cohere.jsonl` | 1 | https://raw.githubusercontent.com/cohere-ai/cohere-python/main/README.md | | `research/parity/judiverse/corporation-for-digital-scholarship.jsonl` | 14 | https://www.zotero.org/support/quick_start_guide | | `research/parity/judiverse/crewai.jsonl` | 25 | https://raw.githubusercontent.com/crewAIInc/crewAI/main/README.md | | `research/parity/judiverse/crossref.jsonl` | 26 | https://api.crossref.org/swagger-docs | | `research/parity/judiverse/customer-io.jsonl` | 19 | https://docs.customer.io/ | | `research/parity/judiverse/daytona.jsonl` | 5 | https://raw.githubusercontent.com/daytonaio/daytona/main/README.md | | `research/parity/judiverse/deepseek.jsonl` | 9 | https://raw.githubusercontent.com/deepseek-ai/DeepSeek-V3/main/README.md | | `research/parity/judiverse/dify.jsonl` | 2 | https://raw.githubusercontent.com/langgenius/dify/main/README.md | | `research/parity/judiverse/elicit.jsonl` | 23 | https://elicit.com/solutions/systematic-review | | `research/parity/judiverse/evernote.jsonl` | 7 | https://evernote.com/features | | `research/parity/judiverse/fathom.jsonl` | 8 | https://fathom.video/ | | `research/parity/judiverse/figma.jsonl` | 251 | https://raw.githubusercontent.com/figma/rest-api-spec/main/openapi/openapi.yaml | | `research/parity/judiverse/fireflies-ai.jsonl` | 84 | https://fireflies.ai/features | | `research/parity/judiverse/google.jsonl` | 422 | https://workspace.google.com/products/gmail/ | | `research/parity/judiverse/guardrails-ai.jsonl` | 1 | https://raw.githubusercontent.com/guardrails-ai/guardrails/main/README.md | | `research/parity/judiverse/helicone.jsonl` | 3 | https://raw.githubusercontent.com/Helicone/helicone/main/README.md | | `research/parity/judiverse/kestra.jsonl` | 36 | https://raw.githubusercontent.com/kestra-io/kestra/develop/README.md | | `research/parity/judiverse/langchain.jsonl` | 37 | https://raw.githubusercontent.com/langchain-ai/deepagents/master/README.md | | `research/parity/judiverse/langflow.jsonl` | 5 | https://raw.githubusercontent.com/langflow-ai/langflow/main/README.md | | `research/parity/judiverse/langfuse.jsonl` | 7 | https://raw.githubusercontent.com/langfuse/langfuse/main/README.md | | `research/parity/judiverse/linear.jsonl` | 42 | https://linear.app/features | | `research/parity/judiverse/llamaindex.jsonl` | 9 | https://raw.githubusercontent.com/run-llama/llama_index/main/README.md | | `research/parity/judiverse/mastra.jsonl` | 8 | https://raw.githubusercontent.com/mastra-ai/mastra/main/README.md | | `research/parity/judiverse/microsoft.jsonl` | 808 | https://mobbin.com/flows/7bb08323-2ede-4abe-bf61-cbd07d0fc273 | | `research/parity/judiverse/miro.jsonl` | 11 | https://miro.com/online-whiteboard/ | | `research/parity/judiverse/mistral-ai.jsonl` | 52 | https://raw.githubusercontent.com/mistralai/client-python/main/README.md | | `research/parity/judiverse/mlflow.jsonl` | 4 | https://raw.githubusercontent.com/mlflow/mlflow/master/README.md | | `research/parity/judiverse/mobbin.jsonl` | 3 | https://mobbin.com/ | | `research/parity/judiverse/monday-com.jsonl` | 18 | https://monday.com/features | | `research/parity/judiverse/n8n.jsonl` | 7 | https://raw.githubusercontent.com/n8n-io/n8n/master/README.md | | `research/parity/judiverse/notion.jsonl` | 62 | https://www.notion.com/help/sharing-and-permissions | | `research/parity/judiverse/nvidia.jsonl` | 8 | https://raw.githubusercontent.com/NVIDIA/NeMo-Guardrails/develop/README.md | | `research/parity/judiverse/obsidian.jsonl` | 4 | https://obsidian.md/ | | `research/parity/judiverse/ollama.jsonl` | 21 | https://raw.githubusercontent.com/ollama/ollama/main/README.md | | `research/parity/judiverse/openpipe.jsonl` | 5 | https://raw.githubusercontent.com/openpipe/openpipe/main/README.md | | `research/parity/judiverse/otter-ai.jsonl` | 4 | https://otter.ai/ | | `research/parity/judiverse/ourresearch.jsonl` | 10 | https://docs.openalex.org/api-entities/works | | `research/parity/judiverse/palantir-aip.jsonl` | 354 | https://www.palantir.com/docs/foundry/agent-builder | | `research/parity/judiverse/palantir-foundry-01.jsonl` | 1654 | https://www.palantir.com/docs/foundry | | `research/parity/judiverse/palantir-foundry-02.jsonl` | 1664 | https://www.palantir.com/docs/foundry/foundry-rules/author-and-run-a-rule | | `research/parity/judiverse/palantir-foundry-03.jsonl` | 1280 | https://www.palantir.com/docs/foundry/pipeline-builder/datasets-generated | | `research/parity/judiverse/palantir-technologies-aip-01.jsonl` | 1447 | https://www.palantir.com/docs/foundry/agent-builder | | `research/parity/judiverse/palantir-technologies-aip-02.jsonl` | 973 | https://www.palantir.com/docs/foundry/assist/adding-documentation-to-aip-assist | | `research/parity/judiverse/palantir-technologies-foundry-01.jsonl` | 1447 | https://www.palantir.com/docs/foundry | | `research/parity/judiverse/palantir-technologies-foundry-02.jsonl` | 1556 | https://www.palantir.com/docs/foundry/announcements/2023-11 | | `research/parity/judiverse/palantir-technologies-foundry-03.jsonl` | 1399 | https://www.palantir.com/docs/foundry/announcements/2022-08 | | `research/parity/judiverse/palantir-technologies-foundry-04.jsonl` | 1434 | https://www.palantir.com/docs/foundry/announcements/2025-04 | | `research/parity/judiverse/palantir-technologies-foundry-05.jsonl` | 1279 | https://www.palantir.com/docs/foundry/announcements/2024-10 | | `research/parity/judiverse/palantir-technologies-foundry-06.jsonl` | 1442 | https://www.palantir.com/docs/foundry/api/sds-v2-resources | | `research/parity/judiverse/palantir-technologies-foundry-07.jsonl` | 1446 | https://www.palantir.com/docs/foundry/available-connectors/databricks | | `research/parity/judiverse/palantir-technologies-foundry-08.jsonl` | 1425 | https://www.palantir.com/docs/foundry/available-connectors/sap-custom-source | | `research/parity/judiverse/palantir-technologies-foundry-09.jsonl` | 1442 | https://www.palantir.com/docs/foundry/carbon/workspaces-navigation | | `research/parity/judiverse/palantir-technologies-foundry-10.jsonl` | 1454 | https://www.palantir.com/docs/foundry/code-workspaces/data | | `research/parity/judiverse/palantir-technologies-foundry-11.jsonl` | 1413 | https://www.palantir.com/docs/foundry/contour/analysis-parameterize | | `research/parity/judiverse/palantir-technologies-foundry-12.jsonl` | 1423 | https://www.palantir.com/docs/foundry/data-connection/listeners-subdomains | | `research/parity/judiverse/palantir-technologies-foundry-13.jsonl` | 1408 | https://www.palantir.com/docs/foundry/dataset-preview | | `research/parity/judiverse/palantir-technologies-foundry-14.jsonl` | 1433 | https://www.palantir.com/docs/foundry/foundry-adoption | | `research/parity/judiverse/palantir-technologies-foundry-15.jsonl` | 1428 | https://www.palantir.com/docs/foundry/functions/typescript-v2-migration | | `research/parity/judiverse/palantir-technologies-foundry-16.jsonl` | 1419 | https://www.palantir.com/docs/foundry/getting-started/foundry-platform-summary-llm | | `research/parity/judiverse/palantir-technologies-foundry-17.jsonl` | 1448 | https://www.palantir.com/docs/foundry/integrate-models/experiment-reference | | `research/parity/judiverse/palantir-technologies-foundry-18.jsonl` | 1441 | https://www.palantir.com/docs/foundry/map/selection | | `research/parity/judiverse/palantir-technologies-foundry-19.jsonl` | 1401 | https://www.palantir.com/docs/foundry/notepad/snapshot-widgets | | `research/parity/judiverse/palantir-technologies-foundry-20.jsonl` | 1415 | https://www.palantir.com/docs/foundry/object-link-types/metadata-statuses | | `research/parity/judiverse/palantir-technologies-foundry-21.jsonl` | 1406 | https://www.palantir.com/docs/foundry/ontology/using-custom-models-to-create-a-semantic-search-workflow | | `research/parity/judiverse/palantir-technologies-foundry-22.jsonl` | 1498 | https://www.palantir.com/docs/foundry/palantir-extension-for-visual-studio-code/transforms-preview | | `research/parity/judiverse/palantir-technologies-foundry-23.jsonl` | 1454 | https://www.palantir.com/docs/foundry/pipeline-builder/management-input-sampling | | `research/parity/judiverse/palantir-technologies-foundry-24.jsonl` | 1471 | https://www.palantir.com/docs/foundry/pipeline-builder/transforms-overview | | `research/parity/judiverse/palantir-technologies-foundry-25.jsonl` | 1526 | https://www.palantir.com/docs/foundry/quiver/objects-filter | | `research/parity/judiverse/palantir-technologies-foundry-26.jsonl` | 1471 | https://www.palantir.com/docs/foundry/reports/change-parameter | | `research/parity/judiverse/palantir-technologies-foundry-27.jsonl` | 1436 | https://www.palantir.com/docs/foundry/slate/best-practices-app-functionality | | `research/parity/judiverse/palantir-technologies-foundry-28.jsonl` | 1405 | https://www.palantir.com/docs/foundry/time-series/time-series-properties | | `research/parity/judiverse/palantir-technologies-foundry-29.jsonl` | 1432 | https://www.palantir.com/docs/foundry/transforms-r/getting-started | | `research/parity/judiverse/palantir-technologies-foundry-30.jsonl` | 1451 | https://www.palantir.com/docs/foundry/workflow-lineage/manage-security | | `research/parity/judiverse/palantir-technologies-foundry-31.jsonl` | 578 | https://www.palantir.com/docs/foundry/workshop/redact-mode | | `research/parity/judiverse/portkey.jsonl` | 12 | https://raw.githubusercontent.com/Portkey-AI/gateway/main/README.md | | `research/parity/judiverse/pydantic.jsonl` | 11 | https://raw.githubusercontent.com/pydantic/pydantic-ai/main/README.md | | `research/parity/judiverse/read-ai.jsonl` | 5 | https://www.read.ai/ | | `research/parity/judiverse/salesforce.jsonl` | 98 | https://slack.com/help/articles/203457187-Customize-data-retention-in-Slack | | `research/parity/judiverse/sglang-project.jsonl` | 2 | https://raw.githubusercontent.com/sgl-project/sglang/main/README.md | | `research/parity/judiverse/swe-agent-maintainers.jsonl` | 7 | https://raw.githubusercontent.com/SWE-agent/mini-swe-agent/main/README.md | | `research/parity/judiverse/tana.jsonl` | 19 | https://tana.inc/ | | `research/parity/judiverse/tl-dv.jsonl` | 7 | https://tldv.io/ | | `research/parity/judiverse/ux-pilot.jsonl` | 10 | https://uxpilot.ai/ | | `research/parity/judiverse/weights-biases.jsonl` | 4 | https://raw.githubusercontent.com/wandb/weave/master/README.md | | `research/parity/judiverse/wikimedia-foundation.jsonl` | 17 | https://www.wikidata.org/wiki/Wikidata:Data_access | | `research/parity/judiverse/zoom.jsonl` | 10 | https://www.zoom.com/en/products/virtual-meetings/ | | `research/parity/public-consortium/15five.jsonl` | 1 | https://mobbin.com/flows/8be79f86-d72c-4768-9a57-ef1a0c0c38b2 | | `research/parity/public-consortium/auditboard.jsonl` | 6 | https://auditboard.com/platform | | `research/parity/public-consortium/brilliant-worldwide.jsonl` | 9 | https://brilliant.org/about/ | | `research/parity/public-consortium/cloudflare-zero-trust.jsonl` | 9 | https://www.cloudflare.com/zero-trust/products/ | | `research/parity/public-consortium/companies-house.jsonl` | 9 | https://developer.company-information.service.gov.uk/how-to-create-an-application | | `research/parity/public-consortium/coursera.jsonl` | 19 | https://www.coursera.org/business/compare-plans | | `research/parity/public-consortium/credo-ai.jsonl` | 13 | https://www.credo.ai/platform | | `research/parity/public-consortium/crowdstrike.jsonl` | 5 | https://www.crowdstrike.com/en-us/platform/ | | `research/parity/public-consortium/customer-io.jsonl` | 2 | https://mobbin.com/flows/a1ebdb08-3cf3-4330-a6c0-370272087f9d | | `research/parity/public-consortium/d2l.jsonl` | 126 | https://docs.valence.desire2learn.com/reference.html | | `research/parity/public-consortium/darktrace.jsonl` | 10 | https://www.darktrace.com/products | | `research/parity/public-consortium/datadog.jsonl` | 6 | https://www.datadoghq.com/product/ | | `research/parity/public-consortium/dependabot.jsonl` | 5 | https://docs.github.com/en/code-security/dependabot/working-with-dependabot | | `research/parity/public-consortium/drata.jsonl` | 13 | https://help.drata.com/ | | `research/parity/public-consortium/elastic.jsonl` | 9 | https://www.elastic.co/security | | `research/parity/public-consortium/futurelearn.jsonl` | 15 | https://www.futurelearn.com/using-futurelearn | | `research/parity/public-consortium/github-advanced-security.jsonl` | 4 | https://docs.github.com/en/get-started/learning-about-github/about-github-advanced-security | | `research/parity/public-consortium/government-digital-service.jsonl` | 55 | https://www.forms.service.gov.uk/features/ | | `research/parity/public-consortium/grafana.jsonl` | 2 | https://grafana.com/docs/grafana/latest/alerting/ | | `research/parity/public-consortium/honeycomb.jsonl` | 6 | https://www.honeycomb.io/platform | | `research/parity/public-consortium/hyperproof.jsonl` | 5 | https://hyperproof.io/ | | `research/parity/public-consortium/ibm.jsonl` | 4 | https://www.ibm.com/products/watsonx-governance | | `research/parity/public-consortium/instructure-01.jsonl` | 1631 | https://canvas.instructure.com/doc/api/all_resources.html | | `research/parity/public-consortium/instructure-02.jsonl` | 657 | https://canvas.instructure.com/doc/api/all_resources.html | | `research/parity/public-consortium/logicgate.jsonl` | 11 | https://www.logicgate.com/platform/ | | `research/parity/public-consortium/metricstream.jsonl` | 7 | https://www.metricstream.com/ | | `research/parity/public-consortium/microsoft.jsonl` | 10 | https://learn.microsoft.com/en-us/defender/ | | `research/parity/public-consortium/moodle.jsonl` | 30 | https://docs.moodle.org/501/en/Features | | `research/parity/public-consortium/mysociety.jsonl` | 6 | https://raw.githubusercontent.com/mysociety/fixmystreet/master/README.md | | `research/parity/public-consortium/onetrust.jsonl` | 9 | https://www.onetrust.com/products/ | | `research/parity/public-consortium/opentelemetry.jsonl` | 7 | https://opentelemetry.io/docs/what-is-opentelemetry/ | | `research/parity/public-consortium/palantir-apollo.jsonl` | 448 | https://www.palantir.com/docs/apollo | | `research/parity/public-consortium/palantir-foundry.jsonl` | 475 | https://www.palantir.com/docs/foundry/administration/configure-code-workbook-profiles | | `research/parity/public-consortium/palantir-gotham.jsonl` | 47 | https://www.palantir.com/docs/gotham | | `research/parity/public-consortium/palantir-technologies-apollo-01.jsonl` | 1420 | https://www.palantir.com/docs/apollo | | `research/parity/public-consortium/palantir-technologies-apollo-02.jsonl` | 1406 | https://www.palantir.com/docs/apollo/core/whats-new | | `research/parity/public-consortium/palantir-technologies-apollo-03.jsonl` | 177 | https://www.palantir.com/docs/apollo/managing-vulnerabilities | | `research/parity/public-consortium/palantir-technologies-foundry-01.jsonl` | 1334 | https://www.palantir.com/docs/foundry/administration/analyze-user-activity-metrics | | `research/parity/public-consortium/palantir-technologies-foundry-02.jsonl` | 1371 | https://www.palantir.com/docs/foundry/authentication/oidc-getting-started | | `research/parity/public-consortium/palantir-technologies-foundry-03.jsonl` | 1061 | https://www.palantir.com/docs/foundry/security/audit-logs-overview | | `research/parity/public-consortium/palantir-technologies-gotham.jsonl` | 360 | https://www.palantir.com/docs/gotham | | `research/parity/public-consortium/prisma-cloud.jsonl` | 5 | https://www.paloaltonetworks.com/prisma/cloud | | `research/parity/public-consortium/prometheus.jsonl` | 7 | https://prometheus.io/docs/introduction/overview/ | | `research/parity/public-consortium/qualys.jsonl` | 7 | https://www.qualys.com/apps/ | | `research/parity/public-consortium/secureframe.jsonl` | 4 | https://secureframe.com/ | | `research/parity/public-consortium/sentinelone.jsonl` | 7 | https://www.sentinelone.com/platform/ | | `research/parity/public-consortium/servicenow-grc.jsonl` | 7 | https://www.servicenow.com/products/integrated-risk-management.html | | `research/parity/public-consortium/snyk.jsonl` | 8 | https://snyk.io/product/ | | `research/parity/public-consortium/sonarqube.jsonl` | 16 | https://www.sonarsource.com/products/sonarqube/ | | `research/parity/public-consortium/splunk.jsonl` | 8 | https://www.splunk.com/en_us/products/enterprise-security.html | | `research/parity/public-consortium/sprinto.jsonl` | 5 | https://sprinto.com/ | | `research/parity/public-consortium/tenable.jsonl` | 17 | https://www.tenable.com/products | | `research/parity/public-consortium/twingate.jsonl` | 9 | https://mobbin.com/flows/1e35abbb-b23d-4398-94ff-b5975007c242 | | `research/parity/public-consortium/udemy.jsonl` | 16 | https://business.udemy.com/plans/ | | `research/parity/public-consortium/uk-research-and-innovation.jsonl` | 17 | https://apply-for-innovation-funding.service.gov.uk/competition/search | | `research/parity/public-consortium/vanta.jsonl` | 19 | https://mobbin.com/flows/94d183df-de59-417d-bc58-44297c44dd5c | | `research/parity/public-consortium/wiz.jsonl` | 12 | https://www.wiz.io/platform | | `research/parity/public-consortium/workiva.jsonl` | 7 | https://www.workiva.com/platform | | `research/parity/public-consortium/zscaler.jsonl` | 7 | https://www.zscaler.com/products-and-solutions/zero-trust-exchange | | `research/parity/unassigned/airbnb.jsonl` | 3 | https://www.airbnb.co.uk/help/article/14 | | `research/parity/unassigned/booking-com.jsonl` | 45 | https://www.booking.com/content/how_we_work.html | | `research/parity/unassigned/citymapper.jsonl` | 6 | https://citymapper.com/ | | `research/parity/unassigned/deliveroo.jsonl` | 22 | https://deliveroo.co.uk/faq | | `research/parity/unassigned/headspace.jsonl` | 4 | https://www.headspace.com/meditation | | `research/parity/unassigned/instacart.jsonl` | 12 | https://www.instacart.com/help | | `research/parity/unassigned/shopify.jsonl` | 77 | https://help.shopify.com/en/manual/online-store/storefront-search/search-and-discovery-filters | | `research/parity/unassigned/strava.jsonl` | 5 | https://support.strava.com/hc/en-us/articles/216917657-Strava-Subscription | | `research/parity/unassigned/too-good-to-go.jsonl` | 4 | https://www.toogoodtogo.com/en-us/how-does-the-app-work | ## 5. Documentation ### README.md `README.md` · sha256 `05245f9617828ced989dd83e7bbcae49e66b9b6d0f4db355a32735394de28bd2` #### JudeAI | EDGE ID — governed intelligence autonomous agent estate **Your data. Your intelligence. Your authority. Our responsibility: work done right, with independently verified outcomes.** **New estate and old estate.** This folder, `new-estate/`, is the **new estate**: the new specification and the only current one. Everything outside it in the repository `governed-intelligence-judeai-edgeid` is the **old estate** (the old specification): research only, never instructions. Its identities and contract bindings are never carried over. Run every command below from inside `new-estate/`. This estate is the living source of truth for the estate: every application, foundation, capability, journey, screen, control, operation, agent, procedure and piece of evidence. Everything in it is an object in TOK, the one ontology; everything else is generated from TOK. This is the only instruction file. Deeper material is read on demand. ##### Truth, agreement, claim Truth is complete and already exists; it can only be observed. Agreement creates laws, rules, standards and plans, and agreements can be disagreed with. A claim is what someone asserts. Every statement here is exactly one of the three, and an agreement or a claim is never presented as an observation. - **OBSERVATION**: what was observed, how, by whom, when, how reliably, on what evidence. - **AGREEMENT**: who agreed, who dissented, with what authority, where and from when it applies, on what basis. - **CLAIM**: who asserts it, on what evidence, who supports and who disputes it. Actions obey agreements: law, permission, policy. Knowledge follows observation: evidence weighted by independence, never by headcount. Where an agreement departs from the evidence, the estate shows both and breaks neither. ##### How to work here 1. **Bind the contract.** `python3 tools/gate.py snapshot` prints the contract version and fingerprint. Bind the exact fingerprint; a stale one authorises nothing. 2. **Everything is a TOK object.** Nothing enters as a loose file. `python3 tools/admit.py` mints a permanent identity, records provenance, digest, licence and admitter, and places the object. Copying is not admitting. 3. **Build from the ontology.** `python3 tools/build.py` regenerates `generated/` byte for byte. Never edit `generated/`. 4. **Change rules by amendment.** `python3 tools/amend.py` records what a change supersedes or narrows and retires it in the same commit. An addition that reconciles nothing is refused. 5. **Gate before merge.** `python3 tools/gate.py` runs every gate. `main` requires the `estate-ci` status; a failing gate cannot merge. 6. **Say only what the status allows.** VERIFIED, OBSERVED, HYPOTHESIS, PREDICTION, SIMULATION, UNKNOWN, REFUTED, or an explicitly labelled ASPIRATION on a claim about the future. An attempt is not a completion; a rehearsal is not an outcome; an intention is not an observation. Nobody fabricates evidence, authority, benchmarks or readiness. 7. **Every number is live.** A count is produced by the gate at the moment it is stated, with its method. No count is written by hand. 8. **Agents are native users.** Every operation has a human path and an agent path to the same outcome, under the agent's own EDGE ID, with the same authority, preview, audit and undo. Agents work from their portal, mailbox and racetrack, and rehearse consequential actions before committing them. 9. **Keep the screen clear.** The essential result first; depth behind a help control as markdown, available to people and agents alike on demand. ##### Layout | Path | Holds | |---|---| | `tok/kernel.json`, `tok/KERNEL.md` | The kernel: statement classes, kinds, relations, identity, admission, gates | | `tok/objects/<kind>/<ulid>.json` | One file per object; the path is the identity | | `contract/estate-contract.json` | The contract: intentions, outcomes, principles, applications, foundations | | `spec/SPECIFICATION.md` | How every object is specified, the required journeys, the outcome ledger, what each application would be, what complete means | | `research/<application>/` | Research written for the estate: competitors, regulators, gaps, reusable systems, real screens and flows, journeys and endpoints. Every finding carries its source link and date, and is marked observation, agreement or claim | | `sources/` | Admitted sources, each with its TOK record | | `src/`, `tests/` | Code modules and their tests, each a TOK object with declared dependencies | | `tools/` | `admit.py`, `amend.py`, `build.py`, `gate.py` | | `generated/` | Plan, views, indexes, diagrams, live counts: build output only | | `design/brand/` | The estate's logos, used as-is and never recreated | | `ci/` | The self-hosted CI runner and its host template | ##### What enters, and what never does Only what this estate produces: TOK objects, generated output, code with its tests, and the estate's original logo set as design assets. Research (statements of intention, regulations, standards, competitor documentation, benchmarks, source code, diagrams, datasets) is held in storage, referenced by digest, and used by the factory to produce new objects; it never enters and is never authority. Every build starts from the current sources and the current agreement. Never: credentials; machine probes; host access maps; personal contact data; raw datasets, scrapes and large generated data; duplicate or stale generated output; any rule, budget, permission, approval, count, identity or readiness level not produced in this estate. ##### Words used here | Word | Meaning | |---|---| | Estate | The whole product: <!-- views:applications:start -->three applications (EDGE Finance, EDGE Clinic, EDGE Public Consortium) plus Judiverse, which comes with all three<!-- views:applications:end -->, their shared foundations, and every person and agent working in them | | TOK | The catalogue of everything in the estate (apps, screens, actions, agents, rules, evidence), each with a permanent ID and links to what it relates to. Also called the ontology | | Object | One entry in TOK: one file under `tok/objects/` | | Kernel | The rules every object must follow (`tok/kernel.json`) | | Gate | An automatic check run on every change; a failing gate blocks the change | | Observation | Something seen or measured about reality, recorded with how, by whom, when and how reliably | | Agreement | Something people agreed (a law, rule, standard or plan), recorded with who agreed, who disagreed and on what basis; it can be disagreed with | | Claim | Something someone asserts, recorded with their evidence and who supports or disputes it | | Role | A named responsibility held by a person or organisation (never by an agent alone) that answers for what it owns | | Definition | The plain-language meaning of a term used anywhere in the estate | | Unknown | The truth status of anything not yet observed; never counted as zero | | Behaviour model | The one description every action must have: who, permission, checks, effect, confirmation, success, failure, recovery, evidence | | Agent | An AI actor with its own identity, owner, mandate and term, acting as a customer or staff member | | Mandate | What an agent may do, for whom, with what limits and until when | | Racetrack | An agent's private rehearsal space, where it runs an action before committing it | | Receiving confirmation | The receiving system's own confirmation that an action really happened | | Procedure | A compiled, deterministic piece of logic that runs without a large AI model and states its error bound | | Design asset | An approved file used in the product exactly as it is, such as a logo, kept under `design/` with its digest | | Generated | Produced automatically by `tools/build.py`; never edited by hand | ##### TOK dependency and state tools The existing `estate-hub` MCP exposes `tok_overview`, `tok_search`, `tok_get`, `tok_dependencies`, `tok_issues`, `tok_context`, `tok_runtime_state` and `tok_diagram`. These are source-linked views of TOK and the specification, not a second ontology. Generate the index with `python3 tools/dependency_map.py --output .cache/tok-dependencies.sqlite`. The scope and parser gaps are part of the result; a declared dependency is not evidence that it works. `tok_context` supplies the source fingerprint, subject, dependencies and unresolved links. Hub operation dispatch checks the same fingerprint, index freshness and exact declared operation automatically, then retains the existing engine's identity, authority and receiving-system checks. A missing or stale map is refused; no new human approval is introduced. Private runtime observations are supplied outside Git through `TOK_RUNTIME_OBSERVATIONS` or `.cache/tok-runtime-observations.json`. Their evidence and observation times remain visible; a configured key or service label does not prove an outcome. Local stdio uses the caller's filesystem access. HTTP queries require the existing verified identity; private runtime observations additionally require an authorised operational role. ##### Where to read more `tok/KERNEL.md` explains the ontology. `spec/SPECIFICATION.md` explains how things are specified and what complete means. `contract/estate-contract.json` is the contract. `ci/README.md` installs the CI. Read them when the task needs them. ### docs/README.md `docs/README.md` · sha256 `080ebe08a86280093a824e6c651b62bf52581eee20d3e51011cf342333d9ec2d` #### Documentation Generated by `tools/docs.py` from `spec/units/`. 1616 help articles, 1616 admin guides, 1616 API references, set against 9034 competitor features and articles. How the whole estate is produced: [STANDARD.md](STANDARD.md). | Capability | Help | Admin | API | Journeys | |---|---|---|---|---| | Accept a proxy mandate | [help](help/clinic/accept-a-proxy-mandate.md) | [admin](admin/clinic/accept-a-proxy-mandate.md) | [API](api/clinic/accept-a-proxy-mandate.md) | [journeys](../generated/diagrams/clinic/accept-a-proxy-mandate.md) | | Acquire and send images | [help](help/clinic/acquire-and-send-images.md) | [admin](admin/clinic/acquire-and-send-images.md) | [API](api/clinic/acquire-and-send-images.md) | [journeys](../generated/diagrams/clinic/acquire-and-send-images.md) | | Act on a drug safety alert | [help](help/clinic/act-on-a-drug-safety-alert.md) | [admin](admin/clinic/act-on-a-drug-safety-alert.md) | [API](api/clinic/act-on-a-drug-safety-alert.md) | [journeys](../generated/diagrams/clinic/act-on-a-drug-safety-alert.md) | | Act on a high-risk answer | [help](help/clinic/act-on-a-high-risk-answer.md) | [admin](admin/clinic/act-on-a-high-risk-answer.md) | [API](api/clinic/act-on-a-high-risk-answer.md) | [journeys](../generated/diagrams/clinic/act-on-a-high-risk-answer.md) | | Add an appointment to my calendar | [help](help/clinic/add-an-appointment-to-my-calendar.md) | [admin](admin/clinic/add-an-appointment-to-my-calendar.md) | [API](api/clinic/add-an-appointment-to-my-calendar.md) | [journeys](../generated/diagrams/clinic/add-an-appointment-to-my-calendar.md) | | Adjust a premium at renewal | [help](help/clinic/adjust-a-premium-at-renewal.md) | [admin](admin/clinic/adjust-a-premium-at-renewal.md) | [API](api/clinic/adjust-a-premium-at-renewal.md) | [journeys](../generated/diagrams/clinic/adjust-a-premium-at-renewal.md) | | Administer a medicine | [help](help/clinic/administer-a-medicine.md) | [admin](admin/clinic/administer-a-medicine.md) | [API](api/clinic/administer-a-medicine.md) | [journeys](../generated/diagrams/clinic/administer-a-medicine.md) | | Allocate a breach between trusts | [help](help/clinic/allocate-a-breach-between-trusts.md) | [admin](admin/clinic/allocate-a-breach-between-trusts.md) | [API](api/clinic/allocate-a-breach-between-trusts.md) | [journeys](../generated/diagrams/clinic/allocate-a-breach-between-trusts.md) | | Annotate variants | [help](help/clinic/annotate-variants.md) | [admin](admin/clinic/annotate-variants.md) | [API](api/clinic/annotate-variants.md) | [journeys](../generated/diagrams/clinic/annotate-variants.md) | | Answer a patient call | [help](help/clinic/answer-a-patient-call.md) | [admin](admin/clinic/answer-a-patient-call.md) | [API](api/clinic/answer-a-patient-call.md) | [journeys](../generated/diagrams/clinic/answer-a-patient-call.md) | | Apply a national data opt-out | [help](help/clinic/apply-a-national-data-opt-out.md) | [admin](admin/clinic/apply-a-national-data-opt-out.md) | [API](api/clinic/apply-a-national-data-opt-out.md) | [journeys](../generated/diagrams/clinic/apply-a-national-data-opt-out.md) | | Apply a treatment rule | [help](help/clinic/apply-a-treatment-rule.md) | [admin](admin/clinic/apply-a-treatment-rule.md) | [API](api/clinic/apply-a-treatment-rule.md) | [journeys](../generated/diagrams/clinic/apply-a-treatment-rule.md) | | Apply for HRA and REC approval | [help](help/clinic/apply-for-hra-and-rec-approval.md) | [admin](admin/clinic/apply-for-hra-and-rec-approval.md) | [API](api/clinic/apply-for-hra-and-rec-approval.md) | [journeys](../generated/diagrams/clinic/apply-for-hra-and-rec-approval.md) | | Approve a data release | [help](help/clinic/approve-a-data-release.md) | [admin](admin/clinic/approve-a-data-release.md) | [API](api/clinic/approve-a-data-release.md) | [journeys](../generated/diagrams/clinic/approve-a-data-release.md) | | Approve a leave or swap request | [help](help/clinic/approve-a-leave-or-swap-request.md) | [admin](admin/clinic/approve-a-leave-or-swap-request.md) | [API](api/clinic/approve-a-leave-or-swap-request.md) | [journeys](../generated/diagrams/clinic/approve-a-leave-or-swap-request.md) | | Approve a questionnaire or test threshold | [help](help/clinic/approve-a-questionnaire-or-test-threshold.md) | [admin](admin/clinic/approve-a-questionnaire-or-test-threshold.md) | [API](api/clinic/approve-a-questionnaire-or-test-threshold.md) | [journeys](../generated/diagrams/clinic/approve-a-questionnaire-or-test-threshold.md) | | Approve a staff request | [help](help/clinic/approve-a-staff-request.md) | [admin](admin/clinic/approve-a-staff-request.md) | [API](api/clinic/approve-a-staff-request.md) | [journeys](../generated/diagrams/clinic/approve-a-staff-request.md) | | Arrange a cross-trust MDT | [help](help/clinic/arrange-a-cross-trust-mdt.md) | [admin](admin/clinic/arrange-a-cross-trust-mdt.md) | [API](api/clinic/arrange-a-cross-trust-mdt.md) | [journeys](../generated/diagrams/clinic/arrange-a-cross-trust-mdt.md) | | Ask for advice and guidance | [help](help/clinic/ask-for-advice-and-guidance.md) | [admin](admin/clinic/ask-for-advice-and-guidance.md) | [API](api/clinic/ask-for-advice-and-guidance.md) | [journeys](../generated/diagrams/clinic/ask-for-advice-and-guidance.md) | | Ask for help now | [help](help/clinic/ask-for-help-now.md) | [admin](admin/clinic/ask-for-help-now.md) | [API](api/clinic/ask-for-help-now.md) | [journeys](../generated/diagrams/clinic/ask-for-help-now.md) | | Assess a safeguarding concern | [help](help/clinic/assess-a-safeguarding-concern.md) | [admin](admin/clinic/assess-a-safeguarding-concern.md) | [API](api/clinic/assess-a-safeguarding-concern.md) | [journeys](../generated/diagrams/clinic/assess-a-safeguarding-concern.md) | | Attend an appointment with the person | [help](help/clinic/attend-an-appointment-with-the-person.md) | [admin](admin/clinic/attend-an-appointment-with-the-person.md) | [API](api/clinic/attend-an-appointment-with-the-person.md) | [journeys](../generated/diagrams/clinic/attend-an-appointment-with-the-person.md) | | Audit access to records | [help](help/clinic/audit-access-to-records.md) | [admin](admin/clinic/audit-access-to-records.md) | [API](api/clinic/audit-access-to-records.md) | [journeys](../generated/diagrams/clinic/audit-access-to-records.md) | | Audit antimicrobial use | [help](help/clinic/audit-antimicrobial-use.md) | [admin](admin/clinic/audit-antimicrobial-use.md) | [API](api/clinic/audit-antimicrobial-use.md) | [journeys](../generated/diagrams/clinic/audit-antimicrobial-use.md) | | Audit consent records | [help](help/clinic/audit-consent-records.md) | [admin](admin/clinic/audit-consent-records.md) | [API](api/clinic/audit-consent-records.md) | [journeys](../generated/diagrams/clinic/audit-consent-records.md) | | Audit data use against mandates | [help](help/clinic/audit-data-use-against-mandates.md) | [admin](admin/clinic/audit-data-use-against-mandates.md) | [API](api/clinic/audit-data-use-against-mandates.md) | [journeys](../generated/diagrams/clinic/audit-data-use-against-mandates.md) | | Audit MDT attendance | [help](help/clinic/audit-mdt-attendance.md) | [admin](admin/clinic/audit-mdt-attendance.md) | [API](api/clinic/audit-mdt-attendance.md) | [journeys](../generated/diagrams/clinic/audit-mdt-attendance.md) | | Audit reporting quality | [help](help/clinic/audit-reporting-quality.md) | [admin](admin/clinic/audit-reporting-quality.md) | [API](api/clinic/audit-reporting-quality.md) | [journeys](../generated/diagrams/clinic/audit-reporting-quality.md) | | Audit response times | [help](help/clinic/audit-response-times.md) | [admin](admin/clinic/audit-response-times.md) | [API](api/clinic/audit-response-times.md) | [journeys](../generated/diagrams/clinic/audit-response-times.md) | | Audit waiting time data | [help](help/clinic/audit-waiting-time-data.md) | [admin](admin/clinic/audit-waiting-time-data.md) | [API](api/clinic/audit-waiting-time-data.md) | [journeys](../generated/diagrams/clinic/audit-waiting-time-data.md) | | Authorise a result | [help](help/clinic/authorise-a-result.md) | [admin](admin/clinic/authorise-a-result.md) | [API](api/clinic/authorise-a-result.md) | [journeys](../generated/diagrams/clinic/authorise-a-result.md) | | Book a pathway test | [help](help/clinic/book-a-pathway-test.md) | [admin](admin/clinic/book-a-pathway-test.md) | [API](api/clinic/book-a-pathway-test.md) | [journeys](../generated/diagrams/clinic/book-a-pathway-test.md) | | Book an appointment | [help](help/clinic/book-an-appointment.md) | [admin](admin/clinic/book-an-appointment.md) | [API](api/clinic/book-an-appointment.md) | [journeys](../generated/diagrams/clinic/book-an-appointment.md) | | Book an interpreter | [help](help/clinic/book-an-interpreter.md) | [admin](admin/clinic/book-an-interpreter.md) | [API](api/clinic/book-an-interpreter.md) | [journeys](../generated/diagrams/clinic/book-an-interpreter.md) | | Book for the person I care for | [help](help/clinic/book-for-the-person-i-care-for.md) | [admin](admin/clinic/book-for-the-person-i-care-for.md) | [API](api/clinic/book-for-the-person-i-care-for.md) | [journeys](../generated/diagrams/clinic/book-for-the-person-i-care-for.md) | | Book from the referral queue | [help](help/clinic/book-from-the-referral-queue.md) | [admin](admin/clinic/book-from-the-referral-queue.md) | [API](api/clinic/book-from-the-referral-queue.md) | [journeys](../generated/diagrams/clinic/book-from-the-referral-queue.md) | | Book transport | [help](help/clinic/book-transport.md) | [admin](admin/clinic/book-transport.md) | [API](api/clinic/book-transport.md) | [journeys](../generated/diagrams/clinic/book-transport.md) | | Build and publish the rota | [help](help/clinic/build-and-publish-the-rota.md) | [admin](admin/clinic/build-and-publish-the-rota.md) | [API](api/clinic/build-and-publish-the-rota.md) | [journeys](../generated/diagrams/clinic/build-and-publish-the-rota.md) | | Build the clinic template | [help](help/clinic/build-the-clinic-template.md) | [admin](admin/clinic/build-the-clinic-template.md) | [API](api/clinic/build-the-clinic-template.md) | [journeys](../generated/diagrams/clinic/build-the-clinic-template.md) | | Build the MDT list | [help](help/clinic/build-the-mdt-list.md) | [admin](admin/clinic/build-the-mdt-list.md) | [API](api/clinic/build-the-mdt-list.md) | [journeys](../generated/diagrams/clinic/build-the-mdt-list.md) | | Build the variant to trial graph | [help](help/clinic/build-the-variant-to-trial-graph.md) | [admin](admin/clinic/build-the-variant-to-trial-graph.md) | [API](api/clinic/build-the-variant-to-trial-graph.md) | [journeys](../generated/diagrams/clinic/build-the-variant-to-trial-graph.md) | | Cancel a clinic session | [help](help/clinic/cancel-a-clinic-session.md) | [admin](admin/clinic/cancel-a-clinic-session.md) | [API](api/clinic/cancel-a-clinic-session.md) | [journeys](../generated/diagrams/clinic/cancel-a-clinic-session.md) | | Change my pharmacy | [help](help/clinic/change-my-pharmacy.md) | [admin](admin/clinic/change-my-pharmacy.md) | [API](api/clinic/change-my-pharmacy.md) | [journeys](../generated/diagrams/clinic/change-my-pharmacy.md) | | Chase a pathway clock | [help](help/clinic/chase-a-pathway-clock.md) | [admin](admin/clinic/chase-a-pathway-clock.md) | [API](api/clinic/chase-a-pathway-clock.md) | [journeys](../generated/diagrams/clinic/chase-a-pathway-clock.md) | | Chase missing results | [help](help/clinic/chase-missing-results.md) | [admin](admin/clinic/chase-missing-results.md) | [API](api/clinic/chase-missing-results.md) | [journeys](../generated/diagrams/clinic/chase-missing-results.md) | | Check a chemotherapy prescription | [help](help/clinic/check-a-chemotherapy-prescription.md) | [admin](admin/clinic/check-a-chemotherapy-prescription.md) | [API](api/clinic/check-a-chemotherapy-prescription.md) | [journeys](../generated/diagrams/clinic/check-a-chemotherapy-prescription.md) | | Check access controls | [help](help/clinic/check-access-controls.md) | [admin](admin/clinic/check-access-controls.md) | [API](api/clinic/check-access-controls.md) | [journeys](../generated/diagrams/clinic/check-access-controls.md) | | Check fragment size before sequencing | [help](help/clinic/check-fragment-size-before-sequencing.md) | [admin](admin/clinic/check-fragment-size-before-sequencing.md) | [API](api/clinic/check-fragment-size-before-sequencing.md) | [journeys](../generated/diagrams/clinic/check-fragment-size-before-sequencing.md) | | Check identity and consent before scanning | [help](help/clinic/check-identity-and-consent-before-scanning.md) | [admin](admin/clinic/check-identity-and-consent-before-scanning.md) | [API](api/clinic/check-identity-and-consent-before-scanning.md) | [journeys](../generated/diagrams/clinic/check-identity-and-consent-before-scanning.md) | | Check in an arriving patient | [help](help/clinic/check-in-an-arriving-patient.md) | [admin](admin/clinic/check-in-an-arriving-patient.md) | [API](api/clinic/check-in-an-arriving-patient.md) | [journeys](../generated/diagrams/clinic/check-in-an-arriving-patient.md) | | Check in on arrival | [help](help/clinic/check-in-on-arrival.md) | [admin](admin/clinic/check-in-on-arrival.md) | [API](api/clinic/check-in-on-arrival.md) | [journeys](../generated/diagrams/clinic/check-in-on-arrival.md) | | Check laboratory submissions for a variant | [help](help/clinic/check-laboratory-submissions-for-a-variant.md) | [admin](admin/clinic/check-laboratory-submissions-for-a-variant.md) | [API](api/clinic/check-laboratory-submissions-for-a-variant.md) | [journeys](../generated/diagrams/clinic/check-laboratory-submissions-for-a-variant.md) | | Check record accuracy | [help](help/clinic/check-record-accuracy.md) | [admin](admin/clinic/check-record-accuracy.md) | [API](api/clinic/check-record-accuracy.md) | [journeys](../generated/diagrams/clinic/check-record-accuracy.md) | | Check research integrity | [help](help/clinic/check-research-integrity.md) | [admin](admin/clinic/check-research-integrity.md) | [API](api/clinic/check-research-integrity.md) | [journeys](../generated/diagrams/clinic/check-research-integrity.md) | | Check staff registration and training | [help](help/clinic/check-staff-registration-and-training.md) | [admin](admin/clinic/check-staff-registration-and-training.md) | [API](api/clinic/check-staff-registration-and-training.md) | [journeys](../generated/diagrams/clinic/check-staff-registration-and-training.md) | | Claim my EDGE ID and record | [help](help/clinic/claim-my-edge-id-and-record.md) | [admin](admin/clinic/claim-my-edge-id-and-record.md) | [API](api/clinic/claim-my-edge-id-and-record.md) | [journeys](../generated/diagrams/clinic/claim-my-edge-id-and-record.md) | | Classify a variant | [help](help/clinic/classify-a-variant.md) | [admin](admin/clinic/classify-a-variant.md) | [API](api/clinic/classify-a-variant.md) | [journeys](../generated/diagrams/clinic/classify-a-variant.md) | | Classify software as a medical device | [help](help/clinic/classify-software-as-a-medical-device.md) | [admin](admin/clinic/classify-software-as-a-medical-device.md) | [API](api/clinic/classify-software-as-a-medical-device.md) | [journeys](../generated/diagrams/clinic/classify-software-as-a-medical-device.md) | | Clinically check a prescription | [help](help/clinic/clinically-check-a-prescription.md) | [admin](admin/clinic/clinically-check-a-prescription.md) | [API](api/clinic/clinically-check-a-prescription.md) | [journeys](../generated/diagrams/clinic/clinically-check-a-prescription.md) | | Close a referral | [help](help/clinic/close-a-referral.md) | [admin](admin/clinic/close-a-referral.md) | [API](api/clinic/close-a-referral.md) | [journeys](../generated/diagrams/clinic/close-a-referral.md) | | Close a study | [help](help/clinic/close-a-study.md) | [admin](admin/clinic/close-a-study.md) | [API](api/clinic/close-a-study.md) | [journeys](../generated/diagrams/clinic/close-a-study.md) | | Collect case report data | [help](help/clinic/collect-case-report-data.md) | [admin](admin/clinic/collect-case-report-data.md) | [API](api/clinic/collect-case-report-data.md) | [journeys](../generated/diagrams/clinic/collect-case-report-data.md) | | Compare two pipeline runs | [help](help/clinic/compare-two-pipeline-runs.md) | [admin](admin/clinic/compare-two-pipeline-runs.md) | [API](api/clinic/compare-two-pipeline-runs.md) | [journeys](../generated/diagrams/clinic/compare-two-pipeline-runs.md) | | Complete a mental health check-in | [help](help/clinic/complete-a-mental-health-check-in.md) | [admin](admin/clinic/complete-a-mental-health-check-in.md) | [API](api/clinic/complete-a-mental-health-check-in.md) | [journeys](../generated/diagrams/clinic/complete-a-mental-health-check-in.md) | | Complete a pre-clinic questionnaire | [help](help/clinic/complete-a-pre-clinic-questionnaire.md) | [admin](admin/clinic/complete-a-pre-clinic-questionnaire.md) | [API](api/clinic/complete-a-pre-clinic-questionnaire.md) | [journeys](../generated/diagrams/clinic/complete-a-pre-clinic-questionnaire.md) | | Complete a risk formulation | [help](help/clinic/complete-a-risk-formulation.md) | [admin](admin/clinic/complete-a-risk-formulation.md) | [API](api/clinic/complete-a-risk-formulation.md) | [journeys](../generated/diagrams/clinic/complete-a-risk-formulation.md) | | Complete mandatory training | [help](help/clinic/complete-mandatory-training.md) | [admin](admin/clinic/complete-mandatory-training.md) | [API](api/clinic/complete-mandatory-training.md) | [journeys](../generated/diagrams/clinic/complete-mandatory-training.md) | | Complete the Data Security and Protection Toolkit | [help](help/clinic/complete-the-data-security-and-protection-toolkit.md) | [admin](admin/clinic/complete-the-data-security-and-protection-toolkit.md) | [API](api/clinic/complete-the-data-security-and-protection-toolkit.md) | [journeys](../generated/diagrams/clinic/complete-the-data-security-and-protection-toolkit.md) | | Confirm attendance and prepare | [help](help/clinic/confirm-attendance-and-prepare.md) | [admin](admin/clinic/confirm-attendance-and-prepare.md) | [API](api/clinic/confirm-attendance-and-prepare.md) | [journeys](../generated/diagrams/clinic/confirm-attendance-and-prepare.md) | | Connect a LIMS | [help](help/clinic/connect-a-lims.md) | [admin](admin/clinic/connect-a-lims.md) | [API](api/clinic/connect-a-lims.md) | [journeys](../generated/diagrams/clinic/connect-a-lims.md) | | Connect a PACS or imaging system | [help](help/clinic/connect-a-pacs-or-imaging-system.md) | [admin](admin/clinic/connect-a-pacs-or-imaging-system.md) | [API](api/clinic/connect-a-pacs-or-imaging-system.md) | [journeys](../generated/diagrams/clinic/connect-a-pacs-or-imaging-system.md) | | Counsel a patient on a new medicine | [help](help/clinic/counsel-a-patient-on-a-new-medicine.md) | [admin](admin/clinic/counsel-a-patient-on-a-new-medicine.md) | [API](api/clinic/counsel-a-patient-on-a-new-medicine.md) | [journeys](../generated/diagrams/clinic/counsel-a-patient-on-a-new-medicine.md) | | Decide a re-run | [help](help/clinic/decide-a-re-run.md) | [admin](admin/clinic/decide-a-re-run.md) | [API](api/clinic/decide-a-re-run.md) | [journeys](../generated/diagrams/clinic/decide-a-re-run.md) | | Deliver cancer nurse specialist support | [help](help/clinic/deliver-cancer-nurse-specialist-support.md) | [admin](admin/clinic/deliver-cancer-nurse-specialist-support.md) | [API](api/clinic/deliver-cancer-nurse-specialist-support.md) | [journeys](../generated/diagrams/clinic/deliver-cancer-nurse-specialist-support.md) | | Deploy a release | [help](help/clinic/deploy-a-release.md) | [admin](admin/clinic/deploy-a-release.md) | [API](api/clinic/deploy-a-release.md) | [journeys](../generated/diagrams/clinic/deploy-a-release.md) | | Design a smart questionnaire | [help](help/clinic/design-a-smart-questionnaire.md) | [admin](admin/clinic/design-a-smart-questionnaire.md) | [API](api/clinic/design-a-smart-questionnaire.md) | [journeys](../generated/diagrams/clinic/design-a-smart-questionnaire.md) | | Design a trial | [help](help/clinic/design-a-trial.md) | [admin](admin/clinic/design-a-trial.md) | [API](api/clinic/design-a-trial.md) | [journeys](../generated/diagrams/clinic/design-a-trial.md) | | Detect claim anomalies | [help](help/clinic/detect-claim-anomalies.md) | [admin](admin/clinic/detect-claim-anomalies.md) | [API](api/clinic/detect-claim-anomalies.md) | [journeys](../generated/diagrams/clinic/detect-claim-anomalies.md) | | Discharge from the pathway | [help](help/clinic/discharge-from-the-pathway.md) | [admin](admin/clinic/discharge-from-the-pathway.md) | [API](api/clinic/discharge-from-the-pathway.md) | [journeys](../generated/diagrams/clinic/discharge-from-the-pathway.md) | | Discharge from the stream | [help](help/clinic/discharge-from-the-stream.md) | [admin](admin/clinic/discharge-from-the-stream.md) | [API](api/clinic/discharge-from-the-stream.md) | [journeys](../generated/diagrams/clinic/discharge-from-the-stream.md) | | Dispense a medicine | [help](help/clinic/dispense-a-medicine.md) | [admin](admin/clinic/dispense-a-medicine.md) | [API](api/clinic/dispense-a-medicine.md) | [journeys](../generated/diagrams/clinic/dispense-a-medicine.md) | | Escalate to emergency services | [help](help/clinic/escalate-to-emergency-services.md) | [admin](admin/clinic/escalate-to-emergency-services.md) | [API](api/clinic/escalate-to-emergency-services.md) | [journeys](../generated/diagrams/clinic/escalate-to-emergency-services.md) | | Explain a result in plain language | [help](help/clinic/explain-a-result-in-plain-language.md) | [admin](admin/clinic/explain-a-result-in-plain-language.md) | [API](api/clinic/explain-a-result-in-plain-language.md) | [journeys](../generated/diagrams/clinic/explain-a-result-in-plain-language.md) | | Explain a result to the patient | [help](help/clinic/explain-a-result-to-the-patient.md) | [admin](admin/clinic/explain-a-result-to-the-patient.md) | [API](api/clinic/explain-a-result-to-the-patient.md) | [journeys](../generated/diagrams/clinic/explain-a-result-to-the-patient.md) | | Find a recruiting trial near me | [help](help/clinic/find-a-recruiting-trial-near-me.md) | [admin](admin/clinic/find-a-recruiting-trial-near-me.md) | [API](api/clinic/find-a-recruiting-trial-near-me.md) | [journeys](../generated/diagrams/clinic/find-a-recruiting-trial-near-me.md) | | Flag a critical finding | [help](help/clinic/flag-a-critical-finding.md) | [admin](admin/clinic/flag-a-critical-finding.md) | [API](api/clinic/flag-a-critical-finding.md) | [journeys](../generated/diagrams/clinic/flag-a-critical-finding.md) | | Flag, never block | [help](help/clinic/flag-never-block.md) | [admin](admin/clinic/flag-never-block.md) | [API](api/clinic/flag-never-block.md) | [journeys](../generated/diagrams/clinic/flag-never-block.md) | | Follow up after discharge from crisis care | [help](help/clinic/follow-up-after-discharge-from-crisis-care.md) | [admin](admin/clinic/follow-up-after-discharge-from-crisis-care.md) | [API](api/clinic/follow-up-after-discharge-from-crisis-care.md) | [journeys](../generated/diagrams/clinic/follow-up-after-discharge-from-crisis-care.md) | | Follow up an action plan | [help](help/clinic/follow-up-an-action-plan.md) | [admin](admin/clinic/follow-up-an-action-plan.md) | [API](api/clinic/follow-up-an-action-plan.md) | [journeys](../generated/diagrams/clinic/follow-up-an-action-plan.md) | | Forecast compliance | [help](help/clinic/forecast-compliance.md) | [admin](admin/clinic/forecast-compliance.md) | [API](api/clinic/forecast-compliance.md) | [journeys](../generated/diagrams/clinic/forecast-compliance.md) | | Forecast days to treatment | [help](help/clinic/forecast-days-to-treatment.md) | [admin](admin/clinic/forecast-days-to-treatment.md) | [API](api/clinic/forecast-days-to-treatment.md) | [journeys](../generated/diagrams/clinic/forecast-days-to-treatment.md) | | Front door check-in | [help](help/clinic/front-door-check-in.md) | [admin](admin/clinic/front-door-check-in.md) | [API](api/clinic/front-door-check-in.md) | [journeys](../generated/diagrams/clinic/front-door-check-in.md) | | Fulfil a subject access request | [help](help/clinic/fulfil-a-subject-access-request.md) | [admin](admin/clinic/fulfil-a-subject-access-request.md) | [API](api/clinic/fulfil-a-subject-access-request.md) | [journeys](../generated/diagrams/clinic/fulfil-a-subject-access-request.md) | | Give a vaccination | [help](help/clinic/give-a-vaccination.md) | [admin](admin/clinic/give-a-vaccination.md) | [API](api/clinic/give-a-vaccination.md) | [journeys](../generated/diagrams/clinic/give-a-vaccination.md) | | Give consent on behalf | [help](help/clinic/give-consent-on-behalf.md) | [admin](admin/clinic/give-consent-on-behalf.md) | [API](api/clinic/give-consent-on-behalf.md) | [journeys](../generated/diagrams/clinic/give-consent-on-behalf.md) | | Give or withdraw consent | [help](help/clinic/give-or-withdraw-consent.md) | [admin](admin/clinic/give-or-withdraw-consent.md) | [API](api/clinic/give-or-withdraw-consent.md) | [journeys](../generated/diagrams/clinic/give-or-withdraw-consent.md) | | Grade a retinal image | [help](help/clinic/grade-a-retinal-image.md) | [admin](admin/clinic/grade-a-retinal-image.md) | [API](api/clinic/grade-a-retinal-image.md) | [journeys](../generated/diagrams/clinic/grade-a-retinal-image.md) | | Hand off to Codex or Jude | [help](help/clinic/hand-off-to-codex-or-jude.md) | [admin](admin/clinic/hand-off-to-codex-or-jude.md) | [API](api/clinic/hand-off-to-codex-or-jude.md) | [journeys](../generated/diagrams/clinic/hand-off-to-codex-or-jude.md) | | Handle a complaint | [help](help/clinic/handle-a-complaint.md) | [admin](admin/clinic/handle-a-complaint.md) | [API](api/clinic/handle-a-complaint.md) | [journeys](../generated/diagrams/clinic/handle-a-complaint.md) | | Handle a controlled drug | [help](help/clinic/handle-a-controlled-drug.md) | [admin](admin/clinic/handle-a-controlled-drug.md) | [API](api/clinic/handle-a-controlled-drug.md) | [journeys](../generated/diagrams/clinic/handle-a-controlled-drug.md) | | Handle a did-not-attend | [help](help/clinic/handle-a-did-not-attend.md) | [admin](admin/clinic/handle-a-did-not-attend.md) | [API](api/clinic/handle-a-did-not-attend.md) | [journeys](../generated/diagrams/clinic/handle-a-did-not-attend.md) | | Handle a disclosure of abuse | [help](help/clinic/handle-a-disclosure-of-abuse.md) | [admin](admin/clinic/handle-a-disclosure-of-abuse.md) | [API](api/clinic/handle-a-disclosure-of-abuse.md) | [journeys](../generated/diagrams/clinic/handle-a-disclosure-of-abuse.md) | | Handle a freedom of information request | [help](help/clinic/handle-a-freedom-of-information-request.md) | [admin](admin/clinic/handle-a-freedom-of-information-request.md) | [API](api/clinic/handle-a-freedom-of-information-request.md) | [journeys](../generated/diagrams/clinic/handle-a-freedom-of-information-request.md) | | Handle a policyholder complaint | [help](help/clinic/handle-a-policyholder-complaint.md) | [admin](admin/clinic/handle-a-policyholder-complaint.md) | [API](api/clinic/handle-a-policyholder-complaint.md) | [journeys](../generated/diagrams/clinic/handle-a-policyholder-complaint.md) | | Handle a stock shortage | [help](help/clinic/handle-a-stock-shortage.md) | [admin](admin/clinic/handle-a-stock-shortage.md) | [API](api/clinic/handle-a-stock-shortage.md) | [journeys](../generated/diagrams/clinic/handle-a-stock-shortage.md) | | Handle the record after death | [help](help/clinic/handle-the-record-after-death.md) | [admin](admin/clinic/handle-the-record-after-death.md) | [API](api/clinic/handle-the-record-after-death.md) | [journeys](../generated/diagrams/clinic/handle-the-record-after-death.md) | | Hold a remote session | [help](help/clinic/hold-a-remote-session.md) | [admin](admin/clinic/hold-a-remote-session.md) | [API](api/clinic/hold-a-remote-session.md) | [journeys](../generated/diagrams/clinic/hold-a-remote-session.md) | | Hold a slot for a tertiary handover | [help](help/clinic/hold-a-slot-for-a-tertiary-handover.md) | [admin](admin/clinic/hold-a-slot-for-a-tertiary-handover.md) | [API](api/clinic/hold-a-slot-for-a-tertiary-handover.md) | [journeys](../generated/diagrams/clinic/hold-a-slot-for-a-tertiary-handover.md) | | Hold the duty rota | [help](help/clinic/hold-the-duty-rota.md) | [admin](admin/clinic/hold-the-duty-rota.md) | [API](api/clinic/hold-the-duty-rota.md) | [journeys](../generated/diagrams/clinic/hold-the-duty-rota.md) | | Import my practice register | [help](help/clinic/import-my-practice-register.md) | [admin](admin/clinic/import-my-practice-register.md) | [API](api/clinic/import-my-practice-register.md) | [journeys](../generated/diagrams/clinic/import-my-practice-register.md) | | Inspect the hazard log and safety case | [help](help/clinic/inspect-the-hazard-log-and-safety-case.md) | [admin](admin/clinic/inspect-the-hazard-log-and-safety-case.md) | [API](api/clinic/inspect-the-hazard-log-and-safety-case.md) | [journeys](../generated/diagrams/clinic/inspect-the-hazard-log-and-safety-case.md) | | Investigate a clinical safety incident | [help](help/clinic/investigate-a-clinical-safety-incident.md) | [admin](admin/clinic/investigate-a-clinical-safety-incident.md) | [API](api/clinic/investigate-a-clinical-safety-incident.md) | [journeys](../generated/diagrams/clinic/investigate-a-clinical-safety-incident.md) | | Issue a clinical safety case report | [help](help/clinic/issue-a-clinical-safety-case-report.md) | [admin](admin/clinic/issue-a-clinical-safety-case-report.md) | [API](api/clinic/issue-a-clinical-safety-case-report.md) | [journeys](../generated/diagrams/clinic/issue-a-clinical-safety-case-report.md) | | Issue a field safety notice | [help](help/clinic/issue-a-field-safety-notice.md) | [admin](admin/clinic/issue-a-field-safety-notice.md) | [API](api/clinic/issue-a-field-safety-notice.md) | [journeys](../generated/diagrams/clinic/issue-a-field-safety-notice.md) | | Issue a genomic report | [help](help/clinic/issue-a-genomic-report.md) | [admin](admin/clinic/issue-a-genomic-report.md) | [API](api/clinic/issue-a-genomic-report.md) | [journeys](../generated/diagrams/clinic/issue-a-genomic-report.md) | | Issue a policy | [help](help/clinic/issue-a-policy.md) | [admin](admin/clinic/issue-a-policy.md) | [API](api/clinic/issue-a-policy.md) | [journeys](../generated/diagrams/clinic/issue-a-policy.md) | | Maintain laboratory accreditation evidence | [help](help/clinic/maintain-laboratory-accreditation-evidence.md) | [admin](admin/clinic/maintain-laboratory-accreditation-evidence.md) | [API](api/clinic/maintain-laboratory-accreditation-evidence.md) | [journeys](../generated/diagrams/clinic/maintain-laboratory-accreditation-evidence.md) | | Maintain the hazard log | [help](help/clinic/maintain-the-hazard-log.md) | [admin](admin/clinic/maintain-the-hazard-log.md) | [API](api/clinic/maintain-the-hazard-log.md) | [journeys](../generated/diagrams/clinic/maintain-the-hazard-log.md) | | Maintain the suicide prevention plan | [help](help/clinic/maintain-the-suicide-prevention-plan.md) | [admin](admin/clinic/maintain-the-suicide-prevention-plan.md) | [API](api/clinic/maintain-the-suicide-prevention-plan.md) | [journeys](../generated/diagrams/clinic/maintain-the-suicide-prevention-plan.md) | | Make a follow-up call | [help](help/clinic/make-a-follow-up-call.md) | [admin](admin/clinic/make-a-follow-up-call.md) | [API](api/clinic/make-a-follow-up-call.md) | [journeys](../generated/diagrams/clinic/make-a-follow-up-call.md) | | Make an urgent same-day referral | [help](help/clinic/make-an-urgent-same-day-referral.md) | [admin](admin/clinic/make-an-urgent-same-day-referral.md) | [API](api/clinic/make-an-urgent-same-day-referral.md) | [journeys](../generated/diagrams/clinic/make-an-urgent-same-day-referral.md) | | Manage an incident | [help](help/clinic/manage-an-incident.md) | [admin](admin/clinic/manage-an-incident.md) | [API](api/clinic/manage-an-incident.md) | [journeys](../generated/diagrams/clinic/manage-an-incident.md) | | Manage freezer alarms | [help](help/clinic/manage-freezer-alarms.md) | [admin](admin/clinic/manage-freezer-alarms.md) | [API](api/clinic/manage-freezer-alarms.md) | [journeys](../generated/diagrams/clinic/manage-freezer-alarms.md) | | Manage patient group directions | [help](help/clinic/manage-patient-group-directions.md) | [admin](admin/clinic/manage-patient-group-directions.md) | [API](api/clinic/manage-patient-group-directions.md) | [journeys](../generated/diagrams/clinic/manage-patient-group-directions.md) | | Manage patient transport and interpreter contracts | [help](help/clinic/manage-patient-transport-and-interpreter-contracts.md) | [admin](admin/clinic/manage-patient-transport-and-interpreter-contracts.md) | [API](api/clinic/manage-patient-transport-and-interpreter-contracts.md) | [journeys](../generated/diagrams/clinic/manage-patient-transport-and-interpreter-contracts.md) | | Manage retention and destruction | [help](help/clinic/manage-retention-and-destruction.md) | [admin](admin/clinic/manage-retention-and-destruction.md) | [API](api/clinic/manage-retention-and-destruction.md) | [journeys](../generated/diagrams/clinic/manage-retention-and-destruction.md) | | Manage room allocation | [help](help/clinic/manage-room-allocation.md) | [admin](admin/clinic/manage-room-allocation.md) | [API](api/clinic/manage-room-allocation.md) | [journeys](../generated/diagrams/clinic/manage-room-allocation.md) | | Manage the information asset register | [help](help/clinic/manage-the-information-asset-register.md) | [admin](admin/clinic/manage-the-information-asset-register.md) | [API](api/clinic/manage-the-information-asset-register.md) | [journeys](../generated/diagrams/clinic/manage-the-information-asset-register.md) | | Manage the outbound allow-list | [help](help/clinic/manage-the-outbound-allow-list.md) | [admin](admin/clinic/manage-the-outbound-allow-list.md) | [API](api/clinic/manage-the-outbound-allow-list.md) | [journeys](../generated/diagrams/clinic/manage-the-outbound-allow-list.md) | | Manage the reporting backlog | [help](help/clinic/manage-the-reporting-backlog.md) | [admin](admin/clinic/manage-the-reporting-backlog.md) | [API](api/clinic/manage-the-reporting-backlog.md) | [journeys](../generated/diagrams/clinic/manage-the-reporting-backlog.md) | | Manage the room and equipment register | [help](help/clinic/manage-the-room-and-equipment-register.md) | [admin](admin/clinic/manage-the-room-and-equipment-register.md) | [API](api/clinic/manage-the-room-and-equipment-register.md) | [journeys](../generated/diagrams/clinic/manage-the-room-and-equipment-register.md) | | Manage the waiting list | [help](help/clinic/manage-the-waiting-list.md) | [admin](admin/clinic/manage-the-waiting-list.md) | [API](api/clinic/manage-the-waiting-list.md) | [journeys](../generated/diagrams/clinic/manage-the-waiting-list.md) | | Measure a nodule and its growth | [help](help/clinic/measure-a-nodule-and-its-growth.md) | [admin](admin/clinic/measure-a-nodule-and-its-growth.md) | [API](api/clinic/measure-a-nodule-and-its-growth.md) | [journeys](../generated/diagrams/clinic/measure-a-nodule-and-its-growth.md) | | Measure compute, energy and carbon | [help](help/clinic/measure-compute-energy-and-carbon.md) | [admin](admin/clinic/measure-compute-energy-and-carbon.md) | [API](api/clinic/measure-compute-energy-and-carbon.md) | [journeys](../generated/diagrams/clinic/measure-compute-energy-and-carbon.md) | | Measure outcomes over time | [help](help/clinic/measure-outcomes-over-time.md) | [admin](admin/clinic/measure-outcomes-over-time.md) | [API](api/clinic/measure-outcomes-over-time.md) | [journeys](../generated/diagrams/clinic/measure-outcomes-over-time.md) | | Medication safety check | [help](help/clinic/medication-safety-check.md) | [admin](admin/clinic/medication-safety-check.md) | [API](api/clinic/medication-safety-check.md) | [journeys](../generated/diagrams/clinic/medication-safety-check.md) | | Mental health care and crisis support | [help](help/clinic/mental-health-care-and-crisis.md) | [admin](admin/clinic/mental-health-care-and-crisis.md) | [API](api/clinic/mental-health-care-and-crisis.md) | [journeys](../generated/diagrams/clinic/mental-health-care-and-crisis.md) | | Message my care team | [help](help/clinic/message-my-care-team.md) | [admin](admin/clinic/message-my-care-team.md) | [API](api/clinic/message-my-care-team.md) | [journeys](../generated/diagrams/clinic/message-my-care-team.md) | | Monitor alerts and escalate | [help](help/clinic/monitor-alerts-and-escalate.md) | [admin](admin/clinic/monitor-alerts-and-escalate.md) | [API](api/clinic/monitor-alerts-and-escalate.md) | [journeys](../generated/diagrams/clinic/monitor-alerts-and-escalate.md) | | Monitor data feeds | [help](help/clinic/monitor-data-feeds.md) | [admin](admin/clinic/monitor-data-feeds.md) | [API](api/clinic/monitor-data-feeds.md) | [journeys](../generated/diagrams/clinic/monitor-data-feeds.md) | | Monitor post-deployment safety | [help](help/clinic/monitor-post-deployment-safety.md) | [admin](admin/clinic/monitor-post-deployment-safety.md) | [API](api/clinic/monitor-post-deployment-safety.md) | [journeys](../generated/diagrams/clinic/monitor-post-deployment-safety.md) | | Nominate a carer or proxy | [help](help/clinic/nominate-a-carer-or-proxy.md) | [admin](admin/clinic/nominate-a-carer-or-proxy.md) | [API](api/clinic/nominate-a-carer-or-proxy.md) | [journeys](../generated/diagrams/clinic/nominate-a-carer-or-proxy.md) | | Obtain confirmation of capacity and capability | [help](help/clinic/obtain-confirmation-of-capacity-and-capability.md) | [admin](admin/clinic/obtain-confirmation-of-capacity-and-capability.md) | [API](api/clinic/obtain-confirmation-of-capacity-and-capability.md) | [journeys](../generated/diagrams/clinic/obtain-confirmation-of-capacity-and-capability.md) | | Offer a cancellation slot | [help](help/clinic/offer-a-cancellation-slot.md) | [admin](admin/clinic/offer-a-cancellation-slot.md) | [API](api/clinic/offer-a-cancellation-slot.md) | [journeys](../generated/diagrams/clinic/offer-a-cancellation-slot.md) | | Offer self-help resources | [help](help/clinic/offer-self-help-resources.md) | [admin](admin/clinic/offer-self-help-resources.md) | [API](api/clinic/offer-self-help-resources.md) | [journeys](../generated/diagrams/clinic/offer-self-help-resources.md) | | Onboard a staff member | [help](help/clinic/onboard-a-staff-member.md) | [admin](admin/clinic/onboard-a-staff-member.md) | [API](api/clinic/onboard-a-staff-member.md) | [journeys](../generated/diagrams/clinic/onboard-a-staff-member.md) | | Onboard a wearable device type | [help](help/clinic/onboard-a-wearable-device-type.md) | [admin](admin/clinic/onboard-a-wearable-device-type.md) | [API](api/clinic/onboard-a-wearable-device-type.md) | [journeys](../generated/diagrams/clinic/onboard-a-wearable-device-type.md) | | Online urgent triage | [help](help/clinic/online-urgent-triage.md) | [admin](admin/clinic/online-urgent-triage.md) | [API](api/clinic/online-urgent-triage.md) | [journeys](../generated/diagrams/clinic/online-urgent-triage.md) | | Order a diagnostic test | [help](help/clinic/order-a-diagnostic-test.md) | [admin](admin/clinic/order-a-diagnostic-test.md) | [API](api/clinic/order-a-diagnostic-test.md) | [journeys](../generated/diagrams/clinic/order-a-diagnostic-test.md) | | Order a direct access test | [help](help/clinic/order-a-direct-access-test.md) | [admin](admin/clinic/order-a-direct-access-test.md) | [API](api/clinic/order-a-direct-access-test.md) | [journeys](../generated/diagrams/clinic/order-a-direct-access-test.md) | | Order a genomic test | [help](help/clinic/order-a-genomic-test.md) | [admin](admin/clinic/order-a-genomic-test.md) | [API](api/clinic/order-a-genomic-test.md) | [journeys](../generated/diagrams/clinic/order-a-genomic-test.md) | | Order a repeat medicine | [help](help/clinic/order-a-repeat-medicine.md) | [admin](admin/clinic/order-a-repeat-medicine.md) | [API](api/clinic/order-a-repeat-medicine.md) | [journeys](../generated/diagrams/clinic/order-a-repeat-medicine.md) | | Order repeat medicines on behalf | [help](help/clinic/order-repeat-medicines-on-behalf.md) | [admin](admin/clinic/order-repeat-medicines-on-behalf.md) | [API](api/clinic/order-repeat-medicines-on-behalf.md) | [journeys](../generated/diagrams/clinic/order-repeat-medicines-on-behalf.md) | | Pair my wearable device | [help](help/clinic/pair-my-wearable-device.md) | [admin](admin/clinic/pair-my-wearable-device.md) | [API](api/clinic/pair-my-wearable-device.md) | [journeys](../generated/diagrams/clinic/pair-my-wearable-device.md) | | Person-held health record | [help](help/clinic/person-held-health-record.md) | [admin](admin/clinic/person-held-health-record.md) | [API](api/clinic/person-held-health-record.md) | [journeys](../generated/diagrams/clinic/person-held-health-record.md) | | Phone a critical result | [help](help/clinic/phone-a-critical-result.md) | [admin](admin/clinic/phone-a-critical-result.md) | [API](api/clinic/phone-a-critical-result.md) | [journeys](../generated/diagrams/clinic/phone-a-critical-result.md) | | Plan weekly capacity | [help](help/clinic/plan-weekly-capacity.md) | [admin](admin/clinic/plan-weekly-capacity.md) | [API](api/clinic/plan-weekly-capacity.md) | [journeys](../generated/diagrams/clinic/plan-weekly-capacity.md) | | Post a receipt to the Agent Channel | [help](help/clinic/post-a-receipt-to-the-agent-channel.md) | [admin](admin/clinic/post-a-receipt-to-the-agent-channel.md) | [API](api/clinic/post-a-receipt-to-the-agent-channel.md) | [journeys](../generated/diagrams/clinic/post-a-receipt-to-the-agent-channel.md) | | Predict fields in a form | [help](help/clinic/predict-fields-in-a-form.md) | [admin](admin/clinic/predict-fields-in-a-form.md) | [API](api/clinic/predict-fields-in-a-form.md) | [journeys](../generated/diagrams/clinic/predict-fields-in-a-form.md) | | Prepare a clinic letter for a clinician | [help](help/clinic/prepare-a-clinic-letter-for-a-clinician.md) | [admin](admin/clinic/prepare-a-clinic-letter-for-a-clinician.md) | [API](api/clinic/prepare-a-clinic-letter-for-a-clinician.md) | [journeys](../generated/diagrams/clinic/prepare-a-clinic-letter-for-a-clinician.md) | | Prepare a patient for imaging | [help](help/clinic/prepare-a-patient-for-imaging.md) | [admin](admin/clinic/prepare-a-patient-for-imaging.md) | [API](api/clinic/prepare-a-patient-for-imaging.md) | [journeys](../generated/diagrams/clinic/prepare-a-patient-for-imaging.md) | | Prepare for a CQC inspection | [help](help/clinic/prepare-for-a-cqc-inspection.md) | [admin](admin/clinic/prepare-for-a-cqc-inspection.md) | [API](api/clinic/prepare-for-a-cqc-inspection.md) | [journeys](../generated/diagrams/clinic/prepare-for-a-cqc-inspection.md) | | Prescribe | [help](help/clinic/prescribe.md) | [admin](admin/clinic/prescribe.md) | [API](api/clinic/prescribe.md) | [journeys](../generated/diagrams/clinic/prescribe.md) | | Prescribe under the pathway protocol | [help](help/clinic/prescribe-under-the-pathway-protocol.md) | [admin](admin/clinic/prescribe-under-the-pathway-protocol.md) | [API](api/clinic/prescribe-under-the-pathway-protocol.md) | [journeys](../generated/diagrams/clinic/prescribe-under-the-pathway-protocol.md) | | Price a policy from clinic data | [help](help/clinic/price-a-policy-from-clinic-data.md) | [admin](admin/clinic/price-a-policy-from-clinic-data.md) | [API](api/clinic/price-a-policy-from-clinic-data.md) | [journeys](../generated/diagrams/clinic/price-a-policy-from-clinic-data.md) | | Process a claim | [help](help/clinic/process-a-claim.md) | [admin](admin/clinic/process-a-claim.md) | [API](api/clinic/process-a-claim.md) | [journeys](../generated/diagrams/clinic/process-a-claim.md) | | Pseudonymise a dataset | [help](help/clinic/pseudonymise-a-dataset.md) | [admin](admin/clinic/pseudonymise-a-dataset.md) | [API](api/clinic/pseudonymise-a-dataset.md) | [journeys](../generated/diagrams/clinic/pseudonymise-a-dataset.md) | | Publish results | [help](help/clinic/publish-results.md) | [admin](admin/clinic/publish-results.md) | [API](api/clinic/publish-results.md) | [journeys](../generated/diagrams/clinic/publish-results.md) | | Raise a complaint or feedback | [help](help/clinic/raise-a-complaint-or-feedback.md) | [admin](admin/clinic/raise-a-complaint-or-feedback.md) | [API](api/clinic/raise-a-complaint-or-feedback.md) | [journeys](../generated/diagrams/clinic/raise-a-complaint-or-feedback.md) | | Raise a safeguarding concern from a response | [help](help/clinic/raise-a-safeguarding-concern-from-a-response.md) | [admin](admin/clinic/raise-a-safeguarding-concern-from-a-response.md) | [API](api/clinic/raise-a-safeguarding-concern-from-a-response.md) | [journeys](../generated/diagrams/clinic/raise-a-safeguarding-concern-from-a-response.md) | | Randomise a participant | [help](help/clinic/randomise-a-participant.md) | [admin](admin/clinic/randomise-a-participant.md) | [API](api/clinic/randomise-a-participant.md) | [journeys](../generated/diagrams/clinic/randomise-a-participant.md) | | Read assistive output on an image | [help](help/clinic/read-assistive-output-on-an-image.md) | [admin](admin/clinic/read-assistive-output-on-an-image.md) | [API](api/clinic/read-assistive-output-on-an-image.md) | [journeys](../generated/diagrams/clinic/read-assistive-output-on-an-image.md) | | Rebook or cancel an appointment | [help](help/clinic/rebook-or-cancel-an-appointment.md) | [admin](admin/clinic/rebook-or-cancel-an-appointment.md) | [API](api/clinic/rebook-or-cancel-an-appointment.md) | [journeys](../generated/diagrams/clinic/rebook-or-cancel-an-appointment.md) | | Receive a crisis alert | [help](help/clinic/receive-a-crisis-alert.md) | [admin](admin/clinic/receive-a-crisis-alert.md) | [API](api/clinic/receive-a-crisis-alert.md) | [journeys](../generated/diagrams/clinic/receive-a-crisis-alert.md) | | Receive a crisis alert for the person | [help](help/clinic/receive-a-crisis-alert-for-the-person.md) | [admin](admin/clinic/receive-a-crisis-alert-for-the-person.md) | [API](api/clinic/receive-a-crisis-alert-for-the-person.md) | [journeys](../generated/diagrams/clinic/receive-a-crisis-alert-for-the-person.md) | | Receive a critical result for my patient | [help](help/clinic/receive-a-critical-result-for-my-patient.md) | [admin](admin/clinic/receive-a-critical-result-for-my-patient.md) | [API](api/clinic/receive-a-critical-result-for-my-patient.md) | [journeys](../generated/diagrams/clinic/receive-a-critical-result-for-my-patient.md) | | Receive a did-not-attend notice | [help](help/clinic/receive-a-did-not-attend-notice.md) | [admin](admin/clinic/receive-a-did-not-attend-notice.md) | [API](api/clinic/receive-a-did-not-attend-notice.md) | [journeys](../generated/diagrams/clinic/receive-a-did-not-attend-notice.md) | | Receive a pre-authorisation request | [help](help/clinic/receive-a-pre-authorisation-request.md) | [admin](admin/clinic/receive-a-pre-authorisation-request.md) | [API](api/clinic/receive-a-pre-authorisation-request.md) | [journeys](../generated/diagrams/clinic/receive-a-pre-authorisation-request.md) | | Receive a statutory notification | [help](help/clinic/receive-a-statutory-notification.md) | [admin](admin/clinic/receive-a-statutory-notification.md) | [API](api/clinic/receive-a-statutory-notification.md) | [journeys](../generated/diagrams/clinic/receive-a-statutory-notification.md) | | Receive and book in a sample | [help](help/clinic/receive-and-book-in-a-sample.md) | [admin](admin/clinic/receive-and-book-in-a-sample.md) | [API](api/clinic/receive-and-book-in-a-sample.md) | [journeys](../generated/diagrams/clinic/receive-and-book-in-a-sample.md) | | Receive the outcome letter | [help](help/clinic/receive-the-outcome-letter.md) | [admin](admin/clinic/receive-the-outcome-letter.md) | [API](api/clinic/receive-the-outcome-letter.md) | [journeys](../generated/diagrams/clinic/receive-the-outcome-letter.md) | | Reclassify a variant | [help](help/clinic/reclassify-a-variant.md) | [admin](admin/clinic/reclassify-a-variant.md) | [API](api/clinic/reclassify-a-variant.md) | [journeys](../generated/diagrams/clinic/reclassify-a-variant.md) | | Reconcile medicines | [help](help/clinic/reconcile-medicines.md) | [admin](admin/clinic/reconcile-medicines.md) | [API](api/clinic/reconcile-medicines.md) | [journeys](../generated/diagrams/clinic/reconcile-medicines.md) | | Reconcile medicines at admission | [help](help/clinic/reconcile-medicines-at-admission.md) | [admin](admin/clinic/reconcile-medicines-at-admission.md) | [API](api/clinic/reconcile-medicines-at-admission.md) | [journeys](../generated/diagrams/clinic/reconcile-medicines-at-admission.md) | | Record a clinic consultation | [help](help/clinic/record-a-clinic-consultation.md) | [admin](admin/clinic/record-a-clinic-consultation.md) | [API](api/clinic/record-a-clinic-consultation.md) | [journeys](../generated/diagrams/clinic/record-a-clinic-consultation.md) | | Record a dose incident | [help](help/clinic/record-a-dose-incident.md) | [admin](admin/clinic/record-a-dose-incident.md) | [API](api/clinic/record-a-dose-incident.md) | [journeys](../generated/diagrams/clinic/record-a-dose-incident.md) | | Record a patient deferred | [help](help/clinic/record-a-patient-deferred.md) | [admin](admin/clinic/record-a-patient-deferred.md) | [API](api/clinic/record-a-patient-deferred.md) | [journeys](../generated/diagrams/clinic/record-a-patient-deferred.md) | | Record a patient's own view | [help](help/clinic/record-a-patient-s-own-view.md) | [admin](admin/clinic/record-a-patient-s-own-view.md) | [API](api/clinic/record-a-patient-s-own-view.md) | [journeys](../generated/diagrams/clinic/record-a-patient-s-own-view.md) | | Record a point-of-care urine result | [help](help/clinic/record-a-point-of-care-urine-result.md) | [admin](admin/clinic/record-a-point-of-care-urine-result.md) | [API](api/clinic/record-a-point-of-care-urine-result.md) | [journeys](../generated/diagrams/clinic/record-a-point-of-care-urine-result.md) | | Record a safeguarding concern | [help](help/clinic/record-a-safeguarding-concern.md) | [admin](admin/clinic/record-a-safeguarding-concern.md) | [API](api/clinic/record-a-safeguarding-concern.md) | [journeys](../generated/diagrams/clinic/record-a-safeguarding-concern.md) | | Record a self check-in exception | [help](help/clinic/record-a-self-check-in-exception.md) | [admin](admin/clinic/record-a-self-check-in-exception.md) | [API](api/clinic/record-a-self-check-in-exception.md) | [journeys](../generated/diagrams/clinic/record-a-self-check-in-exception.md) | | Record a wound or skin check | [help](help/clinic/record-a-wound-or-skin-check.md) | [admin](admin/clinic/record-a-wound-or-skin-check.md) | [API](api/clinic/record-a-wound-or-skin-check.md) | [journeys](../generated/diagrams/clinic/record-a-wound-or-skin-check.md) | | Record an adverse drug reaction | [help](help/clinic/record-an-adverse-drug-reaction.md) | [admin](admin/clinic/record-an-adverse-drug-reaction.md) | [API](api/clinic/record-an-adverse-drug-reaction.md) | [journeys](../generated/diagrams/clinic/record-an-adverse-drug-reaction.md) | | Record attendance outcome | [help](help/clinic/record-attendance-outcome.md) | [admin](admin/clinic/record-attendance-outcome.md) | [API](api/clinic/record-attendance-outcome.md) | [journeys](../generated/diagrams/clinic/record-attendance-outcome.md) | | Record capacity and best interests | [help](help/clinic/record-capacity-and-best-interests.md) | [admin](admin/clinic/record-capacity-and-best-interests.md) | [API](api/clinic/record-capacity-and-best-interests.md) | [journeys](../generated/diagrams/clinic/record-capacity-and-best-interests.md) | | Record decision to treat | [help](help/clinic/record-decision-to-treat.md) | [admin](admin/clinic/record-decision-to-treat.md) | [API](api/clinic/record-decision-to-treat.md) | [journeys](../generated/diagrams/clinic/record-decision-to-treat.md) | | Record sample consent version | [help](help/clinic/record-sample-consent-version.md) | [admin](admin/clinic/record-sample-consent-version.md) | [API](api/clinic/record-sample-consent-version.md) | [journeys](../generated/diagrams/clinic/record-sample-consent-version.md) | | Record the faster diagnosis outcome | [help](help/clinic/record-the-faster-diagnosis-outcome.md) | [admin](admin/clinic/record-the-faster-diagnosis-outcome.md) | [API](api/clinic/record-the-faster-diagnosis-outcome.md) | [journeys](../generated/diagrams/clinic/record-the-faster-diagnosis-outcome.md) | | Record the lawful basis | [help](help/clinic/record-the-lawful-basis.md) | [admin](admin/clinic/record-the-lawful-basis.md) | [API](api/clinic/record-the-lawful-basis.md) | [journeys](../generated/diagrams/clinic/record-the-lawful-basis.md) | | Record the MDT recommendation | [help](help/clinic/record-the-mdt-recommendation.md) | [admin](admin/clinic/record-the-mdt-recommendation.md) | [API](api/clinic/record-the-mdt-recommendation.md) | [journeys](../generated/diagrams/clinic/record-the-mdt-recommendation.md) | | Record trial eligibility at the MDT | [help](help/clinic/record-trial-eligibility-at-the-mdt.md) | [admin](admin/clinic/record-trial-eligibility-at-the-mdt.md) | [API](api/clinic/record-trial-eligibility-at-the-mdt.md) | [journeys](../generated/diagrams/clinic/record-trial-eligibility-at-the-mdt.md) | | Refer on a suspected cancer pathway | [help](help/clinic/refer-on-a-suspected-cancer-pathway.md) | [admin](admin/clinic/refer-on-a-suspected-cancer-pathway.md) | [API](api/clinic/refer-on-a-suspected-cancer-pathway.md) | [journeys](../generated/diagrams/clinic/refer-on-a-suspected-cancer-pathway.md) | | Refer to a crisis team | [help](help/clinic/refer-to-a-crisis-team.md) | [admin](admin/clinic/refer-to-a-crisis-team.md) | [API](api/clinic/refer-to-a-crisis-team.md) | [journeys](../generated/diagrams/clinic/refer-to-a-crisis-team.md) | | Refer to tertiary care | [help](help/clinic/refer-to-tertiary-care.md) | [admin](admin/clinic/refer-to-tertiary-care.md) | [API](api/clinic/refer-to-tertiary-care.md) | [journeys](../generated/diagrams/clinic/refer-to-tertiary-care.md) | | Refer to the mental health stream | [help](help/clinic/refer-to-the-mental-health-stream.md) | [admin](admin/clinic/refer-to-the-mental-health-stream.md) | [API](api/clinic/refer-to-the-mental-health-stream.md) | [journeys](../generated/diagrams/clinic/refer-to-the-mental-health-stream.md) | | Refer to the multidisciplinary team | [help](help/clinic/refer-to-the-multidisciplinary-team.md) | [admin](admin/clinic/refer-to-the-multidisciplinary-team.md) | [API](api/clinic/refer-to-the-multidisciplinary-team.md) | [journeys](../generated/diagrams/clinic/refer-to-the-multidisciplinary-team.md) | | Refresh public data caches | [help](help/clinic/refresh-public-data-caches.md) | [admin](admin/clinic/refresh-public-data-caches.md) | [API](api/clinic/refresh-public-data-caches.md) | [journeys](../generated/diagrams/clinic/refresh-public-data-caches.md) | | Register a child as a patient | [help](help/clinic/register-a-child-as-a-patient.md) | [admin](admin/clinic/register-a-child-as-a-patient.md) | [API](api/clinic/register-a-child-as-a-patient.md) | [journeys](../generated/diagrams/clinic/register-a-child-as-a-patient.md) | | Register a device fault | [help](help/clinic/register-a-device-fault.md) | [admin](admin/clinic/register-a-device-fault.md) | [API](api/clinic/register-a-device-fault.md) | [journeys](../generated/diagrams/clinic/register-a-device-fault.md) | | Register a walk-in patient | [help](help/clinic/register-a-walk-in-patient.md) | [admin](admin/clinic/register-a-walk-in-patient.md) | [API](api/clinic/register-a-walk-in-patient.md) | [journeys](../generated/diagrams/clinic/register-a-walk-in-patient.md) | | Register an imaging device | [help](help/clinic/register-an-imaging-device.md) | [admin](admin/clinic/register-an-imaging-device.md) | [API](api/clinic/register-an-imaging-device.md) | [journeys](../generated/diagrams/clinic/register-an-imaging-device.md) | | Register my practice with the clinic | [help](help/clinic/register-my-practice-with-the-clinic.md) | [admin](admin/clinic/register-my-practice-with-the-clinic.md) | [API](api/clinic/register-my-practice-with-the-clinic.md) | [journeys](../generated/diagrams/clinic/register-my-practice-with-the-clinic.md) | | Register the trial | [help](help/clinic/register-the-trial.md) | [admin](admin/clinic/register-the-trial.md) | [API](api/clinic/register-the-trial.md) | [journeys](../generated/diagrams/clinic/register-the-trial.md) | | Reject an unlabelled sample | [help](help/clinic/reject-an-unlabelled-sample.md) | [admin](admin/clinic/reject-an-unlabelled-sample.md) | [API](api/clinic/reject-an-unlabelled-sample.md) | [journeys](../generated/diagrams/clinic/reject-an-unlabelled-sample.md) | | Report a change in the person's condition | [help](help/clinic/report-a-change-in-the-person-s-condition.md) | [admin](admin/clinic/report-a-change-in-the-person-s-condition.md) | [API](api/clinic/report-a-change-in-the-person-s-condition.md) | [journeys](../generated/diagrams/clinic/report-a-change-in-the-person-s-condition.md) | | Report a data breach | [help](help/clinic/report-a-data-breach.md) | [admin](admin/clinic/report-a-data-breach.md) | [API](api/clinic/report-a-data-breach.md) | [journeys](../generated/diagrams/clinic/report-a-data-breach.md) | | Report a dispensing error | [help](help/clinic/report-a-dispensing-error.md) | [admin](admin/clinic/report-a-dispensing-error.md) | [API](api/clinic/report-a-dispensing-error.md) | [journeys](../generated/diagrams/clinic/report-a-dispensing-error.md) | | Report a safety incident | [help](help/clinic/report-a-safety-incident.md) | [admin](admin/clinic/report-a-safety-incident.md) | [API](api/clinic/report-a-safety-incident.md) | [journeys](../generated/diagrams/clinic/report-a-safety-incident.md) | | Report a side effect | [help](help/clinic/report-a-side-effect.md) | [admin](admin/clinic/report-a-side-effect.md) | [API](api/clinic/report-a-side-effect.md) | [journeys](../generated/diagrams/clinic/report-a-side-effect.md) | | Report a study | [help](help/clinic/report-a-study.md) | [admin](admin/clinic/report-a-study.md) | [API](api/clinic/report-a-study.md) | [journeys](../generated/diagrams/clinic/report-a-study.md) | | Report an adverse event | [help](help/clinic/report-an-adverse-event.md) | [admin](admin/clinic/report-an-adverse-event.md) | [API](api/clinic/report-an-adverse-event.md) | [journeys](../generated/diagrams/clinic/report-an-adverse-event.md) | | Report an incident | [help](help/clinic/report-an-incident.md) | [admin](admin/clinic/report-an-incident.md) | [API](api/clinic/report-an-incident.md) | [journeys](../generated/diagrams/clinic/report-an-incident.md) | | Report histology | [help](help/clinic/report-histology.md) | [admin](admin/clinic/report-histology.md) | [API](api/clinic/report-histology.md) | [journeys](../generated/diagrams/clinic/report-histology.md) | | Report monthly performance | [help](help/clinic/report-monthly-performance.md) | [admin](admin/clinic/report-monthly-performance.md) | [API](api/clinic/report-monthly-performance.md) | [journeys](../generated/diagrams/clinic/report-monthly-performance.md) | | Report to the regulator | [help](help/clinic/report-to-the-regulator.md) | [admin](admin/clinic/report-to-the-regulator.md) | [API](api/clinic/report-to-the-regulator.md) | [journeys](../generated/diagrams/clinic/report-to-the-regulator.md) | | Report usage and efficiency | [help](help/clinic/report-usage-and-efficiency.md) | [admin](admin/clinic/report-usage-and-efficiency.md) | [API](api/clinic/report-usage-and-efficiency.md) | [journeys](../generated/diagrams/clinic/report-usage-and-efficiency.md) | | Request a copy of my record | [help](help/clinic/request-a-copy-of-my-record.md) | [admin](admin/clinic/request-a-copy-of-my-record.md) | [API](api/clinic/request-a-copy-of-my-record.md) | [journeys](../generated/diagrams/clinic/request-a-copy-of-my-record.md) | | Request a rota swap or leave | [help](help/clinic/request-a-rota-swap-or-leave.md) | [admin](admin/clinic/request-a-rota-swap-or-leave.md) | [API](api/clinic/request-a-rota-swap-or-leave.md) | [journeys](../generated/diagrams/clinic/request-a-rota-swap-or-leave.md) | | Request a second opinion | [help](help/clinic/request-a-second-opinion.md) | [admin](admin/clinic/request-a-second-opinion.md) | [API](api/clinic/request-a-second-opinion.md) | [journeys](../generated/diagrams/clinic/request-a-second-opinion.md) | | Request a second read | [help](help/clinic/request-a-second-read.md) | [admin](admin/clinic/request-a-second-read.md) | [API](api/clinic/request-a-second-read.md) | [journeys](../generated/diagrams/clinic/request-a-second-read.md) | | Request a second read or discrepancy review | [help](help/clinic/request-a-second-read-or-discrepancy-review.md) | [admin](admin/clinic/request-a-second-read-or-discrepancy-review.md) | [API](api/clinic/request-a-second-read-or-discrepancy-review.md) | [journeys](../generated/diagrams/clinic/request-a-second-read-or-discrepancy-review.md) | | Request a young person's own access | [help](help/clinic/request-a-young-person-s-own-access.md) | [admin](admin/clinic/request-a-young-person-s-own-access.md) | [API](api/clinic/request-a-young-person-s-own-access.md) | [journeys](../generated/diagrams/clinic/request-a-young-person-s-own-access.md) | | Request clinic data for a quote | [help](help/clinic/request-clinic-data-for-a-quote.md) | [admin](admin/clinic/request-clinic-data-for-a-quote.md) | [API](api/clinic/request-clinic-data-for-a-quote.md) | [journeys](../generated/diagrams/clinic/request-clinic-data-for-a-quote.md) | | Request evidence against fundamental standards | [help](help/clinic/request-evidence-against-fundamental-standards.md) | [admin](admin/clinic/request-evidence-against-fundamental-standards.md) | [API](api/clinic/request-evidence-against-fundamental-standards.md) | [journeys](../generated/diagrams/clinic/request-evidence-against-fundamental-standards.md) | | Retinal screening | [help](help/clinic/retinal-screening.md) | [admin](admin/clinic/retinal-screening.md) | [API](api/clinic/retinal-screening.md) | [journeys](../generated/diagrams/clinic/retinal-screening.md) | | Review a breach notification | [help](help/clinic/review-a-breach-notification.md) | [admin](admin/clinic/review-a-breach-notification.md) | [API](api/clinic/review-a-breach-notification.md) | [journeys](../generated/diagrams/clinic/review-a-breach-notification.md) | | Review a death by suicide | [help](help/clinic/review-a-death-by-suicide.md) | [admin](admin/clinic/review-a-death-by-suicide.md) | [API](api/clinic/review-a-death-by-suicide.md) | [journeys](../generated/diagrams/clinic/review-a-death-by-suicide.md) | | Review a real result | [help](help/clinic/review-a-real-result.md) | [admin](admin/clinic/review-a-real-result.md) | [API](api/clinic/review-a-real-result.md) | [journeys](../generated/diagrams/clinic/review-a-real-result.md) | | Review a release for clinical safety | [help](help/clinic/review-a-release-for-clinical-safety.md) | [admin](admin/clinic/review-a-release-for-clinical-safety.md) | [API](api/clinic/review-a-release-for-clinical-safety.md) | [journeys](../generated/diagrams/clinic/review-a-release-for-clinical-safety.md) | | Review a repeat request | [help](help/clinic/review-a-repeat-request.md) | [admin](admin/clinic/review-a-repeat-request.md) | [API](api/clinic/review-a-repeat-request.md) | [journeys](../generated/diagrams/clinic/review-a-repeat-request.md) | | Review a trial for compliance | [help](help/clinic/review-a-trial-for-compliance.md) | [admin](admin/clinic/review-a-trial-for-compliance.md) | [API](api/clinic/review-a-trial-for-compliance.md) | [journeys](../generated/diagrams/clinic/review-a-trial-for-compliance.md) | | Review a wearable alert | [help](help/clinic/review-a-wearable-alert.md) | [admin](admin/clinic/review-a-wearable-alert.md) | [API](api/clinic/review-a-wearable-alert.md) | [journeys](../generated/diagrams/clinic/review-a-wearable-alert.md) | | Review assistive output labelling | [help](help/clinic/review-assistive-output-labelling.md) | [admin](admin/clinic/review-assistive-output-labelling.md) | [API](api/clinic/review-assistive-output-labelling.md) | [journeys](../generated/diagrams/clinic/review-assistive-output-labelling.md) | | Review payslips and expenses approvals | [help](help/clinic/review-payslips-and-expenses-approvals.md) | [admin](admin/clinic/review-payslips-and-expenses-approvals.md) | [API](api/clinic/review-payslips-and-expenses-approvals.md) | [journeys](../generated/diagrams/clinic/review-payslips-and-expenses-approvals.md) | | Review the caseload | [help](help/clinic/review-the-caseload.md) | [admin](admin/clinic/review-the-caseload.md) | [API](api/clinic/review-the-caseload.md) | [journeys](../generated/diagrams/clinic/review-the-caseload.md) | | Review the dashboard | [help](help/clinic/review-the-dashboard.md) | [admin](admin/clinic/review-the-dashboard.md) | [API](api/clinic/review-the-dashboard.md) | [journeys](../generated/diagrams/clinic/review-the-dashboard.md) | | Review the waiting list position | [help](help/clinic/review-the-waiting-list-position.md) | [admin](admin/clinic/review-the-waiting-list-position.md) | [API](api/clinic/review-the-waiting-list-position.md) | [journeys](../generated/diagrams/clinic/review-the-waiting-list-position.md) | | Review the worklist | [help](help/clinic/review-the-worklist.md) | [admin](admin/clinic/review-the-worklist.md) | [API](api/clinic/review-the-worklist.md) | [journeys](../generated/diagrams/clinic/review-the-worklist.md) | | Revoke access when a mandate ends | [help](help/clinic/revoke-access-when-a-mandate-ends.md) | [admin](admin/clinic/revoke-access-when-a-mandate-ends.md) | [API](api/clinic/revoke-access-when-a-mandate-ends.md) | [journeys](../generated/diagrams/clinic/revoke-access-when-a-mandate-ends.md) | | Revoke my own proxy role | [help](help/clinic/revoke-my-own-proxy-role.md) | [admin](admin/clinic/revoke-my-own-proxy-role.md) | [API](api/clinic/revoke-my-own-proxy-role.md) | [journeys](../generated/diagrams/clinic/revoke-my-own-proxy-role.md) | | Reward early detection participation | [help](help/clinic/reward-early-detection-participation.md) | [admin](admin/clinic/reward-early-detection-participation.md) | [API](api/clinic/reward-early-detection-participation.md) | [journeys](../generated/diagrams/clinic/reward-early-detection-participation.md) | | Roll back a release | [help](help/clinic/roll-back-a-release.md) | [admin](admin/clinic/roll-back-a-release.md) | [API](api/clinic/roll-back-a-release.md) | [journeys](../generated/diagrams/clinic/roll-back-a-release.md) | | Rotate keys and certificates | [help](help/clinic/rotate-keys-and-certificates.md) | [admin](admin/clinic/rotate-keys-and-certificates.md) | [API](api/clinic/rotate-keys-and-certificates.md) | [journeys](../generated/diagrams/clinic/rotate-keys-and-certificates.md) | | Run a business continuity drill | [help](help/clinic/run-a-business-continuity-drill.md) | [admin](admin/clinic/run-a-business-continuity-drill.md) | [API](api/clinic/run-a-business-continuity-drill.md) | [journeys](../generated/diagrams/clinic/run-a-business-continuity-drill.md) | | Run a catalogue workload | [help](help/clinic/run-a-catalogue-workload.md) | [admin](admin/clinic/run-a-catalogue-workload.md) | [API](api/clinic/run-a-catalogue-workload.md) | [journeys](../generated/diagrams/clinic/run-a-catalogue-workload.md) | | Run a clinical safety training check | [help](help/clinic/run-a-clinical-safety-training-check.md) | [admin](admin/clinic/run-a-clinical-safety-training-check.md) | [API](api/clinic/run-a-clinical-safety-training-check.md) | [journeys](../generated/diagrams/clinic/run-a-clinical-safety-training-check.md) | | Run a data protection impact assessment | [help](help/clinic/run-a-data-protection-impact-assessment.md) | [admin](admin/clinic/run-a-data-protection-impact-assessment.md) | [API](api/clinic/run-a-data-protection-impact-assessment.md) | [journeys](../generated/diagrams/clinic/run-a-data-protection-impact-assessment.md) | | Run a pipeline | [help](help/clinic/run-a-pipeline.md) | [admin](admin/clinic/run-a-pipeline.md) | [API](api/clinic/run-a-pipeline.md) | [journeys](../generated/diagrams/clinic/run-a-pipeline.md) | | Run a pre-assessment | [help](help/clinic/run-a-pre-assessment.md) | [admin](admin/clinic/run-a-pre-assessment.md) | [API](api/clinic/run-a-pre-assessment.md) | [journeys](../generated/diagrams/clinic/run-a-pre-assessment.md) | | Run a safeguarding review | [help](help/clinic/run-a-safeguarding-review.md) | [admin](admin/clinic/run-a-safeguarding-review.md) | [API](api/clinic/run-a-safeguarding-review.md) | [journeys](../generated/diagrams/clinic/run-a-safeguarding-review.md) | | Run a synthetic pathway drill | [help](help/clinic/run-a-synthetic-pathway-drill.md) | [admin](admin/clinic/run-a-synthetic-pathway-drill.md) | [API](api/clinic/run-a-synthetic-pathway-drill.md) | [journeys](../generated/diagrams/clinic/run-a-synthetic-pathway-drill.md) | | Run a urine test likelihood ledger | [help](help/clinic/run-a-urine-test-likelihood-ledger.md) | [admin](admin/clinic/run-a-urine-test-likelihood-ledger.md) | [API](api/clinic/run-a-urine-test-likelihood-ledger.md) | [journeys](../generated/diagrams/clinic/run-a-urine-test-likelihood-ledger.md) | | Run an interim analysis | [help](help/clinic/run-an-interim-analysis.md) | [admin](admin/clinic/run-an-interim-analysis.md) | [API](api/clinic/run-an-interim-analysis.md) | [journeys](../generated/diagrams/clinic/run-an-interim-analysis.md) | | Run an online early symptom test | [help](help/clinic/run-an-online-early-symptom-test.md) | [admin](admin/clinic/run-an-online-early-symptom-test.md) | [API](api/clinic/run-an-online-early-symptom-test.md) | [journeys](../generated/diagrams/clinic/run-an-online-early-symptom-test.md) | | Run quality control | [help](help/clinic/run-quality-control.md) | [admin](admin/clinic/run-quality-control.md) | [API](api/clinic/run-quality-control.md) | [journeys](../generated/diagrams/clinic/run-quality-control.md) | | Run scanner quality assurance | [help](help/clinic/run-scanner-quality-assurance.md) | [admin](admin/clinic/run-scanner-quality-assurance.md) | [API](api/clinic/run-scanner-quality-assurance.md) | [journeys](../generated/diagrams/clinic/run-scanner-quality-assurance.md) | | Run sequencing QC | [help](help/clinic/run-sequencing-qc.md) | [admin](admin/clinic/run-sequencing-qc.md) | [API](api/clinic/run-sequencing-qc.md) | [journeys](../generated/diagrams/clinic/run-sequencing-qc.md) | | Run the day's capacity check | [help](help/clinic/run-the-day-s-capacity-check.md) | [admin](admin/clinic/run-the-day-s-capacity-check.md) | [API](api/clinic/run-the-day-s-capacity-check.md) | [journeys](../generated/diagrams/clinic/run-the-day-s-capacity-check.md) | | Run the MDT meeting | [help](help/clinic/run-the-mdt-meeting.md) | [admin](admin/clinic/run-the-mdt-meeting.md) | [API](api/clinic/run-the-mdt-meeting.md) | [journeys](../generated/diagrams/clinic/run-the-mdt-meeting.md) | | Run the training controller | [help](help/clinic/run-the-training-controller.md) | [admin](admin/clinic/run-the-training-controller.md) | [API](api/clinic/run-the-training-controller.md) | [journeys](../generated/diagrams/clinic/run-the-training-controller.md) | | Run vulnerability scans | [help](help/clinic/run-vulnerability-scans.md) | [admin](admin/clinic/run-vulnerability-scans.md) | [API](api/clinic/run-vulnerability-scans.md) | [journeys](../generated/diagrams/clinic/run-vulnerability-scans.md) | | Schedule the scan | [help](help/clinic/schedule-the-scan.md) | [admin](admin/clinic/schedule-the-scan.md) | [API](api/clinic/schedule-the-scan.md) | [journeys](../generated/diagrams/clinic/schedule-the-scan.md) | | Score a model on a held split | [help](help/clinic/score-a-model-on-a-held-split.md) | [admin](admin/clinic/score-a-model-on-a-held-split.md) | [API](api/clinic/score-a-model-on-a-held-split.md) | [journeys](../generated/diagrams/clinic/score-a-model-on-a-held-split.md) | | Score and interpret a response | [help](help/clinic/score-and-interpret-a-response.md) | [admin](admin/clinic/score-and-interpret-a-response.md) | [API](api/clinic/score-and-interpret-a-response.md) | [journeys](../generated/diagrams/clinic/score-and-interpret-a-response.md) | | Screen a patient for eligibility | [help](help/clinic/screen-a-patient-for-eligibility.md) | [admin](admin/clinic/screen-a-patient-for-eligibility.md) | [API](api/clinic/screen-a-patient-for-eligibility.md) | [journeys](../generated/diagrams/clinic/screen-a-patient-for-eligibility.md) | | Search recruiting trials | [help](help/clinic/search-recruiting-trials.md) | [admin](admin/clinic/search-recruiting-trials.md) | [API](api/clinic/search-recruiting-trials.md) | [journeys](../generated/diagrams/clinic/search-recruiting-trials.md) | | See my day | [help](help/clinic/see-my-day.md) | [admin](admin/clinic/see-my-day.md) | [API](api/clinic/see-my-day.md) | [journeys](../generated/diagrams/clinic/see-my-day.md) | | Send a check-in to a caseload | [help](help/clinic/send-a-check-in-to-a-caseload.md) | [admin](admin/clinic/send-a-check-in-to-a-caseload.md) | [API](api/clinic/send-a-check-in-to-a-caseload.md) | [journeys](../generated/diagrams/clinic/send-a-check-in-to-a-caseload.md) | | Send a results-ready notice | [help](help/clinic/send-a-results-ready-notice.md) | [admin](admin/clinic/send-a-results-ready-notice.md) | [API](api/clinic/send-a-results-ready-notice.md) | [journeys](../generated/diagrams/clinic/send-a-results-ready-notice.md) | | Send appointment letters and reminders | [help](help/clinic/send-appointment-letters-and-reminders.md) | [admin](admin/clinic/send-appointment-letters-and-reminders.md) | [API](api/clinic/send-appointment-letters-and-reminders.md) | [journeys](../generated/diagrams/clinic/send-appointment-letters-and-reminders.md) | | Serve a vision model | [help](help/clinic/serve-a-vision-model.md) | [admin](admin/clinic/serve-a-vision-model.md) | [API](api/clinic/serve-a-vision-model.md) | [journeys](../generated/diagrams/clinic/serve-a-vision-model.md) | | Set a surveillance plan | [help](help/clinic/set-a-surveillance-plan.md) | [admin](admin/clinic/set-a-surveillance-plan.md) | [API](api/clinic/set-a-surveillance-plan.md) | [journeys](../generated/diagrams/clinic/set-a-surveillance-plan.md) | | Set my communication preferences | [help](help/clinic/set-my-communication-preferences.md) | [admin](admin/clinic/set-my-communication-preferences.md) | [API](api/clinic/set-my-communication-preferences.md) | [journeys](../generated/diagrams/clinic/set-my-communication-preferences.md) | | Set my data sharing choices | [help](help/clinic/set-my-data-sharing-choices.md) | [admin](admin/clinic/set-my-data-sharing-choices.md) | [API](api/clinic/set-my-data-sharing-choices.md) | [journeys](../generated/diagrams/clinic/set-my-data-sharing-choices.md) | | Set role permissions and expiry | [help](help/clinic/set-role-permissions-and-expiry.md) | [admin](admin/clinic/set-role-permissions-and-expiry.md) | [API](api/clinic/set-role-permissions-and-expiry.md) | [journeys](../generated/diagrams/clinic/set-role-permissions-and-expiry.md) | | Settle with the clinic | [help](help/clinic/settle-with-the-clinic.md) | [admin](admin/clinic/settle-with-the-clinic.md) | [API](api/clinic/settle-with-the-clinic.md) | [journeys](../generated/diagrams/clinic/settle-with-the-clinic.md) | | Share a medicines list with the clinic | [help](help/clinic/share-a-medicines-list-with-the-clinic.md) | [admin](admin/clinic/share-a-medicines-list-with-the-clinic.md) | [API](api/clinic/share-a-medicines-list-with-the-clinic.md) | [journeys](../generated/diagrams/clinic/share-a-medicines-list-with-the-clinic.md) | | Share a wearable signal | [help](help/clinic/share-a-wearable-signal.md) | [admin](admin/clinic/share-a-wearable-signal.md) | [API](api/clinic/share-a-wearable-signal.md) | [journeys](../generated/diagrams/clinic/share-a-wearable-signal.md) | | Share clinic data with my insurer | [help](help/clinic/share-clinic-data-with-my-insurer.md) | [admin](admin/clinic/share-clinic-data-with-my-insurer.md) | [API](api/clinic/share-clinic-data-with-my-insurer.md) | [journeys](../generated/diagrams/clinic/share-clinic-data-with-my-insurer.md) | | Share images with another trust | [help](help/clinic/share-images-with-another-trust.md) | [admin](admin/clinic/share-images-with-another-trust.md) | [API](api/clinic/share-images-with-another-trust.md) | [journeys](../generated/diagrams/clinic/share-images-with-another-trust.md) | | Share risk information with partners | [help](help/clinic/share-risk-information-with-partners.md) | [admin](admin/clinic/share-risk-information-with-partners.md) | [API](api/clinic/share-risk-information-with-partners.md) | [journeys](../generated/diagrams/clinic/share-risk-information-with-partners.md) | | Sign a clinic letter | [help](help/clinic/sign-a-clinic-letter.md) | [admin](admin/clinic/sign-a-clinic-letter.md) | [API](api/clinic/sign-a-clinic-letter.md) | [journeys](../generated/diagrams/clinic/sign-a-clinic-letter.md) | | Sign the clinical risk management plan | [help](help/clinic/sign-the-clinical-risk-management-plan.md) | [admin](admin/clinic/sign-the-clinical-risk-management-plan.md) | [API](api/clinic/sign-the-clinical-risk-management-plan.md) | [journeys](../generated/diagrams/clinic/sign-the-clinical-risk-management-plan.md) | | Simulate trial monitoring | [help](help/clinic/simulate-trial-monitoring.md) | [admin](admin/clinic/simulate-trial-monitoring.md) | [API](api/clinic/simulate-trial-monitoring.md) | [journeys](../generated/diagrams/clinic/simulate-trial-monitoring.md) | | Social prescribing referral | [help](help/clinic/social-prescribing-referral.md) | [admin](admin/clinic/social-prescribing-referral.md) | [API](api/clinic/social-prescribing-referral.md) | [journeys](../generated/diagrams/clinic/social-prescribing-referral.md) | | Stop on revoked authority | [help](help/clinic/stop-on-revoked-authority.md) | [admin](admin/clinic/stop-on-revoked-authority.md) | [API](api/clinic/stop-on-revoked-authority.md) | [journeys](../generated/diagrams/clinic/stop-on-revoked-authority.md) | | Store and retrieve genomic data | [help](help/clinic/store-and-retrieve-genomic-data.md) | [admin](admin/clinic/store-and-retrieve-genomic-data.md) | [API](api/clinic/store-and-retrieve-genomic-data.md) | [journeys](../generated/diagrams/clinic/store-and-retrieve-genomic-data.md) | | Submit a cancer registration | [help](help/clinic/submit-a-cancer-registration.md) | [admin](admin/clinic/submit-a-cancer-registration.md) | [API](api/clinic/submit-a-cancer-registration.md) | [journeys](../generated/diagrams/clinic/submit-a-cancer-registration.md) | | Submit cancer waiting times data | [help](help/clinic/submit-cancer-waiting-times-data.md) | [admin](admin/clinic/submit-cancer-waiting-times-data.md) | [API](api/clinic/submit-cancer-waiting-times-data.md) | [journeys](../generated/diagrams/clinic/submit-cancer-waiting-times-data.md) | | Submit COSD and other datasets | [help](help/clinic/submit-cosd-and-other-datasets.md) | [admin](admin/clinic/submit-cosd-and-other-datasets.md) | [API](api/clinic/submit-cosd-and-other-datasets.md) | [journeys](../generated/diagrams/clinic/submit-cosd-and-other-datasets.md) | | Submit MDT data | [help](help/clinic/submit-mdt-data.md) | [admin](admin/clinic/submit-mdt-data.md) | [API](api/clinic/submit-mdt-data.md) | [journeys](../generated/diagrams/clinic/submit-mdt-data.md) | | Submit the mental health dataset | [help](help/clinic/submit-the-mental-health-dataset.md) | [admin](admin/clinic/submit-the-mental-health-dataset.md) | [API](api/clinic/submit-the-mental-health-dataset.md) | [journeys](../generated/diagrams/clinic/submit-the-mental-health-dataset.md) | | Supervise and sign off a trainee | [help](help/clinic/supervise-and-sign-off-a-trainee.md) | [admin](admin/clinic/supervise-and-sign-off-a-trainee.md) | [API](api/clinic/supervise-and-sign-off-a-trainee.md) | [journeys](../generated/diagrams/clinic/supervise-and-sign-off-a-trainee.md) | | Support deploying organisations | [help](help/clinic/support-deploying-organisations.md) | [admin](admin/clinic/support-deploying-organisations.md) | [API](api/clinic/support-deploying-organisations.md) | [journeys](../generated/diagrams/clinic/support-deploying-organisations.md) | | Symptom check and triage | [help](help/clinic/symptom-check-and-triage.md) | [admin](admin/clinic/symptom-check-and-triage.md) | [API](api/clinic/symptom-check-and-triage.md) | [journeys](../generated/diagrams/clinic/symptom-check-and-triage.md) | | Take a blood or urine sample | [help](help/clinic/take-a-blood-or-urine-sample.md) | [admin](admin/clinic/take-a-blood-or-urine-sample.md) | [API](api/clinic/take-a-blood-or-urine-sample.md) | [journeys](../generated/diagrams/clinic/take-a-blood-or-urine-sample.md) | | Take a mandate ticket | [help](help/clinic/take-a-mandate-ticket.md) | [admin](admin/clinic/take-a-mandate-ticket.md) | [API](api/clinic/take-a-mandate-ticket.md) | [journeys](../generated/diagrams/clinic/take-a-mandate-ticket.md) | | Take an online early symptom test | [help](help/clinic/take-an-online-early-symptom-test.md) | [admin](admin/clinic/take-an-online-early-symptom-test.md) | [API](api/clinic/take-an-online-early-symptom-test.md) | [journeys](../generated/diagrams/clinic/take-an-online-early-symptom-test.md) | | Take informed consent for research | [help](help/clinic/take-informed-consent-for-research.md) | [admin](admin/clinic/take-informed-consent-for-research.md) | [API](api/clinic/take-informed-consent-for-research.md) | [journeys](../generated/diagrams/clinic/take-informed-consent-for-research.md) | | Take observations | [help](help/clinic/take-observations.md) | [admin](admin/clinic/take-observations.md) | [API](api/clinic/take-observations.md) | [journeys](../generated/diagrams/clinic/take-observations.md) | | Take part in external quality assessment | [help](help/clinic/take-part-in-external-quality-assessment.md) | [admin](admin/clinic/take-part-in-external-quality-assessment.md) | [API](api/clinic/take-part-in-external-quality-assessment.md) | [journeys](../generated/diagrams/clinic/take-part-in-external-quality-assessment.md) | | Tell the patient a result | [help](help/clinic/tell-the-patient-a-result.md) | [admin](admin/clinic/tell-the-patient-a-result.md) | [API](api/clinic/tell-the-patient-a-result.md) | [journeys](../generated/diagrams/clinic/tell-the-patient-a-result.md) | | Test a hypothesis on an open cohort | [help](help/clinic/test-a-hypothesis-on-an-open-cohort.md) | [admin](admin/clinic/test-a-hypothesis-on-an-open-cohort.md) | [API](api/clinic/test-a-hypothesis-on-an-open-cohort.md) | [journeys](../generated/diagrams/clinic/test-a-hypothesis-on-an-open-cohort.md) | | Test a pre-registered hypothesis | [help](help/clinic/test-a-pre-registered-hypothesis.md) | [admin](admin/clinic/test-a-pre-registered-hypothesis.md) | [API](api/clinic/test-a-pre-registered-hypothesis.md) | [journeys](../generated/diagrams/clinic/test-a-pre-registered-hypothesis.md) | | Trace ingress and egress | [help](help/clinic/trace-ingress-and-egress.md) | [admin](admin/clinic/trace-ingress-and-egress.md) | [API](api/clinic/trace-ingress-and-egress.md) | [journeys](../generated/diagrams/clinic/trace-ingress-and-egress.md) | | Track a referral | [help](help/clinic/track-a-referral.md) | [admin](admin/clinic/track-a-referral.md) | [API](api/clinic/track-a-referral.md) | [journeys](../generated/diagrams/clinic/track-a-referral.md) | | Track actions from the MDT | [help](help/clinic/track-actions-from-the-mdt.md) | [admin](admin/clinic/track-actions-from-the-mdt.md) | [API](api/clinic/track-actions-from-the-mdt.md) | [journeys](../generated/diagrams/clinic/track-actions-from-the-mdt.md) | | Track histology turnaround | [help](help/clinic/track-histology-turnaround.md) | [admin](admin/clinic/track-histology-turnaround.md) | [API](api/clinic/track-histology-turnaround.md) | [journeys](../generated/diagrams/clinic/track-histology-turnaround.md) | | Track my cancer pathway | [help](help/clinic/track-my-cancer-pathway.md) | [admin](admin/clinic/track-my-cancer-pathway.md) | [API](api/clinic/track-my-cancer-pathway.md) | [journeys](../generated/diagrams/clinic/track-my-cancer-pathway.md) | | Track sample custody | [help](help/clinic/track-sample-custody.md) | [admin](admin/clinic/track-sample-custody.md) | [API](api/clinic/track-sample-custody.md) | [journeys](../generated/diagrams/clinic/track-sample-custody.md) | | Track the 62-day clock across the MDT | [help](help/clinic/track-the-62-day-clock-across-the-mdt.md) | [admin](admin/clinic/track-the-62-day-clock-across-the-mdt.md) | [API](api/clinic/track-the-62-day-clock-across-the-mdt.md) | [journeys](../generated/diagrams/clinic/track-the-62-day-clock-across-the-mdt.md) | | Train staff in information governance | [help](help/clinic/train-staff-in-information-governance.md) | [admin](admin/clinic/train-staff-in-information-governance.md) | [API](api/clinic/train-staff-in-information-governance.md) | [journeys](../generated/diagrams/clinic/train-staff-in-information-governance.md) | | Train staff in suicide awareness | [help](help/clinic/train-staff-in-suicide-awareness.md) | [admin](admin/clinic/train-staff-in-suicide-awareness.md) | [API](api/clinic/train-staff-in-suicide-awareness.md) | [journeys](../generated/diagrams/clinic/train-staff-in-suicide-awareness.md) | | Triage a mental health referral | [help](help/clinic/triage-a-mental-health-referral.md) | [admin](admin/clinic/triage-a-mental-health-referral.md) | [API](api/clinic/triage-a-mental-health-referral.md) | [journeys](../generated/diagrams/clinic/triage-a-mental-health-referral.md) | | Triage a patient message | [help](help/clinic/triage-a-patient-message.md) | [admin](admin/clinic/triage-a-patient-message.md) | [API](api/clinic/triage-a-patient-message.md) | [journeys](../generated/diagrams/clinic/triage-a-patient-message.md) | | Triage a referral | [help](help/clinic/triage-a-referral.md) | [admin](admin/clinic/triage-a-referral.md) | [API](api/clinic/triage-a-referral.md) | [journeys](../generated/diagrams/clinic/triage-a-referral.md) | | Triage inbox items for a clinician | [help](help/clinic/triage-inbox-items-for-a-clinician.md) | [admin](admin/clinic/triage-inbox-items-for-a-clinician.md) | [API](api/clinic/triage-inbox-items-for-a-clinician.md) | [journeys](../generated/diagrams/clinic/triage-inbox-items-for-a-clinician.md) | | Tune alert thresholds | [help](help/clinic/tune-alert-thresholds.md) | [admin](admin/clinic/tune-alert-thresholds.md) | [API](api/clinic/tune-alert-thresholds.md) | [journeys](../generated/diagrams/clinic/tune-alert-thresholds.md) | | Update patient demographics | [help](help/clinic/update-patient-demographics.md) | [admin](admin/clinic/update-patient-demographics.md) | [API](api/clinic/update-patient-demographics.md) | [journeys](../generated/diagrams/clinic/update-patient-demographics.md) | | Validate a new assay | [help](help/clinic/validate-a-new-assay.md) | [admin](admin/clinic/validate-a-new-assay.md) | [API](api/clinic/validate-a-new-assay.md) | [journeys](../generated/diagrams/clinic/validate-a-new-assay.md) | | Validate the test against outcomes | [help](help/clinic/validate-the-test-against-outcomes.md) | [admin](admin/clinic/validate-the-test-against-outcomes.md) | [API](api/clinic/validate-the-test-against-outcomes.md) | [journeys](../generated/diagrams/clinic/validate-the-test-against-outcomes.md) | | Verify a device classification | [help](help/clinic/verify-a-device-classification.md) | [admin](admin/clinic/verify-a-device-classification.md) | [API](api/clinic/verify-a-device-classification.md) | [journeys](../generated/diagrams/clinic/verify-a-device-classification.md) | | Verify an applicant's identity | [help](help/clinic/verify-an-applicant-s-identity.md) | [admin](admin/clinic/verify-an-applicant-s-identity.md) | [API](api/clinic/verify-an-applicant-s-identity.md) | [journeys](../generated/diagrams/clinic/verify-an-applicant-s-identity.md) | | Verify an evidence statement | [help](help/clinic/verify-an-evidence-statement.md) | [admin](admin/clinic/verify-an-evidence-statement.md) | [API](api/clinic/verify-an-evidence-statement.md) | [journeys](../generated/diagrams/clinic/verify-an-evidence-statement.md) | | Verify backups and restore | [help](help/clinic/verify-backups-and-restore.md) | [admin](admin/clinic/verify-backups-and-restore.md) | [API](api/clinic/verify-backups-and-restore.md) | [journeys](../generated/diagrams/clinic/verify-backups-and-restore.md) | | Vet an imaging request | [help](help/clinic/vet-an-imaging-request.md) | [admin](admin/clinic/vet-an-imaging-request.md) | [API](api/clinic/vet-an-imaging-request.md) | [journeys](../generated/diagrams/clinic/vet-an-imaging-request.md) | | View my medicines list | [help](help/clinic/view-my-medicines-list.md) | [admin](admin/clinic/view-my-medicines-list.md) | [API](api/clinic/view-my-medicines-list.md) | [journeys](../generated/diagrams/clinic/view-my-medicines-list.md) | | View my results | [help](help/clinic/view-my-results.md) | [admin](admin/clinic/view-my-results.md) | [API](api/clinic/view-my-results.md) | [journeys](../generated/diagrams/clinic/view-my-results.md) | | View results in scope | [help](help/clinic/view-results-in-scope.md) | [admin](admin/clinic/view-results-in-scope.md) | [API](api/clinic/view-results-in-scope.md) | [journeys](../generated/diagrams/clinic/view-results-in-scope.md) | | Wearable and home device signals | [help](help/clinic/wearable-signals.md) | [admin](admin/clinic/wearable-signals.md) | [API](api/clinic/wearable-signals.md) | [journeys](../generated/diagrams/clinic/wearable-signals.md) | | Wellbeing questionnaires | [help](help/clinic/wellbeing-questionnaires.md) | [admin](admin/clinic/wellbeing-questionnaires.md) | [API](api/clinic/wellbeing-questionnaires.md) | [journeys](../generated/diagrams/clinic/wellbeing-questionnaires.md) | | Write a care plan | [help](help/clinic/write-a-care-plan.md) | [admin](admin/clinic/write-a-care-plan.md) | [API](api/clinic/write-a-care-plan.md) | [journeys](../generated/diagrams/clinic/write-a-care-plan.md) | | Accept a review assignment | [help](help/edge-consortium/accept-a-review-assignment.md) | [admin](admin/edge-consortium/accept-a-review-assignment.md) | [API](api/edge-consortium/accept-a-review-assignment.md) | [journeys](../generated/diagrams/edge-consortium/accept-a-review-assignment.md) | | Accept supervision of a learner | [help](help/edge-consortium/accept-supervision-of-a-learner.md) | [admin](admin/edge-consortium/accept-supervision-of-a-learner.md) | [API](api/edge-consortium/accept-supervision-of-a-learner.md) | [journeys](../generated/diagrams/edge-consortium/accept-supervision-of-a-learner.md) | | Acknowledge the principles version | [help](help/edge-consortium/acknowledge-the-principles-version.md) | [admin](admin/edge-consortium/acknowledge-the-principles-version.md) | [API](api/edge-consortium/acknowledge-the-principles-version.md) | [journeys](../generated/diagrams/edge-consortium/acknowledge-the-principles-version.md) | | Add Jude to a channel under mandate | [help](help/edge-consortium/add-jude-to-a-channel-under-mandate.md) | [admin](admin/edge-consortium/add-jude-to-a-channel-under-mandate.md) | [API](api/edge-consortium/add-jude-to-a-channel-under-mandate.md) | [journeys](../generated/diagrams/edge-consortium/add-jude-to-a-channel-under-mandate.md) | | Add members and guests with expiry | [help](help/edge-consortium/add-members-and-guests-with-expiry.md) | [admin](admin/edge-consortium/add-members-and-guests-with-expiry.md) | [API](api/edge-consortium/add-members-and-guests-with-expiry.md) | [journeys](../generated/diagrams/edge-consortium/add-members-and-guests-with-expiry.md) | | Administer a database | [help](help/edge-consortium/administer-a-database.md) | [admin](admin/edge-consortium/administer-a-database.md) | [API](api/edge-consortium/administer-a-database.md) | [journeys](../generated/diagrams/edge-consortium/administer-a-database.md) | | Admit or refuse a guest | [help](help/edge-consortium/admit-or-refuse-a-guest.md) | [admin](admin/edge-consortium/admit-or-refuse-a-guest.md) | [API](api/edge-consortium/admit-or-refuse-a-guest.md) | [journeys](../generated/diagrams/edge-consortium/admit-or-refuse-a-guest.md) | | Agree a data sharing agreement | [help](help/edge-consortium/agree-a-data-sharing-agreement.md) | [admin](admin/edge-consortium/agree-a-data-sharing-agreement.md) | [API](api/edge-consortium/agree-a-data-sharing-agreement.md) | [journeys](../generated/diagrams/edge-consortium/agree-a-data-sharing-agreement.md) | | Allocate indirect costs | [help](help/edge-consortium/allocate-indirect-costs.md) | [admin](admin/edge-consortium/allocate-indirect-costs.md) | [API](api/edge-consortium/allocate-indirect-costs.md) | [journeys](../generated/diagrams/edge-consortium/allocate-indirect-costs.md) | | Answer a reference request | [help](help/edge-consortium/answer-a-reference-request.md) | [admin](admin/edge-consortium/answer-a-reference-request.md) | [API](api/edge-consortium/answer-a-reference-request.md) | [journeys](../generated/diagrams/edge-consortium/answer-a-reference-request.md) | | Answer reviewer questions on a contribution | [help](help/edge-consortium/answer-reviewer-questions-on-a-contribution.md) | [admin](admin/edge-consortium/answer-reviewer-questions-on-a-contribution.md) | [API](api/edge-consortium/answer-reviewer-questions-on-a-contribution.md) | [journeys](../generated/diagrams/edge-consortium/answer-reviewer-questions-on-a-contribution.md) | | Answer with sources in a channel | [help](help/edge-consortium/answer-with-sources-in-a-channel.md) | [admin](admin/edge-consortium/answer-with-sources-in-a-channel.md) | [API](api/edge-consortium/answer-with-sources-in-a-channel.md) | [journeys](../generated/diagrams/edge-consortium/answer-with-sources-in-a-channel.md) | | Apply a non-disclosure agreement to a room | [help](help/edge-consortium/apply-a-non-disclosure-agreement-to-a-room.md) | [admin](admin/edge-consortium/apply-a-non-disclosure-agreement-to-a-room.md) | [API](api/edge-consortium/apply-a-non-disclosure-agreement-to-a-room.md) | [journeys](../generated/diagrams/edge-consortium/apply-a-non-disclosure-agreement-to-a-room.md) | | Apply an open licence | [help](help/edge-consortium/apply-an-open-licence.md) | [admin](admin/edge-consortium/apply-an-open-licence.md) | [API](api/edge-consortium/apply-an-open-licence.md) | [journeys](../generated/diagrams/edge-consortium/apply-an-open-licence.md) | | Apply for ethics approval | [help](help/edge-consortium/apply-for-ethics-approval.md) | [admin](admin/edge-consortium/apply-for-ethics-approval.md) | [API](api/edge-consortium/apply-for-ethics-approval.md) | [journeys](../generated/diagrams/edge-consortium/apply-for-ethics-approval.md) | | Apply retention and deletion to records | [help](help/edge-consortium/apply-retention-and-deletion-to-records.md) | [admin](admin/edge-consortium/apply-retention-and-deletion-to-records.md) | [API](api/edge-consortium/apply-retention-and-deletion-to-records.md) | [journeys](../generated/diagrams/edge-consortium/apply-retention-and-deletion-to-records.md) | | Apply to become a verified reviewer | [help](help/edge-consortium/apply-to-become-a-verified-reviewer.md) | [admin](admin/edge-consortium/apply-to-become-a-verified-reviewer.md) | [API](api/edge-consortium/apply-to-become-a-verified-reviewer.md) | [journeys](../generated/diagrams/edge-consortium/apply-to-become-a-verified-reviewer.md) | | Approve a compute request | [help](help/edge-consortium/approve-a-compute-request.md) | [admin](admin/edge-consortium/approve-a-compute-request.md) | [API](api/edge-consortium/approve-a-compute-request.md) | [journeys](../generated/diagrams/edge-consortium/approve-a-compute-request.md) | | Approve a connector or integration | [help](help/edge-consortium/approve-a-connector-or-integration.md) | [admin](admin/edge-consortium/approve-a-connector-or-integration.md) | [API](api/edge-consortium/approve-a-connector-or-integration.md) | [journeys](../generated/diagrams/edge-consortium/approve-a-connector-or-integration.md) | | Approve a contribution before submission | [help](help/edge-consortium/approve-a-contribution-before-submission.md) | [admin](admin/edge-consortium/approve-a-contribution-before-submission.md) | [API](api/edge-consortium/approve-a-contribution-before-submission.md) | [journeys](../generated/diagrams/edge-consortium/approve-a-contribution-before-submission.md) | | Approve a director passkey action | [help](help/edge-consortium/approve-a-director-passkey-action.md) | [admin](admin/edge-consortium/approve-a-director-passkey-action.md) | [API](api/edge-consortium/approve-a-director-passkey-action.md) | [journeys](../generated/diagrams/edge-consortium/approve-a-director-passkey-action.md) | | Approve a federated query request | [help](help/edge-consortium/approve-a-federated-query-request.md) | [admin](admin/edge-consortium/approve-a-federated-query-request.md) | [API](api/edge-consortium/approve-a-federated-query-request.md) | [journeys](../generated/diagrams/edge-consortium/approve-a-federated-query-request.md) | | Approve a guest account | [help](help/edge-consortium/approve-a-guest-account.md) | [admin](admin/edge-consortium/approve-a-guest-account.md) | [API](api/edge-consortium/approve-a-guest-account.md) | [journeys](../generated/diagrams/edge-consortium/approve-a-guest-account.md) | | Approve a reviewer payment batch | [help](help/edge-consortium/approve-a-reviewer-payment-batch.md) | [admin](admin/edge-consortium/approve-a-reviewer-payment-batch.md) | [API](api/edge-consortium/approve-a-reviewer-payment-batch.md) | [journeys](../generated/diagrams/edge-consortium/approve-a-reviewer-payment-batch.md) | | Approve a team member data access request | [help](help/edge-consortium/approve-a-team-member-data-access-request.md) | [admin](admin/edge-consortium/approve-a-team-member-data-access-request.md) | [API](api/edge-consortium/approve-a-team-member-data-access-request.md) | [journeys](../generated/diagrams/edge-consortium/approve-a-team-member-data-access-request.md) | | Approve a visiting researcher affiliation | [help](help/edge-consortium/approve-a-visiting-researcher-affiliation.md) | [admin](admin/edge-consortium/approve-a-visiting-researcher-affiliation.md) | [API](api/edge-consortium/approve-a-visiting-researcher-affiliation.md) | [journeys](../generated/diagrams/edge-consortium/approve-a-visiting-researcher-affiliation.md) | | Approve an agent consequential action | [help](help/edge-consortium/approve-an-agent-consequential-action.md) | [admin](admin/edge-consortium/approve-an-agent-consequential-action.md) | [API](api/edge-consortium/approve-an-agent-consequential-action.md) | [journeys](../generated/diagrams/edge-consortium/approve-an-agent-consequential-action.md) | | Approve an open-source release | [help](help/edge-consortium/approve-an-open-source-release.md) | [admin](admin/edge-consortium/approve-an-open-source-release.md) | [API](api/edge-consortium/approve-an-open-source-release.md) | [journeys](../generated/diagrams/edge-consortium/approve-an-open-source-release.md) | | Approve and mark a contribution verified | [help](help/edge-consortium/approve-and-mark-a-contribution-verified.md) | [admin](admin/edge-consortium/approve-and-mark-a-contribution-verified.md) | [API](api/edge-consortium/approve-and-mark-a-contribution-verified.md) | [journeys](../generated/diagrams/edge-consortium/approve-and-mark-a-contribution-verified.md) | | Approve compute spend against a grant | [help](help/edge-consortium/approve-compute-spend-against-a-grant.md) | [admin](admin/edge-consortium/approve-compute-spend-against-a-grant.md) | [API](api/edge-consortium/approve-compute-spend-against-a-grant.md) | [journeys](../generated/diagrams/edge-consortium/approve-compute-spend-against-a-grant.md) | | Approve expenditure against a grant | [help](help/edge-consortium/approve-expenditure-against-a-grant.md) | [admin](admin/edge-consortium/approve-expenditure-against-a-grant.md) | [API](api/edge-consortium/approve-expenditure-against-a-grant.md) | [journeys](../generated/diagrams/edge-consortium/approve-expenditure-against-a-grant.md) | | Archive a channel | [help](help/edge-consortium/archive-a-channel.md) | [admin](admin/edge-consortium/archive-a-channel.md) | [API](api/edge-consortium/archive-a-channel.md) | [journeys](../generated/diagrams/edge-consortium/archive-a-channel.md) | | Archive a module version | [help](help/edge-consortium/archive-a-module-version.md) | [admin](admin/edge-consortium/archive-a-module-version.md) | [API](api/edge-consortium/archive-a-module-version.md) | [journeys](../generated/diagrams/edge-consortium/archive-a-module-version.md) | | Ask a clarification question | [help](help/edge-consortium/ask-a-clarification-question.md) | [admin](admin/edge-consortium/ask-a-clarification-question.md) | [API](api/edge-consortium/ask-a-clarification-question.md) | [journeys](../generated/diagrams/edge-consortium/ask-a-clarification-question.md) | | Ask Jude for a sourced answer | [help](help/edge-consortium/ask-jude-for-a-sourced-answer.md) | [admin](admin/edge-consortium/ask-jude-for-a-sourced-answer.md) | [API](api/edge-consortium/ask-jude-for-a-sourced-answer.md) | [journeys](../generated/diagrams/edge-consortium/ask-jude-for-a-sourced-answer.md) | | Assess ownership of an output | [help](help/edge-consortium/assess-ownership-of-an-output.md) | [admin](admin/edge-consortium/assess-ownership-of-an-output.md) | [API](api/edge-consortium/assess-ownership-of-an-output.md) | [journeys](../generated/diagrams/edge-consortium/assess-ownership-of-an-output.md) | | Assign a persistent identifier | [help](help/edge-consortium/assign-a-persistent-identifier.md) | [admin](admin/edge-consortium/assign-a-persistent-identifier.md) | [API](api/edge-consortium/assign-a-persistent-identifier.md) | [journeys](../generated/diagrams/edge-consortium/assign-a-persistent-identifier.md) | | Attribute carbon to a project | [help](help/edge-consortium/attribute-carbon-to-a-project.md) | [admin](admin/edge-consortium/attribute-carbon-to-a-project.md) | [API](api/edge-consortium/attribute-carbon-to-a-project.md) | [journeys](../generated/diagrams/edge-consortium/attribute-carbon-to-a-project.md) | | Attribute cost to a grant and project | [help](help/edge-consortium/attribute-cost-to-a-grant-and-project.md) | [admin](admin/edge-consortium/attribute-cost-to-a-grant-and-project.md) | [API](api/edge-consortium/attribute-cost-to-a-grant-and-project.md) | [journeys](../generated/diagrams/edge-consortium/attribute-cost-to-a-grant-and-project.md) | | Audit reviewer quality by re-review | [help](help/edge-consortium/audit-reviewer-quality-by-re-review.md) | [admin](admin/edge-consortium/audit-reviewer-quality-by-re-review.md) | [API](api/edge-consortium/audit-reviewer-quality-by-re-review.md) | [journeys](../generated/diagrams/edge-consortium/audit-reviewer-quality-by-re-review.md) | | Audit the signing log | [help](help/edge-consortium/audit-the-signing-log.md) | [admin](admin/edge-consortium/audit-the-signing-log.md) | [API](api/edge-consortium/audit-the-signing-log.md) | [journeys](../generated/diagrams/edge-consortium/audit-the-signing-log.md) | | Award and publish the contract award notice | [help](help/edge-consortium/award-and-publish-the-contract-award-notice.md) | [admin](admin/edge-consortium/award-and-publish-the-contract-award-notice.md) | [API](api/edge-consortium/award-and-publish-the-contract-award-notice.md) | [journeys](../generated/diagrams/edge-consortium/award-and-publish-the-contract-award-notice.md) | | Bind a device | [help](help/edge-consortium/bind-a-device.md) | [admin](admin/edge-consortium/bind-a-device.md) | [API](api/edge-consortium/bind-a-device.md) | [journeys](../generated/diagrams/edge-consortium/bind-a-device.md) | | Book a room or resource | [help](help/edge-consortium/book-a-room-or-resource.md) | [admin](admin/edge-consortium/book-a-room-or-resource.md) | [API](api/edge-consortium/book-a-room-or-resource.md) | [journeys](../generated/diagrams/edge-consortium/book-a-room-or-resource.md) | | Build a project team across universities | [help](help/edge-consortium/build-a-project-team-across-universities.md) | [admin](admin/edge-consortium/build-a-project-team-across-universities.md) | [API](api/edge-consortium/build-a-project-team-across-universities.md) | [journeys](../generated/diagrams/edge-consortium/build-a-project-team-across-universities.md) | | Build an interactive prototype | [help](help/edge-consortium/build-an-interactive-prototype.md) | [admin](admin/edge-consortium/build-an-interactive-prototype.md) | [API](api/edge-consortium/build-an-interactive-prototype.md) | [journeys](../generated/diagrams/edge-consortium/build-an-interactive-prototype.md) | | Build the research profile and skills record | [help](help/edge-consortium/build-the-research-profile-and-skills-record.md) | [admin](admin/edge-consortium/build-the-research-profile-and-skills-record.md) | [API](api/edge-consortium/build-the-research-profile-and-skills-record.md) | [journeys](../generated/diagrams/edge-consortium/build-the-research-profile-and-skills-record.md) | | Bulk enrol students and staff | [help](help/edge-consortium/bulk-enrol-students-and-staff.md) | [admin](admin/edge-consortium/bulk-enrol-students-and-staff.md) | [API](api/edge-consortium/bulk-enrol-students-and-staff.md) | [journeys](../generated/diagrams/edge-consortium/bulk-enrol-students-and-staff.md) | | Catalogue a library item | [help](help/edge-consortium/catalogue-a-library-item.md) | [admin](admin/edge-consortium/catalogue-a-library-item.md) | [API](api/edge-consortium/catalogue-a-library-item.md) | [journeys](../generated/diagrams/edge-consortium/catalogue-a-library-item.md) | | Charge compute and storage to grant codes | [help](help/edge-consortium/charge-compute-and-storage-to-grant-codes.md) | [admin](admin/edge-consortium/charge-compute-and-storage-to-grant-codes.md) | [API](api/edge-consortium/charge-compute-and-storage-to-grant-codes.md) | [journeys](../generated/diagrams/edge-consortium/charge-compute-and-storage-to-grant-codes.md) | | Chase an overdue review | [help](help/edge-consortium/chase-an-overdue-review.md) | [admin](admin/edge-consortium/chase-an-overdue-review.md) | [API](api/edge-consortium/chase-an-overdue-review.md) | [journeys](../generated/diagrams/edge-consortium/chase-an-overdue-review.md) | | Check exclusion grounds | [help](help/edge-consortium/check-exclusion-grounds.md) | [admin](admin/edge-consortium/check-exclusion-grounds.md) | [API](api/edge-consortium/check-exclusion-grounds.md) | [journeys](../generated/diagrams/edge-consortium/check-exclusion-grounds.md) | | Choose a specialism after the whole map | [help](help/edge-consortium/choose-a-specialism.md) | [admin](admin/edge-consortium/choose-a-specialism.md) | [API](api/edge-consortium/choose-a-specialism.md) | [journeys](../generated/diagrams/edge-consortium/choose-a-specialism.md) | | Claim payment for a completed review | [help](help/edge-consortium/claim-payment-for-a-completed-review.md) | [admin](admin/edge-consortium/claim-payment-for-a-completed-review.md) | [API](api/edge-consortium/claim-payment-for-a-completed-review.md) | [journeys](../generated/diagrams/edge-consortium/claim-payment-for-a-completed-review.md) | | Classify data sensitivity | [help](help/edge-consortium/classify-data-sensitivity.md) | [admin](admin/edge-consortium/classify-data-sensitivity.md) | [API](api/edge-consortium/classify-data-sensitivity.md) | [journeys](../generated/diagrams/edge-consortium/classify-data-sensitivity.md) | | Close a grant | [help](help/edge-consortium/close-a-grant.md) | [admin](admin/edge-consortium/close-a-grant.md) | [API](api/edge-consortium/close-a-grant.md) | [journeys](../generated/diagrams/edge-consortium/close-a-grant.md) | | Close a project and archive its records | [help](help/edge-consortium/close-a-project-and-archive-its-records.md) | [admin](admin/edge-consortium/close-a-project-and-archive-its-records.md) | [API](api/edge-consortium/close-a-project-and-archive-its-records.md) | [journeys](../generated/diagrams/edge-consortium/close-a-project-and-archive-its-records.md) | | Co-edit a Markdown in real time | [help](help/edge-consortium/co-edit-a-markdown-in-real-time.md) | [admin](admin/edge-consortium/co-edit-a-markdown-in-real-time.md) | [API](api/edge-consortium/co-edit-a-markdown-in-real-time.md) | [journeys](../generated/diagrams/edge-consortium/co-edit-a-markdown-in-real-time.md) | | Complete a right-to-work check | [help](help/edge-consortium/complete-a-right-to-work-check.md) | [admin](admin/edge-consortium/complete-a-right-to-work-check.md) | [API](api/edge-consortium/complete-a-right-to-work-check.md) | [journeys](../generated/diagrams/edge-consortium/complete-a-right-to-work-check.md) | | Complete research integrity training | [help](help/edge-consortium/complete-research-integrity-training.md) | [admin](admin/edge-consortium/complete-research-integrity-training.md) | [API](api/edge-consortium/complete-research-integrity-training.md) | [journeys](../generated/diagrams/edge-consortium/complete-research-integrity-training.md) | | Compose an image in the imaging studio | [help](help/edge-consortium/compose-an-image-in-the-imaging-studio.md) | [admin](admin/edge-consortium/compose-an-image-in-the-imaging-studio.md) | [API](api/edge-consortium/compose-an-image-in-the-imaging-studio.md) | [journeys](../generated/diagrams/edge-consortium/compose-an-image-in-the-imaging-studio.md) | | Compose audio in Judio | [help](help/edge-consortium/compose-audio-in-judio.md) | [admin](admin/edge-consortium/compose-audio-in-judio.md) | [API](api/edge-consortium/compose-audio-in-judio.md) | [journeys](../generated/diagrams/edge-consortium/compose-audio-in-judio.md) | | Conduct a probation review | [help](help/edge-consortium/conduct-a-probation-review.md) | [admin](admin/edge-consortium/conduct-a-probation-review.md) | [API](api/edge-consortium/conduct-a-probation-review.md) | [journeys](../generated/diagrams/edge-consortium/conduct-a-probation-review.md) | | Confirm a delivery | [help](help/edge-consortium/confirm-a-delivery.md) | [admin](admin/edge-consortium/confirm-a-delivery.md) | [API](api/edge-consortium/confirm-a-delivery.md) | [journeys](../generated/diagrams/edge-consortium/confirm-a-delivery.md) | | Confirm a person's enrolment status | [help](help/edge-consortium/confirm-a-person-s-enrolment-status.md) | [admin](admin/edge-consortium/confirm-a-person-s-enrolment-status.md) | [API](api/edge-consortium/confirm-a-person-s-enrolment-status.md) | [journeys](../generated/diagrams/edge-consortium/confirm-a-person-s-enrolment-status.md) | | Confirm a time and send the invitation | [help](help/edge-consortium/confirm-a-time-and-send-the-invitation.md) | [admin](admin/edge-consortium/confirm-a-time-and-send-the-invitation.md) | [API](api/edge-consortium/confirm-a-time-and-send-the-invitation.md) | [journeys](../generated/diagrams/edge-consortium/confirm-a-time-and-send-the-invitation.md) | | Contribute a library entry as an outsider | [help](help/edge-consortium/contribute-a-library-entry-as-an-outsider.md) | [admin](admin/edge-consortium/contribute-a-library-entry-as-an-outsider.md) | [API](api/edge-consortium/contribute-a-library-entry-as-an-outsider.md) | [journeys](../generated/diagrams/edge-consortium/contribute-a-library-entry-as-an-outsider.md) | | Correct a student record | [help](help/edge-consortium/correct-a-student-record.md) | [admin](admin/edge-consortium/correct-a-student-record.md) | [API](api/edge-consortium/correct-a-student-record.md) | [journeys](../generated/diagrams/edge-consortium/correct-a-student-record.md) | | Cost a grant application | [help](help/edge-consortium/cost-a-grant-application.md) | [admin](admin/edge-consortium/cost-a-grant-application.md) | [API](api/edge-consortium/cost-a-grant-application.md) | [journeys](../generated/diagrams/edge-consortium/cost-a-grant-application.md) | | Count agents and subagents per task | [help](help/edge-consortium/count-agents-and-subagents-per-task.md) | [admin](admin/edge-consortium/count-agents-and-subagents-per-task.md) | [API](api/edge-consortium/count-agents-and-subagents-per-task.md) | [journeys](../generated/diagrams/edge-consortium/count-agents-and-subagents-per-task.md) | | Create a contract from a template | [help](help/edge-consortium/create-a-contract-from-a-template.md) | [admin](admin/edge-consortium/create-a-contract-from-a-template.md) | [API](api/edge-consortium/create-a-contract-from-a-template.md) | [journeys](../generated/diagrams/edge-consortium/create-a-contract-from-a-template.md) | | Create a job role and vacancy | [help](help/edge-consortium/create-a-job-role-and-vacancy.md) | [admin](admin/edge-consortium/create-a-job-role-and-vacancy.md) | [API](api/edge-consortium/create-a-job-role-and-vacancy.md) | [journeys](../generated/diagrams/edge-consortium/create-a-job-role-and-vacancy.md) | | Create a Markdown document | [help](help/edge-consortium/create-a-markdown-document.md) | [admin](admin/edge-consortium/create-a-markdown-document.md) | [API](api/edge-consortium/create-a-markdown-document.md) | [journeys](../generated/diagrams/edge-consortium/create-a-markdown-document.md) | | Create a module | [help](help/edge-consortium/create-a-module.md) | [admin](admin/edge-consortium/create-a-module.md) | [API](api/edge-consortium/create-a-module.md) | [journeys](../generated/diagrams/edge-consortium/create-a-module.md) | | Create a team and its channels | [help](help/edge-consortium/create-a-team-and-its-channels.md) | [admin](admin/edge-consortium/create-a-team-and-its-channels.md) | [API](api/edge-consortium/create-a-team-and-its-channels.md) | [journeys](../generated/diagrams/edge-consortium/create-a-team-and-its-channels.md) | | Create an availability poll | [help](help/edge-consortium/create-an-availability-poll.md) | [admin](admin/edge-consortium/create-an-availability-poll.md) | [API](api/edge-consortium/create-an-availability-poll.md) | [journeys](../generated/diagrams/edge-consortium/create-an-availability-poll.md) | | Curate a collection | [help](help/edge-consortium/curate-a-collection.md) | [admin](admin/edge-consortium/curate-a-collection.md) | [API](api/edge-consortium/curate-a-collection.md) | [journeys](../generated/diagrams/edge-consortium/curate-a-collection.md) | | Declare a conflict of interest | [help](help/edge-consortium/declare-a-conflict-of-interest.md) | [admin](admin/edge-consortium/declare-a-conflict-of-interest.md) | [API](api/edge-consortium/declare-a-conflict-of-interest.md) | [journeys](../generated/diagrams/edge-consortium/declare-a-conflict-of-interest.md) | | Declare exclusion grounds and conflicts | [help](help/edge-consortium/declare-exclusion-grounds-and-conflicts.md) | [admin](admin/edge-consortium/declare-exclusion-grounds-and-conflicts.md) | [API](api/edge-consortium/declare-exclusion-grounds-and-conflicts.md) | [journeys](../generated/diagrams/edge-consortium/declare-exclusion-grounds-and-conflicts.md) | | Decline an assignment | [help](help/edge-consortium/decline-an-assignment.md) | [admin](admin/edge-consortium/decline-an-assignment.md) | [API](api/edge-consortium/decline-an-assignment.md) | [journeys](../generated/diagrams/edge-consortium/decline-an-assignment.md) | | Define a purpose view | [help](help/edge-consortium/define-a-purpose-view.md) | [admin](admin/edge-consortium/define-a-purpose-view.md) | [API](api/edge-consortium/define-a-purpose-view.md) | [journeys](../generated/diagrams/edge-consortium/define-a-purpose-view.md) | | Define or change a schema | [help](help/edge-consortium/define-or-change-a-schema.md) | [admin](admin/edge-consortium/define-or-change-a-schema.md) | [API](api/edge-consortium/define-or-change-a-schema.md) | [journeys](../generated/diagrams/edge-consortium/define-or-change-a-schema.md) | | Delete data at retention end | [help](help/edge-consortium/delete-data-at-retention-end.md) | [admin](admin/edge-consortium/delete-data-at-retention-end.md) | [API](api/edge-consortium/delete-data-at-retention-end.md) | [journeys](../generated/diagrams/edge-consortium/delete-data-at-retention-end.md) | | Deploy a release | [help](help/edge-consortium/deploy-a-release.md) | [admin](admin/edge-consortium/deploy-a-release.md) | [API](api/edge-consortium/deploy-a-release.md) | [journeys](../generated/diagrams/edge-consortium/deploy-a-release.md) | | Deposit a dataset with a persistent identifier | [help](help/edge-consortium/deposit-a-dataset-with-a-persistent-identifier.md) | [admin](admin/edge-consortium/deposit-a-dataset-with-a-persistent-identifier.md) | [API](api/edge-consortium/deposit-a-dataset-with-a-persistent-identifier.md) | [journeys](../generated/diagrams/edge-consortium/deposit-a-dataset-with-a-persistent-identifier.md) | | Deprecate a feature | [help](help/edge-consortium/deprecate-a-feature.md) | [admin](admin/edge-consortium/deprecate-a-feature.md) | [API](api/edge-consortium/deprecate-a-feature.md) | [journeys](../generated/diagrams/edge-consortium/deprecate-a-feature.md) | | Detect unlicensed reuse | [help](help/edge-consortium/detect-unlicensed-reuse.md) | [admin](admin/edge-consortium/detect-unlicensed-reuse.md) | [API](api/edge-consortium/detect-unlicensed-reuse.md) | [journeys](../generated/diagrams/edge-consortium/detect-unlicensed-reuse.md) | | Draw on JudeCanvas | [help](help/edge-consortium/draw-on-judecanvas.md) | [admin](admin/edge-consortium/draw-on-judecanvas.md) | [API](api/edge-consortium/draw-on-judecanvas.md) | [journeys](../generated/diagrams/edge-consortium/draw-on-judecanvas.md) | | Edit a video | [help](help/edge-consortium/edit-a-video.md) | [admin](admin/edge-consortium/edit-a-video.md) | [API](api/edge-consortium/edit-a-video.md) | [journeys](../generated/diagrams/edge-consortium/edit-a-video.md) | | End employment and revoke access | [help](help/edge-consortium/end-employment-and-revoke-access.md) | [admin](admin/edge-consortium/end-employment-and-revoke-access.md) | [API](api/edge-consortium/end-employment-and-revoke-access.md) | [journeys](../generated/diagrams/edge-consortium/end-employment-and-revoke-access.md) | | Enforce the code of conduct | [help](help/edge-consortium/enforce-the-code-of-conduct.md) | [admin](admin/edge-consortium/enforce-the-code-of-conduct.md) | [API](api/edge-consortium/enforce-the-code-of-conduct.md) | [journeys](../generated/diagrams/edge-consortium/enforce-the-code-of-conduct.md) | | Enrol on a doctoral training module | [help](help/edge-consortium/enrol-on-a-doctoral-training-module.md) | [admin](admin/edge-consortium/enrol-on-a-doctoral-training-module.md) | [API](api/edge-consortium/enrol-on-a-doctoral-training-module.md) | [journeys](../generated/diagrams/edge-consortium/enrol-on-a-doctoral-training-module.md) | | Escalate a dispute to the editor | [help](help/edge-consortium/escalate-a-dispute-to-the-editor.md) | [admin](admin/edge-consortium/escalate-a-dispute-to-the-editor.md) | [API](api/edge-consortium/escalate-a-dispute-to-the-editor.md) | [journeys](../generated/diagrams/edge-consortium/escalate-a-dispute-to-the-editor.md) | | Evaluate bids with a scoring panel | [help](help/edge-consortium/evaluate-bids-with-a-scoring-panel.md) | [admin](admin/edge-consortium/evaluate-bids-with-a-scoring-panel.md) | [API](api/edge-consortium/evaluate-bids-with-a-scoring-panel.md) | [journeys](../generated/diagrams/edge-consortium/evaluate-bids-with-a-scoring-panel.md) | | Expire or revoke a licence | [help](help/edge-consortium/expire-or-revoke-a-licence.md) | [admin](admin/edge-consortium/expire-or-revoke-a-licence.md) | [API](api/edge-consortium/expire-or-revoke-a-licence.md) | [journeys](../generated/diagrams/edge-consortium/expire-or-revoke-a-licence.md) | | Explore the knowledge graph visually | [help](help/edge-consortium/explore-the-knowledge-graph-visually.md) | [admin](admin/edge-consortium/explore-the-knowledge-graph-visually.md) | [API](api/edge-consortium/explore-the-knowledge-graph-visually.md) | [journeys](../generated/diagrams/edge-consortium/explore-the-knowledge-graph-visually.md) | | Export a creation with licence and attribution | [help](help/edge-consortium/export-a-creation-with-licence-and-attribution.md) | [admin](admin/edge-consortium/export-a-creation-with-licence-and-attribution.md) | [API](api/edge-consortium/export-a-creation-with-licence-and-attribution.md) | [journeys](../generated/diagrams/edge-consortium/export-a-creation-with-licence-and-attribution.md) | | Export a meeting record and audit chain | [help](help/edge-consortium/export-a-meeting-record-and-audit-chain.md) | [admin](admin/edge-consortium/export-a-meeting-record-and-audit-chain.md) | [API](api/edge-consortium/export-a-meeting-record-and-audit-chain.md) | [journeys](../generated/diagrams/edge-consortium/export-a-meeting-record-and-audit-chain.md) | | Export and verify an audit chain | [help](help/edge-consortium/export-and-verify-an-audit-chain.md) | [admin](admin/edge-consortium/export-and-verify-an-audit-chain.md) | [API](api/edge-consortium/export-and-verify-an-audit-chain.md) | [journeys](../generated/diagrams/edge-consortium/export-and-verify-an-audit-chain.md) | | Find a time across attendees | [help](help/edge-consortium/find-a-time-across-attendees.md) | [admin](admin/edge-consortium/find-a-time-across-attendees.md) | [API](api/edge-consortium/find-a-time-across-attendees.md) | [journeys](../generated/diagrams/edge-consortium/find-a-time-across-attendees.md) | | Find and follow tender opportunities | [help](help/edge-consortium/find-and-follow-tender-opportunities.md) | [admin](admin/edge-consortium/find-and-follow-tender-opportunities.md) | [API](api/edge-consortium/find-and-follow-tender-opportunities.md) | [journeys](../generated/diagrams/edge-consortium/find-and-follow-tender-opportunities.md) | | Find funding calls | [help](help/edge-consortium/find-funding-calls.md) | [admin](admin/edge-consortium/find-funding-calls.md) | [API](api/edge-consortium/find-funding-calls.md) | [journeys](../generated/diagrams/edge-consortium/find-funding-calls.md) | | Flag a concern without blocking | [help](help/edge-consortium/flag-a-concern-without-blocking.md) | [admin](admin/edge-consortium/flag-a-concern-without-blocking.md) | [API](api/edge-consortium/flag-a-concern-without-blocking.md) | [journeys](../generated/diagrams/edge-consortium/flag-a-concern-without-blocking.md) | | Flag plagiarism or an integrity concern | [help](help/edge-consortium/flag-plagiarism-or-an-integrity-concern.md) | [admin](admin/edge-consortium/flag-plagiarism-or-an-integrity-concern.md) | [API](api/edge-consortium/flag-plagiarism-or-an-integrity-concern.md) | [journeys](../generated/diagrams/edge-consortium/flag-plagiarism-or-an-integrity-concern.md) | | Study the governed intelligence degree | [help](help/edge-consortium/governed-intelligence-degree.md) | [admin](admin/edge-consortium/governed-intelligence-degree.md) | [API](api/edge-consortium/governed-intelligence-degree.md) | [journeys](../generated/diagrams/edge-consortium/governed-intelligence-degree.md) | | Grant a clearance with expiry | [help](help/edge-consortium/grant-a-clearance-with-expiry.md) | [admin](admin/edge-consortium/grant-a-clearance-with-expiry.md) | [API](api/edge-consortium/grant-a-clearance-with-expiry.md) | [journeys](../generated/diagrams/edge-consortium/grant-a-clearance-with-expiry.md) | | Grant a licence to a third party | [help](help/edge-consortium/grant-a-licence-to-a-third-party.md) | [admin](admin/edge-consortium/grant-a-licence-to-a-third-party.md) | [API](api/edge-consortium/grant-a-licence-to-a-third-party.md) | [journeys](../generated/diagrams/edge-consortium/grant-a-licence-to-a-third-party.md) | | Grant a maintainer role with expiry | [help](help/edge-consortium/grant-a-maintainer-role-with-expiry.md) | [admin](admin/edge-consortium/grant-a-maintainer-role-with-expiry.md) | [API](api/edge-consortium/grant-a-maintainer-role-with-expiry.md) | [journeys](../generated/diagrams/edge-consortium/grant-a-maintainer-role-with-expiry.md) | | Hand over to another agent | [help](help/edge-consortium/hand-over-to-another-agent.md) | [admin](admin/edge-consortium/hand-over-to-another-agent.md) | [API](api/edge-consortium/hand-over-to-another-agent.md) | [journeys](../generated/diagrams/edge-consortium/hand-over-to-another-agent.md) | | Hand over work with the Baton | [help](help/edge-consortium/hand-over-work-with-the-baton.md) | [admin](admin/edge-consortium/hand-over-work-with-the-baton.md) | [API](api/edge-consortium/hand-over-work-with-the-baton.md) | [journeys](../generated/diagrams/edge-consortium/hand-over-work-with-the-baton.md) | | Handle a contract dispute | [help](help/edge-consortium/handle-a-contract-dispute.md) | [admin](admin/edge-consortium/handle-a-contract-dispute.md) | [API](api/edge-consortium/handle-a-contract-dispute.md) | [journeys](../generated/diagrams/edge-consortium/handle-a-contract-dispute.md) | | Handle a data subject access request | [help](help/edge-consortium/handle-a-data-subject-access-request.md) | [admin](admin/edge-consortium/handle-a-data-subject-access-request.md) | [API](api/edge-consortium/handle-a-data-subject-access-request.md) | [journeys](../generated/diagrams/edge-consortium/handle-a-data-subject-access-request.md) | | Handle an academic appeal on marks | [help](help/edge-consortium/handle-an-academic-appeal-on-marks.md) | [admin](admin/edge-consortium/handle-an-academic-appeal-on-marks.md) | [API](api/edge-consortium/handle-an-academic-appeal-on-marks.md) | [journeys](../generated/diagrams/edge-consortium/handle-an-academic-appeal-on-marks.md) | | Handle an appeal | [help](help/edge-consortium/handle-an-appeal.md) | [admin](admin/edge-consortium/handle-an-appeal.md) | [API](api/edge-consortium/handle-an-appeal.md) | [journeys](../generated/diagrams/edge-consortium/handle-an-appeal.md) | | Hold an audit interview by secure call | [help](help/edge-consortium/hold-an-audit-interview-by-secure-call.md) | [admin](admin/edge-consortium/hold-an-audit-interview-by-secure-call.md) | [API](api/edge-consortium/hold-an-audit-interview-by-secure-call.md) | [journeys](../generated/diagrams/edge-consortium/hold-an-audit-interview-by-secure-call.md) | | Import external records | [help](help/edge-consortium/import-external-records.md) | [admin](admin/edge-consortium/import-external-records.md) | [API](api/edge-consortium/import-external-records.md) | [journeys](../generated/diagrams/edge-consortium/import-external-records.md) | | Inspect carbon and energy records | [help](help/edge-consortium/inspect-carbon-and-energy-records.md) | [admin](admin/edge-consortium/inspect-carbon-and-energy-records.md) | [API](api/edge-consortium/inspect-carbon-and-energy-records.md) | [journeys](../generated/diagrams/edge-consortium/inspect-carbon-and-energy-records.md) | | Inspect release gates and evidence | [help](help/edge-consortium/inspect-release-gates-and-evidence.md) | [admin](admin/edge-consortium/inspect-release-gates-and-evidence.md) | [API](api/edge-consortium/inspect-release-gates-and-evidence.md) | [journeys](../generated/diagrams/edge-consortium/inspect-release-gates-and-evidence.md) | | Inspect retention schedule compliance | [help](help/edge-consortium/inspect-retention-schedule-compliance.md) | [admin](admin/edge-consortium/inspect-retention-schedule-compliance.md) | [API](api/edge-consortium/inspect-retention-schedule-compliance.md) | [journeys](../generated/diagrams/edge-consortium/inspect-retention-schedule-compliance.md) | | Inspect reviewer payment records | [help](help/edge-consortium/inspect-reviewer-payment-records.md) | [admin](admin/edge-consortium/inspect-reviewer-payment-records.md) | [API](api/edge-consortium/inspect-reviewer-payment-records.md) | [journeys](../generated/diagrams/edge-consortium/inspect-reviewer-payment-records.md) | | International technical students | [help](help/edge-consortium/international-technical-students.md) | [admin](admin/edge-consortium/international-technical-students.md) | [API](api/edge-consortium/international-technical-students.md) | [journeys](../generated/diagrams/edge-consortium/international-technical-students.md) | | Issue a certificate of membership | [help](help/edge-consortium/issue-a-certificate-of-membership.md) | [admin](admin/edge-consortium/issue-a-certificate-of-membership.md) | [API](api/edge-consortium/issue-a-certificate-of-membership.md) | [journeys](../generated/diagrams/edge-consortium/issue-a-certificate-of-membership.md) | | Issue a takedown or attribution request | [help](help/edge-consortium/issue-a-takedown-or-attribution-request.md) | [admin](admin/edge-consortium/issue-a-takedown-or-attribution-request.md) | [API](api/edge-consortium/issue-a-takedown-or-attribution-request.md) | [journeys](../generated/diagrams/edge-consortium/issue-a-takedown-or-attribution-request.md) | | Issue a transcript of completed modules | [help](help/edge-consortium/issue-a-transcript-of-completed-modules.md) | [admin](admin/edge-consortium/issue-a-transcript-of-completed-modules.md) | [API](api/edge-consortium/issue-a-transcript-of-completed-modules.md) | [journeys](../generated/diagrams/edge-consortium/issue-a-transcript-of-completed-modules.md) | | Issue an employment contract | [help](help/edge-consortium/issue-an-employment-contract.md) | [admin](admin/edge-consortium/issue-an-employment-contract.md) | [API](api/edge-consortium/issue-an-employment-contract.md) | [journeys](../generated/diagrams/edge-consortium/issue-an-employment-contract.md) | | Issue an offer letter | [help](help/edge-consortium/issue-an-offer-letter.md) | [admin](admin/edge-consortium/issue-an-offer-letter.md) | [API](api/edge-consortium/issue-an-offer-letter.md) | [journeys](../generated/diagrams/edge-consortium/issue-an-offer-letter.md) | | Issue findings and track remediation | [help](help/edge-consortium/issue-findings-and-track-remediation.md) | [admin](admin/edge-consortium/issue-findings-and-track-remediation.md) | [API](api/edge-consortium/issue-findings-and-track-remediation.md) | [journeys](../generated/diagrams/edge-consortium/issue-findings-and-track-remediation.md) | | Issue the decision letter | [help](help/edge-consortium/issue-the-decision-letter.md) | [admin](admin/edge-consortium/issue-the-decision-letter.md) | [API](api/edge-consortium/issue-the-decision-letter.md) | [journeys](../generated/diagrams/edge-consortium/issue-the-decision-letter.md) | | Issue the written statement of particulars | [help](help/edge-consortium/issue-the-written-statement-of-particulars.md) | [admin](admin/edge-consortium/issue-the-written-statement-of-particulars.md) | [API](api/edge-consortium/issue-the-written-statement-of-particulars.md) | [journeys](../generated/diagrams/edge-consortium/issue-the-written-statement-of-particulars.md) | | Join a secure call through the lobby | [help](help/edge-consortium/join-a-secure-call-through-the-lobby.md) | [admin](admin/edge-consortium/join-a-secure-call-through-the-lobby.md) | [API](api/edge-consortium/join-a-secure-call-through-the-lobby.md) | [journeys](../generated/diagrams/edge-consortium/join-a-secure-call-through-the-lobby.md) | | Join Consortium through a university | [help](help/edge-consortium/join-consortium-through-a-university.md) | [admin](admin/edge-consortium/join-consortium-through-a-university.md) | [API](api/edge-consortium/join-consortium-through-a-university.md) | [journeys](../generated/diagrams/edge-consortium/join-consortium-through-a-university.md) | | Judiverse included in Edge Consortium tuition | [help](help/edge-consortium/judiverse-included-in-tuition.md) | [admin](admin/edge-consortium/judiverse-included-in-tuition.md) | [API](api/edge-consortium/judiverse-included-in-tuition.md) | [journeys](../generated/diagrams/edge-consortium/judiverse-included-in-tuition.md) | | Link ORCID to the profile | [help](help/edge-consortium/link-orcid-to-the-profile.md) | [admin](admin/edge-consortium/link-orcid-to-the-profile.md) | [API](api/edge-consortium/link-orcid-to-the-profile.md) | [journeys](../generated/diagrams/edge-consortium/link-orcid-to-the-profile.md) | | Log a receipt for every action | [help](help/edge-consortium/log-a-receipt-for-every-action.md) | [admin](admin/edge-consortium/log-a-receipt-for-every-action.md) | [API](api/edge-consortium/log-a-receipt-for-every-action.md) | [journeys](../generated/diagrams/edge-consortium/log-a-receipt-for-every-action.md) | | Make a below-threshold direct award | [help](help/edge-consortium/make-a-below-threshold-direct-award.md) | [admin](admin/edge-consortium/make-a-below-threshold-direct-award.md) | [API](api/edge-consortium/make-a-below-threshold-direct-award.md) | [journeys](../generated/diagrams/edge-consortium/make-a-below-threshold-direct-award.md) | | Manage a controlled vocabulary | [help](help/edge-consortium/manage-a-controlled-vocabulary.md) | [admin](admin/edge-consortium/manage-a-controlled-vocabulary.md) | [API](api/edge-consortium/manage-a-controlled-vocabulary.md) | [journeys](../generated/diagrams/edge-consortium/manage-a-controlled-vocabulary.md) | | Manage a grant extension or variation | [help](help/edge-consortium/manage-a-grant-extension-or-variation.md) | [admin](admin/edge-consortium/manage-a-grant-extension-or-variation.md) | [API](api/edge-consortium/manage-a-grant-extension-or-variation.md) | [journeys](../generated/diagrams/edge-consortium/manage-a-grant-extension-or-variation.md) | | Manage an embargo and release date | [help](help/edge-consortium/manage-an-embargo-and-release-date.md) | [admin](admin/edge-consortium/manage-an-embargo-and-release-date.md) | [API](api/edge-consortium/manage-an-embargo-and-release-date.md) | [journeys](../generated/diagrams/edge-consortium/manage-an-embargo-and-release-date.md) | | Manage consent records | [help](help/edge-consortium/manage-consent-records.md) | [admin](admin/edge-consortium/manage-consent-records.md) | [API](api/edge-consortium/manage-consent-records.md) | [journeys](../generated/diagrams/edge-consortium/manage-consent-records.md) | | Manage institutional roles and delegates | [help](help/edge-consortium/manage-institutional-roles-and-delegates.md) | [admin](admin/edge-consortium/manage-institutional-roles-and-delegates.md) | [API](api/edge-consortium/manage-institutional-roles-and-delegates.md) | [journeys](../generated/diagrams/edge-consortium/manage-institutional-roles-and-delegates.md) | | Map a dataset to the knowledge graph | [help](help/edge-consortium/map-a-dataset-to-the-knowledge-graph.md) | [admin](admin/edge-consortium/map-a-dataset-to-the-knowledge-graph.md) | [API](api/edge-consortium/map-a-dataset-to-the-knowledge-graph.md) | [journeys](../generated/diagrams/edge-consortium/map-a-dataset-to-the-knowledge-graph.md) | | Map a module to a challenge specification | [help](help/edge-consortium/map-a-module-to-a-challenge-specification.md) | [admin](admin/edge-consortium/map-a-module-to-a-challenge-specification.md) | [API](api/edge-consortium/map-a-module-to-a-challenge-specification.md) | [journeys](../generated/diagrams/edge-consortium/map-a-module-to-a-challenge-specification.md) | | Map a module to programme and standard | [help](help/edge-consortium/map-a-module-to-programme-and-standard.md) | [admin](admin/edge-consortium/map-a-module-to-programme-and-standard.md) | [API](api/edge-consortium/map-a-module-to-programme-and-standard.md) | [journeys](../generated/diagrams/edge-consortium/map-a-module-to-programme-and-standard.md) | | Mark and moderate submissions | [help](help/edge-consortium/mark-and-moderate-submissions.md) | [admin](admin/edge-consortium/mark-and-moderate-submissions.md) | [API](api/edge-consortium/mark-and-moderate-submissions.md) | [journeys](../generated/diagrams/edge-consortium/mark-and-moderate-submissions.md) | | Match reviewers to a submission | [help](help/edge-consortium/match-reviewers-to-a-submission.md) | [admin](admin/edge-consortium/match-reviewers-to-a-submission.md) | [API](api/edge-consortium/match-reviewers-to-a-submission.md) | [journeys](../generated/diagrams/edge-consortium/match-reviewers-to-a-submission.md) | | Measure live energy per job and service | [help](help/edge-consortium/measure-live-energy-per-job-and-service.md) | [admin](admin/edge-consortium/measure-live-energy-per-job-and-service.md) | [API](api/edge-consortium/measure-live-energy-per-job-and-service.md) | [journeys](../generated/diagrams/edge-consortium/measure-live-energy-per-job-and-service.md) | | Merge duplicate entities | [help](help/edge-consortium/merge-duplicate-entities.md) | [admin](admin/edge-consortium/merge-duplicate-entities.md) | [API](api/edge-consortium/merge-duplicate-entities.md) | [journeys](../generated/diagrams/edge-consortium/merge-duplicate-entities.md) | | Merge with required checks | [help](help/edge-consortium/merge-with-required-checks.md) | [admin](admin/edge-consortium/merge-with-required-checks.md) | [API](api/edge-consortium/merge-with-required-checks.md) | [journeys](../generated/diagrams/edge-consortium/merge-with-required-checks.md) | | Mirror a repository to the source repository | [help](help/edge-consortium/mirror-a-repository-to-the-source-repository.md) | [admin](admin/edge-consortium/mirror-a-repository-to-the-source-repository.md) | [API](api/edge-consortium/mirror-a-repository-to-the-source-repository.md) | [journeys](../generated/diagrams/edge-consortium/mirror-a-repository-to-the-source-repository.md) | | Moderate a post | [help](help/edge-consortium/moderate-a-post.md) | [admin](admin/edge-consortium/moderate-a-post.md) | [API](api/edge-consortium/moderate-a-post.md) | [journeys](../generated/diagrams/edge-consortium/moderate-a-post.md) | | Monitor data quality | [help](help/edge-consortium/monitor-data-quality.md) | [admin](admin/edge-consortium/monitor-data-quality.md) | [API](api/edge-consortium/monitor-data-quality.md) | [journeys](../generated/diagrams/edge-consortium/monitor-data-quality.md) | | Monitor job progress and output | [help](help/edge-consortium/monitor-job-progress-and-output.md) | [admin](admin/edge-consortium/monitor-job-progress-and-output.md) | [API](api/edge-consortium/monitor-job-progress-and-output.md) | [journeys](../generated/diagrams/edge-consortium/monitor-job-progress-and-output.md) | | Monitor tokens, prompts and time to evidence | [help](help/edge-consortium/monitor-tokens-prompts-and-time-to-evidence.md) | [admin](admin/edge-consortium/monitor-tokens-prompts-and-time-to-evidence.md) | [API](api/edge-consortium/monitor-tokens-prompts-and-time-to-evidence.md) | [journeys](../generated/diagrams/edge-consortium/monitor-tokens-prompts-and-time-to-evidence.md) | | Negotiate clauses with the counterparty | [help](help/edge-consortium/negotiate-clauses-with-the-counterparty.md) | [admin](admin/edge-consortium/negotiate-clauses-with-the-counterparty.md) | [API](api/edge-consortium/negotiate-clauses-with-the-counterparty.md) | [journeys](../generated/diagrams/edge-consortium/negotiate-clauses-with-the-counterparty.md) | | Nominate a learner for reviewer training | [help](help/edge-consortium/nominate-a-learner-for-reviewer-training.md) | [admin](admin/edge-consortium/nominate-a-learner-for-reviewer-training.md) | [API](api/edge-consortium/nominate-a-learner-for-reviewer-training.md) | [journeys](../generated/diagrams/edge-consortium/nominate-a-learner-for-reviewer-training.md) | | Offer the degree with a partner institution | [help](help/edge-consortium/offer-the-degree-with-a-partner-institution.md) | [admin](admin/edge-consortium/offer-the-degree-with-a-partner-institution.md) | [API](api/edge-consortium/offer-the-degree-with-a-partner-institution.md) | [journeys](../generated/diagrams/edge-consortium/offer-the-degree-with-a-partner-institution.md) | | Onboard a new starter | [help](help/edge-consortium/onboard-a-new-starter.md) | [admin](admin/edge-consortium/onboard-a-new-starter.md) | [API](api/edge-consortium/onboard-a-new-starter.md) | [journeys](../generated/diagrams/edge-consortium/onboard-a-new-starter.md) | | Open an issue | [help](help/edge-consortium/open-an-issue.md) | [admin](admin/edge-consortium/open-an-issue.md) | [API](api/edge-consortium/open-an-issue.md) | [journeys](../generated/diagrams/edge-consortium/open-an-issue.md) | | Paid student work, offset against tuition | [help](help/edge-consortium/paid-student-work-with-tuition-offset.md) | [admin](admin/edge-consortium/paid-student-work-with-tuition-offset.md) | [API](api/edge-consortium/paid-student-work-with-tuition-offset.md) | [journeys](../generated/diagrams/edge-consortium/paid-student-work-with-tuition-offset.md) | | Pass reviewer credential verification | [help](help/edge-consortium/pass-reviewer-credential-verification.md) | [admin](admin/edge-consortium/pass-reviewer-credential-verification.md) | [API](api/edge-consortium/pass-reviewer-credential-verification.md) | [journeys](../generated/diagrams/edge-consortium/pass-reviewer-credential-verification.md) | | Pause or pre-empt a job | [help](help/edge-consortium/pause-or-pre-empt-a-job.md) | [admin](admin/edge-consortium/pause-or-pre-empt-a-job.md) | [API](api/edge-consortium/pause-or-pre-empt-a-job.md) | [journeys](../generated/diagrams/edge-consortium/pause-or-pre-empt-a-job.md) | | Pay reviewer fees | [help](help/edge-consortium/pay-reviewer-fees.md) | [admin](admin/edge-consortium/pay-reviewer-fees.md) | [API](api/edge-consortium/pay-reviewer-fees.md) | [journeys](../generated/diagrams/edge-consortium/pay-reviewer-fees.md) | | Pay salary | [help](help/edge-consortium/pay-salary.md) | [admin](admin/edge-consortium/pay-salary.md) | [API](api/edge-consortium/pay-salary.md) | [journeys](../generated/diagrams/edge-consortium/pay-salary.md) | | Petition for student finance | [help](help/edge-consortium/petition-for-student-finance.md) | [admin](admin/edge-consortium/petition-for-student-finance.md) | [API](api/edge-consortium/petition-for-student-finance.md) | [journeys](../generated/diagrams/edge-consortium/petition-for-student-finance.md) | | Pin, edit or delete a post | [help](help/edge-consortium/pin-edit-or-delete-a-post.md) | [admin](admin/edge-consortium/pin-edit-or-delete-a-post.md) | [API](api/edge-consortium/pin-edit-or-delete-a-post.md) | [journeys](../generated/diagrams/edge-consortium/pin-edit-or-delete-a-post.md) | | Plan a maintenance window | [help](help/edge-consortium/plan-a-maintenance-window.md) | [admin](admin/edge-consortium/plan-a-maintenance-window.md) | [API](api/edge-consortium/plan-a-maintenance-window.md) | [journeys](../generated/diagrams/edge-consortium/plan-a-maintenance-window.md) | | Plan a procurement and publish pipeline notice | [help](help/edge-consortium/plan-a-procurement-and-publish-pipeline-notice.md) | [admin](admin/edge-consortium/plan-a-procurement-and-publish-pipeline-notice.md) | [API](api/edge-consortium/plan-a-procurement-and-publish-pipeline-notice.md) | [journeys](../generated/diagrams/edge-consortium/plan-a-procurement-and-publish-pipeline-notice.md) | | Post an announcement | [help](help/edge-consortium/post-an-announcement.md) | [admin](admin/edge-consortium/post-an-announcement.md) | [API](api/edge-consortium/post-an-announcement.md) | [journeys](../generated/diagrams/edge-consortium/post-an-announcement.md) | | Pre-register a study | [help](help/edge-consortium/pre-register-a-study.md) | [admin](admin/edge-consortium/pre-register-a-study.md) | [API](api/edge-consortium/pre-register-a-study.md) | [journeys](../generated/diagrams/edge-consortium/pre-register-a-study.md) | | Prepare a review summary for a reviewer | [help](help/edge-consortium/prepare-a-review-summary-for-a-reviewer.md) | [admin](admin/edge-consortium/prepare-a-review-summary-for-a-reviewer.md) | [API](api/edge-consortium/prepare-a-review-summary-for-a-reviewer.md) | [journeys](../generated/diagrams/edge-consortium/prepare-a-review-summary-for-a-reviewer.md) | | Prepare an employment contract template | [help](help/edge-consortium/prepare-an-employment-contract-template.md) | [admin](admin/edge-consortium/prepare-an-employment-contract-template.md) | [API](api/edge-consortium/prepare-an-employment-contract-template.md) | [journeys](../generated/diagrams/edge-consortium/prepare-an-employment-contract-template.md) | | Preserve and migrate file formats | [help](help/edge-consortium/preserve-and-migrate-file-formats.md) | [admin](admin/edge-consortium/preserve-and-migrate-file-formats.md) | [API](api/edge-consortium/preserve-and-migrate-file-formats.md) | [journeys](../generated/diagrams/edge-consortium/preserve-and-migrate-file-formats.md) | | Process a supplier invoice | [help](help/edge-consortium/process-a-supplier-invoice.md) | [admin](admin/edge-consortium/process-a-supplier-invoice.md) | [API](api/edge-consortium/process-a-supplier-invoice.md) | [journeys](../generated/diagrams/edge-consortium/process-a-supplier-invoice.md) | | Produce a carbon reduction plan | [help](help/edge-consortium/produce-a-carbon-reduction-plan.md) | [admin](admin/edge-consortium/produce-a-carbon-reduction-plan.md) | [API](api/edge-consortium/produce-a-carbon-reduction-plan.md) | [journeys](../generated/diagrams/edge-consortium/produce-a-carbon-reduction-plan.md) | | Produce a graphic in Visual Studio | [help](help/edge-consortium/produce-a-graphic-in-visual-studio.md) | [admin](admin/edge-consortium/produce-a-graphic-in-visual-studio.md) | [API](api/edge-consortium/produce-a-graphic-in-visual-studio.md) | [journeys](../generated/diagrams/edge-consortium/produce-a-graphic-in-visual-studio.md) | | Produce a statutory data return | [help](help/edge-consortium/produce-a-statutory-data-return.md) | [admin](admin/edge-consortium/produce-a-statutory-data-return.md) | [API](api/edge-consortium/produce-a-statutory-data-return.md) | [journeys](../generated/diagrams/edge-consortium/produce-a-statutory-data-return.md) | | Propose a change from a fork | [help](help/edge-consortium/propose-a-change-from-a-fork.md) | [admin](admin/edge-consortium/propose-a-change-from-a-fork.md) | [API](api/edge-consortium/propose-a-change-from-a-fork.md) | [journeys](../generated/diagrams/edge-consortium/propose-a-change-from-a-fork.md) | | Propose a real-world problem project | [help](help/edge-consortium/propose-a-real-world-problem-project.md) | [admin](admin/edge-consortium/propose-a-real-world-problem-project.md) | [API](api/edge-consortium/propose-a-real-world-problem-project.md) | [journeys](../generated/diagrams/edge-consortium/propose-a-real-world-problem-project.md) | | Provide a funder audit trail | [help](help/edge-consortium/provide-a-funder-audit-trail.md) | [admin](admin/edge-consortium/provide-a-funder-audit-trail.md) | [API](api/edge-consortium/provide-a-funder-audit-trail.md) | [journeys](../generated/diagrams/edge-consortium/provide-a-funder-audit-trail.md) | | Provide a reference | [help](help/edge-consortium/provide-a-reference.md) | [admin](admin/edge-consortium/provide-a-reference.md) | [API](api/edge-consortium/provide-a-reference.md) | [journeys](../generated/diagrams/edge-consortium/provide-a-reference.md) | | Publish a data dictionary | [help](help/edge-consortium/publish-a-data-dictionary.md) | [admin](admin/edge-consortium/publish-a-data-dictionary.md) | [API](api/edge-consortium/publish-a-data-dictionary.md) | [journeys](../generated/diagrams/edge-consortium/publish-a-data-dictionary.md) | | Publish a Markdown as deck or book | [help](help/edge-consortium/publish-a-markdown-as-deck-or-book.md) | [admin](admin/edge-consortium/publish-a-markdown-as-deck-or-book.md) | [API](api/edge-consortium/publish-a-markdown-as-deck-or-book.md) | [journeys](../generated/diagrams/edge-consortium/publish-a-markdown-as-deck-or-book.md) | | Publish a new dataset version | [help](help/edge-consortium/publish-a-new-dataset-version.md) | [admin](admin/edge-consortium/publish-a-new-dataset-version.md) | [API](api/edge-consortium/publish-a-new-dataset-version.md) | [journeys](../generated/diagrams/edge-consortium/publish-a-new-dataset-version.md) | | Publish a tender notice | [help](help/edge-consortium/publish-a-tender-notice.md) | [admin](admin/edge-consortium/publish-a-tender-notice.md) | [API](api/edge-consortium/publish-a-tender-notice.md) | [journeys](../generated/diagrams/edge-consortium/publish-a-tender-notice.md) | | Publish a terms version and require re-acceptance | [help](help/edge-consortium/publish-a-terms-version-and-require-re-acceptance.md) | [admin](admin/edge-consortium/publish-a-terms-version-and-require-re-acceptance.md) | [API](api/edge-consortium/publish-a-terms-version-and-require-re-acceptance.md) | [journeys](../generated/diagrams/edge-consortium/publish-a-terms-version-and-require-re-acceptance.md) | | Publish an open access version | [help](help/edge-consortium/publish-an-open-access-version.md) | [admin](admin/edge-consortium/publish-an-open-access-version.md) | [API](api/edge-consortium/publish-an-open-access-version.md) | [journeys](../generated/diagrams/edge-consortium/publish-an-open-access-version.md) | | Publish documentation | [help](help/edge-consortium/publish-documentation.md) | [admin](admin/edge-consortium/publish-documentation.md) | [API](api/edge-consortium/publish-documentation.md) | [journeys](../generated/diagrams/edge-consortium/publish-documentation.md) | | Publish module materials to the library | [help](help/edge-consortium/publish-module-materials-to-the-library.md) | [admin](admin/edge-consortium/publish-module-materials-to-the-library.md) | [API](api/edge-consortium/publish-module-materials-to-the-library.md) | [journeys](../generated/diagrams/edge-consortium/publish-module-materials-to-the-library.md) | | Publish the catalogue to open indexes | [help](help/edge-consortium/publish-the-catalogue-to-open-indexes.md) | [admin](admin/edge-consortium/publish-the-catalogue-to-open-indexes.md) | [API](api/edge-consortium/publish-the-catalogue-to-open-indexes.md) | [journeys](../generated/diagrams/edge-consortium/publish-the-catalogue-to-open-indexes.md) | | Publish the contract details notice | [help](help/edge-consortium/publish-the-contract-details-notice.md) | [admin](admin/edge-consortium/publish-the-contract-details-notice.md) | [API](api/edge-consortium/publish-the-contract-details-notice.md) | [journeys](../generated/diagrams/edge-consortium/publish-the-contract-details-notice.md) | | Publish the monthly carbon and cost statement | [help](help/edge-consortium/publish-the-monthly-carbon-and-cost-statement.md) | [admin](admin/edge-consortium/publish-the-monthly-carbon-and-cost-statement.md) | [API](api/edge-consortium/publish-the-monthly-carbon-and-cost-statement.md) | [journeys](../generated/diagrams/edge-consortium/publish-the-monthly-carbon-and-cost-statement.md) | | Raise a challenge or complaint | [help](help/edge-consortium/raise-a-challenge-or-complaint.md) | [admin](admin/edge-consortium/raise-a-challenge-or-complaint.md) | [API](api/edge-consortium/raise-a-challenge-or-complaint.md) | [journeys](../generated/diagrams/edge-consortium/raise-a-challenge-or-complaint.md) | | Raise a grievance | [help](help/edge-consortium/raise-a-grievance.md) | [admin](admin/edge-consortium/raise-a-grievance.md) | [API](api/edge-consortium/raise-a-grievance.md) | [journeys](../generated/diagrams/edge-consortium/raise-a-grievance.md) | | Raise a progress or wellbeing concern | [help](help/edge-consortium/raise-a-progress-or-wellbeing-concern.md) | [admin](admin/edge-consortium/raise-a-progress-or-wellbeing-concern.md) | [API](api/edge-consortium/raise-a-progress-or-wellbeing-concern.md) | [journeys](../generated/diagrams/edge-consortium/raise-a-progress-or-wellbeing-concern.md) | | Rank agents and models by efficiency | [help](help/edge-consortium/rank-agents-and-models-by-efficiency.md) | [admin](admin/edge-consortium/rank-agents-and-models-by-efficiency.md) | [API](api/edge-consortium/rank-agents-and-models-by-efficiency.md) | [journeys](../generated/diagrams/edge-consortium/rank-agents-and-models-by-efficiency.md) | | Read a verified entry with its provenance | [help](help/edge-consortium/read-a-verified-entry-with-its-provenance.md) | [admin](admin/edge-consortium/read-a-verified-entry-with-its-provenance.md) | [API](api/edge-consortium/read-a-verified-entry-with-its-provenance.md) | [journeys](../generated/diagrams/edge-consortium/read-a-verified-entry-with-its-provenance.md) | | Assignments that are real projects, sandboxed | [help](help/edge-consortium/real-project-assignments.md) | [admin](admin/edge-consortium/real-project-assignments.md) | [API](api/edge-consortium/real-project-assignments.md) | [journeys](../generated/diagrams/edge-consortium/real-project-assignments.md) | | Receive a mandate | [help](help/edge-consortium/receive-a-mandate.md) | [admin](admin/edge-consortium/receive-a-mandate.md) | [API](api/edge-consortium/receive-a-mandate.md) | [journeys](../generated/diagrams/edge-consortium/receive-a-mandate.md) | | Receive bids in a sealed queue | [help](help/edge-consortium/receive-bids-in-a-sealed-queue.md) | [admin](admin/edge-consortium/receive-bids-in-a-sealed-queue.md) | [API](api/edge-consortium/receive-bids-in-a-sealed-queue.md) | [journeys](../generated/diagrams/edge-consortium/receive-bids-in-a-sealed-queue.md) | | Receive the award outcome | [help](help/edge-consortium/receive-the-award-outcome.md) | [admin](admin/edge-consortium/receive-the-award-outcome.md) | [API](api/edge-consortium/receive-the-award-outcome.md) | [journeys](../generated/diagrams/edge-consortium/receive-the-award-outcome.md) | | Receive time-boxed auditor access | [help](help/edge-consortium/receive-time-boxed-auditor-access.md) | [admin](admin/edge-consortium/receive-time-boxed-auditor-access.md) | [API](api/edge-consortium/receive-time-boxed-auditor-access.md) | [journeys](../generated/diagrams/edge-consortium/receive-time-boxed-auditor-access.md) | | Reconcile monthly spend | [help](help/edge-consortium/reconcile-monthly-spend.md) | [admin](admin/edge-consortium/reconcile-monthly-spend.md) | [API](api/edge-consortium/reconcile-monthly-spend.md) | [journeys](../generated/diagrams/edge-consortium/reconcile-monthly-spend.md) | | Reconcile provider invoices with metered use | [help](help/edge-consortium/reconcile-provider-invoices-with-metered-use.md) | [admin](admin/edge-consortium/reconcile-provider-invoices-with-metered-use.md) | [API](api/edge-consortium/reconcile-provider-invoices-with-metered-use.md) | [journeys](../generated/diagrams/edge-consortium/reconcile-provider-invoices-with-metered-use.md) | | Reconcile with the student record system | [help](help/edge-consortium/reconcile-with-the-student-record-system.md) | [admin](admin/edge-consortium/reconcile-with-the-student-record-system.md) | [API](api/edge-consortium/reconcile-with-the-student-record-system.md) | [journeys](../generated/diagrams/edge-consortium/reconcile-with-the-student-record-system.md) | | Record a call with consent | [help](help/edge-consortium/record-a-call-with-consent.md) | [admin](admin/edge-consortium/record-a-call-with-consent.md) | [API](api/edge-consortium/record-a-call-with-consent.md) | [journeys](../generated/diagrams/edge-consortium/record-a-call-with-consent.md) | | Record a conflict of interest assessment | [help](help/edge-consortium/record-a-conflict-of-interest-assessment.md) | [admin](admin/edge-consortium/record-a-conflict-of-interest-assessment.md) | [API](api/edge-consortium/record-a-conflict-of-interest-assessment.md) | [journeys](../generated/diagrams/edge-consortium/record-a-conflict-of-interest-assessment.md) | | Record a data processing agreement | [help](help/edge-consortium/record-a-data-processing-agreement.md) | [admin](admin/edge-consortium/record-a-data-processing-agreement.md) | [API](api/edge-consortium/record-a-data-processing-agreement.md) | [journeys](../generated/diagrams/edge-consortium/record-a-data-processing-agreement.md) | | Record a priority date | [help](help/edge-consortium/record-a-priority-date.md) | [admin](admin/edge-consortium/record-a-priority-date.md) | [API](api/edge-consortium/record-a-priority-date.md) | [journeys](../generated/diagrams/edge-consortium/record-a-priority-date.md) | | Record a trademark and brand use rule | [help](help/edge-consortium/record-a-trademark-and-brand-use-rule.md) | [admin](admin/edge-consortium/record-a-trademark-and-brand-use-rule.md) | [API](api/edge-consortium/record-a-trademark-and-brand-use-rule.md) | [journeys](../generated/diagrams/edge-consortium/record-a-trademark-and-brand-use-rule.md) | | Record an invention disclosure | [help](help/edge-consortium/record-an-invention-disclosure.md) | [admin](admin/edge-consortium/record-an-invention-disclosure.md) | [API](api/edge-consortium/record-an-invention-disclosure.md) | [journeys](../generated/diagrams/edge-consortium/record-an-invention-disclosure.md) | | Record applications and outcomes in a portfolio | [help](help/edge-consortium/record-applications-and-outcomes-in-a-portfolio.md) | [admin](admin/edge-consortium/record-applications-and-outcomes-in-a-portfolio.md) | [API](api/edge-consortium/record-applications-and-outcomes-in-a-portfolio.md) | [journeys](../generated/diagrams/edge-consortium/record-applications-and-outcomes-in-a-portfolio.md) | | Record attendance and engagement | [help](help/edge-consortium/record-attendance-and-engagement.md) | [admin](admin/edge-consortium/record-attendance-and-engagement.md) | [API](api/edge-consortium/record-attendance-and-engagement.md) | [journeys](../generated/diagrams/edge-consortium/record-attendance-and-engagement.md) | | Record authorship and contributor roles | [help](help/edge-consortium/record-authorship-and-contributor-roles.md) | [admin](admin/edge-consortium/record-authorship-and-contributor-roles.md) | [API](api/edge-consortium/record-authorship-and-contributor-roles.md) | [journeys](../generated/diagrams/edge-consortium/record-authorship-and-contributor-roles.md) | | Record background and foreground IP | [help](help/edge-consortium/record-background-and-foreground-ip.md) | [admin](admin/edge-consortium/record-background-and-foreground-ip.md) | [API](api/edge-consortium/record-background-and-foreground-ip.md) | [journeys](../generated/diagrams/edge-consortium/record-background-and-foreground-ip.md) | | Record copyright and attribution | [help](help/edge-consortium/record-copyright-and-attribution.md) | [admin](admin/edge-consortium/record-copyright-and-attribution.md) | [API](api/edge-consortium/record-copyright-and-attribution.md) | [journeys](../generated/diagrams/edge-consortium/record-copyright-and-attribution.md) | | Record energy per job | [help](help/edge-consortium/record-energy-per-job.md) | [admin](admin/edge-consortium/record-energy-per-job.md) | [API](api/edge-consortium/record-energy-per-job.md) | [journeys](../generated/diagrams/edge-consortium/record-energy-per-job.md) | | Record payment compliance | [help](help/edge-consortium/record-payment-compliance.md) | [admin](admin/edge-consortium/record-payment-compliance.md) | [API](api/edge-consortium/record-payment-compliance.md) | [journeys](../generated/diagrams/edge-consortium/record-payment-compliance.md) | | Record provenance and lineage | [help](help/edge-consortium/record-provenance-and-lineage.md) | [admin](admin/edge-consortium/record-provenance-and-lineage.md) | [API](api/edge-consortium/record-provenance-and-lineage.md) | [journeys](../generated/diagrams/edge-consortium/record-provenance-and-lineage.md) | | Record working hours and leave | [help](help/edge-consortium/record-working-hours-and-leave.md) | [admin](admin/edge-consortium/record-working-hours-and-leave.md) | [API](api/edge-consortium/record-working-hours-and-leave.md) | [journeys](../generated/diagrams/edge-consortium/record-working-hours-and-leave.md) | | Refuse an action outside the mandate | [help](help/edge-consortium/refuse-an-action-outside-the-mandate.md) | [admin](admin/edge-consortium/refuse-an-action-outside-the-mandate.md) | [API](api/edge-consortium/refuse-an-action-outside-the-mandate.md) | [journeys](../generated/diagrams/edge-consortium/refuse-an-action-outside-the-mandate.md) | | Register a compute resource | [help](help/edge-consortium/register-a-compute-resource.md) | [admin](admin/edge-consortium/register-a-compute-resource.md) | [API](api/edge-consortium/register-a-compute-resource.md) | [journeys](../generated/diagrams/edge-consortium/register-a-compute-resource.md) | | Register a dataset in the catalogue | [help](help/edge-consortium/register-a-dataset-in-the-catalogue.md) | [admin](admin/edge-consortium/register-a-dataset-in-the-catalogue.md) | [API](api/edge-consortium/register-a-dataset-in-the-catalogue.md) | [journeys](../generated/diagrams/edge-consortium/register-a-dataset-in-the-catalogue.md) | | Register a delegated authority | [help](help/edge-consortium/register-a-delegated-authority.md) | [admin](admin/edge-consortium/register-a-delegated-authority.md) | [API](api/edge-consortium/register-a-delegated-authority.md) | [journeys](../generated/diagrams/edge-consortium/register-a-delegated-authority.md) | | Register a research project with an ID | [help](help/edge-consortium/register-a-research-project-with-an-id.md) | [admin](admin/edge-consortium/register-a-research-project-with-an-id.md) | [API](api/edge-consortium/register-a-research-project-with-an-id.md) | [journeys](../generated/diagrams/edge-consortium/register-a-research-project-with-an-id.md) | | Register a university as a member institution | [help](help/edge-consortium/register-a-university-as-a-member-institution.md) | [admin](admin/edge-consortium/register-a-university-as-a-member-institution.md) | [API](api/edge-consortium/register-a-university-as-a-member-institution.md) | [journeys](../generated/diagrams/edge-consortium/register-a-university-as-a-member-institution.md) | | Register as a supplier | [help](help/edge-consortium/register-as-a-supplier.md) | [admin](admin/edge-consortium/register-as-a-supplier.md) | [API](api/edge-consortium/register-as-a-supplier.md) | [journeys](../generated/diagrams/edge-consortium/register-as-a-supplier.md) | | Register with the Office for Students | [help](help/edge-consortium/register-with-the-office-for-students.md) | [admin](admin/edge-consortium/register-with-the-office-for-students.md) | [API](api/edge-consortium/register-with-the-office-for-students.md) | [journeys](../generated/diagrams/edge-consortium/register-with-the-office-for-students.md) | | Reject a contribution with reasons | [help](help/edge-consortium/reject-a-contribution-with-reasons.md) | [admin](admin/edge-consortium/reject-a-contribution-with-reasons.md) | [API](api/edge-consortium/reject-a-contribution-with-reasons.md) | [journeys](../generated/diagrams/edge-consortium/reject-a-contribution-with-reasons.md) | | Release a version | [help](help/edge-consortium/release-a-version.md) | [admin](admin/edge-consortium/release-a-version.md) | [API](api/edge-consortium/release-a-version.md) | [journeys](../generated/diagrams/edge-consortium/release-a-version.md) | | Release marks and feedback | [help](help/edge-consortium/release-marks-and-feedback.md) | [admin](admin/edge-consortium/release-marks-and-feedback.md) | [API](api/edge-consortium/release-marks-and-feedback.md) | [journeys](../generated/diagrams/edge-consortium/release-marks-and-feedback.md) | | Release results with disclosure control | [help](help/edge-consortium/release-results-with-disclosure-control.md) | [admin](admin/edge-consortium/release-results-with-disclosure-control.md) | [API](api/edge-consortium/release-results-with-disclosure-control.md) | [journeys](../generated/diagrams/edge-consortium/release-results-with-disclosure-control.md) | | Renew annual institutional membership | [help](help/edge-consortium/renew-annual-institutional-membership.md) | [admin](admin/edge-consortium/renew-annual-institutional-membership.md) | [API](api/edge-consortium/renew-annual-institutional-membership.md) | [journeys](../generated/diagrams/edge-consortium/renew-annual-institutional-membership.md) | | Renew reviewer verification | [help](help/edge-consortium/renew-reviewer-verification.md) | [admin](admin/edge-consortium/renew-reviewer-verification.md) | [API](api/edge-consortium/renew-reviewer-verification.md) | [journeys](../generated/diagrams/edge-consortium/renew-reviewer-verification.md) | | Report a personal data breach | [help](help/edge-consortium/report-a-personal-data-breach.md) | [admin](admin/edge-consortium/report-a-personal-data-breach.md) | [API](api/edge-consortium/report-a-personal-data-breach.md) | [journeys](../generated/diagrams/edge-consortium/report-a-personal-data-breach.md) | | Report a research misconduct concern | [help](help/edge-consortium/report-a-research-misconduct-concern.md) | [admin](admin/edge-consortium/report-a-research-misconduct-concern.md) | [API](api/edge-consortium/report-a-research-misconduct-concern.md) | [journeys](../generated/diagrams/edge-consortium/report-a-research-misconduct-concern.md) | | Report carbon intensity per accepted outcome | [help](help/edge-consortium/report-carbon-intensity-per-accepted-outcome.md) | [admin](admin/edge-consortium/report-carbon-intensity-per-accepted-outcome.md) | [API](api/edge-consortium/report-carbon-intensity-per-accepted-outcome.md) | [journeys](../generated/diagrams/edge-consortium/report-carbon-intensity-per-accepted-outcome.md) | | Report expenditure to the funder | [help](help/edge-consortium/report-expenditure-to-the-funder.md) | [admin](admin/edge-consortium/report-expenditure-to-the-funder.md) | [API](api/edge-consortium/report-expenditure-to-the-funder.md) | [journeys](../generated/diagrams/edge-consortium/report-expenditure-to-the-funder.md) | | Report monitoring measures | [help](help/edge-consortium/report-monitoring-measures.md) | [admin](admin/edge-consortium/report-monitoring-measures.md) | [API](api/edge-consortium/report-monitoring-measures.md) | [journeys](../generated/diagrams/edge-consortium/report-monitoring-measures.md) | | Report performance indicator evidence | [help](help/edge-consortium/report-performance-indicator-evidence.md) | [admin](admin/edge-consortium/report-performance-indicator-evidence.md) | [API](api/edge-consortium/report-performance-indicator-evidence.md) | [journeys](../generated/diagrams/edge-consortium/report-performance-indicator-evidence.md) | | Report review throughput and times | [help](help/edge-consortium/report-review-throughput-and-times.md) | [admin](admin/edge-consortium/report-review-throughput-and-times.md) | [API](api/edge-consortium/report-review-throughput-and-times.md) | [journeys](../generated/diagrams/edge-consortium/report-review-throughput-and-times.md) | | Report tokens, prompts and time to evidence | [help](help/edge-consortium/report-tokens-prompts-and-time-to-evidence.md) | [admin](admin/edge-consortium/report-tokens-prompts-and-time-to-evidence.md) | [API](api/edge-consortium/report-tokens-prompts-and-time-to-evidence.md) | [journeys](../generated/diagrams/edge-consortium/report-tokens-prompts-and-time-to-evidence.md) | | Report utilisation | [help](help/edge-consortium/report-utilisation.md) | [admin](admin/edge-consortium/report-utilisation.md) | [API](api/edge-consortium/report-utilisation.md) | [journeys](../generated/diagrams/edge-consortium/report-utilisation.md) | | Request a compute allowance for a project | [help](help/edge-consortium/request-a-compute-allowance-for-a-project.md) | [admin](admin/edge-consortium/request-a-compute-allowance-for-a-project.md) | [API](api/edge-consortium/request-a-compute-allowance-for-a-project.md) | [journeys](../generated/diagrams/edge-consortium/request-a-compute-allowance-for-a-project.md) | | Request a purpose view on federated data | [help](help/edge-consortium/request-a-purpose-view-on-federated-data.md) | [admin](admin/edge-consortium/request-a-purpose-view-on-federated-data.md) | [API](api/edge-consortium/request-a-purpose-view-on-federated-data.md) | [journeys](../generated/diagrams/edge-consortium/request-a-purpose-view-on-federated-data.md) | | Request a Sandbox with GPU and data | [help](help/edge-consortium/request-a-sandbox-with-gpu-and-data.md) | [admin](admin/edge-consortium/request-a-sandbox-with-gpu-and-data.md) | [API](api/edge-consortium/request-a-sandbox-with-gpu-and-data.md) | [journeys](../generated/diagrams/edge-consortium/request-a-sandbox-with-gpu-and-data.md) | | Request approval for a consequential action | [help](help/edge-consortium/request-approval-for-a-consequential-action.md) | [admin](admin/edge-consortium/request-approval-for-a-consequential-action.md) | [API](api/edge-consortium/request-approval-for-a-consequential-action.md) | [journeys](../generated/diagrams/edge-consortium/request-approval-for-a-consequential-action.md) | | Request changes from the contributor | [help](help/edge-consortium/request-changes-from-the-contributor.md) | [admin](admin/edge-consortium/request-changes-from-the-contributor.md) | [API](api/edge-consortium/request-changes-from-the-contributor.md) | [journeys](../generated/diagrams/edge-consortium/request-changes-from-the-contributor.md) | | Request supervisor allocation | [help](help/edge-consortium/request-supervisor-allocation.md) | [admin](admin/edge-consortium/request-supervisor-allocation.md) | [API](api/edge-consortium/request-supervisor-allocation.md) | [journeys](../generated/diagrams/edge-consortium/request-supervisor-allocation.md) | | Research and development partnership | [help](help/edge-consortium/research-and-development-partnership.md) | [admin](admin/edge-consortium/research-and-development-partnership.md) | [API](api/edge-consortium/research-and-development-partnership.md) | [journeys](../generated/diagrams/edge-consortium/research-and-development-partnership.md) | | Research consortium | [help](help/edge-consortium/research-consortium.md) | [admin](admin/edge-consortium/research-consortium.md) | [API](api/edge-consortium/research-consortium.md) | [journeys](../generated/diagrams/edge-consortium/research-consortium.md) | | Reserve capacity for a deadline | [help](help/edge-consortium/reserve-capacity-for-a-deadline.md) | [admin](admin/edge-consortium/reserve-capacity-for-a-deadline.md) | [API](api/edge-consortium/reserve-capacity-for-a-deadline.md) | [journeys](../generated/diagrams/edge-consortium/reserve-capacity-for-a-deadline.md) | | Resolve conflicting reviews | [help](help/edge-consortium/resolve-conflicting-reviews.md) | [admin](admin/edge-consortium/resolve-conflicting-reviews.md) | [API](api/edge-consortium/resolve-conflicting-reviews.md) | [journeys](../generated/diagrams/edge-consortium/resolve-conflicting-reviews.md) | | Respond to a security incident | [help](help/edge-consortium/respond-to-a-security-incident.md) | [admin](admin/edge-consortium/respond-to-a-security-incident.md) | [API](api/edge-consortium/respond-to-a-security-incident.md) | [journeys](../generated/diagrams/edge-consortium/respond-to-a-security-incident.md) | | Respond to an invitation | [help](help/edge-consortium/respond-to-an-invitation.md) | [admin](admin/edge-consortium/respond-to-an-invitation.md) | [API](api/edge-consortium/respond-to-an-invitation.md) | [journeys](../generated/diagrams/edge-consortium/respond-to-an-invitation.md) | | Retract or correct a verified entry | [help](help/edge-consortium/retract-or-correct-a-verified-entry.md) | [admin](admin/edge-consortium/retract-or-correct-a-verified-entry.md) | [API](api/edge-consortium/retract-or-correct-a-verified-entry.md) | [journeys](../generated/diagrams/edge-consortium/retract-or-correct-a-verified-entry.md) | | Retry a failed job | [help](help/edge-consortium/retry-a-failed-job.md) | [admin](admin/edge-consortium/retry-a-failed-job.md) | [API](api/edge-consortium/retry-a-failed-job.md) | [journeys](../generated/diagrams/edge-consortium/retry-a-failed-job.md) | | Review a contribution against criteria | [help](help/edge-consortium/review-a-contribution-against-criteria.md) | [admin](admin/edge-consortium/review-a-contribution-against-criteria.md) | [API](api/edge-consortium/review-a-contribution-against-criteria.md) | [journeys](../generated/diagrams/edge-consortium/review-a-contribution-against-criteria.md) | | Review a funder expenditure statement | [help](help/edge-consortium/review-a-funder-expenditure-statement.md) | [admin](admin/edge-consortium/review-a-funder-expenditure-statement.md) | [API](api/edge-consortium/review-a-funder-expenditure-statement.md) | [journeys](../generated/diagrams/edge-consortium/review-a-funder-expenditure-statement.md) | | Review a pull request | [help](help/edge-consortium/review-a-pull-request.md) | [admin](admin/edge-consortium/review-a-pull-request.md) | [API](api/edge-consortium/review-a-pull-request.md) | [journeys](../generated/diagrams/edge-consortium/review-a-pull-request.md) | | Review access logs and anomalies | [help](help/edge-consortium/review-access-logs-and-anomalies.md) | [admin](admin/edge-consortium/review-access-logs-and-anomalies.md) | [API](api/edge-consortium/review-access-logs-and-anomalies.md) | [journeys](../generated/diagrams/edge-consortium/review-access-logs-and-anomalies.md) | | Review agent receipts | [help](help/edge-consortium/review-agent-receipts.md) | [admin](admin/edge-consortium/review-agent-receipts.md) | [API](api/edge-consortium/review-agent-receipts.md) | [journeys](../generated/diagrams/edge-consortium/review-agent-receipts.md) | | Review module evaluation feedback | [help](help/edge-consortium/review-module-evaluation-feedback.md) | [admin](admin/edge-consortium/review-module-evaluation-feedback.md) | [API](api/edge-consortium/review-module-evaluation-feedback.md) | [journeys](../generated/diagrams/edge-consortium/review-module-evaluation-feedback.md) | | Review quality and standards evidence | [help](help/edge-consortium/review-quality-and-standards-evidence.md) | [admin](admin/edge-consortium/review-quality-and-standards-evidence.md) | [API](api/edge-consortium/review-quality-and-standards-evidence.md) | [journeys](../generated/diagrams/edge-consortium/review-quality-and-standards-evidence.md) | | Review the DPIA and record of processing | [help](help/edge-consortium/review-the-dpia-and-record-of-processing.md) | [admin](admin/edge-consortium/review-the-dpia-and-record-of-processing.md) | [API](api/edge-consortium/review-the-dpia-and-record-of-processing.md) | [journeys](../generated/diagrams/edge-consortium/review-the-dpia-and-record-of-processing.md) | | Revoke a purpose view | [help](help/edge-consortium/revoke-a-purpose-view.md) | [admin](admin/edge-consortium/revoke-a-purpose-view.md) | [API](api/edge-consortium/revoke-a-purpose-view.md) | [journeys](../generated/diagrams/edge-consortium/revoke-a-purpose-view.md) | | Revoke access for a leaver | [help](help/edge-consortium/revoke-access-for-a-leaver.md) | [admin](admin/edge-consortium/revoke-access-for-a-leaver.md) | [API](api/edge-consortium/revoke-access-for-a-leaver.md) | [journeys](../generated/diagrams/edge-consortium/revoke-access-for-a-leaver.md) | | Rotate keys | [help](help/edge-consortium/rotate-keys.md) | [admin](admin/edge-consortium/rotate-keys.md) | [API](api/edge-consortium/rotate-keys.md) | [journeys](../generated/diagrams/edge-consortium/rotate-keys.md) | | Run a clarification round | [help](help/edge-consortium/run-a-clarification-round.md) | [admin](admin/edge-consortium/run-a-clarification-round.md) | [API](api/edge-consortium/run-a-clarification-round.md) | [journeys](../generated/diagrams/edge-consortium/run-a-clarification-round.md) | | Run a compute job under mandate | [help](help/edge-consortium/run-a-compute-job-under-mandate.md) | [admin](admin/edge-consortium/run-a-compute-job-under-mandate.md) | [API](api/edge-consortium/run-a-compute-job-under-mandate.md) | [journeys](../generated/diagrams/edge-consortium/run-a-compute-job-under-mandate.md) | | Run a cyber recovery drill | [help](help/edge-consortium/run-a-cyber-recovery-drill.md) | [admin](admin/edge-consortium/run-a-cyber-recovery-drill.md) | [API](api/edge-consortium/run-a-cyber-recovery-drill.md) | [journeys](../generated/diagrams/edge-consortium/run-a-cyber-recovery-drill.md) | | Run a live assessment with two examiners and a recording | [help](help/edge-consortium/run-a-live-assessment.md) | [admin](admin/edge-consortium/run-a-live-assessment.md) | [API](api/edge-consortium/run-a-live-assessment.md) | [journeys](../generated/diagrams/edge-consortium/run-a-live-assessment.md) | | Run a live seminar | [help](help/edge-consortium/run-a-live-seminar.md) | [admin](admin/edge-consortium/run-a-live-seminar.md) | [API](api/edge-consortium/run-a-live-seminar.md) | [journeys](../generated/diagrams/edge-consortium/run-a-live-seminar.md) | | Run a security-classified procurement | [help](help/edge-consortium/run-a-security-classified-procurement.md) | [admin](admin/edge-consortium/run-a-security-classified-procurement.md) | [API](api/edge-consortium/run-a-security-classified-procurement.md) | [journeys](../generated/diagrams/edge-consortium/run-a-security-classified-procurement.md) | | Run a training job on scheduled compute | [help](help/edge-consortium/run-a-training-job-on-scheduled-compute.md) | [admin](admin/edge-consortium/run-a-training-job-on-scheduled-compute.md) | [API](api/edge-consortium/run-a-training-job-on-scheduled-compute.md) | [journeys](../generated/diagrams/edge-consortium/run-a-training-job-on-scheduled-compute.md) | | Run a vulnerability scan on a release | [help](help/edge-consortium/run-a-vulnerability-scan-on-a-release.md) | [admin](admin/edge-consortium/run-a-vulnerability-scan-on-a-release.md) | [API](api/edge-consortium/run-a-vulnerability-scan-on-a-release.md) | [journeys](../generated/diagrams/edge-consortium/run-a-vulnerability-scan-on-a-release.md) | | Run an approved query where data lives | [help](help/edge-consortium/run-an-approved-query-where-data-lives.md) | [admin](admin/edge-consortium/run-an-approved-query-where-data-lives.md) | [API](api/edge-consortium/run-an-approved-query-where-data-lives.md) | [journeys](../generated/diagrams/edge-consortium/run-an-approved-query-where-data-lives.md) | | Run code in the Sandbox | [help](help/edge-consortium/run-code-in-the-sandbox.md) | [admin](admin/edge-consortium/run-code-in-the-sandbox.md) | [API](api/edge-consortium/run-code-in-the-sandbox.md) | [journeys](../generated/diagrams/edge-consortium/run-code-in-the-sandbox.md) | | Run data quality checks on a database | [help](help/edge-consortium/run-data-quality-checks-on-a-database.md) | [admin](admin/edge-consortium/run-data-quality-checks-on-a-database.md) | [API](api/edge-consortium/run-data-quality-checks-on-a-database.md) | [journeys](../generated/diagrams/edge-consortium/run-data-quality-checks-on-a-database.md) | | Sample verified entries for quality | [help](help/edge-consortium/sample-verified-entries-for-quality.md) | [admin](admin/edge-consortium/sample-verified-entries-for-quality.md) | [API](api/edge-consortium/sample-verified-entries-for-quality.md) | [journeys](../generated/diagrams/edge-consortium/sample-verified-entries-for-quality.md) | | Save sources to a reading list | [help](help/edge-consortium/save-sources-to-a-reading-list.md) | [admin](admin/edge-consortium/save-sources-to-a-reading-list.md) | [API](api/edge-consortium/save-sources-to-a-reading-list.md) | [journeys](../generated/diagrams/edge-consortium/save-sources-to-a-reading-list.md) | | Scale with an external provider | [help](help/edge-consortium/scale-with-an-external-provider.md) | [admin](admin/edge-consortium/scale-with-an-external-provider.md) | [API](api/edge-consortium/scale-with-an-external-provider.md) | [journeys](../generated/diagrams/edge-consortium/scale-with-an-external-provider.md) | | Schedule a job | [help](help/edge-consortium/schedule-a-job.md) | [admin](admin/edge-consortium/schedule-a-job.md) | [API](api/edge-consortium/schedule-a-job.md) | [journeys](../generated/diagrams/edge-consortium/schedule-a-job.md) | | Schedule a meeting on request | [help](help/edge-consortium/schedule-a-meeting-on-request.md) | [admin](admin/edge-consortium/schedule-a-meeting-on-request.md) | [API](api/edge-consortium/schedule-a-meeting-on-request.md) | [journeys](../generated/diagrams/edge-consortium/schedule-a-meeting-on-request.md) | | Schedule a secure government resource | [help](help/edge-consortium/schedule-a-secure-government-resource.md) | [admin](admin/edge-consortium/schedule-a-secure-government-resource.md) | [API](api/edge-consortium/schedule-a-secure-government-resource.md) | [journeys](../generated/diagrams/edge-consortium/schedule-a-secure-government-resource.md) | | Schedule a send and set a signature | [help](help/edge-consortium/schedule-a-send-and-set-a-signature.md) | [admin](admin/edge-consortium/schedule-a-send-and-set-a-signature.md) | [API](api/edge-consortium/schedule-a-send-and-set-a-signature.md) | [journeys](../generated/diagrams/edge-consortium/schedule-a-send-and-set-a-signature.md) | | Schedule and record supervision meetings | [help](help/edge-consortium/schedule-and-record-supervision-meetings.md) | [admin](admin/edge-consortium/schedule-and-record-supervision-meetings.md) | [API](api/edge-consortium/schedule-and-record-supervision-meetings.md) | [journeys](../generated/diagrams/edge-consortium/schedule-and-record-supervision-meetings.md) | | Schedule deliveries and logistics | [help](help/edge-consortium/schedule-deliveries-and-logistics.md) | [admin](admin/edge-consortium/schedule-deliveries-and-logistics.md) | [API](api/edge-consortium/schedule-deliveries-and-logistics.md) | [journeys](../generated/diagrams/edge-consortium/schedule-deliveries-and-logistics.md) | | Search across mail, channels and meetings | [help](help/edge-consortium/search-across-mail-channels-and-meetings.md) | [admin](admin/edge-consortium/search-across-mail-channels-and-meetings.md) | [API](api/edge-consortium/search-across-mail-channels-and-meetings.md) | [journeys](../generated/diagrams/edge-consortium/search-across-mail-channels-and-meetings.md) | | Search libraries and the knowledge graph | [help](help/edge-consortium/search-libraries-and-the-knowledge-graph.md) | [admin](admin/edge-consortium/search-libraries-and-the-knowledge-graph.md) | [API](api/edge-consortium/search-libraries-and-the-knowledge-graph.md) | [journeys](../generated/diagrams/edge-consortium/search-libraries-and-the-knowledge-graph.md) | | Send a contract for signature | [help](help/edge-consortium/send-a-contract-for-signature.md) | [admin](admin/edge-consortium/send-a-contract-for-signature.md) | [API](api/edge-consortium/send-a-contract-for-signature.md) | [journeys](../generated/diagrams/edge-consortium/send-a-contract-for-signature.md) | | Send mail from an estate mailbox | [help](help/edge-consortium/send-mail-from-an-estate-mailbox.md) | [admin](admin/edge-consortium/send-mail-from-an-estate-mailbox.md) | [API](api/edge-consortium/send-mail-from-an-estate-mailbox.md) | [journeys](../generated/diagrams/edge-consortium/send-mail-from-an-estate-mailbox.md) | | Set a carbon budget and threshold | [help](help/edge-consortium/set-a-carbon-budget-and-threshold.md) | [admin](admin/edge-consortium/set-a-carbon-budget-and-threshold.md) | [API](api/edge-consortium/set-a-carbon-budget-and-threshold.md) | [journeys](../generated/diagrams/edge-consortium/set-a-carbon-budget-and-threshold.md) | | Set a project access control policy | [help](help/edge-consortium/set-a-project-access-control-policy.md) | [admin](admin/edge-consortium/set-a-project-access-control-policy.md) | [API](api/edge-consortium/set-a-project-access-control-policy.md) | [journeys](../generated/diagrams/edge-consortium/set-a-project-access-control-policy.md) | | Set agent authority scope | [help](help/edge-consortium/set-agent-authority-scope.md) | [admin](admin/edge-consortium/set-agent-authority-scope.md) | [API](api/edge-consortium/set-agent-authority-scope.md) | [journeys](../generated/diagrams/edge-consortium/set-agent-authority-scope.md) | | Set and publish key performance indicators | [help](help/edge-consortium/set-and-publish-key-performance-indicators.md) | [admin](admin/edge-consortium/set-and-publish-key-performance-indicators.md) | [API](api/edge-consortium/set-and-publish-key-performance-indicators.md) | [journeys](../generated/diagrams/edge-consortium/set-and-publish-key-performance-indicators.md) | | Set assessment and marking scheme | [help](help/edge-consortium/set-assessment-and-marking-scheme.md) | [admin](admin/edge-consortium/set-assessment-and-marking-scheme.md) | [API](api/edge-consortium/set-assessment-and-marking-scheme.md) | [journeys](../generated/diagrams/edge-consortium/set-assessment-and-marking-scheme.md) | | Set contract parties, permissions and expiry | [help](help/edge-consortium/set-contract-parties-permissions-and-expiry.md) | [admin](admin/edge-consortium/set-contract-parties-permissions-and-expiry.md) | [API](api/edge-consortium/set-contract-parties-permissions-and-expiry.md) | [journeys](../generated/diagrams/edge-consortium/set-contract-parties-permissions-and-expiry.md) | | Set evaluation criteria and weightings | [help](help/edge-consortium/set-evaluation-criteria-and-weightings.md) | [admin](admin/edge-consortium/set-evaluation-criteria-and-weightings.md) | [API](api/edge-consortium/set-evaluation-criteria-and-weightings.md) | [journeys](../generated/diagrams/edge-consortium/set-evaluation-criteria-and-weightings.md) | | Set mail rules, flags and snooze | [help](help/edge-consortium/set-mail-rules-flags-and-snooze.md) | [admin](admin/edge-consortium/set-mail-rules-flags-and-snooze.md) | [API](api/edge-consortium/set-mail-rules-flags-and-snooze.md) | [journeys](../generated/diagrams/edge-consortium/set-mail-rules-flags-and-snooze.md) | | Set mention and tag rules | [help](help/edge-consortium/set-mention-and-tag-rules.md) | [admin](admin/edge-consortium/set-mention-and-tag-rules.md) | [API](api/edge-consortium/set-mention-and-tag-rules.md) | [journeys](../generated/diagrams/edge-consortium/set-mention-and-tag-rules.md) | | Set out of office and working hours | [help](help/edge-consortium/set-out-of-office-and-working-hours.md) | [admin](admin/edge-consortium/set-out-of-office-and-working-hours.md) | [API](api/edge-consortium/set-out-of-office-and-working-hours.md) | [journeys](../generated/diagrams/edge-consortium/set-out-of-office-and-working-hours.md) | | Set permissions on project records | [help](help/edge-consortium/set-permissions-on-project-records.md) | [admin](admin/edge-consortium/set-permissions-on-project-records.md) | [API](api/edge-consortium/set-permissions-on-project-records.md) | [journeys](../generated/diagrams/edge-consortium/set-permissions-on-project-records.md) | | Set presence | [help](help/edge-consortium/set-presence.md) | [admin](admin/edge-consortium/set-presence.md) | [API](api/edge-consortium/set-presence.md) | [journeys](../generated/diagrams/edge-consortium/set-presence.md) | | Set project governance | [help](help/edge-consortium/set-project-governance.md) | [admin](admin/edge-consortium/set-project-governance.md) | [API](api/edge-consortium/set-project-governance.md) | [journeys](../generated/diagrams/edge-consortium/set-project-governance.md) | | Set quotas per project | [help](help/edge-consortium/set-quotas-per-project.md) | [admin](admin/edge-consortium/set-quotas-per-project.md) | [API](api/edge-consortium/set-quotas-per-project.md) | [journeys](../generated/diagrams/edge-consortium/set-quotas-per-project.md) | | Set renewal and expiry alerts | [help](help/edge-consortium/set-renewal-and-expiry-alerts.md) | [admin](admin/edge-consortium/set-renewal-and-expiry-alerts.md) | [API](api/edge-consortium/set-renewal-and-expiry-alerts.md) | [journeys](../generated/diagrams/edge-consortium/set-renewal-and-expiry-alerts.md) | | Set retention and export a channel | [help](help/edge-consortium/set-retention-and-export-a-channel.md) | [admin](admin/edge-consortium/set-retention-and-export-a-channel.md) | [API](api/edge-consortium/set-retention-and-export-a-channel.md) | [journeys](../generated/diagrams/edge-consortium/set-retention-and-export-a-channel.md) | | Set review criteria for a field | [help](help/edge-consortium/set-review-criteria-for-a-field.md) | [admin](admin/edge-consortium/set-review-criteria-for-a-field.md) | [API](api/edge-consortium/set-review-criteria-for-a-field.md) | [journeys](../generated/diagrams/edge-consortium/set-review-criteria-for-a-field.md) | | Set review fields and availability | [help](help/edge-consortium/set-review-fields-and-availability.md) | [admin](admin/edge-consortium/set-review-fields-and-availability.md) | [API](api/edge-consortium/set-review-fields-and-availability.md) | [journeys](../generated/diagrams/edge-consortium/set-review-fields-and-availability.md) | | Set scheduling policy and queues | [help](help/edge-consortium/set-scheduling-policy-and-queues.md) | [admin](admin/edge-consortium/set-scheduling-policy-and-queues.md) | [API](api/edge-consortium/set-scheduling-policy-and-queues.md) | [journeys](../generated/diagrams/edge-consortium/set-scheduling-policy-and-queues.md) | | Set the access tier on an item | [help](help/edge-consortium/set-the-access-tier-on-an-item.md) | [admin](admin/edge-consortium/set-the-access-tier-on-an-item.md) | [API](api/edge-consortium/set-the-access-tier-on-an-item.md) | [journeys](../generated/diagrams/edge-consortium/set-the-access-tier-on-an-item.md) | | Set the review deadline and reminders | [help](help/edge-consortium/set-the-review-deadline-and-reminders.md) | [admin](admin/edge-consortium/set-the-review-deadline-and-reminders.md) | [API](api/edge-consortium/set-the-review-deadline-and-reminders.md) | [journeys](../generated/diagrams/edge-consortium/set-the-review-deadline-and-reminders.md) | | Set up an awarded grant | [help](help/edge-consortium/set-up-an-awarded-grant.md) | [admin](admin/edge-consortium/set-up-an-awarded-grant.md) | [API](api/edge-consortium/set-up-an-awarded-grant.md) | [journeys](../generated/diagrams/edge-consortium/set-up-an-awarded-grant.md) | | Share a calendar with a delegate | [help](help/edge-consortium/share-a-calendar-with-a-delegate.md) | [admin](admin/edge-consortium/share-a-calendar-with-a-delegate.md) | [API](api/edge-consortium/share-a-calendar-with-a-delegate.md) | [journeys](../generated/diagrams/edge-consortium/share-a-calendar-with-a-delegate.md) | | Share a campus with a partner university | [help](help/edge-consortium/share-a-campus-with-a-partner-university.md) | [admin](admin/edge-consortium/share-a-campus-with-a-partner-university.md) | [API](api/edge-consortium/share-a-campus-with-a-partner-university.md) | [journeys](../generated/diagrams/edge-consortium/share-a-campus-with-a-partner-university.md) | | Share a creation in the JudeBox exchange | [help](help/edge-consortium/share-a-creation-in-the-judebox-exchange.md) | [admin](admin/edge-consortium/share-a-creation-in-the-judebox-exchange.md) | [API](api/edge-consortium/share-a-creation-in-the-judebox-exchange.md) | [journeys](../generated/diagrams/edge-consortium/share-a-creation-in-the-judebox-exchange.md) | | Share a document in a call | [help](help/edge-consortium/share-a-document-in-a-call.md) | [admin](admin/edge-consortium/share-a-document-in-a-call.md) | [API](api/edge-consortium/share-a-document-in-a-call.md) | [journeys](../generated/diagrams/edge-consortium/share-a-document-in-a-call.md) | | Share a workspace folder across universities | [help](help/edge-consortium/share-a-workspace-folder-across-universities.md) | [admin](admin/edge-consortium/share-a-workspace-folder-across-universities.md) | [API](api/edge-consortium/share-a-workspace-folder-across-universities.md) | [journeys](../generated/diagrams/edge-consortium/share-a-workspace-folder-across-universities.md) | | Shift jobs to low-carbon windows | [help](help/edge-consortium/shift-jobs-to-low-carbon-windows.md) | [admin](admin/edge-consortium/shift-jobs-to-low-carbon-windows.md) | [API](api/edge-consortium/shift-jobs-to-low-carbon-windows.md) | [journeys](../generated/diagrams/edge-consortium/shift-jobs-to-low-carbon-windows.md) | | Show undisclosed figures as gaps | [help](help/edge-consortium/show-undisclosed-figures-as-gaps.md) | [admin](admin/edge-consortium/show-undisclosed-figures-as-gaps.md) | [API](api/edge-consortium/show-undisclosed-figures-as-gaps.md) | [journeys](../generated/diagrams/edge-consortium/show-undisclosed-figures-as-gaps.md) | | Sign a collaboration agreement | [help](help/edge-consortium/sign-a-collaboration-agreement.md) | [admin](admin/edge-consortium/sign-a-collaboration-agreement.md) | [API](api/edge-consortium/sign-a-collaboration-agreement.md) | [journeys](../generated/diagrams/edge-consortium/sign-a-collaboration-agreement.md) | | Sign a contract with EDGE ID | [help](help/edge-consortium/sign-a-contract-with-edge-id.md) | [admin](admin/edge-consortium/sign-a-contract-with-edge-id.md) | [API](api/edge-consortium/sign-a-contract-with-edge-id.md) | [journeys](../generated/diagrams/edge-consortium/sign-a-contract-with-edge-id.md) | | Sign off a progression milestone | [help](help/edge-consortium/sign-off-a-progression-milestone.md) | [admin](admin/edge-consortium/sign-off-a-progression-milestone.md) | [API](api/edge-consortium/sign-off-a-progression-milestone.md) | [journeys](../generated/diagrams/edge-consortium/sign-off-a-progression-milestone.md) | | Sign the contributor agreement | [help](help/edge-consortium/sign-the-contributor-agreement.md) | [admin](admin/edge-consortium/sign-the-contributor-agreement.md) | [API](api/edge-consortium/sign-the-contributor-agreement.md) | [journeys](../generated/diagrams/edge-consortium/sign-the-contributor-agreement.md) | | Sign the employment contract | [help](help/edge-consortium/sign-the-employment-contract.md) | [admin](admin/edge-consortium/sign-the-employment-contract.md) | [API](api/edge-consortium/sign-the-employment-contract.md) | [journeys](../generated/diagrams/edge-consortium/sign-the-employment-contract.md) | | Sign the public contract | [help](help/edge-consortium/sign-the-public-contract.md) | [admin](admin/edge-consortium/sign-the-public-contract.md) | [API](api/edge-consortium/sign-the-public-contract.md) | [journeys](../generated/diagrams/edge-consortium/sign-the-public-contract.md) | | Stop or kill a job | [help](help/edge-consortium/stop-or-kill-a-job.md) | [admin](admin/edge-consortium/stop-or-kill-a-job.md) | [API](api/edge-consortium/stop-or-kill-a-job.md) | [journeys](../generated/diagrams/edge-consortium/stop-or-kill-a-job.md) | | Store the signed contract in the Vault | [help](help/edge-consortium/store-the-signed-contract-in-the-vault.md) | [admin](admin/edge-consortium/store-the-signed-contract-in-the-vault.md) | [API](api/edge-consortium/store-the-signed-contract-in-the-vault.md) | [journeys](../generated/diagrams/edge-consortium/store-the-signed-contract-in-the-vault.md) | | From student to long-term contractor | [help](help/edge-consortium/student-to-contractor-pathway.md) | [admin](admin/edge-consortium/student-to-contractor-pathway.md) | [API](api/edge-consortium/student-to-contractor-pathway.md) | [journeys](../generated/diagrams/edge-consortium/student-to-contractor-pathway.md) | | Submit a bid | [help](help/edge-consortium/submit-a-bid.md) | [admin](admin/edge-consortium/submit-a-bid.md) | [API](api/edge-consortium/submit-a-bid.md) | [journeys](../generated/diagrams/edge-consortium/submit-a-bid.md) | | Submit a contribution for review | [help](help/edge-consortium/submit-a-contribution-for-review.md) | [admin](admin/edge-consortium/submit-a-contribution-for-review.md) | [API](api/edge-consortium/submit-a-contribution-for-review.md) | [journeys](../generated/diagrams/edge-consortium/submit-a-contribution-for-review.md) | | Submit a grant application | [help](help/edge-consortium/submit-a-grant-application.md) | [admin](admin/edge-consortium/submit-a-grant-application.md) | [API](api/edge-consortium/submit-a-grant-application.md) | [journeys](../generated/diagrams/edge-consortium/submit-a-grant-application.md) | | Submit a paper to Consortium review | [help](help/edge-consortium/submit-a-paper-to-consortium-review.md) | [admin](admin/edge-consortium/submit-a-paper-to-consortium-review.md) | [API](api/edge-consortium/submit-a-paper-to-consortium-review.md) | [journeys](../generated/diagrams/edge-consortium/submit-a-paper-to-consortium-review.md) | | Submit an annual report to the funder | [help](help/edge-consortium/submit-an-annual-report-to-the-funder.md) | [admin](admin/edge-consortium/submit-an-annual-report-to-the-funder.md) | [API](api/edge-consortium/submit-an-annual-report-to-the-funder.md) | [journeys](../generated/diagrams/edge-consortium/submit-an-annual-report-to-the-funder.md) | | Submit an invoice and track payment | [help](help/edge-consortium/submit-an-invoice-and-track-payment.md) | [admin](admin/edge-consortium/submit-an-invoice-and-track-payment.md) | [API](api/edge-consortium/submit-an-invoice-and-track-payment.md) | [journeys](../generated/diagrams/edge-consortium/submit-an-invoice-and-track-payment.md) | | Submit staff for security vetting | [help](help/edge-consortium/submit-staff-for-security-vetting.md) | [admin](admin/edge-consortium/submit-staff-for-security-vetting.md) | [API](api/edge-consortium/submit-staff-for-security-vetting.md) | [journeys](../generated/diagrams/edge-consortium/submit-staff-for-security-vetting.md) | | Suspend or end a membership | [help](help/edge-consortium/suspend-or-end-a-membership.md) | [admin](admin/edge-consortium/suspend-or-end-a-membership.md) | [API](api/edge-consortium/suspend-or-end-a-membership.md) | [journeys](../generated/diagrams/edge-consortium/suspend-or-end-a-membership.md) | | Terminate a contract | [help](help/edge-consortium/terminate-a-contract.md) | [admin](admin/edge-consortium/terminate-a-contract.md) | [API](api/edge-consortium/terminate-a-contract.md) | [journeys](../generated/diagrams/edge-consortium/terminate-a-contract.md) | | Terminate a contract and publish the notice | [help](help/edge-consortium/terminate-a-contract-and-publish-the-notice.md) | [admin](admin/edge-consortium/terminate-a-contract-and-publish-the-notice.md) | [API](api/edge-consortium/terminate-a-contract-and-publish-the-notice.md) | [journeys](../generated/diagrams/edge-consortium/terminate-a-contract-and-publish-the-notice.md) | | Trace network ingress and egress | [help](help/edge-consortium/trace-network-ingress-and-egress.md) | [admin](admin/edge-consortium/trace-network-ingress-and-egress.md) | [API](api/edge-consortium/trace-network-ingress-and-egress.md) | [journeys](../generated/diagrams/edge-consortium/trace-network-ingress-and-egress.md) | | Track research outputs for a REF return | [help](help/edge-consortium/track-research-outputs-for-a-ref-return.md) | [admin](admin/edge-consortium/track-research-outputs-for-a-ref-return.md) | [API](api/edge-consortium/track-research-outputs-for-a-ref-return.md) | [journeys](../generated/diagrams/edge-consortium/track-research-outputs-for-a-ref-return.md) | | Track royalties and revenue share | [help](help/edge-consortium/track-royalties-and-revenue-share.md) | [admin](admin/edge-consortium/track-royalties-and-revenue-share.md) | [API](api/edge-consortium/track-royalties-and-revenue-share.md) | [journeys](../generated/diagrams/edge-consortium/track-royalties-and-revenue-share.md) | | Triage a security vulnerability report | [help](help/edge-consortium/triage-a-security-vulnerability-report.md) | [admin](admin/edge-consortium/triage-a-security-vulnerability-report.md) | [API](api/edge-consortium/triage-a-security-vulnerability-report.md) | [journeys](../generated/diagrams/edge-consortium/triage-a-security-vulnerability-report.md) | | Triage the submissions queue | [help](help/edge-consortium/triage-the-submissions-queue.md) | [admin](admin/edge-consortium/triage-the-submissions-queue.md) | [API](api/edge-consortium/triage-the-submissions-queue.md) | [journeys](../generated/diagrams/edge-consortium/triage-the-submissions-queue.md) | | University membership and enrolment | [help](help/edge-consortium/university-enrolment.md) | [admin](admin/edge-consortium/university-enrolment.md) | [API](api/edge-consortium/university-enrolment.md) | [journeys](../generated/diagrams/edge-consortium/university-enrolment.md) | | Update supplier information | [help](help/edge-consortium/update-supplier-information.md) | [admin](admin/edge-consortium/update-supplier-information.md) | [API](api/edge-consortium/update-supplier-information.md) | [journeys](../generated/diagrams/edge-consortium/update-supplier-information.md) | | Use the Terminal | [help](help/edge-consortium/use-the-terminal.md) | [admin](admin/edge-consortium/use-the-terminal.md) | [API](api/edge-consortium/use-the-terminal.md) | [journeys](../generated/diagrams/edge-consortium/use-the-terminal.md) | | Vary a contract | [help](help/edge-consortium/vary-a-contract.md) | [admin](admin/edge-consortium/vary-a-contract.md) | [API](api/edge-consortium/vary-a-contract.md) | [journeys](../generated/diagrams/edge-consortium/vary-a-contract.md) | | Verify a fact before stating it | [help](help/edge-consortium/verify-a-fact-before-stating-it.md) | [admin](admin/edge-consortium/verify-a-fact-before-stating-it.md) | [API](api/edge-consortium/verify-a-fact-before-stating-it.md) | [journeys](../generated/diagrams/edge-consortium/verify-a-fact-before-stating-it.md) | | Verify a fact in the knowledge graph | [help](help/edge-consortium/verify-a-fact-in-the-knowledge-graph.md) | [admin](admin/edge-consortium/verify-a-fact-in-the-knowledge-graph.md) | [API](api/edge-consortium/verify-a-fact-in-the-knowledge-graph.md) | [journeys](../generated/diagrams/edge-consortium/verify-a-fact-in-the-knowledge-graph.md) | | Verify a person's identity | [help](help/edge-consortium/verify-a-person-s-identity.md) | [admin](admin/edge-consortium/verify-a-person-s-identity.md) | [API](api/edge-consortium/verify-a-person-s-identity.md) | [journeys](../generated/diagrams/edge-consortium/verify-a-person-s-identity.md) | | Verify a signature | [help](help/edge-consortium/verify-a-signature.md) | [admin](admin/edge-consortium/verify-a-signature.md) | [API](api/edge-consortium/verify-a-signature.md) | [journeys](../generated/diagrams/edge-consortium/verify-a-signature.md) | | Verify institutional accreditation status | [help](help/edge-consortium/verify-institutional-accreditation-status.md) | [admin](admin/edge-consortium/verify-institutional-accreditation-status.md) | [API](api/edge-consortium/verify-institutional-accreditation-status.md) | [journeys](../generated/diagrams/edge-consortium/verify-institutional-accreditation-status.md) | | Verify procurement notice publication | [help](help/edge-consortium/verify-procurement-notice-publication.md) | [admin](admin/edge-consortium/verify-procurement-notice-publication.md) | [API](api/edge-consortium/verify-procurement-notice-publication.md) | [journeys](../generated/diagrams/edge-consortium/verify-procurement-notice-publication.md) | | Verify the accessibility statement | [help](help/edge-consortium/verify-the-accessibility-statement.md) | [admin](admin/edge-consortium/verify-the-accessibility-statement.md) | [API](api/edge-consortium/verify-the-accessibility-statement.md) | [journeys](../generated/diagrams/edge-consortium/verify-the-accessibility-statement.md) | | Verify the audit chain offline | [help](help/edge-consortium/verify-the-audit-chain-offline.md) | [admin](admin/edge-consortium/verify-the-audit-chain-offline.md) | [API](api/edge-consortium/verify-the-audit-chain-offline.md) | [journeys](../generated/diagrams/edge-consortium/verify-the-audit-chain-offline.md) | | Version and roll back a creation | [help](help/edge-consortium/version-and-roll-back-a-creation.md) | [admin](admin/edge-consortium/version-and-roll-back-a-creation.md) | [API](api/edge-consortium/version-and-roll-back-a-creation.md) | [journeys](../generated/diagrams/edge-consortium/version-and-roll-back-a-creation.md) | | Vet a reviewer or supplier staff member | [help](help/edge-consortium/vet-a-reviewer-or-supplier-staff-member.md) | [admin](admin/edge-consortium/vet-a-reviewer-or-supplier-staff-member.md) | [API](api/edge-consortium/vet-a-reviewer-or-supplier-staff-member.md) | [journeys](../generated/diagrams/edge-consortium/vet-a-reviewer-or-supplier-staff-member.md) | | View the Diary by day, week, month | [help](help/edge-consortium/view-the-diary-by-day-week-month.md) | [admin](admin/edge-consortium/view-the-diary-by-day-week-month.md) | [API](api/edge-consortium/view-the-diary-by-day-week-month.md) | [journeys](../generated/diagrams/edge-consortium/view-the-diary-by-day-week-month.md) | | View the public reviewer track record | [help](help/edge-consortium/view-the-public-reviewer-track-record.md) | [admin](admin/edge-consortium/view-the-public-reviewer-track-record.md) | [API](api/edge-consortium/view-the-public-reviewer-track-record.md) | [journeys](../generated/diagrams/edge-consortium/view-the-public-reviewer-track-record.md) | | Vote on a poll | [help](help/edge-consortium/vote-on-a-poll.md) | [admin](admin/edge-consortium/vote-on-a-poll.md) | [API](api/edge-consortium/vote-on-a-poll.md) | [journeys](../generated/diagrams/edge-consortium/vote-on-a-poll.md) | | Write a data management plan | [help](help/edge-consortium/write-a-data-management-plan.md) | [admin](admin/edge-consortium/write-a-data-management-plan.md) | [API](api/edge-consortium/write-a-data-management-plan.md) | [journeys](../generated/diagrams/edge-consortium/write-a-data-management-plan.md) | | Write a paper in Markdowns with co-authors | [help](help/edge-consortium/write-a-paper-in-markdowns-with-co-authors.md) | [admin](admin/edge-consortium/write-a-paper-in-markdowns-with-co-authors.md) | [API](api/edge-consortium/write-a-paper-in-markdowns-with-co-authors.md) | [journeys](../generated/diagrams/edge-consortium/write-a-paper-in-markdowns-with-co-authors.md) | | Accept a cheque deposit | [help](help/finance/accept-a-cheque-deposit.md) | [admin](admin/finance/accept-a-cheque-deposit.md) | [API](api/finance/accept-a-cheque-deposit.md) | [journeys](../generated/diagrams/finance/accept-a-cheque-deposit.md) | | Access a read only evidence room | [help](help/finance/access-a-read-only-evidence-room.md) | [admin](admin/finance/access-a-read-only-evidence-room.md) | [API](api/finance/access-a-read-only-evidence-room.md) | [journeys](../generated/diagrams/finance/access-a-read-only-evidence-room.md) | | Account for agent resource use | [help](help/finance/account-for-agent-resource-use.md) | [admin](admin/finance/account-for-agent-resource-use.md) | [API](api/finance/account-for-agent-resource-use.md) | [journeys](../generated/diagrams/finance/account-for-agent-resource-use.md) | | Accrue deposit interest expense | [help](help/finance/accrue-deposit-interest-expense.md) | [admin](admin/finance/accrue-deposit-interest-expense.md) | [API](api/finance/accrue-deposit-interest-expense.md) | [journeys](../generated/diagrams/finance/accrue-deposit-interest-expense.md) | | Acknowledge a complaint | [help](help/finance/acknowledge-a-complaint.md) | [admin](admin/finance/acknowledge-a-complaint.md) | [API](api/finance/acknowledge-a-complaint.md) | [journeys](../generated/diagrams/finance/acknowledge-a-complaint.md) | | Act under a power of attorney | [help](help/finance/act-under-a-power-of-attorney.md) | [admin](admin/finance/act-under-a-power-of-attorney.md) | [API](api/finance/act-under-a-power-of-attorney.md) | [journeys](../generated/diagrams/finance/act-under-a-power-of-attorney.md) | | Action an erasure or rectification request | [help](help/finance/action-an-erasure-or-rectification-request.md) | [admin](admin/finance/action-an-erasure-or-rectification-request.md) | [API](api/finance/action-an-erasure-or-rectification-request.md) | [journeys](../generated/diagrams/finance/action-an-erasure-or-rectification-request.md) | | Activate a monitoring rule | [help](help/finance/activate-a-monitoring-rule.md) | [admin](admin/finance/activate-a-monitoring-rule.md) | [API](api/finance/activate-a-monitoring-rule.md) | [journeys](../generated/diagrams/finance/activate-a-monitoring-rule.md) | | Activate a proposed rate | [help](help/finance/activate-a-proposed-rate.md) | [admin](admin/finance/activate-a-proposed-rate.md) | [API](api/finance/activate-a-proposed-rate.md) | [journeys](../generated/diagrams/finance/activate-a-proposed-rate.md) | | Add a joint account holder | [help](help/finance/add-a-joint-account-holder.md) | [admin](admin/finance/add-a-joint-account-holder.md) | [API](api/finance/add-a-joint-account-holder.md) | [journeys](../generated/diagrams/finance/add-a-joint-account-holder.md) | | Add a payee with name check | [help](help/finance/add-a-payee-with-name-check.md) | [admin](admin/finance/add-a-payee-with-name-check.md) | [API](api/finance/add-a-payee-with-name-check.md) | [journeys](../generated/diagrams/finance/add-a-payee-with-name-check.md) | | Add or remove a signatory | [help](help/finance/add-or-remove-a-signatory.md) | [admin](admin/finance/add-or-remove-a-signatory.md) | [API](api/finance/add-or-remove-a-signatory.md) | [journeys](../generated/diagrams/finance/add-or-remove-a-signatory.md) | | Set and control an agent's mandate | [help](help/finance/agent-mandate.md) | [admin](admin/finance/agent-mandate.md) | [API](api/finance/agent-mandate.md) | [journeys](../generated/diagrams/finance/agent-mandate.md) | | Onboard an agent as a customer | [help](help/finance/agent-onboarding.md) | [admin](admin/finance/agent-onboarding.md) | [API](api/finance/agent-onboarding.md) | [journeys](../generated/diagrams/finance/agent-onboarding.md) | | Retire an agent | [help](help/finance/agent-retirement.md) | [admin](admin/finance/agent-retirement.md) | [API](api/finance/agent-retirement.md) | [journeys](../generated/diagrams/finance/agent-retirement.md) | | Agree a forbearance plan | [help](help/finance/agree-a-forbearance-plan.md) | [admin](admin/finance/agree-a-forbearance-plan.md) | [API](api/finance/agree-a-forbearance-plan.md) | [journeys](../generated/diagrams/finance/agree-a-forbearance-plan.md) | | Amend a policy mid term | [help](help/finance/amend-a-policy-mid-term.md) | [admin](admin/finance/amend-a-policy-mid-term.md) | [API](api/finance/amend-a-policy-mid-term.md) | [journeys](../generated/diagrams/finance/amend-a-policy-mid-term.md) | | Amend a repayment schedule | [help](help/finance/amend-a-repayment-schedule.md) | [admin](admin/finance/amend-a-repayment-schedule.md) | [API](api/finance/amend-a-repayment-schedule.md) | [journeys](../generated/diagrams/finance/amend-a-repayment-schedule.md) | | Amend or cancel a standing order | [help](help/finance/amend-or-cancel-a-standing-order.md) | [admin](admin/finance/amend-or-cancel-a-standing-order.md) | [API](api/finance/amend-or-cancel-a-standing-order.md) | [journeys](../generated/diagrams/finance/amend-or-cancel-a-standing-order.md) | | Analyse complaint root causes | [help](help/finance/analyse-complaint-root-causes.md) | [admin](admin/finance/analyse-complaint-root-causes.md) | [API](api/finance/analyse-complaint-root-causes.md) | [journeys](../generated/diagrams/finance/analyse-complaint-root-causes.md) | | Answer a customer question from records | [help](help/finance/answer-a-customer-question-from-records.md) | [admin](admin/finance/answer-a-customer-question-from-records.md) | [API](api/finance/answer-a-customer-question-from-records.md) | [journeys](../generated/diagrams/finance/answer-a-customer-question-from-records.md) | | Answer a subject access request | [help](help/finance/answer-a-subject-access-request.md) | [admin](admin/finance/answer-a-subject-access-request.md) | [API](api/finance/answer-a-subject-access-request.md) | [journeys](../generated/diagrams/finance/answer-a-subject-access-request.md) | | Apply a database migration | [help](help/finance/apply-a-database-migration.md) | [admin](admin/finance/apply-a-database-migration.md) | [API](api/finance/apply-a-database-migration.md) | [journeys](../generated/diagrams/finance/apply-a-database-migration.md) | | Apply a payment holiday | [help](help/finance/apply-a-payment-holiday.md) | [admin](admin/finance/apply-a-payment-holiday.md) | [API](api/finance/apply-a-payment-holiday.md) | [journeys](../generated/diagrams/finance/apply-a-payment-holiday.md) | | Apply enhanced due diligence | [help](help/finance/apply-enhanced-due-diligence.md) | [admin](admin/finance/apply-enhanced-due-diligence.md) | [API](api/finance/apply-enhanced-due-diligence.md) | [journeys](../generated/diagrams/finance/apply-enhanced-due-diligence.md) | | Apply for a business account | [help](help/finance/apply-for-a-business-account.md) | [admin](admin/finance/apply-for-a-business-account.md) | [API](api/finance/apply-for-a-business-account.md) | [journeys](../generated/diagrams/finance/apply-for-a-business-account.md) | | Apply for a business loan or facility | [help](help/finance/apply-for-a-business-loan-or-facility.md) | [admin](admin/finance/apply-for-a-business-loan-or-facility.md) | [API](api/finance/apply-for-a-business-loan-or-facility.md) | [journeys](../generated/diagrams/finance/apply-for-a-business-loan-or-facility.md) | | Apply for a personal loan | [help](help/finance/apply-for-a-personal-loan.md) | [admin](admin/finance/apply-for-a-personal-loan.md) | [API](api/finance/apply-for-a-personal-loan.md) | [journeys](../generated/diagrams/finance/apply-for-a-personal-loan.md) | | Apply for a loan | [help](help/finance/apply-for-loan.md) | [admin](admin/finance/apply-for-loan.md) | [API](api/finance/apply-for-loan.md) | [journeys](../generated/diagrams/finance/apply-for-loan.md) | | Apply reasonable adjustments | [help](help/finance/apply-reasonable-adjustments.md) | [admin](admin/finance/apply-reasonable-adjustments.md) | [API](api/finance/apply-reasonable-adjustments.md) | [journeys](../generated/diagrams/finance/apply-reasonable-adjustments.md) | | Apply retention and deletion | [help](help/finance/apply-retention-and-deletion.md) | [admin](admin/finance/apply-retention-and-deletion.md) | [API](api/finance/apply-retention-and-deletion.md) | [journeys](../generated/diagrams/finance/apply-retention-and-deletion.md) | | Approve a financial promotion | [help](help/finance/approve-a-financial-promotion.md) | [admin](admin/finance/approve-a-financial-promotion.md) | [API](api/finance/approve-a-financial-promotion.md) | [journeys](../generated/diagrams/finance/approve-a-financial-promotion.md) | | Approve a new integration or API consumer | [help](help/finance/approve-a-new-integration-or-api-consumer.md) | [admin](admin/finance/approve-a-new-integration-or-api-consumer.md) | [API](api/finance/approve-a-new-integration-or-api-consumer.md) | [journeys](../generated/diagrams/finance/approve-a-new-integration-or-api-consumer.md) | | Approve or decline an application | [help](help/finance/approve-or-decline-an-application.md) | [admin](admin/finance/approve-or-decline-an-application.md) | [API](api/finance/approve-or-decline-an-application.md) | [journeys](../generated/diagrams/finance/approve-or-decline-an-application.md) | | Approve or reject a payment | [help](help/finance/approve-or-reject-a-payment.md) | [admin](admin/finance/approve-or-reject-a-payment.md) | [API](api/finance/approve-or-reject-a-payment.md) | [journeys](../generated/diagrams/finance/approve-or-reject-a-payment.md) | | Assess a claim | [help](help/finance/assess-a-claim.md) | [admin](admin/finance/assess-a-claim.md) | [API](api/finance/assess-a-claim.md) | [journeys](../generated/diagrams/finance/assess-a-claim.md) | | Assess an outsourcing or third party | [help](help/finance/assess-an-outsourcing-or-third-party.md) | [admin](admin/finance/assess-an-outsourcing-or-third-party.md) | [API](api/finance/assess-an-outsourcing-or-third-party.md) | [journeys](../generated/diagrams/finance/assess-an-outsourcing-or-third-party.md) | | Assess income and expenditure | [help](help/finance/assess-income-and-expenditure.md) | [admin](admin/finance/assess-income-and-expenditure.md) | [API](api/finance/assess-income-and-expenditure.md) | [journeys](../generated/diagrams/finance/assess-income-and-expenditure.md) | | Assess source of funds and wealth | [help](help/finance/assess-source-of-funds-and-wealth.md) | [admin](admin/finance/assess-source-of-funds-and-wealth.md) | [API](api/finance/assess-source-of-funds-and-wealth.md) | [journeys](../generated/diagrams/finance/assess-source-of-funds-and-wealth.md) | | Audit partner usage | [help](help/finance/audit-partner-usage.md) | [admin](admin/finance/audit-partner-usage.md) | [API](api/finance/audit-partner-usage.md) | [journeys](../generated/diagrams/finance/audit-partner-usage.md) | | Authorise card transactions in real time | [help](help/finance/authorise-card-transactions-in-real-time.md) | [admin](admin/finance/authorise-card-transactions-in-real-time.md) | [API](api/finance/authorise-card-transactions-in-real-time.md) | [journeys](../generated/diagrams/finance/authorise-card-transactions-in-real-time.md) | | Back up the database | [help](help/finance/back-up-the-database.md) | [admin](admin/finance/back-up-the-database.md) | [API](api/finance/back-up-the-database.md) | [journeys](../generated/diagrams/finance/back-up-the-database.md) | | Bill a partner | [help](help/finance/bill-a-partner.md) | [admin](admin/finance/bill-a-partner.md) | [API](api/finance/bill-a-partner.md) | [journeys](../generated/diagrams/finance/bill-a-partner.md) | | Block a payment in flight | [help](help/finance/block-a-payment-in-flight.md) | [admin](admin/finance/block-a-payment-in-flight.md) | [API](api/finance/block-a-payment-in-flight.md) | [journeys](../generated/diagrams/finance/block-a-payment-in-flight.md) | | Book an appointment | [help](help/finance/book-an-appointment.md) | [admin](admin/finance/book-an-appointment.md) | [API](api/finance/book-an-appointment.md) | [journeys](../generated/diagrams/finance/book-an-appointment.md) | | Bulk payments | [help](help/finance/bulk-payments.md) | [admin](admin/finance/bulk-payments.md) | [API](api/finance/bulk-payments.md) | [journeys](../generated/diagrams/finance/bulk-payments.md) | | Business team roles, approvals and accountant access | [help](help/finance/business-roles-and-approvals.md) | [admin](admin/finance/business-roles-and-approvals.md) | [API](api/finance/business-roles-and-approvals.md) | [journeys](../generated/diagrams/finance/business-roles-and-approvals.md) | | Calculate and post interest accrual | [help](help/finance/calculate-and-post-interest-accrual.md) | [admin](admin/finance/calculate-and-post-interest-accrual.md) | [API](api/finance/calculate-and-post-interest-accrual.md) | [journeys](../generated/diagrams/finance/calculate-and-post-interest-accrual.md) | | Calculate the liquidity coverage ratio | [help](help/finance/calculate-the-liquidity-coverage-ratio.md) | [admin](admin/finance/calculate-the-liquidity-coverage-ratio.md) | [API](api/finance/calculate-the-liquidity-coverage-ratio.md) | [journeys](../generated/diagrams/finance/calculate-the-liquidity-coverage-ratio.md) | | Cancel a Direct Debit | [help](help/finance/cancel-a-direct-debit.md) | [admin](admin/finance/cancel-a-direct-debit.md) | [API](api/finance/cancel-a-direct-debit.md) | [journeys](../generated/diagrams/finance/cancel-a-direct-debit.md) | | Cancel a policy | [help](help/finance/cancel-a-policy.md) | [admin](admin/finance/cancel-a-policy.md) | [API](api/finance/cancel-a-policy.md) | [journeys](../generated/diagrams/finance/cancel-a-policy.md) | | Capital to outcomes ledger | [help](help/finance/capital-to-outcomes-ledger.md) | [admin](admin/finance/capital-to-outcomes-ledger.md) | [API](api/finance/capital-to-outcomes-ledger.md) | [journeys](../generated/diagrams/finance/capital-to-outcomes-ledger.md) | | Cards | [help](help/finance/cards.md) | [admin](admin/finance/cards.md) | [API](api/finance/cards.md) | [journeys](../generated/diagrams/finance/cards.md) | | Change contact details | [help](help/finance/change-contact-details.md) | [admin](admin/finance/change-contact-details.md) | [API](api/finance/change-contact-details.md) | [journeys](../generated/diagrams/finance/change-contact-details.md) | | Claim a Direct Debit indemnity refund | [help](help/finance/claim-a-direct-debit-indemnity-refund.md) | [admin](admin/finance/claim-a-direct-debit-indemnity-refund.md) | [API](api/finance/claim-a-direct-debit-indemnity-refund.md) | [journeys](../generated/diagrams/finance/claim-a-direct-debit-indemnity-refund.md) | | Close a business account | [help](help/finance/close-a-business-account.md) | [admin](admin/finance/close-a-business-account.md) | [API](api/finance/close-a-business-account.md) | [journeys](../generated/diagrams/finance/close-a-business-account.md) | | Close a period | [help](help/finance/close-a-period.md) | [admin](admin/finance/close-a-period.md) | [API](api/finance/close-a-period.md) | [journeys](../generated/diagrams/finance/close-a-period.md) | | Close an account | [help](help/finance/close-account.md) | [admin](admin/finance/close-account.md) | [API](api/finance/close-account.md) | [journeys](../generated/diagrams/finance/close-account.md) | | Close an account | [help](help/finance/close-an-account.md) | [admin](admin/finance/close-an-account.md) | [API](api/finance/close-an-account.md) | [journeys](../generated/diagrams/finance/close-an-account.md) | | Collect Direct Debits as an originator | [help](help/finance/collect-direct-debits-as-an-originator.md) | [admin](admin/finance/collect-direct-debits-as-an-originator.md) | [API](api/finance/collect-direct-debits-as-an-originator.md) | [journeys](../generated/diagrams/finance/collect-direct-debits-as-an-originator.md) | | Collect premiums | [help](help/finance/collect-premiums.md) | [admin](admin/finance/collect-premiums.md) | [API](api/finance/collect-premiums.md) | [journeys](../generated/diagrams/finance/collect-premiums.md) | | Compare cover options | [help](help/finance/compare-cover-options.md) | [admin](admin/finance/compare-cover-options.md) | [API](api/finance/compare-cover-options.md) | [journeys](../generated/diagrams/finance/compare-cover-options.md) | | Complete entity tax classification | [help](help/finance/complete-entity-tax-classification.md) | [admin](admin/finance/complete-entity-tax-classification.md) | [API](api/finance/complete-entity-tax-classification.md) | [journeys](../generated/diagrams/finance/complete-entity-tax-classification.md) | | Complete tax residency self certification | [help](help/finance/complete-tax-residency-self-certification.md) | [admin](admin/finance/complete-tax-residency-self-certification.md) | [API](api/finance/complete-tax-residency-self-certification.md) | [journeys](../generated/diagrams/finance/complete-tax-residency-self-certification.md) | | Compute expected credit losses | [help](help/finance/compute-expected-credit-losses.md) | [admin](admin/finance/compute-expected-credit-losses.md) | [API](api/finance/compute-expected-credit-losses.md) | [journeys](../generated/diagrams/finance/compute-expected-credit-losses.md) | | Confidence report | [help](help/finance/confidence-report.md) | [admin](admin/finance/confidence-report.md) | [API](api/finance/confidence-report.md) | [journeys](../generated/diagrams/finance/confidence-report.md) | | Configure scheduled jobs | [help](help/finance/configure-scheduled-jobs.md) | [admin](admin/finance/configure-scheduled-jobs.md) | [API](api/finance/configure-scheduled-jobs.md) | [journeys](../generated/diagrams/finance/configure-scheduled-jobs.md) | | Confirm balances as external auditor | [help](help/finance/confirm-balances-as-external-auditor.md) | [admin](admin/finance/confirm-balances-as-external-auditor.md) | [API](api/finance/confirm-balances-as-external-auditor.md) | [journeys](../generated/diagrams/finance/confirm-balances-as-external-auditor.md) | | Connect accounts at other banks | [help](help/finance/connect-external-accounts.md) | [admin](admin/finance/connect-external-accounts.md) | [API](api/finance/connect-external-accounts.md) | [journeys](../generated/diagrams/finance/connect-external-accounts.md) | | Contact a customer in arrears | [help](help/finance/contact-a-customer-in-arrears.md) | [admin](admin/finance/contact-a-customer-in-arrears.md) | [API](api/finance/contact-a-customer-in-arrears.md) | [journeys](../generated/diagrams/finance/contact-a-customer-in-arrears.md) | | Continue a joint account after a death | [help](help/finance/continue-a-joint-account-after-a-death.md) | [admin](admin/finance/continue-a-joint-account-after-a-death.md) | [API](api/finance/continue-a-joint-account-after-a-death.md) | [journeys](../generated/diagrams/finance/continue-a-joint-account-after-a-death.md) | | Create a monitoring rule | [help](help/finance/create-a-monitoring-rule.md) | [admin](admin/finance/create-a-monitoring-rule.md) | [API](api/finance/create-a-monitoring-rule.md) | [journeys](../generated/diagrams/finance/create-a-monitoring-rule.md) | | Create a prospect record | [help](help/finance/create-a-prospect-record.md) | [admin](admin/finance/create-a-prospect-record.md) | [API](api/finance/create-a-prospect-record.md) | [journeys](../generated/diagrams/finance/create-a-prospect-record.md) | | Create a savings product | [help](help/finance/create-a-savings-product.md) | [admin](admin/finance/create-a-savings-product.md) | [API](api/finance/create-a-savings-product.md) | [journeys](../generated/diagrams/finance/create-a-savings-product.md) | | Create and release a hold | [help](help/finance/create-and-release-a-hold.md) | [admin](admin/finance/create-and-release-a-hold.md) | [API](api/finance/create-and-release-a-hold.md) | [journeys](../generated/diagrams/finance/create-and-release-a-hold.md) | | Crypto wallet | [help](help/finance/crypto-wallet.md) | [admin](admin/finance/crypto-wallet.md) | [API](api/finance/crypto-wallet.md) | [journeys](../generated/diagrams/finance/crypto-wallet.md) | | Decide a treasury proposal | [help](help/finance/decide-a-treasury-proposal.md) | [admin](admin/finance/decide-a-treasury-proposal.md) | [API](api/finance/decide-a-treasury-proposal.md) | [journeys](../generated/diagrams/finance/decide-a-treasury-proposal.md) | | Decide an APP reimbursement claim | [help](help/finance/decide-an-app-reimbursement-claim.md) | [admin](admin/finance/decide-an-app-reimbursement-claim.md) | [API](api/finance/decide-an-app-reimbursement-claim.md) | [journeys](../generated/diagrams/finance/decide-an-app-reimbursement-claim.md) | | Decide and file a SAR | [help](help/finance/decide-and-file-a-sar.md) | [admin](admin/finance/decide-and-file-a-sar.md) | [API](api/finance/decide-and-file-a-sar.md) | [journeys](../generated/diagrams/finance/decide-and-file-a-sar.md) | | Decline with reasons | [help](help/finance/decline-with-reasons.md) | [admin](admin/finance/decline-with-reasons.md) | [API](api/finance/decline-with-reasons.md) | [journeys](../generated/diagrams/finance/decline-with-reasons.md) | | Deploy a release as Jude | [help](help/finance/deploy-a-release-as-jude.md) | [admin](admin/finance/deploy-a-release-as-jude.md) | [API](api/finance/deploy-a-release-as-jude.md) | [journeys](../generated/diagrams/finance/deploy-a-release-as-jude.md) | | Deploy a release through deploy.py | [help](help/finance/deploy-a-release-through-deploy-py.md) | [admin](admin/finance/deploy-a-release-through-deploy-py.md) | [API](api/finance/deploy-a-release-through-deploy-py.md) | [journeys](../generated/diagrams/finance/deploy-a-release-through-deploy-py.md) | | Design a new lending product | [help](help/finance/design-a-new-lending-product.md) | [admin](admin/finance/design-a-new-lending-product.md) | [API](api/finance/design-a-new-lending-product.md) | [journeys](../generated/diagrams/finance/design-a-new-lending-product.md) | | Detect a missed repayment | [help](help/finance/detect-a-missed-repayment.md) | [admin](admin/finance/detect-a-missed-repayment.md) | [API](api/finance/detect-a-missed-repayment.md) | [journeys](../generated/diagrams/finance/detect-a-missed-repayment.md) | | Direct Debits | [help](help/finance/direct-debits.md) | [admin](admin/finance/direct-debits.md) | [API](api/finance/direct-debits.md) | [journeys](../generated/diagrams/finance/direct-debits.md) | | Disburse a loan | [help](help/finance/disburse-a-loan.md) | [admin](admin/finance/disburse-a-loan.md) | [API](api/finance/disburse-a-loan.md) | [journeys](../generated/diagrams/finance/disburse-a-loan.md) | | Disclose a vulnerability or need | [help](help/finance/disclose-a-vulnerability-or-need.md) | [admin](admin/finance/disclose-a-vulnerability-or-need.md) | [API](api/finance/disclose-a-vulnerability-or-need.md) | [journeys](../generated/diagrams/finance/disclose-a-vulnerability-or-need.md) | | Disposition an alert | [help](help/finance/disposition-an-alert.md) | [admin](admin/finance/disposition-an-alert.md) | [API](api/finance/disposition-an-alert.md) | [journeys](../generated/diagrams/finance/disposition-an-alert.md) | | Dispute a card transaction | [help](help/finance/dispute-a-card-transaction.md) | [admin](admin/finance/dispute-a-card-transaction.md) | [API](api/finance/dispute-a-card-transaction.md) | [journeys](../generated/diagrams/finance/dispute-a-card-transaction.md) | | Disputes | [help](help/finance/disputes.md) | [admin](admin/finance/disputes.md) | [API](api/finance/disputes.md) | [journeys](../generated/diagrams/finance/disputes.md) | | Draw down from a facility | [help](help/finance/draw-down-from-a-facility.md) | [admin](admin/finance/draw-down-from-a-facility.md) | [API](api/finance/draw-down-from-a-facility.md) | [journeys](../generated/diagrams/finance/draw-down-from-a-facility.md) | | Enforce strong customer authentication | [help](help/finance/enforce-strong-customer-authentication.md) | [admin](admin/finance/enforce-strong-customer-authentication.md) | [API](api/finance/enforce-strong-customer-authentication.md) | [journeys](../generated/diagrams/finance/enforce-strong-customer-authentication.md) | | Escalate a customer concern | [help](help/finance/escalate-a-customer-concern.md) | [admin](admin/finance/escalate-a-customer-concern.md) | [API](api/finance/escalate-a-customer-concern.md) | [journeys](../generated/diagrams/finance/escalate-a-customer-concern.md) | | Execute a money market placement | [help](help/finance/execute-a-money-market-placement.md) | [admin](admin/finance/execute-a-money-market-placement.md) | [API](api/finance/execute-a-money-market-placement.md) | [journeys](../generated/diagrams/finance/execute-a-money-market-placement.md) | | Exit a customer for financial crime | [help](help/finance/exit-a-customer-for-financial-crime.md) | [admin](admin/finance/exit-a-customer-for-financial-crime.md) | [API](api/finance/exit-a-customer-for-financial-crime.md) | [journeys](../generated/diagrams/finance/exit-a-customer-for-financial-crime.md) | | Explain a transaction | [help](help/finance/explain-a-transaction.md) | [admin](admin/finance/explain-a-transaction.md) | [API](api/finance/explain-a-transaction.md) | [journeys](../generated/diagrams/finance/explain-a-transaction.md) | | Export the evidence bundle | [help](help/finance/export-the-evidence-bundle.md) | [admin](admin/finance/export-the-evidence-bundle.md) | [API](api/finance/export-the-evidence-bundle.md) | [journeys](../generated/diagrams/finance/export-the-evidence-bundle.md) | | Export transactions to accounting software | [help](help/finance/export-transactions-to-accounting-software.md) | [admin](admin/finance/export-transactions-to-accounting-software.md) | [API](api/finance/export-transactions-to-accounting-software.md) | [journeys](../generated/diagrams/finance/export-transactions-to-accounting-software.md) | | Financing graph and independent demand | [help](help/finance/financing-graph-independent-demand.md) | [admin](admin/finance/financing-graph-independent-demand.md) | [API](api/finance/financing-graph-independent-demand.md) | [journeys](../generated/diagrams/finance/financing-graph-independent-demand.md) | | Flag a mule account | [help](help/finance/flag-a-mule-account.md) | [admin](admin/finance/flag-a-mule-account.md) | [API](api/finance/flag-a-mule-account.md) | [journeys](../generated/diagrams/finance/flag-a-mule-account.md) | | Fraud-sharing network | [help](help/finance/fraud-sharing-network.md) | [admin](admin/finance/fraud-sharing-network.md) | [API](api/finance/fraud-sharing-network.md) | [journeys](../generated/diagrams/finance/fraud-sharing-network.md) | | Freeze or unfreeze a card | [help](help/finance/freeze-or-unfreeze-a-card.md) | [admin](admin/finance/freeze-or-unfreeze-a-card.md) | [API](api/finance/freeze-or-unfreeze-a-card.md) | [journeys](../generated/diagrams/finance/freeze-or-unfreeze-a-card.md) | | Fund settlement accounts for rails | [help](help/finance/fund-settlement-accounts-for-rails.md) | [admin](admin/finance/fund-settlement-accounts-for-rails.md) | [API](api/finance/fund-settlement-accounts-for-rails.md) | [journeys](../generated/diagrams/finance/fund-settlement-accounts-for-rails.md) | | Give or revoke open banking consent | [help](help/finance/give-or-revoke-open-banking-consent.md) | [admin](admin/finance/give-or-revoke-open-banking-consent.md) | [API](api/finance/give-or-revoke-open-banking-consent.md) | [journeys](../generated/diagrams/finance/give-or-revoke-open-banking-consent.md) | | Grant an operator role | [help](help/finance/grant-an-operator-role.md) | [admin](admin/finance/grant-an-operator-role.md) | [API](api/finance/grant-an-operator-role.md) | [journeys](../generated/diagrams/finance/grant-an-operator-role.md) | | Hand a case to a specialist team | [help](help/finance/hand-a-case-to-a-specialist-team.md) | [admin](admin/finance/hand-a-case-to-a-specialist-team.md) | [API](api/finance/hand-a-case-to-a-specialist-team.md) | [journeys](../generated/diagrams/finance/hand-a-case-to-a-specialist-team.md) | | Hand work between agents | [help](help/finance/hand-work-between-agents.md) | [admin](admin/finance/hand-work-between-agents.md) | [API](api/finance/hand-work-between-agents.md) | [journeys](../generated/diagrams/finance/hand-work-between-agents.md) | | Handle a Banking Protocol concern | [help](help/finance/handle-a-banking-protocol-concern.md) | [admin](admin/finance/handle-a-banking-protocol-concern.md) | [API](api/finance/handle-a-banking-protocol-concern.md) | [journeys](../generated/diagrams/finance/handle-a-banking-protocol-concern.md) | | Handle a breathing space notice | [help](help/finance/handle-a-breathing-space-notice.md) | [admin](admin/finance/handle-a-breathing-space-notice.md) | [API](api/finance/handle-a-breathing-space-notice.md) | [journeys](../generated/diagrams/finance/handle-a-breathing-space-notice.md) | | Handle a debt sale | [help](help/finance/handle-a-debt-sale.md) | [admin](admin/finance/handle-a-debt-sale.md) | [API](api/finance/handle-a-debt-sale.md) | [journeys](../generated/diagrams/finance/handle-a-debt-sale.md) | | Handle a gambling block request | [help](help/finance/handle-a-gambling-block-request.md) | [admin](admin/finance/handle-a-gambling-block-request.md) | [API](api/finance/handle-a-gambling-block-request.md) | [journeys](../generated/diagrams/finance/handle-a-gambling-block-request.md) | | Handle a mental capacity concern | [help](help/finance/handle-a-mental-capacity-concern.md) | [admin](admin/finance/handle-a-mental-capacity-concern.md) | [API](api/finance/handle-a-mental-capacity-concern.md) | [journeys](../generated/diagrams/finance/handle-a-mental-capacity-concern.md) | | Handle a partner data request or deletion | [help](help/finance/handle-a-partner-data-request-or-deletion.md) | [admin](admin/finance/handle-a-partner-data-request-or-deletion.md) | [API](api/finance/handle-a-partner-data-request-or-deletion.md) | [journeys](../generated/diagrams/finance/handle-a-partner-data-request-or-deletion.md) | | Handle a security incident | [help](help/finance/handle-a-security-incident.md) | [admin](admin/finance/handle-a-security-incident.md) | [API](api/finance/handle-a-security-incident.md) | [journeys](../generated/diagrams/finance/handle-a-security-incident.md) | | Handle a trigger event review | [help](help/finance/handle-a-trigger-event-review.md) | [admin](admin/finance/handle-a-trigger-event-review.md) | [API](api/finance/handle-a-trigger-event-review.md) | [journeys](../generated/diagrams/finance/handle-a-trigger-event-review.md) | | Handle a whistleblowing report | [help](help/finance/handle-a-whistleblowing-report.md) | [admin](admin/finance/handle-a-whistleblowing-report.md) | [API](api/finance/handle-a-whistleblowing-report.md) | [journeys](../generated/diagrams/finance/handle-a-whistleblowing-report.md) | | Handle an indemnity claim as paying bank | [help](help/finance/handle-an-indemnity-claim-as-paying-bank.md) | [admin](admin/finance/handle-an-indemnity-claim-as-paying-bank.md) | [API](api/finance/handle-an-indemnity-claim-as-paying-bank.md) | [journeys](../generated/diagrams/finance/handle-an-indemnity-claim-as-paying-bank.md) | | Hedge interest rate exposure | [help](help/finance/hedge-interest-rate-exposure.md) | [admin](admin/finance/hedge-interest-rate-exposure.md) | [API](api/finance/hedge-interest-rate-exposure.md) | [journeys](../generated/diagrams/finance/hedge-interest-rate-exposure.md) | | Import a bank or rail statement | [help](help/finance/import-a-bank-or-rail-statement.md) | [admin](admin/finance/import-a-bank-or-rail-statement.md) | [API](api/finance/import-a-bank-or-rail-statement.md) | [journeys](../generated/diagrams/finance/import-a-bank-or-rail-statement.md) | | Inspect a SAR audit trail | [help](help/finance/inspect-a-sar-audit-trail.md) | [admin](admin/finance/inspect-a-sar-audit-trail.md) | [API](api/finance/inspect-a-sar-audit-trail.md) | [journeys](../generated/diagrams/finance/inspect-a-sar-audit-trail.md) | | Introduce a customer to a specialist | [help](help/finance/introduce-a-customer-to-a-specialist.md) | [admin](admin/finance/introduce-a-customer-to-a-specialist.md) | [API](api/finance/introduce-a-customer-to-a-specialist.md) | [journeys](../generated/diagrams/finance/introduce-a-customer-to-a-specialist.md) | | Investigate a complaint | [help](help/finance/investigate-a-complaint.md) | [admin](admin/finance/investigate-a-complaint.md) | [API](api/finance/investigate-a-complaint.md) | [journeys](../generated/diagrams/finance/investigate-a-complaint.md) | | Investigate a payment exception | [help](help/finance/investigate-a-payment-exception.md) | [admin](admin/finance/investigate-a-payment-exception.md) | [API](api/finance/investigate-a-payment-exception.md) | [journeys](../generated/diagrams/finance/investigate-a-payment-exception.md) | | Investigate a production error | [help](help/finance/investigate-a-production-error.md) | [admin](admin/finance/investigate-a-production-error.md) | [API](api/finance/investigate-a-production-error.md) | [journeys](../generated/diagrams/finance/investigate-a-production-error.md) | | Investigate a suspected fraud | [help](help/finance/investigate-a-suspected-fraud.md) | [admin](admin/finance/investigate-a-suspected-fraud.md) | [API](api/finance/investigate-a-suspected-fraud.md) | [journeys](../generated/diagrams/finance/investigate-a-suspected-fraud.md) | | Invite a user with a role | [help](help/finance/invite-a-user-with-a-role.md) | [admin](admin/finance/invite-a-user-with-a-role.md) | [API](api/finance/invite-a-user-with-a-role.md) | [journeys](../generated/diagrams/finance/invite-a-user-with-a-role.md) | | Issue a default notice | [help](help/finance/issue-a-default-notice.md) | [admin](admin/finance/issue-a-default-notice.md) | [API](api/finance/issue-a-default-notice.md) | [journeys](../generated/diagrams/finance/issue-a-default-notice.md) | | Issue a final response | [help](help/finance/issue-a-final-response.md) | [admin](admin/finance/issue-a-final-response.md) | [API](api/finance/issue-a-final-response.md) | [journeys](../generated/diagrams/finance/issue-a-final-response.md) | | Issue a finding to the bank | [help](help/finance/issue-a-finding-to-the-bank.md) | [admin](admin/finance/issue-a-finding-to-the-bank.md) | [API](api/finance/issue-a-finding-to-the-bank.md) | [journeys](../generated/diagrams/finance/issue-a-finding-to-the-bank.md) | | Issue a policy | [help](help/finance/issue-a-policy.md) | [admin](admin/finance/issue-a-policy.md) | [API](api/finance/issue-a-policy.md) | [journeys](../generated/diagrams/finance/issue-a-policy.md) | | Issue a replacement card in branch | [help](help/finance/issue-a-replacement-card-in-branch.md) | [admin](admin/finance/issue-a-replacement-card-in-branch.md) | [API](api/finance/issue-a-replacement-card-in-branch.md) | [journeys](../generated/diagrams/finance/issue-a-replacement-card-in-branch.md) | | Issue an agent mandate | [help](help/finance/issue-an-agent-mandate.md) | [admin](admin/finance/issue-an-agent-mandate.md) | [API](api/finance/issue-an-agent-mandate.md) | [journeys](../generated/diagrams/finance/issue-an-agent-mandate.md) | | Issue annual loan statements | [help](help/finance/issue-annual-loan-statements.md) | [admin](admin/finance/issue-annual-loan-statements.md) | [API](api/finance/issue-annual-loan-statements.md) | [journeys](../generated/diagrams/finance/issue-annual-loan-statements.md) | | Joint accounts | [help](help/finance/joint-accounts.md) | [admin](admin/finance/joint-accounts.md) | [API](api/finance/joint-accounts.md) | [journeys](../generated/diagrams/finance/joint-accounts.md) | | Lend | [help](help/finance/lend.md) | [admin](admin/finance/lend.md) | [API](api/finance/lend.md) | [journeys](../generated/diagrams/finance/lend.md) | | Lift an account freeze | [help](help/finance/lift-an-account-freeze.md) | [admin](admin/finance/lift-an-account-freeze.md) | [API](api/finance/lift-an-account-freeze.md) | [journeys](../generated/diagrams/finance/lift-an-account-freeze.md) | | Log an operational incident | [help](help/finance/log-an-operational-incident.md) | [admin](admin/finance/log-an-operational-incident.md) | [API](api/finance/log-an-operational-incident.md) | [journeys](../generated/diagrams/finance/log-an-operational-incident.md) | | Log an outbound contact | [help](help/finance/log-an-outbound-contact.md) | [admin](admin/finance/log-an-outbound-contact.md) | [API](api/finance/log-an-outbound-contact.md) | [journeys](../generated/diagrams/finance/log-an-outbound-contact.md) | | Maintain internal and nostro accounts | [help](help/finance/maintain-internal-and-nostro-accounts.md) | [admin](admin/finance/maintain-internal-and-nostro-accounts.md) | [API](api/finance/maintain-internal-and-nostro-accounts.md) | [journeys](../generated/diagrams/finance/maintain-internal-and-nostro-accounts.md) | | Maintain the business wide risk assessment | [help](help/finance/maintain-the-business-wide-risk-assessment.md) | [admin](admin/finance/maintain-the-business-wide-risk-assessment.md) | [API](api/finance/maintain-the-business-wide-risk-assessment.md) | [journeys](../generated/diagrams/finance/maintain-the-business-wide-risk-assessment.md) | | Maintain the chart of accounts | [help](help/finance/maintain-the-chart-of-accounts.md) | [admin](admin/finance/maintain-the-chart-of-accounts.md) | [API](api/finance/maintain-the-chart-of-accounts.md) | [journeys](../generated/diagrams/finance/maintain-the-chart-of-accounts.md) | | Maintain the record of processing | [help](help/finance/maintain-the-record-of-processing.md) | [admin](admin/finance/maintain-the-record-of-processing.md) | [API](api/finance/maintain-the-record-of-processing.md) | [journeys](../generated/diagrams/finance/maintain-the-record-of-processing.md) | | Maintain the regulatory obligations register | [help](help/finance/maintain-the-regulatory-obligations-register.md) | [admin](admin/finance/maintain-the-regulatory-obligations-register.md) | [API](api/finance/maintain-the-regulatory-obligations-register.md) | [journeys](../generated/diagrams/finance/maintain-the-regulatory-obligations-register.md) | | Maintain the risk register | [help](help/finance/maintain-the-risk-register.md) | [admin](admin/finance/maintain-the-risk-register.md) | [API](api/finance/maintain-the-risk-register.md) | [journeys](../generated/diagrams/finance/maintain-the-risk-register.md) | | Make a complaint | [help](help/finance/make-a-complaint.md) | [admin](admin/finance/make-a-complaint.md) | [API](api/finance/make-a-complaint.md) | [journeys](../generated/diagrams/finance/make-a-complaint.md) | | Make a Faster Payment | [help](help/finance/make-a-faster-payment.md) | [admin](admin/finance/make-a-faster-payment.md) | [API](api/finance/make-a-faster-payment.md) | [journeys](../generated/diagrams/finance/make-a-faster-payment.md) | | Make a regulatory notification | [help](help/finance/make-a-regulatory-notification.md) | [admin](admin/finance/make-a-regulatory-notification.md) | [API](api/finance/make-a-regulatory-notification.md) | [journeys](../generated/diagrams/finance/make-a-regulatory-notification.md) | | Make an internal transfer | [help](help/finance/make-an-internal-transfer.md) | [admin](admin/finance/make-an-internal-transfer.md) | [API](api/finance/make-an-internal-transfer.md) | [journeys](../generated/diagrams/finance/make-an-internal-transfer.md) | | Make an international payment | [help](help/finance/make-an-international-payment.md) | [admin](admin/finance/make-an-international-payment.md) | [API](api/finance/make-an-international-payment.md) | [journeys](../generated/diagrams/finance/make-an-international-payment.md) | | Manage a syndicated participation | [help](help/finance/manage-a-syndicated-participation.md) | [admin](admin/finance/manage-a-syndicated-participation.md) | [API](api/finance/manage-a-syndicated-participation.md) | [journeys](../generated/diagrams/finance/manage-a-syndicated-participation.md) | | Manage capacity | [help](help/finance/manage-capacity.md) | [admin](admin/finance/manage-capacity.md) | [API](api/finance/manage-capacity.md) | [journeys](../generated/diagrams/finance/manage-capacity.md) | | Manage encryption and device binding | [help](help/finance/manage-encryption-and-device-binding.md) | [admin](admin/finance/manage-encryption-and-device-binding.md) | [API](api/finance/manage-encryption-and-device-binding.md) | [journeys](../generated/diagrams/finance/manage-encryption-and-device-binding.md) | | Manage environments and secrets free config | [help](help/finance/manage-environments-and-secrets-free-config.md) | [admin](admin/finance/manage-environments-and-secrets-free-config.md) | [API](api/finance/manage-environments-and-secrets-free-config.md) | [journeys](../generated/diagrams/finance/manage-environments-and-secrets-free-config.md) | | Manage foreign exchange positions | [help](help/finance/manage-foreign-exchange-positions.md) | [admin](admin/finance/manage-foreign-exchange-positions.md) | [API](api/finance/manage-foreign-exchange-positions.md) | [journeys](../generated/diagrams/finance/manage-foreign-exchange-positions.md) | | Manage partner user access | [help](help/finance/manage-partner-user-access.md) | [admin](admin/finance/manage-partner-user-access.md) | [API](api/finance/manage-partner-user-access.md) | [journeys](../generated/diagrams/finance/manage-partner-user-access.md) | | Manage payee name check responses | [help](help/finance/manage-payee-name-check-responses.md) | [admin](admin/finance/manage-payee-name-check-responses.md) | [API](api/finance/manage-payee-name-check-responses.md) | [journeys](../generated/diagrams/finance/manage-payee-name-check-responses.md) | | Manage payment limits | [help](help/finance/manage-payment-limits.md) | [admin](admin/finance/manage-payment-limits.md) | [API](api/finance/manage-payment-limits.md) | [journeys](../generated/diagrams/finance/manage-payment-limits.md) | | Manage senior manager responsibilities | [help](help/finance/manage-senior-manager-responsibilities.md) | [admin](admin/finance/manage-senior-manager-responsibilities.md) | [API](api/finance/manage-senior-manager-responsibilities.md) | [journeys](../generated/diagrams/finance/manage-senior-manager-responsibilities.md) | | Manage the central bank reserve account | [help](help/finance/manage-the-central-bank-reserve-account.md) | [admin](admin/finance/manage-the-central-bank-reserve-account.md) | [API](api/finance/manage-the-central-bank-reserve-account.md) | [journeys](../generated/diagrams/finance/manage-the-central-bank-reserve-account.md) | | Map important business services | [help](help/finance/map-important-business-services.md) | [admin](admin/finance/map-important-business-services.md) | [API](api/finance/map-important-business-services.md) | [journeys](../generated/diagrams/finance/map-important-business-services.md) | | Mark a customer as a fraud victim | [help](help/finance/mark-a-customer-as-a-fraud-victim.md) | [admin](admin/finance/mark-a-customer-as-a-fraud-victim.md) | [API](api/finance/mark-a-customer-as-a-fraud-victim.md) | [journeys](../generated/diagrams/finance/mark-a-customer-as-a-fraud-victim.md) | | Match statement rows to ledger legs | [help](help/finance/match-statement-rows-to-ledger-legs.md) | [admin](admin/finance/match-statement-rows-to-ledger-legs.md) | [API](api/finance/match-statement-rows-to-ledger-legs.md) | [journeys](../generated/diagrams/finance/match-statement-rows-to-ledger-legs.md) | | Measure carbon, tokens and time to evidence | [help](help/finance/measure-carbon-tokens-and-time-to-evidence.md) | [admin](admin/finance/measure-carbon-tokens-and-time-to-evidence.md) | [API](api/finance/measure-carbon-tokens-and-time-to-evidence.md) | [journeys](../generated/diagrams/finance/measure-carbon-tokens-and-time-to-evidence.md) | | Message a customer in a case | [help](help/finance/message-a-customer-in-a-case.md) | [admin](admin/finance/message-a-customer-in-a-case.md) | [API](api/finance/message-a-customer-in-a-case.md) | [journeys](../generated/diagrams/finance/message-a-customer-in-a-case.md) | | Money provenance certificate | [help](help/finance/money-provenance-certificate.md) | [admin](admin/finance/money-provenance-certificate.md) | [API](api/finance/money-provenance-certificate.md) | [journeys](../generated/diagrams/finance/money-provenance-certificate.md) | | Monitor CHAPS cut-offs and queue | [help](help/finance/monitor-chaps-cut-offs-and-queue.md) | [admin](admin/finance/monitor-chaps-cut-offs-and-queue.md) | [API](api/finance/monitor-chaps-cut-offs-and-queue.md) | [journeys](../generated/diagrams/finance/monitor-chaps-cut-offs-and-queue.md) | | Monitor conduct and quality | [help](help/finance/monitor-conduct-and-quality.md) | [admin](admin/finance/monitor-conduct-and-quality.md) | [API](api/finance/monitor-conduct-and-quality.md) | [journeys](../generated/diagrams/finance/monitor-conduct-and-quality.md) | | Monitor facility covenants | [help](help/finance/monitor-facility-covenants.md) | [admin](admin/finance/monitor-facility-covenants.md) | [API](api/finance/monitor-facility-covenants.md) | [journeys](../generated/diagrams/finance/monitor-facility-covenants.md) | | Monitor intraday liquidity | [help](help/finance/monitor-intraday-liquidity.md) | [admin](admin/finance/monitor-intraday-liquidity.md) | [API](api/finance/monitor-intraday-liquidity.md) | [journeys](../generated/diagrams/finance/monitor-intraday-liquidity.md) | | Monitor job runs and leases | [help](help/finance/monitor-job-runs-and-leases.md) | [admin](admin/finance/monitor-job-runs-and-leases.md) | [API](api/finance/monitor-job-runs-and-leases.md) | [journeys](../generated/diagrams/finance/monitor-job-runs-and-leases.md) | | Monitor privileged actions | [help](help/finance/monitor-privileged-actions.md) | [admin](admin/finance/monitor-privileged-actions.md) | [API](api/finance/monitor-privileged-actions.md) | [journeys](../generated/diagrams/finance/monitor-privileged-actions.md) | | Monitor product outcomes | [help](help/finance/monitor-product-outcomes.md) | [admin](admin/finance/monitor-product-outcomes.md) | [API](api/finance/monitor-product-outcomes.md) | [journeys](../generated/diagrams/finance/monitor-product-outcomes.md) | | Monitor service health | [help](help/finance/monitor-service-health.md) | [admin](admin/finance/monitor-service-health.md) | [API](api/finance/monitor-service-health.md) | [journeys](../generated/diagrams/finance/monitor-service-health.md) | | Monitor without blocking | [help](help/finance/monitor-without-blocking.md) | [admin](admin/finance/monitor-without-blocking.md) | [API](api/finance/monitor-without-blocking.md) | [journeys](../generated/diagrams/finance/monitor-without-blocking.md) | | Notify a claim | [help](help/finance/notify-a-claim.md) | [admin](admin/finance/notify-a-claim.md) | [API](api/finance/notify-a-claim.md) | [journeys](../generated/diagrams/finance/notify-a-claim.md) | | Notify a death and restrict accounts | [help](help/finance/notify-a-death-and-restrict-accounts.md) | [admin](admin/finance/notify-a-death-and-restrict-accounts.md) | [API](api/finance/notify-a-death-and-restrict-accounts.md) | [journeys](../generated/diagrams/finance/notify-a-death-and-restrict-accounts.md) | | Notify a personal data breach | [help](help/finance/notify-a-personal-data-breach.md) | [admin](admin/finance/notify-a-personal-data-breach.md) | [API](api/finance/notify-a-personal-data-breach.md) | [journeys](../generated/diagrams/finance/notify-a-personal-data-breach.md) | | Offboard a business customer | [help](help/finance/offboard-a-business-customer.md) | [admin](admin/finance/offboard-a-business-customer.md) | [API](api/finance/offboard-a-business-customer.md) | [journeys](../generated/diagrams/finance/offboard-a-business-customer.md) | | Onboard a partner organisation | [help](help/finance/onboard-a-partner-organisation.md) | [admin](admin/finance/onboard-a-partner-organisation.md) | [API](api/finance/onboard-a-partner-organisation.md) | [journeys](../generated/diagrams/finance/onboard-a-partner-organisation.md) | | Open a fixed term deposit | [help](help/finance/open-a-fixed-term-deposit.md) | [admin](admin/finance/open-a-fixed-term-deposit.md) | [API](api/finance/open-a-fixed-term-deposit.md) | [journeys](../generated/diagrams/finance/open-a-fixed-term-deposit.md) | | Open a foreign currency account | [help](help/finance/open-a-foreign-currency-account.md) | [admin](admin/finance/open-a-foreign-currency-account.md) | [API](api/finance/open-a-foreign-currency-account.md) | [journeys](../generated/diagrams/finance/open-a-foreign-currency-account.md) | | Open a personal current account | [help](help/finance/open-a-personal-current-account.md) | [admin](admin/finance/open-a-personal-current-account.md) | [API](api/finance/open-a-personal-current-account.md) | [journeys](../generated/diagrams/finance/open-a-personal-current-account.md) | | Open a savings account | [help](help/finance/open-a-savings-account.md) | [admin](admin/finance/open-a-savings-account.md) | [API](api/finance/open-a-savings-account.md) | [journeys](../generated/diagrams/finance/open-a-savings-account.md) | | Open a service case | [help](help/finance/open-a-service-case.md) | [admin](admin/finance/open-a-service-case.md) | [API](api/finance/open-a-service-case.md) | [journeys](../generated/diagrams/finance/open-a-service-case.md) | | Open a transaction receipt | [help](help/finance/open-a-transaction-receipt.md) | [admin](admin/finance/open-a-transaction-receipt.md) | [API](api/finance/open-a-transaction-receipt.md) | [journeys](../generated/diagrams/finance/open-a-transaction-receipt.md) | | Open a business account | [help](help/finance/open-business-account.md) | [admin](admin/finance/open-business-account.md) | [API](api/finance/open-business-account.md) | [journeys](../generated/diagrams/finance/open-business-account.md) | | Open a personal account | [help](help/finance/open-personal-account.md) | [admin](admin/finance/open-personal-account.md) | [API](api/finance/open-personal-account.md) | [journeys](../generated/diagrams/finance/open-personal-account.md) | | Open virtual accounts under a master | [help](help/finance/open-virtual-accounts-under-a-master.md) | [admin](admin/finance/open-virtual-accounts-under-a-master.md) | [API](api/finance/open-virtual-accounts-under-a-master.md) | [journeys](../generated/diagrams/finance/open-virtual-accounts-under-a-master.md) | | Order business cards for staff | [help](help/finance/order-business-cards-for-staff.md) | [admin](admin/finance/order-business-cards-for-staff.md) | [API](api/finance/order-business-cards-for-staff.md) | [journeys](../generated/diagrams/finance/order-business-cards-for-staff.md) | | Overdraft | [help](help/finance/overdraft.md) | [admin](admin/finance/overdraft.md) | [API](api/finance/overdraft.md) | [journeys](../generated/diagrams/finance/overdraft.md) | | Pay redress | [help](help/finance/pay-redress.md) | [admin](admin/finance/pay-redress.md) | [API](api/finance/pay-redress.md) | [journeys](../generated/diagrams/finance/pay-redress.md) | | Perform a data protection impact assessment | [help](help/finance/perform-a-data-protection-impact-assessment.md) | [admin](admin/finance/perform-a-data-protection-impact-assessment.md) | [API](api/finance/perform-a-data-protection-impact-assessment.md) | [journeys](../generated/diagrams/finance/perform-a-data-protection-impact-assessment.md) | | Perform an internal audit | [help](help/finance/perform-an-internal-audit.md) | [admin](admin/finance/perform-an-internal-audit.md) | [API](api/finance/perform-an-internal-audit.md) | [journeys](../generated/diagrams/finance/perform-an-internal-audit.md) | | Place a temporary account freeze | [help](help/finance/place-a-temporary-account-freeze.md) | [admin](admin/finance/place-a-temporary-account-freeze.md) | [API](api/finance/place-a-temporary-account-freeze.md) | [journeys](../generated/diagrams/finance/place-a-temporary-account-freeze.md) | | Plan a portfolio review | [help](help/finance/plan-a-portfolio-review.md) | [admin](admin/finance/plan-a-portfolio-review.md) | [API](api/finance/plan-a-portfolio-review.md) | [journeys](../generated/diagrams/finance/plan-a-portfolio-review.md) | | Plan the internal audit year | [help](help/finance/plan-the-internal-audit-year.md) | [admin](admin/finance/plan-the-internal-audit-year.md) | [API](api/finance/plan-the-internal-audit-year.md) | [journeys](../generated/diagrams/finance/plan-the-internal-audit-year.md) | | Post a manual journal | [help](help/finance/post-a-manual-journal.md) | [admin](admin/finance/post-a-manual-journal.md) | [API](api/finance/post-a-manual-journal.md) | [journeys](../generated/diagrams/finance/post-a-manual-journal.md) | | Prepare a business proposition | [help](help/finance/prepare-a-business-proposition.md) | [admin](admin/finance/prepare-a-business-proposition.md) | [API](api/finance/prepare-a-business-proposition.md) | [journeys](../generated/diagrams/finance/prepare-a-business-proposition.md) | | Prepare a compliance return | [help](help/finance/prepare-a-compliance-return.md) | [admin](admin/finance/prepare-a-compliance-return.md) | [API](api/finance/prepare-a-compliance-return.md) | [journeys](../generated/diagrams/finance/prepare-a-compliance-return.md) | | Prepare a payment for approval | [help](help/finance/prepare-a-payment-for-approval.md) | [admin](admin/finance/prepare-a-payment-for-approval.md) | [API](api/finance/prepare-a-payment-for-approval.md) | [journeys](../generated/diagrams/finance/prepare-a-payment-for-approval.md) | | Prepare a payment for human approval | [help](help/finance/prepare-a-payment-for-human-approval.md) | [admin](admin/finance/prepare-a-payment-for-human-approval.md) | [API](api/finance/prepare-a-payment-for-human-approval.md) | [journeys](../generated/diagrams/finance/prepare-a-payment-for-human-approval.md) | | Prepare funds transfer pricing | [help](help/finance/prepare-funds-transfer-pricing.md) | [admin](admin/finance/prepare-funds-transfer-pricing.md) | [API](api/finance/prepare-funds-transfer-pricing.md) | [journeys](../generated/diagrams/finance/prepare-funds-transfer-pricing.md) | | Prepare regulatory returns | [help](help/finance/prepare-regulatory-returns.md) | [admin](admin/finance/prepare-regulatory-returns.md) | [API](api/finance/prepare-regulatory-returns.md) | [journeys](../generated/diagrams/finance/prepare-regulatory-returns.md) | | Prepare statutory accounts | [help](help/finance/prepare-statutory-accounts.md) | [admin](admin/finance/prepare-statutory-accounts.md) | [API](api/finance/prepare-statutory-accounts.md) | [journeys](../generated/diagrams/finance/prepare-statutory-accounts.md) | | Process a Bacs input and output cycle | [help](help/finance/process-a-bacs-input-and-output-cycle.md) | [admin](admin/finance/process-a-bacs-input-and-output-cycle.md) | [API](api/finance/process-a-bacs-input-and-output-cycle.md) | [journeys](../generated/diagrams/finance/process-a-bacs-input-and-output-cycle.md) | | Process a cheque clearing exception | [help](help/finance/process-a-cheque-clearing-exception.md) | [admin](admin/finance/process-a-cheque-clearing-exception.md) | [API](api/finance/process-a-cheque-clearing-exception.md) | [journeys](../generated/diagrams/finance/process-a-cheque-clearing-exception.md) | | Process a counter cash deposit | [help](help/finance/process-a-counter-cash-deposit.md) | [admin](admin/finance/process-a-counter-cash-deposit.md) | [API](api/finance/process-a-counter-cash-deposit.md) | [journeys](../generated/diagrams/finance/process-a-counter-cash-deposit.md) | | Process a counter cash withdrawal | [help](help/finance/process-a-counter-cash-withdrawal.md) | [admin](admin/finance/process-a-counter-cash-withdrawal.md) | [API](api/finance/process-a-counter-cash-withdrawal.md) | [journeys](../generated/diagrams/finance/process-a-counter-cash-withdrawal.md) | | Process a loan repayment manually | [help](help/finance/process-a-loan-repayment-manually.md) | [admin](admin/finance/process-a-loan-repayment-manually.md) | [API](api/finance/process-a-loan-repayment-manually.md) | [journeys](../generated/diagrams/finance/process-a-loan-repayment-manually.md) | | Process an inbound Faster Payment | [help](help/finance/process-an-inbound-faster-payment.md) | [admin](admin/finance/process-an-inbound-faster-payment.md) | [API](api/finance/process-an-inbound-faster-payment.md) | [journeys](../generated/diagrams/finance/process-an-inbound-faster-payment.md) | | Process card clearing and settlement | [help](help/finance/process-card-clearing-and-settlement.md) | [admin](admin/finance/process-card-clearing-and-settlement.md) | [API](api/finance/process-card-clearing-and-settlement.md) | [journeys](../generated/diagrams/finance/process-card-clearing-and-settlement.md) | | Produce a KYC file for review | [help](help/finance/produce-a-kyc-file-for-review.md) | [admin](admin/finance/produce-a-kyc-file-for-review.md) | [API](api/finance/produce-a-kyc-file-for-review.md) | [journeys](../generated/diagrams/finance/produce-a-kyc-file-for-review.md) | | Produce a trial balance | [help](help/finance/produce-a-trial-balance.md) | [admin](admin/finance/produce-a-trial-balance.md) | [API](api/finance/produce-a-trial-balance.md) | [journeys](../generated/diagrams/finance/produce-a-trial-balance.md) | | Produce the annual MLRO report | [help](help/finance/produce-the-annual-mlro-report.md) | [admin](admin/finance/produce-the-annual-mlro-report.md) | [API](api/finance/produce-the-annual-mlro-report.md) | [journeys](../generated/diagrams/finance/produce-the-annual-mlro-report.md) | | Produce the FSCS single customer view | [help](help/finance/produce-the-fscs-single-customer-view.md) | [admin](admin/finance/produce-the-fscs-single-customer-view.md) | [API](api/finance/produce-the-fscs-single-customer-view.md) | [journeys](../generated/diagrams/finance/produce-the-fscs-single-customer-view.md) | | Propose a future rate | [help](help/finance/propose-a-future-rate.md) | [admin](admin/finance/propose-a-future-rate.md) | [API](api/finance/propose-a-future-rate.md) | [journeys](../generated/diagrams/finance/propose-a-future-rate.md) | | Propose a treasury credit | [help](help/finance/propose-a-treasury-credit.md) | [admin](admin/finance/propose-a-treasury-credit.md) | [API](api/finance/propose-a-treasury-credit.md) | [journeys](../generated/diagrams/finance/propose-a-treasury-credit.md) | | Provenance coverage report | [help](help/finance/provenance-coverage-report.md) | [admin](admin/finance/provenance-coverage-report.md) | [API](api/finance/provenance-coverage-report.md) | [journeys](../generated/diagrams/finance/provenance-coverage-report.md) | | Provide extra evidence on request | [help](help/finance/provide-extra-evidence-on-request.md) | [admin](admin/finance/provide-extra-evidence-on-request.md) | [API](api/finance/provide-extra-evidence-on-request.md) | [journeys](../generated/diagrams/finance/provide-extra-evidence-on-request.md) | | Verify identity in person | [help](help/finance/proximity-verification.md) | [admin](admin/finance/proximity-verification.md) | [API](api/finance/proximity-verification.md) | [journeys](../generated/diagrams/finance/proximity-verification.md) | | Publish terms and conditions | [help](help/finance/publish-terms-and-conditions.md) | [admin](admin/finance/publish-terms-and-conditions.md) | [API](api/finance/publish-terms-and-conditions.md) | [journeys](../generated/diagrams/finance/publish-terms-and-conditions.md) | | Pull and record a credit reference | [help](help/finance/pull-and-record-a-credit-reference.md) | [admin](admin/finance/pull-and-record-a-credit-reference.md) | [API](api/finance/pull-and-record-a-credit-reference.md) | [journeys](../generated/diagrams/finance/pull-and-record-a-credit-reference.md) | | Query a mule account indicator | [help](help/finance/query-a-mule-account-indicator.md) | [admin](admin/finance/query-a-mule-account-indicator.md) | [API](api/finance/query-a-mule-account-indicator.md) | [journeys](../generated/diagrams/finance/query-a-mule-account-indicator.md) | | Raise an internal suspicion report | [help](help/finance/raise-an-internal-suspicion-report.md) | [admin](admin/finance/raise-an-internal-suspicion-report.md) | [API](api/finance/raise-an-internal-suspicion-report.md) | [journeys](../generated/diagrams/finance/raise-an-internal-suspicion-report.md) | | Receive a fraud intelligence signal | [help](help/finance/receive-a-fraud-intelligence-signal.md) | [admin](admin/finance/receive-a-fraud-intelligence-signal.md) | [API](api/finance/receive-a-fraud-intelligence-signal.md) | [journeys](../generated/diagrams/finance/receive-a-fraud-intelligence-signal.md) | | Receive a rate change notice | [help](help/finance/receive-a-rate-change-notice.md) | [admin](admin/finance/receive-a-rate-change-notice.md) | [API](api/finance/receive-a-rate-change-notice.md) | [journeys](../generated/diagrams/finance/receive-a-rate-change-notice.md) | | Recognise loan interest income | [help](help/finance/recognise-loan-interest-income.md) | [admin](admin/finance/recognise-loan-interest-income.md) | [API](api/finance/recognise-loan-interest-income.md) | [journeys](../generated/diagrams/finance/recognise-loan-interest-income.md) | | Reconcile rail settlement | [help](help/finance/reconcile-rail-settlement.md) | [admin](admin/finance/reconcile-rail-settlement.md) | [API](api/finance/reconcile-rail-settlement.md) | [journeys](../generated/diagrams/finance/reconcile-rail-settlement.md) | | Reconcile suspense accounts | [help](help/finance/reconcile-suspense-accounts.md) | [admin](admin/finance/reconcile-suspense-accounts.md) | [API](api/finance/reconcile-suspense-accounts.md) | [journeys](../generated/diagrams/finance/reconcile-suspense-accounts.md) | | Record a credit decision override | [help](help/finance/record-a-credit-decision-override.md) | [admin](admin/finance/record-a-credit-decision-override.md) | [API](api/finance/record-a-credit-decision-override.md) | [journeys](../generated/diagrams/finance/record-a-credit-decision-override.md) | | Record a customer meeting | [help](help/finance/record-a-customer-meeting.md) | [admin](admin/finance/record-a-customer-meeting.md) | [API](api/finance/record-a-customer-meeting.md) | [journeys](../generated/diagrams/finance/record-a-customer-meeting.md) | | Record a customer risk assessment | [help](help/finance/record-a-customer-risk-assessment.md) | [admin](admin/finance/record-a-customer-risk-assessment.md) | [API](api/finance/record-a-customer-risk-assessment.md) | [journeys](../generated/diagrams/finance/record-a-customer-risk-assessment.md) | | Record a declined application reason | [help](help/finance/record-a-declined-application-reason.md) | [admin](admin/finance/record-a-declined-application-reason.md) | [API](api/finance/record-a-declined-application-reason.md) | [journeys](../generated/diagrams/finance/record-a-declined-application-reason.md) | | Record a regulatory breach | [help](help/finance/record-a-regulatory-breach.md) | [admin](admin/finance/record-a-regulatory-breach.md) | [API](api/finance/record-a-regulatory-breach.md) | [journeys](../generated/diagrams/finance/record-a-regulatory-breach.md) | | Record a support need | [help](help/finance/record-a-support-need.md) | [admin](admin/finance/record-a-support-need.md) | [API](api/finance/record-a-support-need.md) | [journeys](../generated/diagrams/finance/record-a-support-need.md) | | Record a vulnerability disclosure | [help](help/finance/record-a-vulnerability-disclosure.md) | [admin](admin/finance/record-a-vulnerability-disclosure.md) | [API](api/finance/record-a-vulnerability-disclosure.md) | [journeys](../generated/diagrams/finance/record-a-vulnerability-disclosure.md) | | Record AML training completion | [help](help/finance/record-aml-training-completion.md) | [admin](admin/finance/record-aml-training-completion.md) | [API](api/finance/record-aml-training-completion.md) | [journeys](../generated/diagrams/finance/record-aml-training-completion.md) | | Record and test a control | [help](help/finance/record-and-test-a-control.md) | [admin](admin/finance/record-and-test-a-control.md) | [API](api/finance/record-and-test-a-control.md) | [journeys](../generated/diagrams/finance/record-and-test-a-control.md) | | Recover account access | [help](help/finance/recover-account-access.md) | [admin](admin/finance/recover-account-access.md) | [API](api/finance/recover-account-access.md) | [journeys](../generated/diagrams/finance/recover-account-access.md) | | Recover funds from a receiving account | [help](help/finance/recover-funds-from-a-receiving-account.md) | [admin](admin/finance/recover-funds-from-a-receiving-account.md) | [API](api/finance/recover-funds-from-a-receiving-account.md) | [journeys](../generated/diagrams/finance/recover-funds-from-a-receiving-account.md) | | Refer a complaint to FOS | [help](help/finance/refer-a-complaint-to-fos.md) | [admin](admin/finance/refer-a-complaint-to-fos.md) | [API](api/finance/refer-a-complaint-to-fos.md) | [journeys](../generated/diagrams/finance/refer-a-complaint-to-fos.md) | | Refer to a debt collection agency | [help](help/finance/refer-to-a-debt-collection-agency.md) | [admin](admin/finance/refer-to-a-debt-collection-agency.md) | [API](api/finance/refer-to-a-debt-collection-agency.md) | [journeys](../generated/diagrams/finance/refer-to-a-debt-collection-agency.md) | | Refer to debt advice | [help](help/finance/refer-to-debt-advice.md) | [admin](admin/finance/refer-to-debt-advice.md) | [API](api/finance/refer-to-debt-advice.md) | [journeys](../generated/diagrams/finance/refer-to-debt-advice.md) | | Refund an unauthorised transaction | [help](help/finance/refund-an-unauthorised-transaction.md) | [admin](admin/finance/refund-an-unauthorised-transaction.md) | [API](api/finance/refund-an-unauthorised-transaction.md) | [journeys](../generated/diagrams/finance/refund-an-unauthorised-transaction.md) | | Register a court appointed deputy | [help](help/finance/register-a-court-appointed-deputy.md) | [admin](admin/finance/register-a-court-appointed-deputy.md) | [API](api/finance/register-a-court-appointed-deputy.md) | [journeys](../generated/diagrams/finance/register-a-court-appointed-deputy.md) | | Register a default with agencies | [help](help/finance/register-a-default-with-agencies.md) | [admin](admin/finance/register-a-default-with-agencies.md) | [API](api/finance/register-a-default-with-agencies.md) | [journeys](../generated/diagrams/finance/register-a-default-with-agencies.md) | | Register a lasting power of attorney | [help](help/finance/register-a-lasting-power-of-attorney.md) | [admin](admin/finance/register-a-lasting-power-of-attorney.md) | [API](api/finance/register-a-lasting-power-of-attorney.md) | [journeys](../generated/diagrams/finance/register-a-lasting-power-of-attorney.md) | | Register a new device | [help](help/finance/register-a-new-device.md) | [admin](admin/finance/register-a-new-device.md) | [API](api/finance/register-a-new-device.md) | [journeys](../generated/diagrams/finance/register-a-new-device.md) | | Register a third party mandate | [help](help/finance/register-a-third-party-mandate.md) | [admin](admin/finance/register-a-third-party-mandate.md) | [API](api/finance/register-a-third-party-mandate.md) | [journeys](../generated/diagrams/finance/register-a-third-party-mandate.md) | | Register and value collateral | [help](help/finance/register-and-value-collateral.md) | [admin](admin/finance/register-and-value-collateral.md) | [API](api/finance/register-and-value-collateral.md) | [journeys](../generated/diagrams/finance/register-and-value-collateral.md) | | Register beneficial owners | [help](help/finance/register-beneficial-owners.md) | [admin](admin/finance/register-beneficial-owners.md) | [API](api/finance/register-beneficial-owners.md) | [journeys](../generated/diagrams/finance/register-beneficial-owners.md) | | Regulator inspection | [help](help/finance/regulator-inspection.md) | [admin](admin/finance/regulator-inspection.md) | [API](api/finance/regulator-inspection.md) | [journeys](../generated/diagrams/finance/regulator-inspection.md) | | Release estate funds | [help](help/finance/release-estate-funds.md) | [admin](admin/finance/release-estate-funds.md) | [API](api/finance/release-estate-funds.md) | [journeys](../generated/diagrams/finance/release-estate-funds.md) | | Rely on another firm's due diligence | [help](help/finance/rely-on-another-firm-s-due-diligence.md) | [admin](admin/finance/rely-on-another-firm-s-due-diligence.md) | [API](api/finance/rely-on-another-firm-s-due-diligence.md) | [journeys](../generated/diagrams/finance/rely-on-another-firm-s-due-diligence.md) | | Remove a joint holder | [help](help/finance/remove-a-joint-holder.md) | [admin](admin/finance/remove-a-joint-holder.md) | [API](api/finance/remove-a-joint-holder.md) | [journeys](../generated/diagrams/finance/remove-a-joint-holder.md) | | Rename an account | [help](help/finance/rename-an-account.md) | [admin](admin/finance/rename-an-account.md) | [API](api/finance/rename-an-account.md) | [journeys](../generated/diagrams/finance/rename-an-account.md) | | Renew a policy | [help](help/finance/renew-a-policy.md) | [admin](admin/finance/renew-a-policy.md) | [API](api/finance/renew-a-policy.md) | [journeys](../generated/diagrams/finance/renew-a-policy.md) | | Reopen a period | [help](help/finance/reopen-a-period.md) | [admin](admin/finance/reopen-a-period.md) | [API](api/finance/reopen-a-period.md) | [journeys](../generated/diagrams/finance/reopen-a-period.md) | | Repair a misdirected payment | [help](help/finance/repair-a-misdirected-payment.md) | [admin](admin/finance/repair-a-misdirected-payment.md) | [API](api/finance/repair-a-misdirected-payment.md) | [journeys](../generated/diagrams/finance/repair-a-misdirected-payment.md) | | Repay or overpay a loan | [help](help/finance/repay-or-overpay-a-loan.md) | [admin](admin/finance/repay-or-overpay-a-loan.md) | [API](api/finance/repay-or-overpay-a-loan.md) | [journeys](../generated/diagrams/finance/repay-or-overpay-a-loan.md) | | Report a bereavement | [help](help/finance/report-a-bereavement.md) | [admin](admin/finance/report-a-bereavement.md) | [API](api/finance/report-a-bereavement.md) | [journeys](../generated/diagrams/finance/report-a-bereavement.md) | | Report a change of company details | [help](help/finance/report-a-change-of-company-details.md) | [admin](admin/finance/report-a-change-of-company-details.md) | [API](api/finance/report-a-change-of-company-details.md) | [journeys](../generated/diagrams/finance/report-a-change-of-company-details.md) | | Report a lost or stolen card | [help](help/finance/report-a-lost-or-stolen-card.md) | [admin](admin/finance/report-a-lost-or-stolen-card.md) | [API](api/finance/report-a-lost-or-stolen-card.md) | [journeys](../generated/diagrams/finance/report-a-lost-or-stolen-card.md) | | Report an APP scam | [help](help/finance/report-an-app-scam.md) | [admin](admin/finance/report-an-app-scam.md) | [API](api/finance/report-an-app-scam.md) | [journeys](../generated/diagrams/finance/report-an-app-scam.md) | | Report an unauthorised payment | [help](help/finance/report-an-unauthorised-payment.md) | [admin](admin/finance/report-an-unauthorised-payment.md) | [API](api/finance/report-an-unauthorised-payment.md) | [journeys](../generated/diagrams/finance/report-an-unauthorised-payment.md) | | Report complaints data | [help](help/finance/report-complaints-data.md) | [admin](admin/finance/report-complaints-data.md) | [API](api/finance/report-complaints-data.md) | [journeys](../generated/diagrams/finance/report-complaints-data.md) | | Report fraud losses to the regulator | [help](help/finance/report-fraud-losses-to-the-regulator.md) | [admin](admin/finance/report-fraud-losses-to-the-regulator.md) | [API](api/finance/report-fraud-losses-to-the-regulator.md) | [journeys](../generated/diagrams/finance/report-fraud-losses-to-the-regulator.md) | | Report fraud to Action Fraud or police | [help](help/finance/report-fraud-to-action-fraud-or-police.md) | [admin](admin/finance/report-fraud-to-action-fraud-or-police.md) | [API](api/finance/report-fraud-to-action-fraud-or-police.md) | [journeys](../generated/diagrams/finance/report-fraud-to-action-fraud-or-police.md) | | Report insurance data to the regulator | [help](help/finance/report-insurance-data-to-the-regulator.md) | [admin](admin/finance/report-insurance-data-to-the-regulator.md) | [API](api/finance/report-insurance-data-to-the-regulator.md) | [journeys](../generated/diagrams/finance/report-insurance-data-to-the-regulator.md) | | Report payment statistics | [help](help/finance/report-payment-statistics.md) | [admin](admin/finance/report-payment-statistics.md) | [API](api/finance/report-payment-statistics.md) | [journeys](../generated/diagrams/finance/report-payment-statistics.md) | | Report to credit reference agencies | [help](help/finance/report-to-credit-reference-agencies.md) | [admin](admin/finance/report-to-credit-reference-agencies.md) | [API](api/finance/report-to-credit-reference-agencies.md) | [journeys](../generated/diagrams/finance/report-to-credit-reference-agencies.md) | | Request a bank statement letter | [help](help/finance/request-a-bank-statement-letter.md) | [admin](admin/finance/request-a-bank-statement-letter.md) | [API](api/finance/request-a-bank-statement-letter.md) | [journeys](../generated/diagrams/finance/request-a-bank-statement-letter.md) | | Request a company audit letter | [help](help/finance/request-a-company-audit-letter.md) | [admin](admin/finance/request-a-company-audit-letter.md) | [API](api/finance/request-a-company-audit-letter.md) | [journeys](../generated/diagrams/finance/request-a-company-audit-letter.md) | | Request a covenant waiver | [help](help/finance/request-a-covenant-waiver.md) | [admin](admin/finance/request-a-covenant-waiver.md) | [API](api/finance/request-a-covenant-waiver.md) | [journeys](../generated/diagrams/finance/request-a-covenant-waiver.md) | | Request a date of death balance letter | [help](help/finance/request-a-date-of-death-balance-letter.md) | [admin](admin/finance/request-a-date-of-death-balance-letter.md) | [API](api/finance/request-a-date-of-death-balance-letter.md) | [journeys](../generated/diagrams/finance/request-a-date-of-death-balance-letter.md) | | Request a defence against money laundering | [help](help/finance/request-a-defence-against-money-laundering.md) | [admin](admin/finance/request-a-defence-against-money-laundering.md) | [API](api/finance/request-a-defence-against-money-laundering.md) | [journeys](../generated/diagrams/finance/request-a-defence-against-money-laundering.md) | | Request a loan settlement quote | [help](help/finance/request-a-loan-settlement-quote.md) | [admin](admin/finance/request-a-loan-settlement-quote.md) | [API](api/finance/request-a-loan-settlement-quote.md) | [journeys](../generated/diagrams/finance/request-a-loan-settlement-quote.md) | | Request a trade finance instrument | [help](help/finance/request-a-trade-finance-instrument.md) | [admin](admin/finance/request-a-trade-finance-instrument.md) | [API](api/finance/request-a-trade-finance-instrument.md) | [journeys](../generated/diagrams/finance/request-a-trade-finance-instrument.md) | | Request affordability evidence | [help](help/finance/request-affordability-evidence.md) | [admin](admin/finance/request-affordability-evidence.md) | [API](api/finance/request-affordability-evidence.md) | [journeys](../generated/diagrams/finance/request-affordability-evidence.md) | | Request an arranged overdraft | [help](help/finance/request-an-arranged-overdraft.md) | [admin](admin/finance/request-an-arranged-overdraft.md) | [API](api/finance/request-an-arranged-overdraft.md) | [journeys](../generated/diagrams/finance/request-an-arranged-overdraft.md) | | Request an insurance quote from KYC | [help](help/finance/request-an-insurance-quote-from-kyc.md) | [admin](admin/finance/request-an-insurance-quote-from-kyc.md) | [API](api/finance/request-an-insurance-quote-from-kyc.md) | [journeys](../generated/diagrams/finance/request-an-insurance-quote-from-kyc.md) | | Request further information | [help](help/finance/request-further-information.md) | [admin](admin/finance/request-further-information.md) | [API](api/finance/request-further-information.md) | [journeys](../generated/diagrams/finance/request-further-information.md) | | Request money | [help](help/finance/request-money.md) | [admin](admin/finance/request-money.md) | [API](api/finance/request-money.md) | [journeys](../generated/diagrams/finance/request-money.md) | | Request my personal data | [help](help/finance/request-my-personal-data.md) | [admin](admin/finance/request-my-personal-data.md) | [API](api/finance/request-my-personal-data.md) | [journeys](../generated/diagrams/finance/request-my-personal-data.md) | | Request Owner approval for a consequential action | [help](help/finance/request-owner-approval-for-a-consequential-action.md) | [admin](admin/finance/request-owner-approval-for-a-consequential-action.md) | [API](api/finance/request-owner-approval-for-a-consequential-action.md) | [journeys](../generated/diagrams/finance/request-owner-approval-for-a-consequential-action.md) | | Request records from the bank | [help](help/finance/request-records-from-the-bank.md) | [admin](admin/finance/request-records-from-the-bank.md) | [API](api/finance/request-records-from-the-bank.md) | [journeys](../generated/diagrams/finance/request-records-from-the-bank.md) | | Reset access for a customer | [help](help/finance/reset-access-for-a-customer.md) | [admin](admin/finance/reset-access-for-a-customer.md) | [API](api/finance/reset-access-for-a-customer.md) | [journeys](../generated/diagrams/finance/reset-access-for-a-customer.md) | | Resolve within 3 business days | [help](help/finance/resolve-within-3-business-days.md) | [admin](admin/finance/resolve-within-3-business-days.md) | [API](api/finance/resolve-within-3-business-days.md) | [journeys](../generated/diagrams/finance/resolve-within-3-business-days.md) | | Respond to a production order | [help](help/finance/respond-to-a-production-order.md) | [admin](admin/finance/respond-to-a-production-order.md) | [API](api/finance/respond-to-a-production-order.md) | [journeys](../generated/diagrams/finance/respond-to-a-production-order.md) | | Respond to an open banking payment request | [help](help/finance/respond-to-an-open-banking-payment-request.md) | [admin](admin/finance/respond-to-an-open-banking-payment-request.md) | [API](api/finance/respond-to-an-open-banking-payment-request.md) | [journeys](../generated/diagrams/finance/respond-to-an-open-banking-payment-request.md) | | Responsible lending decision | [help](help/finance/responsible-lending-decision.md) | [admin](admin/finance/responsible-lending-decision.md) | [API](api/finance/responsible-lending-decision.md) | [journeys](../generated/diagrams/finance/responsible-lending-decision.md) | | Restore from a backup | [help](help/finance/restore-from-a-backup.md) | [admin](admin/finance/restore-from-a-backup.md) | [API](api/finance/restore-from-a-backup.md) | [journeys](../generated/diagrams/finance/restore-from-a-backup.md) | | Retry a failed job step | [help](help/finance/retry-a-failed-job-step.md) | [admin](admin/finance/retry-a-failed-job-step.md) | [API](api/finance/retry-a-failed-job-step.md) | [journeys](../generated/diagrams/finance/retry-a-failed-job-step.md) | | Retry a failed occurrence | [help](help/finance/retry-a-failed-occurrence.md) | [admin](admin/finance/retry-a-failed-occurrence.md) | [API](api/finance/retry-a-failed-occurrence.md) | [journeys](../generated/diagrams/finance/retry-a-failed-occurrence.md) | | Return an unapplied payment | [help](help/finance/return-an-unapplied-payment.md) | [admin](admin/finance/return-an-unapplied-payment.md) | [API](api/finance/return-an-unapplied-payment.md) | [journeys](../generated/diagrams/finance/return-an-unapplied-payment.md) | | Reverse a journal with a linked entry | [help](help/finance/reverse-a-journal-with-a-linked-entry.md) | [admin](admin/finance/reverse-a-journal-with-a-linked-entry.md) | [API](api/finance/reverse-a-journal-with-a-linked-entry.md) | [journeys](../generated/diagrams/finance/reverse-a-journal-with-a-linked-entry.md) | | Review a complaint file as FOS | [help](help/finance/review-a-complaint-file-as-fos.md) | [admin](admin/finance/review-a-complaint-file-as-fos.md) | [API](api/finance/review-a-complaint-file-as-fos.md) | [journeys](../generated/diagrams/finance/review-a-complaint-file-as-fos.md) | | Review a loan application | [help](help/finance/review-a-loan-application.md) | [admin](admin/finance/review-a-loan-application.md) | [API](api/finance/review-a-loan-application.md) | [journeys](../generated/diagrams/finance/review-a-loan-application.md) | | Review access rights | [help](help/finance/review-access-rights.md) | [admin](admin/finance/review-access-rights.md) | [API](api/finance/review-access-rights.md) | [journeys](../generated/diagrams/finance/review-access-rights.md) | | Review an application queue | [help](help/finance/review-an-application-queue.md) | [admin](admin/finance/review-an-application-queue.md) | [API](api/finance/review-an-application-queue.md) | [journeys](../generated/diagrams/finance/review-an-application-queue.md) | | Review APP reimbursement performance | [help](help/finance/review-app-reimbursement-performance.md) | [admin](admin/finance/review-app-reimbursement-performance.md) | [API](api/finance/review-app-reimbursement-performance.md) | [journeys](../generated/diagrams/finance/review-app-reimbursement-performance.md) | | Review customer financial statements | [help](help/finance/review-customer-financial-statements.md) | [admin](admin/finance/review-customer-financial-statements.md) | [API](api/finance/review-customer-financial-statements.md) | [journeys](../generated/diagrams/finance/review-customer-financial-statements.md) | | Review device and session signals | [help](help/finance/review-device-and-session-signals.md) | [admin](admin/finance/review-device-and-session-signals.md) | [API](api/finance/review-device-and-session-signals.md) | [journeys](../generated/diagrams/finance/review-device-and-session-signals.md) | | Review operational resilience evidence | [help](help/finance/review-operational-resilience-evidence.md) | [admin](admin/finance/review-operational-resilience-evidence.md) | [API](api/finance/review-operational-resilience-evidence.md) | [journeys](../generated/diagrams/finance/review-operational-resilience-evidence.md) | | Review SM&CR responsibilities | [help](help/finance/review-sm-cr-responsibilities.md) | [admin](admin/finance/review-sm-cr-responsibilities.md) | [API](api/finance/review-sm-cr-responsibilities.md) | [journeys](../generated/diagrams/finance/review-sm-cr-responsibilities.md) | | Review the vulnerability register | [help](help/finance/review-the-vulnerability-register.md) | [admin](admin/finance/review-the-vulnerability-register.md) | [API](api/finance/review-the-vulnerability-register.md) | [journeys](../generated/diagrams/finance/review-the-vulnerability-register.md) | | Revoke a role on leaving | [help](help/finance/revoke-a-role-on-leaving.md) | [admin](admin/finance/revoke-a-role-on-leaving.md) | [API](api/finance/revoke-a-role-on-leaving.md) | [journeys](../generated/diagrams/finance/revoke-a-role-on-leaving.md) | | Revoke an agent mandate | [help](help/finance/revoke-an-agent-mandate.md) | [admin](admin/finance/revoke-an-agent-mandate.md) | [API](api/finance/revoke-an-agent-mandate.md) | [journeys](../generated/diagrams/finance/revoke-an-agent-mandate.md) | | Roll back a release | [help](help/finance/roll-back-a-release.md) | [admin](admin/finance/roll-back-a-release.md) | [API](api/finance/roll-back-a-release.md) | [journeys](../generated/diagrams/finance/roll-back-a-release.md) | | Rotate keys and device bindings | [help](help/finance/rotate-keys-and-device-bindings.md) | [admin](admin/finance/rotate-keys-and-device-bindings.md) | [API](api/finance/rotate-keys-and-device-bindings.md) | [journeys](../generated/diagrams/finance/rotate-keys-and-device-bindings.md) | | Run a Consumer Duty outcomes review | [help](help/finance/run-a-consumer-duty-outcomes-review.md) | [admin](admin/finance/run-a-consumer-duty-outcomes-review.md) | [API](api/finance/run-a-consumer-duty-outcomes-review.md) | [journeys](../generated/diagrams/finance/run-a-consumer-duty-outcomes-review.md) | | Run a disaster recovery test | [help](help/finance/run-a-disaster-recovery-test.md) | [admin](admin/finance/run-a-disaster-recovery-test.md) | [API](api/finance/run-a-disaster-recovery-test.md) | [journeys](../generated/diagrams/finance/run-a-disaster-recovery-test.md) | | Run a fair value assessment | [help](help/finance/run-a-fair-value-assessment.md) | [admin](admin/finance/run-a-fair-value-assessment.md) | [API](api/finance/run-a-fair-value-assessment.md) | [journeys](../generated/diagrams/finance/run-a-fair-value-assessment.md) | | Run a KYC check on a subject | [help](help/finance/run-a-kyc-check-on-a-subject.md) | [admin](admin/finance/run-a-kyc-check-on-a-subject.md) | [API](api/finance/run-a-kyc-check-on-a-subject.md) | [journeys](../generated/diagrams/finance/run-a-kyc-check-on-a-subject.md) | | Run a ledger report for a period | [help](help/finance/run-a-ledger-report-for-a-period.md) | [admin](admin/finance/run-a-ledger-report-for-a-period.md) | [API](api/finance/run-a-ledger-report-for-a-period.md) | [journeys](../generated/diagrams/finance/run-a-ledger-report-for-a-period.md) | | Run a liquidity stress test | [help](help/finance/run-a-liquidity-stress-test.md) | [admin](admin/finance/run-a-liquidity-stress-test.md) | [API](api/finance/run-a-liquidity-stress-test.md) | [journeys](../generated/diagrams/finance/run-a-liquidity-stress-test.md) | | Run an annual relationship review | [help](help/finance/run-an-annual-relationship-review.md) | [admin](admin/finance/run-an-annual-relationship-review.md) | [API](api/finance/run-an-annual-relationship-review.md) | [journeys](../generated/diagrams/finance/run-an-annual-relationship-review.md) | | Run due jobs as service actor | [help](help/finance/run-due-jobs-as-service-actor.md) | [admin](admin/finance/run-due-jobs-as-service-actor.md) | [API](api/finance/run-due-jobs-as-service-actor.md) | [journeys](../generated/diagrams/finance/run-due-jobs-as-service-actor.md) | | Run due standing orders | [help](help/finance/run-due-standing-orders.md) | [admin](admin/finance/run-due-standing-orders.md) | [API](api/finance/run-due-standing-orders.md) | [journeys](../generated/diagrams/finance/run-due-standing-orders.md) | | Run end of day payment reconciliation | [help](help/finance/run-end-of-day-payment-reconciliation.md) | [admin](admin/finance/run-end-of-day-payment-reconciliation.md) | [API](api/finance/run-end-of-day-payment-reconciliation.md) | [journeys](../generated/diagrams/finance/run-end-of-day-payment-reconciliation.md) | | Run payroll payments | [help](help/finance/run-payroll-payments.md) | [admin](admin/finance/run-payroll-payments.md) | [API](api/finance/run-payroll-payments.md) | [journeys](../generated/diagrams/finance/run-payroll-payments.md) | | Run penetration testing | [help](help/finance/run-penetration-testing.md) | [admin](admin/finance/run-penetration-testing.md) | [API](api/finance/run-penetration-testing.md) | [journeys](../generated/diagrams/finance/run-penetration-testing.md) | | Run periodic KYC refresh | [help](help/finance/run-periodic-kyc-refresh.md) | [admin](admin/finance/run-periodic-kyc-refresh.md) | [API](api/finance/run-periodic-kyc-refresh.md) | [journeys](../generated/diagrams/finance/run-periodic-kyc-refresh.md) | | Run protected reveals | [help](help/finance/run-protected-reveals.md) | [admin](admin/finance/run-protected-reveals.md) | [API](api/finance/run-protected-reveals.md) | [journeys](../generated/diagrams/finance/run-protected-reveals.md) | | Run sanctions screening | [help](help/finance/run-sanctions-screening.md) | [admin](admin/finance/run-sanctions-screening.md) | [API](api/finance/run-sanctions-screening.md) | [journeys](../generated/diagrams/finance/run-sanctions-screening.md) | | Run transaction monitoring | [help](help/finance/run-transaction-monitoring.md) | [admin](admin/finance/run-transaction-monitoring.md) | [API](api/finance/run-transaction-monitoring.md) | [journeys](../generated/diagrams/finance/run-transaction-monitoring.md) | | Sample test receipts as external auditor | [help](help/finance/sample-test-receipts-as-external-auditor.md) | [admin](admin/finance/sample-test-receipts-as-external-auditor.md) | [API](api/finance/sample-test-receipts-as-external-auditor.md) | [journeys](../generated/diagrams/finance/sample-test-receipts-as-external-auditor.md) | | Sanctions and PEP screening | [help](help/finance/sanctions-and-pep-screening.md) | [admin](admin/finance/sanctions-and-pep-screening.md) | [API](api/finance/sanctions-and-pep-screening.md) | [journeys](../generated/diagrams/finance/sanctions-and-pep-screening.md) | | Save an unfinished application | [help](help/finance/save-an-unfinished-application.md) | [admin](admin/finance/save-an-unfinished-application.md) | [API](api/finance/save-an-unfinished-application.md) | [journeys](../generated/diagrams/finance/save-an-unfinished-application.md) | | Savings pots and goals | [help](help/finance/savings-pots.md) | [admin](admin/finance/savings-pots.md) | [API](api/finance/savings-pots.md) | [journeys](../generated/diagrams/finance/savings-pots.md) | | Schedule a future payment | [help](help/finance/schedule-a-future-payment.md) | [admin](admin/finance/schedule-a-future-payment.md) | [API](api/finance/schedule-a-future-payment.md) | [journeys](../generated/diagrams/finance/schedule-a-future-payment.md) | | Scheduled payments | [help](help/finance/scheduled-payments.md) | [admin](admin/finance/scheduled-payments.md) | [API](api/finance/scheduled-payments.md) | [journeys](../generated/diagrams/finance/scheduled-payments.md) | | Screen a list against sanctions | [help](help/finance/screen-a-list-against-sanctions.md) | [admin](admin/finance/screen-a-list-against-sanctions.md) | [API](api/finance/screen-a-list-against-sanctions.md) | [journeys](../generated/diagrams/finance/screen-a-list-against-sanctions.md) | | Screen existing customers against list updates | [help](help/finance/screen-existing-customers-against-list-updates.md) | [admin](admin/finance/screen-existing-customers-against-list-updates.md) | [API](api/finance/screen-existing-customers-against-list-updates.md) | [journeys](../generated/diagrams/finance/screen-existing-customers-against-list-updates.md) | | Screen for politically exposed persons | [help](help/finance/screen-for-politically-exposed-persons.md) | [admin](admin/finance/screen-for-politically-exposed-persons.md) | [API](api/finance/screen-for-politically-exposed-persons.md) | [journeys](../generated/diagrams/finance/screen-for-politically-exposed-persons.md) | | Search transactions | [help](help/finance/search-transactions.md) | [admin](admin/finance/search-transactions.md) | [API](api/finance/search-transactions.md) | [journeys](../generated/diagrams/finance/search-transactions.md) | | Security and devices | [help](help/finance/security-and-devices.md) | [admin](admin/finance/security-and-devices.md) | [API](api/finance/security-and-devices.md) | [journeys](../generated/diagrams/finance/security-and-devices.md) | | Send a CHAPS payment | [help](help/finance/send-a-chaps-payment.md) | [admin](admin/finance/send-a-chaps-payment.md) | [API](api/finance/send-a-chaps-payment.md) | [journeys](../generated/diagrams/finance/send-a-chaps-payment.md) | | Send a receiving bank APP notice | [help](help/finance/send-a-receiving-bank-app-notice.md) | [admin](admin/finance/send-a-receiving-bank-app-notice.md) | [API](api/finance/send-a-receiving-bank-app-notice.md) | [journeys](../generated/diagrams/finance/send-a-receiving-bank-app-notice.md) | | Send money | [help](help/finance/send-money.md) | [admin](admin/finance/send-money.md) | [API](api/finance/send-money.md) | [journeys](../generated/diagrams/finance/send-money.md) | | Set and monitor exposure limits | [help](help/finance/set-and-monitor-exposure-limits.md) | [admin](admin/finance/set-and-monitor-exposure-limits.md) | [API](api/finance/set-and-monitor-exposure-limits.md) | [journeys](../generated/diagrams/finance/set-and-monitor-exposure-limits.md) | | Set balance and payment alerts | [help](help/finance/set-balance-and-payment-alerts.md) | [admin](admin/finance/set-balance-and-payment-alerts.md) | [API](api/finance/set-balance-and-payment-alerts.md) | [journeys](../generated/diagrams/finance/set-balance-and-payment-alerts.md) | | Set card limits and controls | [help](help/finance/set-card-limits-and-controls.md) | [admin](admin/finance/set-card-limits-and-control _… 191581 more characters in the zip._ ### docs/STANDARD.md `docs/STANDARD.md` · sha256 `cce45c122c9af2a7c6f1ed5705731eca3f15730f779636905611a2d7d8f4a4aa` #### The blueprint standard: produce a whole estate without generating from scratch Everything that already exists is gathered once and recombined by programs. Nothing that exists is drawn or written by hand, by people or by AI. The same inputs always give the same estate, so anyone can reproduce it exactly and nobody wastes effort or compute generating what already exists. ##### The order (each step uses only the steps above it) 1. **Intention.** What the organisation wants, in its own words: `contract/estate-contract.json`. 2. **Customer outcome.** What is saved or made more (time, money, lives, planet, resources, value) and how it is measured: in the contract and in each capability. 3. **Raw materials, gathered once.** - **Role workflows**, each with a trigger, ordered steps, records, permissions, rules and "done when". - **Competitor clauses and documentation:** `research/parity/`. - **Competitor flows and screens** from Refero and Mobbin, and the research reports: `research/`. 4. **Capabilities.** Each role workflow becomes one capability; each step becomes a screen, a control and an operation, joined in click order (`tools/derive.py`). 5. **Parity first.** Every competitor clause, flow, screen and article becomes a control, operation, endpoint, agent tool and definition on the step it fits, or in a capability for its competitor and area. The ledger records where each one is covered (`tools/match.py`, `ledger/parity/`). 6. **Everything mechanical.** For each capability (`tools/expand.py`): - every screen state, its copy, accessibility and help; - every endpoint, agent tool and full behaviour (checks, rehearsal, approval, receiving confirmation, unknown outcome, recovery, evidence, batch, verification, monitoring); - the evidence screen; - every journey for every actor. 7. **Journeys and workflows drawn.** A Mermaid screen map, a click-by-click breakdown of every journey with a sequence diagram, a state diagram per screen and a behaviour diagram per operation (`tools/build.py`, `generated/diagrams/`). These are what the screens are made from. 8. **Documentation.** A help-centre article, an admin guide and an API reference per capability, each set against the competitor documentation it matches (`tools/docs.py`, `docs/`). 9. **Machine-readable estate.** The estate file, the API description (OpenAPI) and the database schema (`tools/emit.py`, `build/`). The runtime, the agent tool server and the screens are generated from these. ##### Reproduce it ```bash python3 tools/derive.py <role workflow registers> ``` ```bash python3 tools/match.py ``` ```bash python3 tools/expand.py ``` ```bash python3 tools/build.py ``` ```bash python3 tools/docs.py ``` ```bash python3 tools/emit.py ``` Each step takes seconds. Change an input and run again; never edit an output. ##### Rules of the standard - **Gather once, combine many times.** If it exists, match it; never regenerate it. - **Parity first, then better.** Every competitor feature is matched by a named control before anything new is added. - **Deterministic.** Programs, not prompts, produce the estate. Agents only gather raw materials. - **One contract.** The intentions and outcomes in the contract drive everything; no part is changed except through it. - **Agents are customers and staff.** Every operation has an agent tool, used under the agent's own identity and mandate with an accountable owner. - **Glass box.** Every effect leaves evidence that anyone authorised can inspect and reproduce. ### docs/MCP.md `docs/MCP.md` · sha256 `31d0b9eae644ddbc6aaa1e7a81e352613e29f31db1e933166e5fa879d87e3311` #### Every capability and workflow as an MCP Generated by `tools/mcps.py`, which writes the servers into the repository's `.mcp.json`. Each server was started, listed and called once (a rehearsal) when this was written. - **`estate-hub`** is where any agent starts. It has tools to: - search for the capability that carries an intention, and describe or call it; - run an intention on the racetrack: rehearse every step, run, and return the evidence; - list, plug, drag and drop, and test external datasets, models and services. - **`estate-<area>`** servers hold one tool per capability. Each tool offers the capability's operations and its whole workflows, which run on the racetrack and are rehearsed first. | Server | Tools | Capabilities | Test call | |---|---|---|---| | `estate-clinic` | 455 | 455 | listed in-process | | `estate-clinic-match` | 164 | 164 | listed in-process | | `estate-consortium` | 489 | 489 | listed in-process | | `estate-consortium-match` | 76 | 76 | listed in-process | | `estate-evidence` | 67 | 67 | listed in-process | | `estate-finance` | 709 | 709 | started and listed over stdio | | `estate-finance-match` | 310 | 310 | listed in-process | | `estate-id` | 2 | 2 | listed in-process | | `estate-judiverse` | 113 | 113 | listed in-process | | `estate-judiverse-match` | 857 | 857 | listed in-process | | `estate-organisations` | 3 | 3 | listed in-process | | `estate-organisations-match` | 1674 | 1674 | listed in-process | | `estate-public` | 557 | 557 | listed in-process | | `estate-public-match` | 19 | 19 | listed in-process | | `estate-regulators` | 2 | 2 | listed in-process | | `estate-releases` | 1 | 1 | listed in-process | | `estate-research` | 236 | 236 | listed in-process | | `estate-hub` | 8 | 0 | started and listed over stdio | ### docs/PLUGS.md `docs/PLUGS.md` · sha256 `b0cdc203b77a5a6f03529ec0ebc246858f41a212d3a6e89cb36b0675e71fbe4d` #### Plugs: external datasets, models and services Generated by `tools/plugs.py`. Every guard in the estate is bound by its family (`runtime/guards.py`): actor-kind 92, condition 1103, engine 17068, identity 144, input-present 74, inspection-scope 1684, owner-limits 6804, role 6260. Conditions that need facts from outside are answered through 791 slots (dataset 122, estate-record 213, model 9, service 447). Plug one in a single line, then test it: ```bash python3 tools/plugs.py plug <slot> mypackage.module:function ``` ```bash python3 tools/plugs.py test <slot> ``` Use `*` as the slot to bind a default plug for every condition. Credentials come from the environment variable named for each slot; nothing secret is written to the repository. Until a slot is plugged, simulation passes it (marked SIMULATED in the evidence) and live mode refuses it with `PLUG_REQUIRED`. | Slot | Kind | Operations | Guards | Example conditions | Credentials | Status | |---|---|---|---|---|---|---| | `model-detection` | model | 58 | 0 | | `PLUG_MODEL_DETECTION` | UNPLUGGED | | `model-classification` | model | 44 | 0 | | `PLUG_MODEL_CLASSIFICATION` | UNPLUGGED | | `model-recommendation` | model | 42 | 0 | | `PLUG_MODEL_RECOMMENDATION` | UNPLUGGED | | `edge-id-and-agent-mandate` | estate-record | 35 | 35 | the viewer is the subject or acts within a mandate from the subject; caller holds the account or a mandate for it | `PLUG_EDGE_ID_AND_AGENT_MANDATE` | UNPLUGGED | | `model-forecasting` | model | 31 | 0 | | `PLUG_MODEL_FORECASTING` | UNPLUGGED | | `ledger` | estate-record | 16 | 20 | pot exists and is open; linked funding account open and owned by the owner | `PLUG_LEDGER` | UNPLUGGED | | `consent-register` | estate-record | 14 | 15 | the subject's consent for this purpose is current; active consent | `PLUG_CONSENT_REGISTER` | UNPLUGGED | | `model-summarisation` | model | 12 | 0 | | `PLUG_MODEL_SUMMARISATION` | UNPLUGGED | | `mail-inbound` | service | 9 | 0 | | `PLUG_MAIL_INBOUND` | UNPLUGGED | | `joint-account-register` | estate-record | 7 | 8 | caller is a holder or a holder's agent; caller is a holder | `PLUG_JOINT_ACCOUNT_REGISTER` | UNPLUGGED | | `member-role` | estate-record | 7 | 7 | changer is Owner or Admin; granter is Owner or Admin | `PLUG_MEMBER_ROLE` | UNPLUGGED | | `model-transcription` | model | 7 | 0 | | `PLUG_MODEL_TRANSCRIPTION` | UNPLUGGED | | `review-queue` | estate-record | 7 | 7 | case assigned to this reviewer; not already in the queue | `PLUG_REVIEW_QUEUE` | UNPLUGGED | | `ledger-balance` | estate-record | 6 | 6 | amount within balance; enough money for the total and fees | `PLUG_LEDGER_BALANCE` | UNPLUGGED | | `mandate-register` | estate-record | 6 | 6 | no other active mandate; mandate active or paused | `PLUG_MANDATE_REGISTER` | UNPLUGGED | | `model-extraction` | model | 6 | 0 | | `PLUG_MODEL_EXTRACTION` | UNPLUGGED | | `batch-status` | service | 5 | 5 | batch is a draft; batch READY | `PLUG_BATCH_STATUS` | UNPLUGGED | | `check-in-record` | estate-record | 5 | 5 | check-in open; check-in exists | `PLUG_CHECK_IN_RECORD` | UNPLUGGED | | `connection-status` | service | 5 | 6 | active consent; connection not already disconnected | `PLUG_CONNECTION_STATUS` | UNPLUGGED | | `contract-record` | estate-record | 5 | 5 | named contract manager; approver did not accept the work | `PLUG_CONTRACT_RECORD` | UNPLUGGED | | `dispute-store` | service | 5 | 6 | no open dispute for this transaction; dispute not closed | `PLUG_DISPUTE_STORE` | UNPLUGGED | | `edge-id-staff-roles` | estate-record | 5 | 5 | handler role held; granter holds regulator.grant | `PLUG_EDGE_ID_STAFF_ROLES` | UNPLUGGED | | `membership-register` | dataset | 5 | 5 | membership active; status actually changes | `PLUG_MEMBERSHIP_REGISTER` | UNPLUGGED | | `session-store` | service | 5 | 5 | session open with an officer signed in; session open | `PLUG_SESSION_STORE` | UNPLUGGED | | `agent-status` | service | 4 | 4 | keys registered; agent born | `PLUG_AGENT_STATUS` | UNPLUGGED | | `application-status` | service | 4 | 4 | not already decided; account not yet opened | `PLUG_APPLICATION_STATUS` | UNPLUGGED | | `confidence-report-store` | service | 4 | 4 | entry belongs to the subject; entry belongs to the viewer's report | `PLUG_CONFIDENCE_REPORT_STORE` | UNPLUGGED | | `conversation-members` | service | 4 | 4 | member; member count within limit | `PLUG_CONVERSATION_MEMBERS` | UNPLUGGED | | `correction-register` | dataset | 4 | 4 | not already open for this entry; request still open | `PLUG_CORRECTION_REGISTER` | UNPLUGGED | | `edge-id-relations` | estate-record | 4 | 4 | caller owns the agent or is the agent; caller is the subject, its owner or its administrator | `PLUG_EDGE_ID_RELATIONS` | UNPLUGGED | | `request-status` | service | 4 | 4 | request is pending | `PLUG_REQUEST_STATUS` | UNPLUGGED | | `schedule-creator` | service | 4 | 4 | agent created this schedule | `PLUG_SCHEDULE_CREATOR` | UNPLUGGED | | `schedule-status` | service | 4 | 4 | schedule active or paused; schedule active | `PLUG_SCHEDULE_STATUS` | UNPLUGGED | | `search-fields` | service | 4 | 4 | filters valid | `PLUG_SEARCH_FIELDS` | UNPLUGGED | | `share-role` | service | 4 | 4 | caller is an editor; caller may comment | `PLUG_SHARE_ROLE` | UNPLUGGED | | `approval-reason` | service | 3 | 3 | reason given when rejecting; reason given to reject or ask | `PLUG_APPROVAL_REASON` | UNPLUGGED | | `assessment-record` | estate-record | 3 | 3 | assessment open; outcome exists | `PLUG_ASSESSMENT_RECORD` | UNPLUGGED | | `availability` | service | 3 | 3 | slot still free for required attendees; new slot free | `PLUG_AVAILABILITY` | UNPLUGGED | | `batch-content-digest` | service | 3 | 3 | unchanged since checked; unchanged since submitted | `PLUG_BATCH_CONTENT_DIGEST` | UNPLUGGED | | `birth-certificate` | service | 3 | 4 | scope within the agent's capabilities; expiry within the agent's term | `PLUG_BIRTH_CERTIFICATE` | UNPLUGGED | | `brief` | service | 3 | 3 | question long enough; high estimate within the brief's credit limit | `PLUG_BRIEF` | UNPLUGGED | | `catalogue` | dataset | 3 | 3 | track belongs to the artist; track is public | `PLUG_CATALOGUE` | UNPLUGGED | | `challenge-records` | estate-record | 3 | 3 | no open challenge on this check; no open challenge | `PLUG_CHALLENGE_RECORDS` | UNPLUGGED | | `change-register` | dataset | 3 | 4 | no other change waiting; change still live | `PLUG_CHANGE_REGISTER` | UNPLUGGED | | `claim-store` | service | 3 | 3 | not already claimed; claim visible to caller | `PLUG_CLAIM_STORE` | UNPLUGGED | | `consent-given` | estate-record | 3 | 3 | explicit consent; consent ticked | `PLUG_CONSENT_GIVEN` | UNPLUGGED | | `contract-status` | service | 3 | 3 | contract active; contract sent and unsigned | `PLUG_CONTRACT_STATUS` | UNPLUGGED | | `decision-record` | estate-record | 3 | 3 | decision less than 30 days old; offer still valid | `PLUG_DECISION_RECORD` | UNPLUGGED | | `device-register` | dataset | 3 | 3 | registered camera; device not reported lost | `PLUG_DEVICE_REGISTER` | UNPLUGGED | | `directory` | dataset | 3 | 3 | service in directory; record not already claimed | `PLUG_DIRECTORY` | UNPLUGGED | | `edge-id-register` | estate-record | 3 | 3 | agent active; agent has an active EDGE ID | `PLUG_EDGE_ID_REGISTER` | UNPLUGGED | | `evidence-store` | estate-record | 3 | 3 | evidence linked; every evidence document named is on file and unique to this economic event | `PLUG_EVIDENCE_STORE` | UNPLUGGED | | `instruction-status` | service | 3 | 3 | instruction active | `PLUG_INSTRUCTION_STATUS` | UNPLUGGED | | `interest-declarations` | service | 3 | 3 | verifier independent; no conflict of interest | `PLUG_INTEREST_DECLARATIONS` | UNPLUGGED | | `malware-and-type-scan` | service | 3 | 3 | file safe; file safe and allowed type | `PLUG_MALWARE_AND_TYPE_SCAN` | UNPLUGGED | | `occurrence-status` | service | 3 | 3 | next payment not already being sent | `PLUG_OCCURRENCE_STATUS` | UNPLUGGED | | `professional-registration-check` | service | 3 | 3 | registered clinician | `PLUG_PROFESSIONAL_REGISTRATION_CHECK` | UNPLUGGED | | `racetrack-record` | estate-record | 3 | 3 | run rehearsed with these inputs; rehearsed | `PLUG_RACETRACK_RECORD` | UNPLUGGED | | `record-store` | estate-record | 3 | 3 | entry exists; referral exists | `PLUG_RECORD_STORE` | UNPLUGGED | | `regulator-portal-grant` | service | 3 | 3 | within inspection scope | `PLUG_REGULATOR_PORTAL_GRANT` | UNPLUGGED | | `request-store` | service | 3 | 3 | caller sent the request; caller is the payer | `PLUG_REQUEST_STORE` | UNPLUGGED | | `screening-record` | estate-record | 3 | 3 | result has a REFER decision; not already decided | `PLUG_SCREENING_RECORD` | UNPLUGGED | | `task-lease` | service | 3 | 3 | caller holds the current lease and presents the current fencing number; no other live lease | `PLUG_TASK_LEASE` | UNPLUGGED | | `team-register` | dataset | 3 | 3 | not already a member; at least one Owner remains Active | `PLUG_TEAM_REGISTER` | UNPLUGGED | | `account-product-settings` | estate-record | 2 | 2 | account can take Direct Debits | `PLUG_ACCOUNT_PRODUCT_SETTINGS` | UNPLUGGED | | `account-status` | estate-record | 2 | 2 | account in good standing | `PLUG_ACCOUNT_STATUS` | UNPLUGGED | | `address-lookup` | service | 2 | 2 | address deliverable | `PLUG_ADDRESS_LOOKUP` | UNPLUGGED | | `agent-limits` | service | 2 | 2 | within agent limits | `PLUG_AGENT_LIMITS` | UNPLUGGED | | `agent-s-term` | service | 2 | 2 | expiry within the agent's term | `PLUG_AGENT_S_TERM` | UNPLUGGED | | `agreement-date` | service | 2 | 2 | within the withdrawal period; within 14 days of the agreement | `PLUG_AGREEMENT_DATE` | UNPLUGGED | | `agreement-register` | dataset | 2 | 2 | data-sharing agreement signed by all, if data is shared; accepting the current version | `PLUG_AGREEMENT_REGISTER` | UNPLUGGED | | `alert-store` | service | 2 | 2 | alert is open; alert exists | `PLUG_ALERT_STORE` | UNPLUGGED | | `approval-queue` | estate-record | 2 | 2 | agent has an approval reference; agent has approval | `PLUG_APPROVAL_QUEUE` | UNPLUGGED | | `assessment-state` | service | 2 | 2 | answer is for the current question; questions finished | `PLUG_ASSESSMENT_STATE` | UNPLUGGED | | `assessment-store` | service | 2 | 2 | at least one control assessed; at least one assessment exists at that time | `PLUG_ASSESSMENT_STORE` | UNPLUGGED | | `asset-inventory` | service | 2 | 2 | inventory refreshed within its expected interval; affected assets are in the inventory | `PLUG_ASSET_INVENTORY` | UNPLUGGED | | `attendance-record` | estate-record | 2 | 2 | councillor is attending; meeting quorate | `PLUG_ATTENDANCE_RECORD` | UNPLUGGED | | `block-list` | dataset | 2 | 2 | not blocked by any member; block exists | `PLUG_BLOCK_LIST` | UNPLUGGED | | `capture-session` | service | 2 | 2 | live capture session from the applicant's device; live capture session | `PLUG_CAPTURE_SESSION` | UNPLUGGED | | `card-processor` | service | 2 | 2 | card exists; no active cards | `PLUG_CARD_PROCESSOR` | UNPLUGGED | | `card-status` | service | 2 | 3 | card not already frozen; card not cancelled | `PLUG_CARD_STATUS` | UNPLUGGED | | `certificate-registry` | dataset | 2 | 2 | version certified; previous certificate still valid | `PLUG_CERTIFICATE_REGISTRY` | UNPLUGGED | | `check-record` | estate-record | 2 | 2 | outcome is FAIL or REFER; check outcome is FAIL or REFER | `PLUG_CHECK_RECORD` | UNPLUGGED | | `check-records` | estate-record | 2 | 2 | every organisation and document check passed or overturned; every required check passed or overturned on review | `PLUG_CHECK_RECORDS` | UNPLUGGED | | `clinician-queue` | estate-record | 2 | 2 | escalation in queue; in queue | `PLUG_CLINICIAN_QUEUE` | UNPLUGGED | | `column-map` | service | 2 | 2 | display name mapped | `PLUG_COLUMN_MAP` | UNPLUGGED | | `conflict-register` | dataset | 2 | 2 | not the officer who tuned the model for this segment; reviewer has no link to the organisation or claimant | `PLUG_CONFLICT_REGISTER` | UNPLUGGED | | `conversation-roles` | service | 2 | 2 | permission; owner for shared settings | `PLUG_CONVERSATION_ROLES` | UNPLUGGED | | `crisis-alert` | service | 2 | 2 | alert acknowledged by this practitioner; holding the alert | `PLUG_CRISIS_ALERT` | UNPLUGGED | | `decision-status` | service | 2 | 2 | every decision item cleared; paper cleared | `PLUG_DECISION_STATUS` | UNPLUGGED | | `device-list` | dataset | 2 | 2 | device belongs to the account | `PLUG_DEVICE_LIST` | UNPLUGGED | | `document-state` | service | 2 | 2 | document not locked for review | `PLUG_DOCUMENT_STATE` | UNPLUGGED | | `duty-rota` | service | 2 | 2 | a named duty practitioner for this hour; on duty now | `PLUG_DUTY_ROTA` | UNPLUGGED | | `edge-id-access-service` | estate-record | 2 | 2 | share exists; shared with this buyer | `PLUG_EDGE_ID_ACCESS_SERVICE` | UNPLUGGED | | `edge-id-directory` | estate-record | 2 | 2 | invitee has an EDGE ID; payer exists | `PLUG_EDGE_ID_DIRECTORY` | UNPLUGGED | | `edge-id-session-age` | estate-record | 2 | 2 | recent strong sign-in | `PLUG_EDGE_ID_SESSION_AGE` | UNPLUGGED | | `entry-provenance` | service | 2 | 2 | clinical facts go to their author; officer independent of the entry | `PLUG_ENTRY_PROVENANCE` | UNPLUGGED | | `escalation-queue` | estate-record | 2 | 2 | not already escalated | `PLUG_ESCALATION_QUEUE` | UNPLUGGED | | `event` | service | 2 | 2 | end after start; caller is organiser, the booking's invitee, or a mandated agent | `PLUG_EVENT` | UNPLUGGED | | `exception-register` | dataset | 2 | 2 | no open exception of the same type; plan version is the latest | `PLUG_EXCEPTION_REGISTER` | UNPLUGGED | | `face-detection-procedure` | service | 2 | 2 | one face clearly visible; one face visible | `PLUG_FACE_DETECTION_PROCEDURE` | UNPLUGGED | | `file` | service | 2 | 2 | within size limit | `PLUG_FILE` | UNPLUGGED | | `file-scanner` | service | 2 | 2 | files are readable and free of malware; attachments pass the safety scan | `PLUG_FILE_SCANNER` | UNPLUGGED | | `hold-register` | dataset | 2 | 2 | no legal hold; retention covers holds | `PLUG_HOLD_REGISTER` | UNPLUGGED | | `identity-service` | service | 2 | 2 | signed challenge verifies | `PLUG_IDENTITY_SERVICE` | UNPLUGGED | | `institution-register` | dataset | 2 | 2 | the caller is an authorised institution | `PLUG_INSTITUTION_REGISTER` | UNPLUGGED | | `instruction-set-up-by` | service | 2 | 2 | agent set this up | `PLUG_INSTRUCTION_SET_UP_BY` | UNPLUGGED | | `instrument-register` | dataset | 2 | 2 | instrument admitted with its licence and scoring guide; scoring guide admitted | `PLUG_INSTRUMENT_REGISTER` | UNPLUGGED | | `issuer-signature-check` | service | 2 | 2 | chip signed by its issuer | `PLUG_ISSUER_SIGNATURE_CHECK` | UNPLUGGED | | `ledger-balance-on-the-due-date` | estate-record | 2 | 2 | funds available | `PLUG_LEDGER_BALANCE_ON_THE_DUE_DATE` | UNPLUGGED | | `lifecycle-events` | service | 2 | 2 | agent not retired; no death certificate | `PLUG_LIFECYCLE_EVENTS` | UNPLUGGED | | `list-register` | dataset | 2 | 2 | every required list loaded at its current version; new list fetched from the official source and its digest verified | `PLUG_LIST_REGISTER` | UNPLUGGED | | `milestone-status` | service | 2 | 2 | milestone not already accepted; milestone accepted | `PLUG_MILESTONE_STATUS` | UNPLUGGED | | `network-register` | dataset | 2 | 2 | a marker exists for this subject; caller is the filing member | `PLUG_NETWORK_REGISTER` | UNPLUGGED | | `obligation-records` | estate-record | 2 | 2 | no open obligations (owner path); every obligation handed over or cancelled | `PLUG_OBLIGATION_RECORDS` | UNPLUGGED | | `offer-expires-at` | service | 2 | 2 | offer still valid; offer not expired | `PLUG_OFFER_EXPIRES_AT` | UNPLUGGED | | `onboarding-records` | estate-record | 2 | 2 | no other open application for this organisation; application belongs to the caller or its principal | `PLUG_ONBOARDING_RECORDS` | UNPLUGGED | | `opportunity-submission-deadline` | service | 2 | 2 | before the deadline | `PLUG_OPPORTUNITY_SUBMISSION_DEADLINE` | UNPLUGGED | | `organisation-chat-policy` | service | 2 | 2 | outside chat allowed; export allowed | `PLUG_ORGANISATION_CHAT_POLICY` | UNPLUGGED | | `parent-agent-s-current-mandate` | estate-record | 2 | 2 | child capabilities within parent mandate; child design within parent mandate | `PLUG_PARENT_AGENT_S_CURRENT_MANDATE` | UNPLUGGED | | `parser` | service | 2 | 2 | file readable | `PLUG_PARSER` | UNPLUGGED | | `partner-list` | dataset | 2 | 2 | only one lead; every listed partner accepted | `PLUG_PARTNER_LIST` | UNPLUGGED | | `payee-list` | dataset | 2 | 2 | payee still saved; payee details unchanged since the name check | `PLUG_PAYEE_LIST` | UNPLUGGED | | `plan-record` | estate-record | 2 | 2 | plan unchanged since approval or re-approved; plan still valid | `PLUG_PLAN_RECORD` | UNPLUGGED | | `plan-store` | service | 2 | 2 | plan exists; no newer version saved meanwhile | `PLUG_PLAN_STORE` | UNPLUGGED | | `pot-balance` | service | 2 | 2 | enough in the pot | `PLUG_POT_BALANCE` | UNPLUGGED | | `pot-list` | dataset | 2 | 2 | name unique among open pots; name unique | `PLUG_POT_LIST` | UNPLUGGED | | `pot-status` | service | 2 | 2 | pot open | `PLUG_POT_STATUS` | UNPLUGGED | | `prescribing-session` | service | 2 | 2 | a prescriber is drafting for this patient; prescriber owns the draft | `PLUG_PRESCRIBING_SESSION` | UNPLUGGED | | `preview` | service | 2 | 2 | preview still valid; preview still current | `PLUG_PREVIEW` | UNPLUGGED | | `programme-intake-dates` | service | 2 | 2 | programme open for applications; programme still open | `PLUG_PROGRAMME_INTAKE_DATES` | UNPLUGGED | | `register-officers-list` | estate-record | 2 | 2 | claimant appears on the register in the role claimed, where the route needs it; officer listed on the register | `PLUG_REGISTER_OFFICERS_LIST` | UNPLUGGED | | `rehearsal-record` | estate-record | 2 | 2 | rehearsal still valid; rehearsal valid | `PLUG_REHEARSAL_RECORD` | UNPLUGGED | | `research-runtime` | service | 2 | 2 | run still active | `PLUG_RESEARCH_RUNTIME` | UNPLUGGED | | `review-advice` | service | 2 | 2 | advice written; advice written for GIVE_ADVICE and REFER | `PLUG_REVIEW_ADVICE` | UNPLUGGED | | `risk-check` | service | 2 | 2 | eligible; wallet eligible | `PLUG_RISK_CHECK` | UNPLUGGED | | `role-rules` | service | 2 | 2 | role combination is valid; limit and expiry present where required | `PLUG_ROLE_RULES` | UNPLUGGED | | `run-record` | estate-record | 2 | 2 | caller owns the run | `PLUG_RUN_RECORD` | UNPLUGGED | | `run-state` | service | 2 | 2 | run still active; step waiting for approval | `PLUG_RUN_STATE` | UNPLUGGED | | `sealed-bid-store` | service | 2 | 2 | deadline passed and bids opened; supplier bid in this procurement | `PLUG_SEALED_BID_STORE` | UNPLUGGED | | `staff-declarations` | service | 2 | 2 | reviewer has no link to the organisation; reviewer has no link to the applicant | `PLUG_STAFF_DECLARATIONS` | UNPLUGGED | | `staff-register` | dataset | 2 | 2 | practitioner registration current; second approver differs from the first | `PLUG_STAFF_REGISTER` | UNPLUGGED | | `supplier-consent-to-checks` | estate-record | 2 | 2 | consent still in force; supplier still consents | `PLUG_SUPPLIER_CONSENT_TO_CHECKS` | UNPLUGGED | | `supported-asset-list-open` | dataset | 2 | 2 | asset supported | `PLUG_SUPPORTED_ASSET_LIST_OPEN` | UNPLUGGED | | `supported-networks` | service | 2 | 2 | network matches asset; network supported for asset | `PLUG_SUPPORTED_NETWORKS` | UNPLUGGED | | `template-inputs` | service | 2 | 2 | every required input valid; every required answer valid for this template version | `PLUG_TEMPLATE_INPUTS` | UNPLUGGED | | `terms-register` | dataset | 2 | 2 | terms shown are the current version; current upload terms accepted | `PLUG_TERMS_REGISTER` | UNPLUGGED | | `test-set-register` | dataset | 2 | 2 | test set is registered and versioned; backtest set registered | `PLUG_TEST_SET_REGISTER` | UNPLUGGED | | `the-agent-s-action-states` | service | 2 | 2 | no action is executing or unknown | `PLUG_THE_AGENT_S_ACTION_STATES` | UNPLUGGED | | `track-status` | service | 2 | 2 | track is READY; composition licence in force where needed | `PLUG_TRACK_STATUS` | UNPLUGGED | | `triage-procedure-scope-statement` | service | 2 | 2 | age inside certified scope; inside certified scope | `PLUG_TRIAGE_PROCEDURE_SCOPE_STATEMENT` | UNPLUGGED | | `triage-record` | estate-record | 2 | 2 | session open; all protocol questions answered | `PLUG_TRIAGE_RECORD` | UNPLUGGED | | `upload-check` | service | 2 | 2 | files are readable and within size; the file is readable and within size | `PLUG_UPLOAD_CHECK` | UNPLUGGED | | `upload-service` | service | 2 | 2 | files are PDF, JPEG or PNG under the size limit; files are PDF, JPEG or PNG and under the size limit | `PLUG_UPLOAD_SERVICE` | UNPLUGGED | | `vacancy-status` | service | 2 | 2 | not already open; vacancy is OPEN | `PLUG_VACANCY_STATUS` | UNPLUGGED | | `working-day-calendar` | service | 2 | 2 | execution date valid; first date valid | `PLUG_WORKING_DAY_CALENDAR` | UNPLUGGED | | `abuse-and-injection-filter` | service | 1 | 1 | query allowed | `PLUG_ABUSE_AND_INJECTION_FILTER` | UNPLUGGED | | `acceptance-by-the-destination` | service | 1 | 1 | destination accepted every open obligation | `PLUG_ACCEPTANCE_BY_THE_DESTINATION` | UNPLUGGED | | `acceptance-record` | estate-record | 1 | 1 | recipient accepted | `PLUG_ACCEPTANCE_RECORD` | UNPLUGGED | | `accepted-document-list-open-the-identity-verification-provid` | dataset | 1 | 1 | document type accepted | `PLUG_ACCEPTED_DOCUMENT_LIST_OPEN_THE_IDENTITY` | UNPLUGGED | | `access-log` | service | 1 | 1 | event belongs to this record | `PLUG_ACCESS_LOG` | UNPLUGGED | | `account-declarations-signed` | estate-record | 1 | 1 | declarations signed by an officer | `PLUG_ACCOUNT_DECLARATIONS_SIGNED` | UNPLUGGED | | `account-name` | estate-record | 1 | 1 | typed confirmation matches | `PLUG_ACCOUNT_NAME` | UNPLUGGED | | `account-opening-date-and-today` | estate-record | 1 | 1 | range valid | `PLUG_ACCOUNT_OPENING_DATE_AND_TODAY` | UNPLUGGED | | `account-product-rules` | estate-record | 1 | 1 | account can hold an overdraft | `PLUG_ACCOUNT_PRODUCT_RULES` | UNPLUGGED | | `account-provider-account-list` | estate-record | 1 | 1 | at least one eligible account | `PLUG_ACCOUNT_PROVIDER_ACCOUNT_LIST` | UNPLUGGED | | `account-provider-authorisation-response` | estate-record | 1 | 1 | approved at the bank | `PLUG_ACCOUNT_PROVIDER_AUTHORISATION_RESPONSE` | UNPLUGGED | | `account-provider-consent-status` | estate-record | 1 | 1 | consent not rejected | `PLUG_ACCOUNT_PROVIDER_CONSENT_STATUS` | UNPLUGGED | | `account-record` | estate-record | 1 | 1 | account in good standing with us | `PLUG_ACCOUNT_RECORD` | UNPLUGGED | | `account-register` | estate-record | 1 | 1 | invitee has a verified EDGE Finance account | `PLUG_ACCOUNT_REGISTER` | UNPLUGGED | | `account-restrictions-register` | estate-record | 1 | 1 | account not restricted | `PLUG_ACCOUNT_RESTRICTIONS_REGISTER` | UNPLUGGED | | `account-settings` | estate-record | 1 | 1 | within the alert limit | `PLUG_ACCOUNT_SETTINGS` | UNPLUGGED | | `account-terms-accepted` | estate-record | 1 | 1 | terms accepted by the applicant | `PLUG_ACCOUNT_TERMS_ACCEPTED` | UNPLUGGED | | `accountant-grant-expires-on` | estate-record | 1 | 1 | end date within 12 months | `PLUG_ACCOUNTANT_GRANT_EXPIRES_ON` | UNPLUGGED | | `action-undo-until` | service | 1 | 1 | within undo window | `PLUG_ACTION_UNDO_UNTIL` | UNPLUGGED | | `actor-role` | service | 1 | 1 | state allowed for this actor | `PLUG_ACTOR_ROLE` | UNPLUGGED | | `address-checks-and-screening` | service | 1 | 1 | recipients valid and screened | `PLUG_ADDRESS_CHECKS_AND_SCREENING` | UNPLUGGED | | `address-format-and-checksum` | service | 1 | 1 | address valid for network | `PLUG_ADDRESS_FORMAT_AND_CHECKSUM` | UNPLUGGED | | `address-screening-service-provider-open` | service | 1 | 1 | address passes screening | `PLUG_ADDRESS_SCREENING_SERVICE_PROVIDER_OPEN` | UNPLUGGED | | `admitted-framework-source` | service | 1 | 1 | outcome reference exists in the admitted framework text | `PLUG_ADMITTED_FRAMEWORK_SOURCE` | UNPLUGGED | | `admitted-instrument` | service | 1 | 1 | published option | `PLUG_ADMITTED_INSTRUMENT` | UNPLUGGED | | `advance-notice-received-on-and-collection-date` | service | 1 | 1 | dates valid | `PLUG_ADVANCE_NOTICE_RECEIVED_ON_AND_COLLECTIO` | UNPLUGGED | | `affordability-consent-granted` | estate-record | 1 | 1 | consent ticked by the account holder | `PLUG_AFFORDABILITY_CONSENT_GRANTED` | UNPLUGGED | | `affordability-record` | estate-record | 1 | 1 | affordability assessed | `PLUG_AFFORDABILITY_RECORD` | UNPLUGGED | | `agent-display-name` | service | 1 | 1 | typed name matches | `PLUG_AGENT_DISPLAY_NAME` | UNPLUGGED | | `agent-limits-allowed-assets` | service | 1 | 1 | asset allowed for agent | `PLUG_AGENT_LIMITS_ALLOWED_ASSETS` | UNPLUGGED | | `agent-limits-allowed-recipients` | service | 1 | 1 | recipient allowed for agent | `PLUG_AGENT_LIMITS_ALLOWED_RECIPIENTS` | UNPLUGGED | | `agent-limits-and-period-spend` | service | 1 | 1 | within agent limits including this trade | `PLUG_AGENT_LIMITS_AND_PERIOD_SPEND` | UNPLUGGED | | `agent-limits-version` | service | 1 | 1 | version current | `PLUG_AGENT_LIMITS_VERSION` | UNPLUGGED | | `agent-mandate-and-account-holder-approval` | estate-record | 1 | 1 | within agent limit or approved | `PLUG_AGENT_MANDATE_AND_ACCOUNT_HOLDER_APPROVA` | UNPLUGGED | | `agent-mandate-and-chat-policy` | estate-record | 1 | 1 | agent may start chats | `PLUG_AGENT_MANDATE_AND_CHAT_POLICY` | UNPLUGGED | | `agent-mandate-and-document-classification` | estate-record | 1 | 1 | agent allowed to edit this classification | `PLUG_AGENT_MANDATE_AND_DOCUMENT_CLASSIFICATIO` | UNPLUGGED | | `agent-mandate-and-the-patient-s-recorded-instruction` | estate-record | 1 | 1 | agent acts within mandate and on instruction | `PLUG_AGENT_MANDATE_AND_THE_PATIENT_S_RECORDED` | UNPLUGGED | | `agent-mandate-and-warning-severity` | estate-record | 1 | 1 | agent is not continuing past a HIGH warning | `PLUG_AGENT_MANDATE_AND_WARNING_SEVERITY` | UNPLUGGED | | `agent-mandate-limits-and-categories` | estate-record | 1 | 1 | controls within mandate | `PLUG_AGENT_MANDATE_LIMITS_AND_CATEGORIES` | UNPLUGGED | | `agent-nomination-expires-at` | service | 1 | 1 | an expiry is set | `PLUG_AGENT_NOMINATION_EXPIRES_AT` | UNPLUGGED | | `agent-register-and-agent-mandate` | estate-record | 1 | 1 | agent has an accountable owner and an active mandate | `PLUG_AGENT_REGISTER_AND_AGENT_MANDATE` | UNPLUGGED | | `agent-registry` | dataset | 1 | 1 | owner or operator | `PLUG_AGENT_REGISTRY` | UNPLUGGED | | `agent-s-account-record` | estate-record | 1 | 1 | accountable owner named for agent accounts | `PLUG_AGENT_S_ACCOUNT_RECORD` | UNPLUGGED | | `agent-term-ends-on` | service | 1 | 1 | term set and after today | `PLUG_AGENT_TERM_ENDS_ON` | UNPLUGGED | | `agreed-rule-acceptance-criteria` | service | 1 | 1 | measured rates meet the agreed floor | `PLUG_AGREED_RULE_ACCEPTANCE_CRITERIA` | UNPLUGGED | | `agreement` | service | 1 | 1 | lawful basis recorded when personal data is shared | `PLUG_AGREEMENT` | UNPLUGGED | | `agreement-cooling-off-ends` | service | 1 | 1 | within cooling-off | `PLUG_AGREEMENT_COOLING_OFF_ENDS` | UNPLUGGED | | `agreement-data-schedule` | dataset | 1 | 1 | only allowed keys for the subject kind | `PLUG_AGREEMENT_DATA_SCHEDULE` | UNPLUGGED | | `agreement-versions` | service | 1 | 1 | version unchanged since others signed | `PLUG_AGREEMENT_VERSIONS` | UNPLUGGED | | `alerts-settings` | service | 1 | 1 | alerts turned on | `PLUG_ALERTS_SETTINGS` | UNPLUGGED | | `alerts-sweep-max` | service | 1 | 1 | within the set maximum for agents | `PLUG_ALERTS_SWEEP_MAX` | UNPLUGGED | | `allow-list-of-regulator-domains` | dataset | 1 | 1 | source is the regulator's own publication | `PLUG_ALLOW_LIST_OF_REGULATOR_DOMAINS` | UNPLUGGED | | `answer-record` | estate-record | 1 | 1 | answer belongs to the rater | `PLUG_ANSWER_RECORD` | UNPLUGGED | | `appeals-register` | dataset | 1 | 1 | no open appeal against ending | `PLUG_APPEALS_REGISTER` | UNPLUGGED | | `applicant-mobile-confirmation` | service | 1 | 1 | mobile confirmed | `PLUG_APPLICANT_MOBILE_CONFIRMATION` | UNPLUGGED | | `applicant-s-edge-id-consent-record` | estate-record | 1 | 1 | consent given by the applicant | `PLUG_APPLICANT_S_EDGE_ID_CONSENT_RECORD` | UNPLUGGED | | `application-date-of-birth-and-safeguarding-record` | estate-record | 1 | 1 | guardian or safeguarding contact recorded for learners under 18 | `PLUG_APPLICATION_DATE_OF_BIRTH_AND_SAFEGUARDI` | UNPLUGGED | | `applications` | service | 1 | 1 | one application per vacancy | `PLUG_APPLICATIONS` | UNPLUGGED | | `approval-approvers-and-agent-owners` | service | 1 | 1 | approver is neither the author nor an agent acting for the author | `PLUG_APPROVAL_APPROVERS_AND_AGENT_OWNERS` | UNPLUGGED | | `approval-digest` | service | 1 | 1 | learner approved this exact package when an agent submits | `PLUG_APPROVAL_DIGEST` | UNPLUGGED | | `approval-record` | estate-record | 1 | 1 | second approval | `PLUG_APPROVAL_RECORD` | UNPLUGGED | | `approvals` | service | 1 | 1 | second approver | `PLUG_APPROVALS` | UNPLUGGED | | `approved-escalation-protocol-at-minimum-any-answer-showing-t` | service | 1 | 1 | not a risk answer | `PLUG_APPROVED_ESCALATION_PROTOCOL_AT_MINIMUM_` | UNPLUGGED | | `approved-playbook` | service | 1 | 1 | action, target and duration are in the playbook | `PLUG_APPROVED_PLAYBOOK` | UNPLUGGED | | `approved-red-flag-rules-for-this-question` | service | 1 | 1 | not a red-flag answer | `PLUG_APPROVED_RED_FLAG_RULES_FOR_THIS_QUESTIO` | UNPLUGGED | | `approved-red-flag-set` | service | 1 | 1 | every question answered | `PLUG_APPROVED_RED_FLAG_SET` | UNPLUGGED | | `approved-red-flag-set-and-health-worry-check-applied-to-the-` | service | 1 | 1 | no health worry left unrouted | `PLUG_APPROVED_RED_FLAG_SET_AND_HEALTH_WORRY_C` | UNPLUGGED | | `approved-red-flag-set-applied-to-the-description` | service | 1 | 1 | no red flag in the words | `PLUG_APPROVED_RED_FLAG_SET_APPLIED_TO_THE_DES` | UNPLUGGED | | `approved-release-list` | dataset | 1 | 1 | software version approved | `PLUG_APPROVED_RELEASE_LIST` | UNPLUGGED | | `approved-symptom-list` | dataset | 1 | 1 | symptom on the approved list | `PLUG_APPROVED_SYMPTOM_LIST` | UNPLUGGED | | `arrears-and-collections-cases` | estate-record | 1 | 1 | no open arrears case | `PLUG_ARREARS_AND_COLLECTIONS_CASES` | UNPLUGGED | | `artist-confirmation` | service | 1 | 1 | transcribed lyrics confirmed by the artist | `PLUG_ARTIST_CONFIRMATION` | UNPLUGGED | | `assessment-rules-open` | service | 1 | 1 | assessment passed where required | `PLUG_ASSESSMENT_RULES_OPEN` | UNPLUGGED | | `assignment-register` | dataset | 1 | 1 | agent assigned to the problem | `PLUG_ASSIGNMENT_REGISTER` | UNPLUGGED | | `attempt-counter-limit-open` | service | 1 | 1 | attempts within the limit | `PLUG_ATTEMPT_COUNTER_LIMIT_OPEN` | UNPLUGGED | | `authenticator-or-recovery-code-check` | service | 1 | 1 | code correct | `PLUG_AUTHENTICATOR_OR_RECOVERY_CODE_CHECK` | UNPLUGGED | | `authenticator-verification` | service | 1 | 1 | code correct | `PLUG_AUTHENTICATOR_VERIFICATION` | UNPLUGGED | | `authorisation-engine` | service | 1 | 1 | mandate exists | `PLUG_AUTHORISATION_ENGINE` | UNPLUGGED | | `availability-working-hours` | service | 1 | 1 | hours within working hours | `PLUG_AVAILABILITY_WORKING_HOURS` | UNPLUGGED | | `backup-register-and-its-integrity-check` | dataset | 1 | 1 | restore source is known good | `PLUG_BACKUP_REGISTER_AND_ITS_INTEGRITY_CHECK` | UNPLUGGED | | `basis` | service | 1 | 1 | affordability is assessed on the applicant's own figures | `PLUG_BASIS` | UNPLUGGED | | `batch-creator` | service | 1 | 1 | agent created this batch | `PLUG_BATCH_CREATOR` | UNPLUGGED | | `batch-items` | service | 1 | 1 | item exists | `PLUG_BATCH_ITEMS` | UNPLUGGED | | `batch-register` | dataset | 1 | 1 | no re-screen already running for this list version | `PLUG_BATCH_REGISTER` | UNPLUGGED | | `batch-status-and-checkpoint` | service | 1 | 1 | resumable | `PLUG_BATCH_STATUS_AND_CHECKPOINT` | UNPLUGGED | | `batch-status-is-draft-or-ready` | service | 1 | 1 | batch editable | `PLUG_BATCH_STATUS_IS_DRAFT_OR_READY` | UNPLUGGED | | `batch-submitted-by` | service | 1 | 1 | approver is not the submitter | `PLUG_BATCH_SUBMITTED_BY` | UNPLUGGED | | `bereavement-unit-case-record` | estate-record | 1 | 1 | death verified in a bereavement case | `PLUG_BEREAVEMENT_UNIT_CASE_RECORD` | UNPLUGGED | | `bid-declaration` | service | 1 | 1 | declaration made | `PLUG_BID_DECLARATION` | UNPLUGGED | | `bid-register` | dataset | 1 | 1 | milestone belongs to the enactor | `PLUG_BID_REGISTER` | UNPLUGGED | | `birth-certificate-and-lifecycle-events` | service | 1 | 1 | agent born and not retired | `PLUG_BIRTH_CERTIFICATE_AND_LIFECYCLE_EVENTS` | UNPLUGGED | | `birth-certificate-and-lifecycle-events-of-the-successor` | service | 1 | 1 | successor valid | `PLUG_BIRTH_CERTIFICATE_AND_LIFECYCLE_EVENTS_O` | UNPLUGGED | | `birth-certificate-owner-accepts` | service | 1 | 1 | owner accepts accountability | `PLUG_BIRTH_CERTIFICATE_OWNER_ACCEPTS` | UNPLUGGED | | `book-ownership` | service | 1 | 1 | may write to this book | `PLUG_BOOK_OWNERSHIP` | UNPLUGGED | | `book-ownership-and-roles` | service | 1 | 1 | may write to this book | `PLUG_BOOK_OWNERSHIP_AND_ROLES` | UNPLUGGED | | `booking-page-agent-disclosure` | service | 1 | 1 | disclosure on when an agent may handle bookings | `PLUG_BOOKING_PAGE_AGENT_DISCLOSURE` | UNPLUGGED | | `booking-page-state` | service | 1 | 1 | page published | `PLUG_BOOKING_PAGE_STATE` | UNPLUGGED | | `booking-service` | service | 1 | 1 | slot still offered | `PLUG_BOOKING_SERVICE` | UNPLUGGED | | `breathing-space-register` | dataset | 1 | 1 | no breathing space in force | `PLUG_BREATHING_SPACE_REGISTER` | UNPLUGGED | | `build-register` | dataset | 1 | 1 | engine version exists | `PLUG_BUILD_REGISTER` | UNPLUGGED | | `bundle-store` | service | 1 | 1 | bundle exists and in scope | `PLUG_BUNDLE_STORE` | UNPLUGGED | | `buyer-delegation-record` | estate-record | 1 | 1 | second approval above threshold | `PLUG_BUYER_DELEGATION_RECORD` | UNPLUGGED | | `calendar-sharing` | service | 1 | 1 | each calendar shared at least as free and busy with the caller | `PLUG_CALENDAR_SHARING` | UNPLUGGED | | `calibration-run` | service | 1 | 1 | measured bound at or below the required bound | `PLUG_CALIBRATION_RUN` | UNPLUGGED | | `calibration-set-held-out-and-generation-manifest` | service | 1 | 1 | calibration set held out from generation | `PLUG_CALIBRATION_SET_HELD_OUT_AND_GENERATION_` | UNPLUGGED | | `callback-notes` | service | 1 | 1 | notes written | `PLUG_CALLBACK_NOTES` | UNPLUGGED | | `caller-kind` | service | 1 | 1 | only an owner recovers an agent | `PLUG_CALLER_KIND` | UNPLUGGED | | `candidate-agent-mandate` | estate-record | 1 | 1 | agent within candidate's criteria | `PLUG_CANDIDATE_AGENT_MANDATE` | UNPLUGGED | | `capture-requests` | service | 1 | 1 | no unexpired request already open | `PLUG_CAPTURE_REQUESTS` | UNPLUGGED | | `card-form-and-card-holder` | service | 1 | 1 | agents hold virtual cards only | `PLUG_CARD_FORM_AND_CARD_HOLDER` | UNPLUGGED | | `card-processor-merchant-category-list` | dataset | 1 | 1 | categories valid | `PLUG_CARD_PROCESSOR_MERCHANT_CATEGORY_LIST` | UNPLUGGED | | `card-processor-status` | service | 1 | 1 | card exists and is active | `PLUG_CARD_PROCESSOR_STATUS` | UNPLUGGED | | `case-queue` | estate-record | 1 | 1 | no open recall or dispute cases | `PLUG_CASE_QUEUE` | UNPLUGGED | | `case-store` | estate-record | 1 | 1 | case not locked by another investigator | `PLUG_CASE_STORE` | UNPLUGGED | | `caseload-register` | estate-record | 1 | 1 | has a caseload | `PLUG_CASELOAD_REGISTER` | UNPLUGGED | | `cases` | estate-record | 1 | 1 | one open query per line | `PLUG_CASES` | UNPLUGGED | | `certificate-and-build` | service | 1 | 1 | logic digest equals the certified digest | `PLUG_CERTIFICATE_AND_BUILD` | UNPLUGGED | | `certificate-store` | service | 1 | 1 | the document is not relied on by an issued certificate | `PLUG_CERTIFICATE_STORE` | UNPLUGGED | | `certification-expires-on` | service | 1 | 1 | not expired | `PLUG_CERTIFICATION_EXPIRES_ON` | UNPLUGGED | | `certification-recalibrate-by` | service | 1 | 1 | certificate not past its recalibration date | `PLUG_CERTIFICATION_RECALIBRATE_BY` | UNPLUGGED | | `challenge-queue` | estate-record | 1 | 1 | no open challenge on this check | `PLUG_CHALLENGE_QUEUE` | UNPLUGGED | | `challenge-response-reason` | service | 1 | 1 | reason answers each ground | `PLUG_CHALLENGE_RESPONSE_REASON` | UNPLUGGED | | `challenge-store` | service | 1 | 1 | no open challenge on the same item | `PLUG_CHALLENGE_STORE` | UNPLUGGED | | `chat-export` | service | 1 | 1 | range within 5 years | `PLUG_CHAT_EXPORT` | UNPLUGGED | | `chat-policy-delete-for-everyone` | service | 1 | 1 | scope allowed | `PLUG_CHAT_POLICY_DELETE_FOR_EVERYONE` | UNPLUGGED | | `chat-policy-edit-window-minutes` | service | 1 | 1 | inside edit window | `PLUG_CHAT_POLICY_EDIT_WINDOW_MINUTES` | UNPLUGGED | | `check-log` | service | 1 | 1 | the verification scope names checks that were actually run | `PLUG_CHECK_LOG` | UNPLUGGED | | `check-results` | service | 1 | 1 | a REFER, FAIL or UNKNOWN result exists | `PLUG_CHECK_RESULTS` | UNPLUGGED | | `check-service` | service | 1 | 1 | no check running | `PLUG_CHECK_SERVICE` | UNPLUGGED | | `claim-checker` | service | 1 | 1 | every claim has a source and an evidence status no stronger than its evidence | `PLUG_CLAIM_CHECKER` | UNPLUGGED | | `claim-declaration` | service | 1 | 1 | declaration made | `PLUG_CLAIM_DECLARATION` | UNPLUGGED | | `claim-status` | service | 1 | 1 | account holder confirms agent drafts | `PLUG_CLAIM_STATUS` | UNPLUGGED | | `clarification-queue` | estate-record | 1 | 1 | not already answered | `PLUG_CLARIFICATION_QUEUE` | UNPLUGGED | | `clinician-desk-queue` | estate-record | 1 | 1 | clinician assigned | `PLUG_CLINICIAN_DESK_QUEUE` | UNPLUGGED | | `clinician-register-and-review-queue` | estate-record | 1 | 1 | clinician registered and assigned | `PLUG_CLINICIAN_REGISTER_AND_REVIEW_QUEUE` | UNPLUGGED | | `closure-recurring-card-payments-acknowledged` | estate-record | 1 | 1 | recurring card payments acknowledged | `PLUG_CLOSURE_RECURRING_CARD_PAYMENTS_ACKNOWLE` | UNPLUGGED | | `closure-register` | dataset | 1 | 1 | no closure already in progress | `PLUG_CLOSURE_REGISTER` | UNPLUGGED | | `co-writer-acceptances-in-edge-id` | estate-record | 1 | 1 | every co-writer accepted their split | `PLUG_CO_WRITER_ACCEPTANCES_IN_EDGE_ID` | UNPLUGGED | | `collection-status` | service | 1 | 1 | collection was paid | `PLUG_COLLECTION_STATUS` | UNPLUGGED | | `commitment-risk-acknowledged` | estate-record | 1 | 1 | risk acknowledged | `PLUG_COMMITMENT_RISK_ACKNOWLEDGED` | UNPLUGGED | | `companies-house-api-a-production-government-service-whose-li` | dataset | 1 | 1 | company found on the register | `PLUG_COMPANIES_HOUSE_API_A_PRODUCTION_GOVERNM` | UNPLUGGED | | `companies-house-identity-verification` | dataset | 1 | 1 | personal code matches the claimant's verified identity | `PLUG_COMPANIES_HOUSE_IDENTITY_VERIFICATION` | UNPLUGGED | | `companies-house-officers-and-people-with-significant-control` | dataset | 1 | 1 | differences from the register are explained | `PLUG_COMPANIES_HOUSE_OFFICERS_AND_PEOPLE_WITH` | UNPLUGGED | | `concern-summary-confirmed` | service | 1 | 1 | summary confirmed | `PLUG_CONCERN_SUMMARY_CONFIRMED` | UNPLUGGED | | `confirmation-of-payee-and-ledger` | estate-record | 1 | 1 | check services reachable | `PLUG_CONFIRMATION_OF_PAYEE_AND_LEDGER` | UNPLUGGED | | `confirmation-of-payee-or-account-register` | estate-record | 1 | 1 | destination is in the proposing holder's own name | `PLUG_CONFIRMATION_OF_PAYEE_OR_ACCOUNT_REGISTE` | UNPLUGGED | | `conflict-declarations` | service | 1 | 1 | reviewer has no link to artist or claimant | `PLUG_CONFLICT_DECLARATIONS` | UNPLUGGED | | `conflict-of-interest-register` | dataset | 1 | 1 | decider has no declared conflict with the applicant | `PLUG_CONFLICT_OF_INTEREST_REGISTER` | UNPLUGGED | | `connection-register` | dataset | 1 | 1 | source connected | `PLUG_CONNECTION_REGISTER` | UNPLUGGED | | `connection-scopes` | service | 1 | 1 | balances in scope | `PLUG_CONNECTION_SCOPES` | UNPLUGGED | | `connection-scopes-and-connection-status` | service | 1 | 1 | connections active with transactions in scope | `PLUG_CONNECTION_SCOPES_AND_CONNECTION_STATUS` | UNPLUGGED | | `connection-status-and-connection-expires-at` | service | 1 | 1 | consent active | `PLUG_CONNECTION_STATUS_AND_CONNECTION_EXPIRES` | UNPLUGGED | | `connections` | service | 1 | 1 | connection exists | `PLUG_CONNECTIONS` | UNPLUGGED | | `consent-register-and-data-providers` | estate-record | 1 | 1 | enough consented records | `PLUG_CONSENT_REGISTER_AND_DATA_PROVIDERS` | UNPLUGGED | | `consent-register-and-legal-basis` | estate-record | 1 | 1 | the business consented or the law permits this access | `PLUG_CONSENT_REGISTER_AND_LEGAL_BASIS` | UNPLUGGED | | `consented-account-data-any-minimum-period-is-open-the-fca-s-` | estate-record | 1 | 1 | enough account history | `PLUG_CONSENTED_ACCOUNT_DATA_ANY_MINIMUM_PERIO` | UNPLUGGED | | `consignment-custody-plan` | service | 1 | 1 | receiving party is the next expected custodian | `PLUG_CONSIGNMENT_CUSTODY_PLAN` | UNPLUGGED | | `consignment-register` | dataset | 1 | 1 | reference unique | `PLUG_CONSIGNMENT_REGISTER` | UNPLUGGED | | `constitution-data-and-signatures` | dataset | 1 | 1 | enough signatories | `PLUG_CONSTITUTION_DATA_AND_SIGNATURES` | UNPLUGGED | | `contact-deleted-at-within-30-days` | service | 1 | 1 | still restorable | `PLUG_CONTACT_DELETED_AT_WITHIN_30_DAYS` | UNPLUGGED | | `contact-log` | service | 1 | 1 | contact attempted for CRISIS | `PLUG_CONTACT_LOG` | UNPLUGGED | | `contact-outcome` | service | 1 | 1 | closing needs a real contact | `PLUG_CONTACT_OUTCOME` | UNPLUGGED | | `contacts` | service | 1 | 1 | same book | `PLUG_CONTACTS` | UNPLUGGED | | `contest-register` | dataset | 1 | 1 | no open contest on this entry | `PLUG_CONTEST_REGISTER` | UNPLUGGED | | `contract-notice-weeks` | service | 1 | 1 | notice period respected or both agree | `PLUG_CONTRACT_NOTICE_WEEKS` | UNPLUGGED | | `controls-version` | service | 1 | 1 | version current | `PLUG_CONTROLS_VERSION` | UNPLUGGED | | `conversation-expires-at` | service | 1 | 1 | conversation not expired | `PLUG_CONVERSATION_EXPIRES_AT` | UNPLUGGED | | `conversation-history-from` | service | 1 | 1 | within shared history | `PLUG_CONVERSATION_HISTORY_FROM` | UNPLUGGED | | `conversation-kind` | service | 1 | 1 | not a direct chat | `PLUG_CONVERSATION_KIND` | UNPLUGGED | | `cost-estimate-made-before-the-job` | service | 1 | 1 | estimated cost within the owner's or mandate's budget | `PLUG_COST_ESTIMATE_MADE_BEFORE_THE_JOB` | UNPLUGGED | | `council-calendar-and-the-notice-s-publication-time` | service | 1 | 1 | 28 clear days for a key decision | `PLUG_COUNCIL_CALENDAR_AND_THE_NOTICE_S_PUBLIC` | UNPLUGGED | | `court-order-effect` | service | 1 | 1 | effect is one we can carry out | `PLUG_COURT_ORDER_EFFECT` | UNPLUGGED | | `coverage-map-open-no-edge-points-or-travelling-devices-exist` | estate-record | 1 | 1 | address inside device visit coverage | `PLUG_COVERAGE_MAP_OPEN_NO_EDGE_POINTS_OR_TRAV` | UNPLUGGED | | `credential-list` | dataset | 1 | 1 | not the last way to sign in | `PLUG_CREDENTIAL_LIST` | UNPLUGGED | | `credit-applications` | service | 1 | 1 | application open | `PLUG_CREDIT_APPLICATIONS` | UNPLUGGED | | `criteria` | service | 1 | 1 | weights add up to 100 | `PLUG_CRITERIA` | UNPLUGGED | | `current-and-proposed-mandate` | estate-record | 1 | 1 | mandate not widened | `PLUG_CURRENT_AND_PROPOSED_MANDATE` | UNPLUGGED | | `dataset-provenance-records-training-data-rules-in-research-j` | estate-record | 1 | 1 | every dataset item licensed or consented for training | `PLUG_DATASET_PROVENANCE_RECORDS_TRAINING_DATA` | UNPLUGGED | | `debarment-list-and-the-supplier-s-shared-vetting-report` | dataset | 1 | 1 | winner not excluded or debarred | `PLUG_DEBARMENT_LIST_AND_THE_SUPPLIER_S_SHARED` | UNPLUGGED | | `decision-alternatives-rejected` | service | 1 | 1 | alternatives stated | `PLUG_DECISION_ALTERNATIVES_REJECTED` | UNPLUGGED | | `decision-call-in-deadline` | service | 1 | 1 | before the call-in deadline | `PLUG_DECISION_CALL_IN_DEADLINE` | UNPLUGGED | | `decision-check` | service | 1 | 1 | the reasons name the figures relied on and, for a decline, what would change it | `PLUG_DECISION_CHECK` | UNPLUGGED | | `decision-decision-type-and-status-urgent-decisions-excluded-` | service | 1 | 1 | decision can be called in | `PLUG_DECISION_DECISION_TYPE_AND_STATUS_URGENT` | UNPLUGGED | | `decision-feature-register-checked-against-the-equality-act-2` | dataset | 1 | 1 | no protected characteristic or proxy among decision inputs | `PLUG_DECISION_FEATURE_REGISTER_CHECKED_AGAINS` | UNPLUGGED | | `decision-ledger` | estate-record | 1 | 1 | the offer is unexpired and unchanged | `PLUG_DECISION_LEDGER` | UNPLUGGED | | `decision-pre-contract-read` | service | 1 | 1 | pre-contract information read | `PLUG_DECISION_PRE_CONTRACT_READ` | UNPLUGGED | | `decision-register` | dataset | 1 | 1 | paper exists | `PLUG_DECISION_REGISTER` | UNPLUGGED | | `declaration-and-dispensation-register` | dataset | 1 | 1 | withdraw for a disclosable pecuniary interest unless dispensed | `PLUG_DECLARATION_AND_DISPENSATION_REGISTER` | UNPLUGGED | | `declared-interests-register` | dataset | 1 | 1 | no conflict of interest | `PLUG_DECLARED_INTERESTS_REGISTER` | UNPLUGGED | | `delegation` | service | 1 | 1 | caller owns the mailbox or holds an active read delegation | `PLUG_DELEGATION` | UNPLUGGED | | `delegation-approval-required-for-and-the-session-s-data-expo` | dataset | 1 | 1 | approval present when required | `PLUG_DELEGATION_APPROVAL_REQUIRED_FOR_AND_THE` | UNPLUGGED | | `delegation-recipient-scope` | service | 1 | 1 | recipients within scope | `PLUG_DELEGATION_RECIPIENT_SCOPE` | UNPLUGGED | | `delegation-scope` | service | 1 | 1 | caller may organise this mailbox | `PLUG_DELEGATION_SCOPE` | UNPLUGGED | | `delegation-state` | service | 1 | 1 | delegation exists and is active | `PLUG_DELEGATION_STATE` | UNPLUGGED | | `deletion-confirmed` | service | 1 | 1 | confirmed | `PLUG_DELETION_CONFIRMED` | UNPLUGGED | | `deployment-previous-version` | service | 1 | 1 | previous version retained | `PLUG_DEPLOYMENT_PREVIOUS_VERSION` | UNPLUGGED | | `design-rule` | service | 1 | 1 | none that can block emergency guidance | `PLUG_DESIGN_RULE` | UNPLUGGED | | `designation-register` | dataset | 1 | 1 | signer holds the statutory role or a nomination | `PLUG_DESIGNATION_REGISTER` | UNPLUGGED | | `desk-rota` | service | 1 | 1 | clinician desk open | `PLUG_DESK_ROTA` | UNPLUGGED | | `destination-register` | dataset | 1 | 1 | caller is the receiving officer | `PLUG_DESTINATION_REGISTER` | UNPLUGGED | | `device-attestation-status` | service | 1 | 1 | device attested | `PLUG_DEVICE_ATTESTATION_STATUS` | UNPLUGGED | | `device-calibration-record` | estate-record | 1 | 1 | camera calibration current | `PLUG_DEVICE_CALIBRATION_RECORD` | UNPLUGGED | | `device-chip-reader` | service | 1 | 1 | chip readable | `PLUG_DEVICE_CHIP_READER` | UNPLUGGED | | `device-manufacturer-s-attestation-root` | service | 1 | 1 | hardware attestation verifies | `PLUG_DEVICE_MANUFACTURER_S_ATTESTATION_ROOT` | UNPLUGGED | | `difference-between-versions` | service | 1 | 1 | only the owner widens | `PLUG_DIFFERENCE_BETWEEN_VERSIONS` | UNPLUGGED | | `directory-area-list` | dataset | 1 | 1 | area recognised | `PLUG_DIRECTORY_AREA_LIST` | UNPLUGGED | | `directory-entry` | dataset | 1 | 1 | service accepting referrals | `PLUG_DIRECTORY_ENTRY` | UNPLUGGED | | `directory-index` | dataset | 1 | 1 | no likely duplicate | `PLUG_DIRECTORY_INDEX` | UNPLUGGED | | `directory-visibility-rules` | dataset | 1 | 1 | entry visible to caller | `PLUG_DIRECTORY_VISIBILITY_RULES` | UNPLUGGED | | `dispatch-record` | estate-record | 1 | 1 | seal matches dispatch | `PLUG_DISPATCH_RECORD` | UNPLUGGED | | `dispensation-register` | dataset | 1 | 1 | dispensation in date | `PLUG_DISPENSATION_REGISTER` | UNPLUGGED | | `disposal-owner-confirms` | service | 1 | 1 | owner confirmed | `PLUG_DISPOSAL_OWNER_CONFIRMS` | UNPLUGGED | | `disposal-record` | estate-record | 1 | 1 | data disposal recorded | `PLUG_DISPOSAL_RECORD` | UNPLUGGED | | `dispute` | service | 1 | 1 | refund no more than the amount disputed | `PLUG_DISPUTE` | UNPLUGGED | | `dispute-outcome-reason` | service | 1 | 1 | reason answers each ground | `PLUG_DISPUTE_OUTCOME_REASON` | UNPLUGGED | | `dispute-records` | estate-record | 1 | 1 | not already disputed | `PLUG_DISPUTE_RECORDS` | UNPLUGGED | | `dispute-rules-period-open` | service | 1 | 1 | within the appeal period | `PLUG_DISPUTE_RULES_PERIOD_OPEN` | UNPLUGGED | | `dispute-type-and-actor` | service | 1 | 1 | agent not withdrawing a scam claim | `PLUG_DISPUTE_TYPE_AND_ACTOR` | UNPLUGGED | | `dispute-type-rules-time-limits-open` | service | 1 | 1 | within the time limit for this dispute type | `PLUG_DISPUTE_TYPE_RULES_TIME_LIMITS_OPEN` | UNPLUGGED | | `document-and-file-permissions` | service | 1 | 1 | attachments within the sender's data scope | `PLUG_DOCUMENT_AND_FILE_PERMISSIONS` | UNPLUGGED | | `document-classification` | service | 1 | 1 | outside sharing allowed by classification | `PLUG_DOCUMENT_CLASSIFICATION` | UNPLUGGED | | `document-classifier-with-its-error-bound` | service | 1 | 1 | document is the kind stated | `PLUG_DOCUMENT_CLASSIFIER_WITH_ITS_ERROR_BOUND` | UNPLUGGED | | `document-extracted-name-and-dates` | service | 1 | 1 | chip details match the document photo | `PLUG_DOCUMENT_EXTRACTED_NAME_AND_DATES` | UNPLUGGED | | `document-reader` | service | 1 | 1 | file readable | `PLUG_DOCUMENT_READER` | UNPLUGGED | | `domain-settings-sender-rules-in-research-judiverse-competito` | service | 1 | 1 | domain verified for sending with SPF, DKIM and DMARC | `PLUG_DOMAIN_SETTINGS_SENDER_RULES_IN_RESEARCH` | UNPLUGGED | | `edge-id-agent-record` | estate-record | 1 | 1 | agent subject has a verified principal | `PLUG_EDGE_ID_AGENT_RECORD` | UNPLUGGED | | `edge-id-and-the-account-s-mandates` | estate-record | 1 | 1 | caller can see the account | `PLUG_EDGE_ID_AND_THE_ACCOUNT_S_MANDATES` | UNPLUGGED | | `edge-id-challenge-store` | estate-record | 1 | 1 | challenge still valid | `PLUG_EDGE_ID_CHALLENGE_STORE` | UNPLUGGED | | `edge-id-customer-record` | estate-record | 1 | 1 | destination name is the holder's | `PLUG_EDGE_ID_CUSTOMER_RECORD` | UNPLUGGED | | `edge-id-customer-record-kyc-for-people-kyb-for-organisations` | estate-record | 1 | 1 | owner is a verified customer | `PLUG_EDGE_ID_CUSTOMER_RECORD_KYC_FOR_PEOPLE_K` | UNPLUGGED | | `edge-id-directory-or-email-check` | estate-record | 1 | 1 | recipient reachable | `PLUG_EDGE_ID_DIRECTORY_OR_EMAIL_CHECK` | UNPLUGGED | | `edge-id-identity-register-including-retired-identities` | estate-record | 1 | 1 | identity never issued before | `PLUG_EDGE_ID_IDENTITY_REGISTER_INCLUDING_RETI` | UNPLUGGED | | `edge-id-identity-type` | estate-record | 1 | 1 | agents never sign | `PLUG_EDGE_ID_IDENTITY_TYPE` | UNPLUGGED | | `edge-id-key-register-including-revoked-keys` | estate-record | 1 | 1 | public key never registered before | `PLUG_EDGE_ID_KEY_REGISTER_INCLUDING_REVOKED_K` | UNPLUGGED | | `edge-id-key-registration` | estate-record | 1 | 1 | public key registered from the agent's own host | `PLUG_EDGE_ID_KEY_REGISTRATION` | UNPLUGGED | | `edge-id-of-approver-and-planner` | estate-record | 1 | 1 | approver is not the planner | `PLUG_EDGE_ID_OF_APPROVER_AND_PLANNER` | UNPLUGGED | | `edge-id-of-caller` | estate-record | 1 | 1 | caller is the account holder | `PLUG_EDGE_ID_OF_CALLER` | UNPLUGGED | | `edge-id-of-the-session` | estate-record | 1 | 1 | answered by the invitee | `PLUG_EDGE_ID_OF_THE_SESSION` | UNPLUGGED | | `edge-id-organisation-record` | estate-record | 1 | 1 | regulator identity verified | `PLUG_EDGE_ID_ORGANISATION_RECORD` | UNPLUGGED | | `edge-id-organisation-register` | estate-record | 1 | 1 | buyer exists | `PLUG_EDGE_ID_ORGANISATION_REGISTER` | UNPLUGGED | | `edge-id-permissions` | estate-record | 1 | 1 | principal may read every selected private source | `PLUG_EDGE_ID_PERMISSIONS` | UNPLUGGED | | `edge-id-permissions-and-for-agents-the-agent-s-mandate` | estate-record | 1 | 1 | principal may perform every step | `PLUG_EDGE_ID_PERMISSIONS_AND_FOR_AGENTS_THE_A` | UNPLUGGED | | `edge-id-permissions-for-the-asker` | estate-record | 1 | 1 | asker may read every chosen scope | `PLUG_EDGE_ID_PERMISSIONS_FOR_THE_ASKER` | UNPLUGGED | | `edge-id-record` | estate-record | 1 | 1 | every shared claim still verified | `PLUG_EDGE_ID_RECORD` | UNPLUGGED | | `edge-id-registry` | estate-record | 1 | 1 | agent has an EDGE ID with its builder and principal recorded | `PLUG_EDGE_ID_REGISTRY` | UNPLUGGED | | `edge-id-role` | estate-record | 1 | 1 | approver is the clinical safety officer | `PLUG_EDGE_ID_ROLE` | UNPLUGGED | | `edge-id-session` | estate-record | 1 | 1 | trusting needs a strong sign-in on that device | `PLUG_EDGE_ID_SESSION` | UNPLUGGED | | `edge-id-session-age-strong-customer-authentication-is-requir` | estate-record | 1 | 1 | recent strong sign-in | `PLUG_EDGE_ID_SESSION_AGE_STRONG_CUSTOMER_AUTH` | UNPLUGGED | | `edge-id-verified-date-of-birth` | estate-record | 1 | 1 | aged 18 or over | `PLUG_EDGE_ID_VERIFIED_DATE_OF_BIRTH` | UNPLUGGED | | `edge-point-calendar` | estate-record | 1 | 1 | slot still free | `PLUG_EDGE_POINT_CALENDAR` | UNPLUGGED | | `edge-point-directory-open-no-edge-points-or-travelling-devic` | estate-record | 1 | 1 | at least one point or device visit serves the area | `PLUG_EDGE_POINT_DIRECTORY_OPEN_NO_EDGE_POINTS` | UNPLUGGED | | `emergency-access-register` | dataset | 1 | 1 | access live | `PLUG_EMERGENCY_ACCESS_REGISTER` | UNPLUGGED | | `employer-account` | estate-record | 1 | 1 | funds available | `PLUG_EMPLOYER_ACCOUNT` | UNPLUGGED | | `enactor-bid-status` | service | 1 | 1 | bid eligible | `PLUG_ENACTOR_BID_STATUS` | UNPLUGGED | | `entry-date-and-the-date-entered-by-the-councillor` | service | 1 | 1 | within 28 days of becoming aware | `PLUG_ENTRY_DATE_AND_THE_DATE_ENTERED_BY_THE_C` | UNPLUGGED | | `estate-capability-catalogue` | estate-record | 1 | 1 | every capability is in the catalogue | `PLUG_ESTATE_CAPABILITY_CATALOGUE` | UNPLUGGED | | `estate-clock` | estate-record | 1 | 1 | term ends after today | `PLUG_ESTATE_CLOCK` | UNPLUGGED | | `estate-key-policy` | estate-record | 1 | 1 | algorithm approved | `PLUG_ESTATE_KEY_POLICY` | UNPLUGGED | | `evaluation-record` | estate-record | 1 | 1 | every compliant bid fully scored | `PLUG_EVALUATION_RECORD` | UNPLUGGED | | `event-actor-and-agent-identity` | service | 1 | 1 | agent answers only for its own action | `PLUG_EVENT_ACTOR_AND_AGENT_IDENTITY` | UNPLUGGED | | `event-and-scheduling-mandate` | estate-record | 1 | 1 | caller may cancel | `PLUG_EVENT_AND_SCHEDULING_MANDATE` | UNPLUGGED | | `event-starts-at` | service | 1 | 1 | event not in the past | `PLUG_EVENT_STARTS_AT` | UNPLUGGED | | `evidence-policy` | estate-record | 1 | 1 | the payment is above the policy threshold or risk indicates | `PLUG_EVIDENCE_POLICY` | UNPLUGGED | | `expert-queue` | estate-record | 1 | 1 | assigned to this expert | `PLUG_EXPERT_QUEUE` | UNPLUGGED | | `expert-rota-for-the-asker-s-organisation` | service | 1 | 1 | a named human role is on duty for this subject | `PLUG_EXPERT_ROTA_FOR_THE_ASKER_S_ORGANISATION` | UNPLUGGED | | `expiry-date-read-from-the-document` | service | 1 | 1 | document not expired | `PLUG_EXPIRY_DATE_READ_FROM_THE_DOCUMENT` | UNPLUGGED | | `face-match-procedure` | service | 1 | 1 | face matches the chip or document photo | `PLUG_FACE_MATCH_PROCEDURE` | UNPLUGGED | | `face-match-procedure-with-its-error-bound` | service | 1 | 1 | face matches chip photo | `PLUG_FACE_MATCH_PROCEDURE_WITH_ITS_ERROR_BOUN` | UNPLUGGED | | `feedback-records` | estate-record | 1 | 1 | one rating per answer per rater | `PLUG_FEEDBACK_RECORDS` | UNPLUGGED | | `figure-store` | service | 1 | 1 | figure exists | `PLUG_FIGURE_STORE` | UNPLUGGED | | `file-parser` | service | 1 | 1 | file readable | `PLUG_FILE_PARSER` | UNPLUGGED | | `finding-register` | dataset | 1 | 1 | not already recorded | `PLUG_FINDING_REGISTER` | UNPLUGGED | | `fix-record` | estate-record | 1 | 1 | fix not already decided | `PLUG_FIX_RECORD` | UNPLUGGED | | `fix-register` | dataset | 1 | 1 | fix marked done | `PLUG_FIX_REGISTER` | UNPLUGGED | | `fixed-word-list-checking-for-age-date-of-birth-photo-or-mari` | dataset | 1 | 1 | no excluded criterion in the text | `PLUG_FIXED_WORD_LIST_CHECKING_FOR_AGE_DATE_OF` | UNPLUGGED | | `follow-up-due-on` | service | 1 | 1 | date in the future | `PLUG_FOLLOW_UP_DUE_ON` | UNPLUGGED | | `forbearance-history-for-the-loan` | service | 1 | 1 | forbearance offered first | `PLUG_FORBEARANCE_HISTORY_FOR_THE_LOAN` | UNPLUGGED | | `forbearance-policy` | service | 1 | 1 | option within published policy | `PLUG_FORBEARANCE_POLICY` | UNPLUGGED | | `form` | service | 1 | 1 | signed to publish | `PLUG_FORM` | UNPLUGGED | | `form-state` | service | 1 | 1 | form open | `PLUG_FORM_STATE` | UNPLUGGED | | `freeze-frozen-by` | service | 1 | 1 | agent froze it itself | `PLUG_FREEZE_FROZEN_BY` | UNPLUGGED | | `funder-s-award-record` | estate-record | 1 | 1 | award confirmed by the funder | `PLUG_FUNDER_S_AWARD_RECORD` | UNPLUGGED | | `funding-risk-acknowledged` | estate-record | 1 | 1 | risk acknowledged | `PLUG_FUNDING_RISK_ACKNOWLEDGED` | UNPLUGGED | | `funding-status-and-project-status` | service | 1 | 1 | before first release or project stopped | `PLUG_FUNDING_STATUS_AND_PROJECT_STATUS` | UNPLUGGED | | `grader-certification-register` | dataset | 1 | 1 | grader certification current | `PLUG_GRADER_CERTIFICATION_REGISTER` | UNPLUGGED | | `grant-expiry` | service | 1 | 1 | access not expired | `PLUG_GRANT_EXPIRY` | UNPLUGGED | | `grant-register` | dataset | 1 | 1 | grant exists and is active | `PLUG_GRANT_REGISTER` | UNPLUGGED | | `grant-store` | service | 1 | 1 | scope within the opener's own authority | `PLUG_GRANT_STORE` | UNPLUGGED | | `grantor-s-mandate` | estate-record | 1 | 1 | delegation no wider than the grantor's own authority | `PLUG_GRANTOR_S_MANDATE` | UNPLUGGED | | `holdings` | service | 1 | 1 | enough held including network fee | `PLUG_HOLDINGS` | UNPLUGGED | | `identifier-resolver` | service | 1 | 1 | identifier resolves | `PLUG_IDENTIFIER_RESOLVER` | UNPLUGGED | | `identity-check-result` | service | 1 | 1 | identity re-verified as the account holder | `PLUG_IDENTITY_CHECK_RESULT` | UNPLUGGED | | `identity-match` | service | 1 | 1 | officer is not the claimant | `PLUG_IDENTITY_MATCH` | UNPLUGGED | | `identity-provider-claims-or-practice-register` | dataset | 1 | 1 | NHS number passes its check digit and matches | `PLUG_IDENTITY_PROVIDER_CLAIMS_OR_PRACTICE_REG` | UNPLUGGED | | `image-quality-procedure` | service | 1 | 1 | image is readable | `PLUG_IMAGE_QUALITY_PROCEDURE` | UNPLUGGED | | `inbound-screen-quarantined` | service | 1 | 1 | message is quarantined | `PLUG_INBOUND_SCREEN_QUARANTINED` | UNPLUGGED | | `incident-register` | dataset | 1 | 1 | alerts not already in an open incident | `PLUG_INCIDENT_REGISTER` | UNPLUGGED | | `income-and-expenditure-or-account-data` | estate-record | 1 | 1 | plan affordable | `PLUG_INCOME_AND_EXPENDITURE_OR_ACCOUNT_DATA` | UNPLUGGED | | `inputs` | service | 1 | 1 | goal valid | `PLUG_INPUTS` | UNPLUGGED | | `inspection-record` | estate-record | 1 | 1 | any change from the request is recorded with a reason the regulator can see | `PLUG_INSPECTION_RECORD` | UNPLUGGED | | `inspection-register` | dataset | 1 | 1 | inspection still active | `PLUG_INSPECTION_REGISTER` | UNPLUGGED | | `institution-directory` | dataset | 1 | 1 | not already registered | `PLUG_INSTITUTION_DIRECTORY` | UNPLUGGED | | `invitation-register` | dataset | 1 | 1 | invitation still live | `PLUG_INVITATION_REGISTER` | UNPLUGGED | | `invitation-store` | service | 1 | 1 | link valid and unused | `PLUG_INVITATION_STORE` | UNPLUGGED | | `invitation-terms-accepted` | service | 1 | 1 | terms accepted when accepting | `PLUG_INVITATION_TERMS_ACCEPTED` | UNPLUGGED | | `invoice-records` | estate-record | 1 | 1 | not already invoiced | `PLUG_INVOICE_RECORDS` | UNPLUGGED | | `invoice-status` | service | 1 | 1 | invoice matched to an accepted milestone | `PLUG_INVOICE_STATUS` | UNPLUGGED | | `item-count` | service | 1 | 2 | at most 1,000 items; at least one item | `PLUG_ITEM_COUNT` | UNPLUGGED | | `item-flags-including-judebox-uploads` | service | 1 | 1 | no item carries a do-not-train flag | `PLUG_ITEM_FLAGS_INCLUDING_JUDEBOX_UPLOADS` | UNPLUGGED | | `item-record` | estate-record | 1 | 2 | not the preparer; item still waiting | `PLUG_ITEM_RECORD` | UNPLUGGED | | `item-status` | service | 1 | 1 | at least one FAILED or CANCELLED item | `PLUG_ITEM_STATUS` | UNPLUGGED | | `item-warning-acknowledged` | estate-record | 1 | 1 | every warning acknowledged | `PLUG_ITEM_WARNING_ACKNOWLEDGED` | UNPLUGGED | | `joint-account-operating-rule` | estate-record | 1 | 1 | rule differs from the current rule | `PLUG_JOINT_ACCOUNT_OPERATING_RULE` | UNPLUGGED | | `joint-account-policy-two-holders-decision-recorded-in-this-u` | estate-record | 1 | 1 | no more than two holders | `PLUG_JOINT_ACCOUNT_POLICY_TWO_HOLDERS_DECISIO` | UNPLUGGED | | `joint-account-status` | estate-record | 1 | 1 | not already frozen | `PLUG_JOINT_ACCOUNT_STATUS` | UNPLUGGED | | `kya-result` | service | 1 | 1 | checks passed or reviewer approved | `PLUG_KYA_RESULT` | UNPLUGGED | | `lab-catalogue` | dataset | 1 | 1 | procedure exists and is deployed | `PLUG_LAB_CATALOGUE` | UNPLUGGED | | `latest-check-result` | service | 1 | 1 | paper has passed the one-page check | `PLUG_LATEST_CHECK_RESULT` | UNPLUGGED | | `learner-record-of-completed-modules` | estate-record | 1 | 1 | prerequisites passed | `PLUG_LEARNER_RECORD_OF_COMPLETED_MODULES` | UNPLUGGED | | `ledger-and-holdings` | estate-record | 1 | 1 | enough money or holding | `PLUG_LEDGER_AND_HOLDINGS` | UNPLUGGED | | `ledger-and-lending-records` | estate-record | 1 | 1 | nothing owed | `PLUG_LEDGER_AND_LENDING_RECORDS` | UNPLUGGED | | `ledger-balance-checked-again-before-each-item` | estate-record | 1 | 1 | enough money for the remaining items | `PLUG_LEDGER_BALANCE_CHECKED_AGAIN_BEFORE_EACH` | UNPLUGGED | | `ledger-or-sign-in-log` | estate-record | 1 | 1 | event exists | `PLUG_LEDGER_OR_SIGN_IN_LOG` | UNPLUGGED | | `legal-hold-register` | dataset | 1 | 1 | no legal hold | `PLUG_LEGAL_HOLD_REGISTER` | UNPLUGGED | | `legal-minimum-rate-table-in-force-on-the-closing-date` | service | 1 | 1 | pay not below the legal minimum for the hours | `PLUG_LEGAL_MINIMUM_RATE_TABLE_IN_FORCE_ON_THE` | UNPLUGGED | | `legal-minimum-rates-open-source` | service | 1 | 1 | hours, holiday and pay at or above statutory minimums | `PLUG_LEGAL_MINIMUM_RATES_OPEN_SOURCE` | UNPLUGGED | | `lender-eligibility-record` | estate-record | 1 | 1 | eligible to lend | `PLUG_LENDER_ELIGIBILITY_RECORD` | UNPLUGGED | | `list-of-contract-stage-notices-open-until-confirmed-from-the` | dataset | 1 | 1 | notice type recognised | `PLUG_LIST_OF_CONTRACT_STAGE_NOTICES_OPEN_UNTI` | UNPLUGGED | | `list-register-uk-sanctions-list-un-security-council-consolid` | dataset | 1 | 1 | every required list loaded at its current version | `PLUG_LIST_REGISTER_UK_SANCTIONS_LIST_UN_SECUR` | UNPLUGGED | | `loan-ledger` | estate-record | 1 | 1 | agreement belongs to the borrower | `PLUG_LOAN_LEDGER` | UNPLUGGED | | `loan-ledger-and-mandate` | estate-record | 1 | 1 | agreement belongs to the viewer or principal | `PLUG_LOAN_LEDGER_AND_MANDATE` | UNPLUGGED | | `mail-store-and-delegation` | service | 1 | 1 | message exists in a mailbox the caller may read | `PLUG_MAIL_STORE_AND_DELEGATION` | UNPLUGGED | | `mailbox-sending-state` | service | 1 | 2 | agent mailbox vouched for outside sending; mailbox not paused | `PLUG_MAILBOX_SENDING_STATE` | UNPLUGGED | | `mandate-change-revoke-confirmed` | estate-record | 1 | 1 | confirmed | `PLUG_MANDATE_CHANGE_REVOKE_CONFIRMED` | UNPLUGGED | | `mandate-expires-at` | estate-record | 1 | 1 | not expired | `PLUG_MANDATE_EXPIRES_AT` | UNPLUGGED | | `mandate-scope-and-test-catalogue` | estate-record | 1 | 1 | suite covers every tool and includes refusal tests outside the mandate | `PLUG_MANDATE_SCOPE_AND_TEST_CATALOGUE` | UNPLUGGED | | `market-fields` | service | 1 | 1 | filters valid | `PLUG_MARKET_FIELDS` | UNPLUGGED | | `medicine-list` | dataset | 1 | 1 | list size within the limit for one check | `PLUG_MEDICINE_LIST` | UNPLUGGED | | `meeting-start-time-and-council-calendar` | service | 1 | 1 | five clear days | `PLUG_MEETING_START_TIME_AND_COUNCIL_CALENDAR` | UNPLUGGED | | `member-identity` | service | 1 | 1 | not changing own access | `PLUG_MEMBER_IDENTITY` | UNPLUGGED | | `member-register` | dataset | 1 | 1 | caller is a serving councillor | `PLUG_MEMBER_REGISTER` | UNPLUGGED | | `membership-confirmation-purpose` | service | 1 | 1 | purpose chosen | `PLUG_MEMBERSHIP_CONFIRMATION_PURPOSE` | UNPLUGGED | | `membership-register-and-agent-mandate` | estate-record | 1 | 1 | purpose allowed for this member and actor | `PLUG_MEMBERSHIP_REGISTER_AND_AGENT_MANDATE` | UNPLUGGED | | `merge-field-choices` | service | 1 | 1 | every difference resolved | `PLUG_MERGE_FIELD_CHOICES` | UNPLUGGED | | `merge-record` | estate-record | 1 | 1 | within undo window | `PLUG_MERGE_RECORD` | UNPLUGGED | | `message-sender` | service | 1 | 1 | author | `PLUG_MESSAGE_SENDER` | UNPLUGGED | | `message-snooze-until` | service | 1 | 1 | snooze time in the future | `PLUG_MESSAGE_SNOOZE_UNTIL` | UNPLUGGED | | `message-state` | service | 1 | 1 | message still held in the outbox | `PLUG_MESSAGE_STATE` | UNPLUGGED | | `message-status` | service | 1 | 1 | still scheduled | `PLUG_MESSAGE_STATUS` | UNPLUGGED | | `meter-register` | dataset | 1 | 1 | a MEASURED figure names a meter or receiver that exists | `PLUG_METER_REGISTER` | UNPLUGGED | | `milestone-amount` | service | 1 | 1 | amount within milestone | `PLUG_MILESTONE_AMOUNT` | UNPLUGGED | | `milestone-list` | dataset | 1 | 1 | milestone amounts add up to the amount requested | `PLUG_MILESTONE_LIST` | UNPLUGGED | | `milestone-plan` | service | 1 | 1 | milestone is due or overdue | `PLUG_MILESTONE_PLAN` | UNPLUGGED | | `model-licence-record-licence-flags-in-research-section-5-1` | estate-record | 1 | 1 | model licence permits fine-tuning for this use | `PLUG_MODEL_LICENCE_RECORD_LICENCE_FLAGS_IN_RE` | UNPLUGGED | | `model-translation` | model | 1 | 0 | | `PLUG_MODEL_TRANSLATION` | UNPLUGGED | | `module-capacity` | service | 1 | 1 | places left | `PLUG_MODULE_CAPACITY` | UNPLUGGED | | `module-dates` | service | 1 | 1 | before the add-and-drop deadline | `PLUG_MODULE_DATES` | UNPLUGGED | | `module-roster` | service | 1 | 1 | currently enrolled | `PLUG_MODULE_ROSTER` | UNPLUGGED | | `monitor-rule-parser` | service | 1 | 1 | threshold is measurable | `PLUG_MONITOR_RULE_PARSER` | UNPLUGGED | | `network-evidence-standard` | estate-record | 1 | 1 | evidence standard met: fraud confirmed, not suspected | `PLUG_NETWORK_EVIDENCE_STANDARD` | UNPLUGGED | | `notes-classifier-with-stated-error-bound` | service | 1 | 1 | no special-category notes | `PLUG_NOTES_CLASSIFIER_WITH_STATED_ERROR_BOUND` | UNPLUGGED | | `notice-publication-time` | service | 1 | 1 | notice period met for key decisions | `PLUG_NOTICE_PUBLICATION_TIME` | UNPLUGGED | | `number-validation` | service | 1 | 1 | valid phone | `PLUG_NUMBER_VALIDATION` | UNPLUGGED | | `occurrence-status-and-due-date` | service | 1 | 1 | occurrence due and not already sent | `PLUG_OCCURRENCE_STATUS_AND_DUE_DATE` | UNPLUGGED | | `offer-acceptance-record` | estate-record | 1 | 1 | candidate consent | `PLUG_OFFER_ACCEPTANCE_RECORD` | UNPLUGGED | | `offer-conditions` | service | 1 | 1 | right to work checked and other conditions met | `PLUG_OFFER_CONDITIONS` | UNPLUGGED | | `offer-expiry` | service | 1 | 1 | offer still open | `PLUG_OFFER_EXPIRY` | UNPLUGGED | | `offer-status` | service | 1 | 1 | not yet accepted | `PLUG_OFFER_STATUS` | UNPLUGGED | | `officer-list` | dataset | 1 | 1 | caller is an authorised officer | `PLUG_OFFICER_LIST` | UNPLUGGED | | `officer-list-or-caller-edge-id` | estate-record | 1 | 1 | caller is an officer or the agent itself | `PLUG_OFFICER_LIST_OR_CALLER_EDGE_ID` | UNPLUGGED | | `one-time-code-confirmations` | service | 1 | 1 | email and mobile confirmed | `PLUG_ONE_TIME_CODE_CONFIRMATIONS` | UNPLUGGED | | `open-acceptance-limits-for-the-estate-s-own-protocol-largest` | estate-record | 1 | 1 | acceptance limits set | `PLUG_OPEN_ACCEPTANCE_LIMITS_FOR_THE_ESTATE_S_` | UNPLUGGED | | `open-banking-directory` | dataset | 1 | 1 | bank supports account information access | `PLUG_OPEN_BANKING_DIRECTORY` | UNPLUGGED | | `open-contracting-data-standard-schema` | dataset | 1 | 1 | release matches the schema | `PLUG_OPEN_CONTRACTING_DATA_STANDARD_SCHEMA` | UNPLUGGED | | `open-eye-service-directory-and-referral-routes-are-not-agree` | estate-record | 1 | 1 | service accepts referrals of this kind | `PLUG_OPEN_EYE_SERVICE_DIRECTORY_AND_REFERRAL_` | UNPLUGGED | | `open-how-a-check-reaches-a-gp-practice-or-pharmacy-research-` | service | 1 | 1 | recipient reachable | `PLUG_OPEN_HOW_A_CHECK_REACHES_A_GP_PRACTICE_O` | UNPLUGGED | | `open-how-court-orders-are-verified-with-the-issuing-court-an` | estate-record | 1 | 1 | order verified with the issuing court | `PLUG_OPEN_HOW_COURT_ORDERS_ARE_VERIFIED_WITH_` | UNPLUGGED | | `open-judebox-s-licence-with-prs-for-music-and-other-collecti` | service | 1 | 1 | composition licence in force where a collecting society manages the composition | `PLUG_OPEN_JUDEBOX_S_LICENCE_WITH_PRS_FOR_MUSI` | UNPLUGGED | | `open-personal-account-unit-result` | estate-record | 1 | 1 | invitee holds a verified EDGE Finance account or will open one before accepting | `PLUG_OPEN_PERSONAL_ACCOUNT_UNIT_RESULT` | UNPLUGGED | | `open-the-estate-s-approved-error-bound-methods-and-minimum-c` | estate-record | 1 | 1 | bound method approved | `PLUG_OPEN_THE_ESTATE_S_APPROVED_ERROR_BOUND_M` | UNPLUGGED | | `open-the-per-play-royalty-rate-and-how-it-relates-to-listene` | estate-record | 1 | 1 | per-play rate agreed | `PLUG_OPEN_THE_PER_PLAY_ROYALTY_RATE_AND_HOW_I` | UNPLUGGED | | `open-threshold-values-for-each-signal-to-be-set-by-the-revie` | service | 1 | 1 | threshold rules approved | `PLUG_OPEN_THRESHOLD_VALUES_FOR_EACH_SIGNAL_TO` | UNPLUGGED | | `open-which-licensed-urgent-care-triage-protocol-is-used-and-` | estate-record | 1 | 1 | licensed protocol available | `PLUG_OPEN_WHICH_LICENSED_URGENT_CARE_TRIAGE_P` | UNPLUGGED | | `operations-case-queue` | estate-record | 1 | 1 | no open fraud review | `PLUG_OPERATIONS_CASE_QUEUE` | UNPLUGGED | | `opportunity-clarification-deadline` | service | 1 | 1 | before the clarification deadline | `PLUG_OPPORTUNITY_CLARIFICATION_DEADLINE` | UNPLUGGED | | `opportunity-criteria` | service | 1 | 1 | criterion was published | `PLUG_OPPORTUNITY_CRITERIA` | UNPLUGGED | | `opportunity-deadlines` | service | 1 | 1 | deadlines in order and in the future | `PLUG_OPPORTUNITY_DEADLINES` | UNPLUGGED | | `opportunity-register` | dataset | 1 | 1 | opportunity still open | `PLUG_OPPORTUNITY_REGISTER` | UNPLUGGED | | `organisation-legal-name` | service | 1 | 1 | entity name matches the register | `PLUG_ORGANISATION_LEGAL_NAME` | UNPLUGGED | | `organisation-register` | dataset | 1 | 1 | lead institution registered | `PLUG_ORGANISATION_REGISTER` | UNPLUGGED | | `organisation-registry-status` | dataset | 1 | 1 | company is active on the register | `PLUG_ORGANISATION_REGISTRY_STATUS` | UNPLUGGED | | `organisation-roles` | service | 1 | 1 | within the approver's own approval limit | `PLUG_ORGANISATION_ROLES` | UNPLUGGED | | `organisation-roles-and-agent-mandate` | estate-record | 1 | 1 | caller holds the approver role | `PLUG_ORGANISATION_ROLES_AND_AGENT_MANDATE` | UNPLUGGED | | `organisation-s-item-key-register` | dataset | 1 | 1 | item keys not used before | `PLUG_ORGANISATION_S_ITEM_KEY_REGISTER` | UNPLUGGED | | `organisation-s-own-permissions` | service | 1 | 1 | scope within what the organisation holds | `PLUG_ORGANISATION_S_OWN_PERMISSIONS` | UNPLUGGED | | `organisation-s-question-policy` | service | 1 | 1 | no question asks about protected characteristics | `PLUG_ORGANISATION_S_QUESTION_POLICY` | UNPLUGGED | | `organisation-sharing-policy` | service | 1 | 1 | recipient allowed | `PLUG_ORGANISATION_SHARING_POLICY` | UNPLUGGED | | `organisation-signatory-register` | dataset | 1 | 1 | caller is an authorised signatory | `PLUG_ORGANISATION_SIGNATORY_REGISTER` | UNPLUGGED | | `organisation-type-table` | service | 1 | 1 | type is one that has no register | `PLUG_ORGANISATION_TYPE_TABLE` | UNPLUGGED | | `other-pots-rules` | service | 1 | 1 | round-ups go to one pot only | `PLUG_OTHER_POTS_RULES` | UNPLUGGED | | `outcome-health-worry-and-the-red-flag-set` | service | 1 | 1 | health worry routed | `PLUG_OUTCOME_HEALTH_WORRY_AND_THE_RED_FLAG_SE` | UNPLUGGED | | `outcome-record` | estate-record | 1 | 1 | states move one step at a time | `PLUG_OUTCOME_RECORD` | UNPLUGGED | | `overdraft-limit` | service | 1 | 1 | new limit lower than current | `PLUG_OVERDRAFT_LIMIT` | UNPLUGGED | | `overdraft-used` | service | 1 | 1 | new limit not below what is used | `PLUG_OVERDRAFT_USED` | UNPLUGGED | | `owner-approvals` | service | 1 | 1 | owner approved for agent accounts | `PLUG_OWNER_APPROVALS` | UNPLUGGED | | `owner-s-approval` | service | 1 | 1 | owner approves an agent key | `PLUG_OWNER_S_APPROVAL` | UNPLUGGED | | `panel-declarations` | service | 1 | 1 | no conflict of interest | `PLUG_PANEL_DECLARATIONS` | UNPLUGGED | | `panel-record` | estate-record | 1 | 1 | approver is not an evaluator | `PLUG_PANEL_RECORD` | UNPLUGGED | | `panel-rules-quorum-size-open-until-the-solution-bank-s-rules` | service | 1 | 1 | panel quorum reached | `PLUG_PANEL_RULES_QUORUM_SIZE_OPEN_UNTIL_THE_S` | UNPLUGGED | | `paper-digest` | service | 1 | 1 | paper unchanged since the check | `PLUG_PAPER_DIGEST` | UNPLUGGED | | `partner-agent-mandate` | estate-record | 1 | 1 | agent acceptance pre-approved | `PLUG_PARTNER_AGENT_MANDATE` | UNPLUGGED | | `partner-list-and-organisation-register` | dataset | 1 | 1 | verifier independent of every partner | `PLUG_PARTNER_LIST_AND_ORGANISATION_REGISTER` | UNPLUGGED | | `partner-record` | estate-record | 1 | 1 | not already answered | `PLUG_PARTNER_RECORD` | UNPLUGGED | | `patient-record` | estate-record | 1 | 1 | list exists and has at least two matched medicines | `PLUG_PATIENT_RECORD` | UNPLUGGED | | `payee-list-from-the-send-money-capability` | dataset | 1 | 1 | payee saved and checked | `PLUG_PAYEE_LIST_FROM_THE_SEND_MONEY_CAPABILIT` | UNPLUGGED | | `payee-s-bank` | service | 1 | 1 | Confirmation of Payee returns MATCH | `PLUG_PAYEE_S_BANK` | UNPLUGGED | | `payer-s-request-settings` | service | 1 | 1 | payer accepts requests from this requester | `PLUG_PAYER_S_REQUEST_SETTINGS` | UNPLUGGED | | `payrun-status-and-digest` | service | 1 | 1 | rehearsed and unchanged since | `PLUG_PAYRUN_STATUS_AND_DIGEST` | UNPLUGGED | | `payruns` | service | 1 | 1 | period not already paid | `PLUG_PAYRUNS` | UNPLUGGED | | `people-list` | dataset | 1 | 1 | at least one director or equivalent | `PLUG_PEOPLE_LIST` | UNPLUGGED | | `per-identity-rate-counter-60-searches-a-minute` | service | 1 | 1 | rate within limit | `PLUG_PER_IDENTITY_RATE_COUNTER_60_SEARCHES_A_` | UNPLUGGED | | `person-check-status` | service | 1 | 1 | every director and owner passed their identity check | `PLUG_PERSON_CHECK_STATUS` | UNPLUGGED | | `person-control-description` | service | 1 | 1 | every owner's control is explained | `PLUG_PERSON_CONTROL_DESCRIPTION` | UNPLUGGED | | `person-s-confirmation` | service | 1 | 1 | contact agreed | `PLUG_PERSON_S_CONFIRMATION` | UNPLUGGED | | `personal-data-scan` | dataset | 1 | 1 | no personal data about third parties | `PLUG_PERSONAL_DATA_SCAN` | UNPLUGGED | | `phone-s-chip-reader` | service | 1 | 1 | chip readable | `PLUG_PHONE_S_CHIP_READER` | UNPLUGGED | | `plan-approval` | service | 1 | 1 | approved when approval was needed | `PLUG_PLAN_APPROVAL` | UNPLUGGED | | `plan-builder` | service | 1 | 1 | instruction resolves to one plan | `PLUG_PLAN_BUILDER` | UNPLUGGED | | `plan-needs` | service | 1 | 1 | at least one need | `PLUG_PLAN_NEEDS` | UNPLUGGED | | `plans-credits-and-billing-ledger` | estate-record | 1 | 1 | allowance remains | `PLUG_PLANS_CREDITS_AND_BILLING_LEDGER` | UNPLUGGED | | `plausibility-check` | service | 1 | 1 | speed between readings physically plausible | `PLUG_PLAUSIBILITY_CHECK` | UNPLUGGED | | `play-records` | estate-record | 1 | 1 | every counted play accounted for | `PLUG_PLAY_RECORDS` | UNPLUGGED | | `policy-fields` | service | 1 | 1 | policy is complete | `PLUG_POLICY_FIELDS` | UNPLUGGED | | `policy-rules` | service | 1 | 1 | self-approval impossible | `PLUG_POLICY_RULES` | UNPLUGGED | | `policy-version` | service | 1 | 1 | approver is eligible | `PLUG_POLICY_VERSION` | UNPLUGGED | | `pot-created-by` | service | 1 | 1 | agent created this pot | `PLUG_POT_CREATED_BY` | UNPLUGGED | | `pot-goal-amount-and-pot-target-date` | service | 1 | 1 | goal valid | `PLUG_POT_GOAL_AMOUNT_AND_POT_TARGET_DATE` | UNPLUGGED | | `ppn-017` | service | 1 | 1 | AI-use disclosure not scored | `PLUG_PPN_017` | UNPLUGGED | | `practice-identity-policy` | service | 1 | 1 | documents meet the practice's accepted list | `PLUG_PRACTICE_IDENTITY_POLICY` | UNPLUGGED | | `prepare-payrun-procedure` | service | 1 | 1 | final pay calculable | `PLUG_PREPARE_PAYRUN_PROCEDURE` | UNPLUGGED | | `preview-digest` | service | 1 | 1 | preview unchanged since shown | `PLUG_PREVIEW_DIGEST` | UNPLUGGED | | `principal-confirmation-record` | estate-record | 1 | 1 | principal confirmed when an agent prepared it | `PLUG_PRINCIPAL_CONFIRMATION_RECORD` | UNPLUGGED | | `problem-baseline` | service | 1 | 1 | baseline exists or is stated UNKNOWN | `PLUG_PROBLEM_BASELINE` | UNPLUGGED | | `problem-register-similarity-check` | dataset | 1 | 1 | not a duplicate | `PLUG_PROBLEM_REGISTER_SIMILARITY_CHECK` | UNPLUGGED | | `procedure-authorship-record` | estate-record | 1 | 1 | certifier did not author or request it | `PLUG_PROCEDURE_AUTHORSHIP_RECORD` | UNPLUGGED | | `procedure-catalogue` | dataset | 1 | 1 | no certified procedure already covers this scope | `PLUG_PROCEDURE_CATALOGUE` | UNPLUGGED | | `procedure-scope-record` | estate-record | 1 | 1 | camera model and image inside the procedure's certified scope | `PLUG_PROCEDURE_SCOPE_RECORD` | UNPLUGGED | | `professional-register-check` | dataset | 1 | 1 | clinician registration current | `PLUG_PROFESSIONAL_REGISTER_CHECK` | UNPLUGGED | | `programme-entry-requirements` | service | 1 | 1 | a reason for each requirement assessed | `PLUG_PROGRAMME_ENTRY_REQUIREMENTS` | UNPLUGGED | | `programme-rules` | service | 1 | 1 | credit total within the programme's range | `PLUG_PROGRAMME_RULES` | UNPLUGGED | | `programme-schema` | service | 1 | 1 | filters are allowed values | `PLUG_PROGRAMME_SCHEMA` | UNPLUGGED | | `project-membership` | service | 1 | 2 | project visible to creator; assignee can see the project | `PLUG_PROJECT_MEMBERSHIP` | UNPLUGGED | | `project-roles-and-agent-ownership` | service | 1 | 1 | caller may take over from the current holder | `PLUG_PROJECT_ROLES_AND_AGENT_OWNERSHIP` | UNPLUGGED | | `prompt-injection-controls-research-judiverse-competitors-reg` | service | 1 | 1 | content read from files, mail or web pages is treated as data, never as instruct | `PLUG_PROMPT_INJECTION_CONTROLS_RESEARCH_JUDIV` | UNPLUGGED | | `prompt-reason` | service | 1 | 1 | reason given unless the prescription was changed | `PLUG_PROMPT_REASON` | UNPLUGGED | | `protocol` | service | 1 | 1 | answer is one of the question's options | `PLUG_PROTOCOL` | UNPLUGGED | | `public-dns-read-from-two-independent-resolvers` | service | 1 | 1 | proof record published | `PLUG_PUBLIC_DNS_READ_FROM_TWO_INDEPENDENT_RES` | UNPLUGGED | | `public-register` | dataset | 1 | 1 | decision exists and is published | `PLUG_PUBLIC_REGISTER` | UNPLUGGED | | `published-domain-verification-record` | estate-record | 1 | 1 | domain proven | `PLUG_PUBLISHED_DOMAIN_VERIFICATION_RECORD` | UNPLUGGED | | `published-eligibility-rules-lending-to-consumers-needs-fca-c` | dataset | 1 | 1 | eligibility rules met (for example legal capacity to borrow), used as rules only | `PLUG_PUBLISHED_ELIGIBILITY_RULES_LENDING_TO_C` | UNPLUGGED | | `published-sort-code-and-modulus-tables` | service | 1 | 1 | modulus check passes | `PLUG_PUBLISHED_SORT_CODE_AND_MODULUS_TABLES` | UNPLUGGED | | `published-terms` | service | 1 | 1 | current terms version | `PLUG_PUBLISHED_TERMS` | UNPLUGGED | | `purpose-to-scope-table` | service | 1 | 1 | only the data the purpose needs | `PLUG_PURPOSE_TO_SCOPE_TABLE` | UNPLUGGED | | `question-definition` | service | 1 | 1 | answer allowed | `PLUG_QUESTION_DEFINITION` | UNPLUGGED | | `question-record` | estate-record | 1 | 1 | question belongs to the asker | `PLUG_QUESTION_RECORD` | UNPLUGGED | | `quote` | service | 1 | 1 | quote still valid | `PLUG_QUOTE` | UNPLUGGED | | `quote-expiry` | service | 1 | 1 | quote still valid | `PLUG_QUOTE_EXPIRY` | UNPLUGGED | | `quote-risk-level` | service | 1 | 1 | risk acknowledged when warned | `PLUG_QUOTE_RISK_LEVEL` | UNPLUGGED | | `racetrack-run-result` | service | 1 | 1 | latest racetrack run for this build passed | `PLUG_RACETRACK_RUN_RESULT` | UNPLUGGED | | `racetrack-run-result-for-the-candidate-digest` | service | 1 | 1 | candidate passed the racetrack | `PLUG_RACETRACK_RUN_RESULT_FOR_THE_CANDIDATE_D` | UNPLUGGED | | `racetrack-test-runner` | service | 1 | 1 | generated tests pass | `PLUG_RACETRACK_TEST_RUNNER` | UNPLUGGED | | `rate-limiter` | service | 1 | 1 | within rate limit | `PLUG_RATE_LIMITER` | UNPLUGGED | | `rate-register` | dataset | 1 | 1 | within query limit | `PLUG_RATE_REGISTER` | UNPLUGGED | | `reading-store` | service | 1 | 1 | reading not older than the last one | `PLUG_READING_STORE` | UNPLUGGED | | `recommendation-owner` | service | 1 | 1 | caller owns the recommendation | `PLUG_RECOMMENDATION_OWNER` | UNPLUGGED | | `record-register` | estate-record | 1 | 1 | record not already claimed by another identity | `PLUG_RECORD_REGISTER` | UNPLUGGED | | `recovery-code-digests` | service | 1 | 1 | recovery code valid and unused | `PLUG_RECOVERY_CODE_DIGESTS` | UNPLUGGED | | `recovery-records` | estate-record | 1 | 1 | every affected asset recovered or written off | `PLUG_RECOVERY_RECORDS` | UNPLUGGED | | `referral-consent-to-share` | estate-record | 1 | 1 | consent ticked | `PLUG_REFERRAL_CONSENT_TO_SHARE` | UNPLUGGED | | `referral-shared-details-allowed-list` | dataset | 1 | 1 | only allowed details | `PLUG_REFERRAL_SHARED_DETAILS_ALLOWED_LIST` | UNPLUGGED | | `referral-status` | service | 1 | 1 | not already closed | `PLUG_REFERRAL_STATUS` | UNPLUGGED | | `register` | dataset | 1 | 1 | register reachable | `PLUG_REGISTER` | UNPLUGGED | | `register-digest` | dataset | 1 | 1 | list unchanged since shown | `PLUG_REGISTER_DIGEST` | UNPLUGGED | | `register-status` | dataset | 1 | 1 | organisation not closed | `PLUG_REGISTER_STATUS` | UNPLUGGED | | `registrar-s-condition-check` | service | 1 | 1 | conditions met for a conditional offer | `PLUG_REGISTRAR_S_CONDITION_CHECK` | UNPLUGGED | | `regulator-portal-policy-maximum-term-open` | service | 1 | 1 | term within the agreed maximum | `PLUG_REGULATOR_PORTAL_POLICY_MAXIMUM_TERM_OPE` | UNPLUGGED | | `regulator-s-public-register-read-on-the-day` | dataset | 1 | 1 | on the regulator register | `PLUG_REGULATOR_S_PUBLIC_REGISTER_READ_ON_THE_` | UNPLUGGED | | `rehearsal-log` | service | 1 | 1 | a rehearsal of the same file digest within 24 hours before commit | `PLUG_REHEARSAL_LOG` | UNPLUGGED | | `rehearsal-valid-10-minutes` | service | 1 | 1 | rehearsal still valid | `PLUG_REHEARSAL_VALID_10_MINUTES` | UNPLUGGED | | `reminder-last-sent-at` | service | 1 | 1 | no reminder in the last 24 hours | `PLUG_REMINDER_LAST_SENT_AT` | UNPLUGGED | | `report` | service | 1 | 1 | check is in the supplier's report | `PLUG_REPORT` | UNPLUGGED | | `report-include-sources` | service | 1 | 1 | sources included | `PLUG_REPORT_INCLUDE_SOURCES` | UNPLUGGED | | `report-records` | estate-record | 1 | 1 | every required report acknowledged or explained | `PLUG_REPORT_RECORDS` | UNPLUGGED | | `report-shares` | service | 1 | 1 | caller may read the report | `PLUG_REPORT_SHARES` | UNPLUGGED | | `report-store` | service | 1 | 1 | not a duplicate | `PLUG_REPORT_STORE` | UNPLUGGED | | `reproducibility-manifest` | service | 1 | 1 | build reproducible from its manifest | `PLUG_REPRODUCIBILITY_MANIFEST` | UNPLUGGED | | `request-abuse-limit-set-by-the-payments-operations-lead` | service | 1 | 1 | within request rate | `PLUG_REQUEST_ABUSE_LIMIT_SET_BY_THE_PAYMENTS_` | UNPLUGGED | | `request-amount-validation` | service | 1 | 1 | amount valid | `PLUG_REQUEST_AMOUNT_VALIDATION` | UNPLUGGED | | `request-payment-risk-level` | service | 1 | 1 | risk acknowledged when warned | `PLUG_REQUEST_PAYMENT_RISK_LEVEL` | UNPLUGGED | | `request-received-at` | service | 1 | 1 | within statutory time | `PLUG_REQUEST_RECEIVED_AT` | UNPLUGGED | | `request-store-and-staff-role` | service | 1 | 1 | caller sent the request or is operations | `PLUG_REQUEST_STORE_AND_STAFF_ROLE` | UNPLUGGED | | `result-outcome` | service | 1 | 1 | outcome NOT_VERIFIED or REFER | `PLUG_RESULT_OUTCOME` | UNPLUGGED | | `retirement-register` | dataset | 1 | 1 | not already retiring | `PLUG_RETIREMENT_REGISTER` | UNPLUGGED | | `review-changed-level` | service | 1 | 1 | level chosen when changing | `PLUG_REVIEW_CHANGED_LEVEL` | UNPLUGGED | | `review-changed-outcome` | service | 1 | 1 | new outcome chosen when changing | `PLUG_REVIEW_CHANGED_OUTCOME` | UNPLUGGED | | `review-clinician-note` | service | 1 | 1 | note written | `PLUG_REVIEW_CLINICIAN_NOTE` | UNPLUGGED | | `review-note` | service | 1 | 1 | note written | `PLUG_REVIEW_NOTE` | UNPLUGGED | | `review-reasons` | service | 1 | 1 | reasons given when the decision differs from the automated grade | `PLUG_REVIEW_REASONS` | UNPLUGGED | | `review-record` | estate-record | 1 | 1 | assigned to this reviewer | `PLUG_REVIEW_RECORD` | UNPLUGGED | | `reviewer-availability` | service | 1 | 1 | reviewer accepts reviews | `PLUG_REVIEWER_AVAILABILITY` | UNPLUGGED | | `revocation-confirmations` | service | 1 | 1 | keys revoked everywhere | `PLUG_REVOCATION_CONFIRMATIONS` | UNPLUGGED | | `revocation-records` | estate-record | 1 | 1 | keys not yet revoked | `PLUG_REVOCATION_RECORDS` | UNPLUGGED | | `right-to-work-share-code` | service | 1 | 1 | share code format | `PLUG_RIGHT_TO_WORK_SHARE_CODE` | UNPLUGGED | | `risk-check-completed-at` | service | 1 | 1 | any cooling-off period has passed (OPEN whether one applies) | `PLUG_RISK_CHECK_COMPLETED_AT` | UNPLUGGED | | `risk-check-warning-acknowledged` | estate-record | 1 | 1 | warning acknowledged | `PLUG_RISK_CHECK_WARNING_ACKNOWLEDGED` | UNPLUGGED | | `role-table-per-organisation-type` | service | 1 | 1 | role exists for the type | `PLUG_ROLE_TABLE_PER_ORGANISATION_TYPE` | UNPLUGGED | | `roster-schema` | service | 1 | 1 | every required column mapped | `PLUG_ROSTER_SCHEMA` | UNPLUGGED | | `rota` | service | 1 | 1 | every hour covered | `PLUG_ROTA` | UNPLUGGED | | `route-table` | service | 1 | 1 | route offered for this type | `PLUG_ROUTE_TABLE` | UNPLUGGED | | `routing-procedure-scope-statement` | service | 1 | 1 | inside the procedure's certified scope | `PLUG_ROUTING_PROCEDURE_SCOPE_STATEMENT` | UNPLUGGED | | `rule-compiler` | service | 1 | 1 | rule compiles | `PLUG_RULE_COMPILER` | UNPLUGGED | | `rule-register` | dataset | 1 | 1 | active rule set loaded and verified | `PLUG_RULE_REGISTER` | UNPLUGGED | | `run-history` | service | 1 | 1 | approver did not do the work | `PLUG_RUN_HISTORY` | UNPLUGGED | | `run-record-and-shares` | estate-record | 1 | 1 | caller may read the report | `PLUG_RUN_RECORD_AND_SHARES` | UNPLUGGED | | `run-starter-and-project-roles` | service | 1 | 1 | caller may pause or resume | `PLUG_RUN_STARTER_AND_PROJECT_ROLES` | UNPLUGGED | | `sanctions-screening` | dataset | 1 | 1 | sanctions screening clear | `PLUG_SANCTIONS_SCREENING` | UNPLUGGED | | `saving-rules` | service | 1 | 1 | schedule complete | `PLUG_SAVING_RULES` | UNPLUGGED | | `savings-pots` | service | 1 | 1 | pot belongs to the account holder | `PLUG_SAVINGS_POTS` | UNPLUGGED | | `schedule-end-rule` | service | 1 | 1 | end valid | `PLUG_SCHEDULE_END_RULE` | UNPLUGGED | | `scheduling-mandate` | estate-record | 1 | 1 | within the agent's hours, meeting types and attendee scope | `PLUG_SCHEDULING_MANDATE` | UNPLUGGED | | `scheduling-mandate-attendee-scope` | estate-record | 1 | 1 | attendees within the agent's attendee scope | `PLUG_SCHEDULING_MANDATE_ATTENDEE_SCOPE` | UNPLUGGED | | `scheduling-mandate-may-move-existing` | estate-record | 1 | 1 | agent allowed to move accepted events | `PLUG_SCHEDULING_MANDATE_MAY_MOVE_EXISTING` | UNPLUGGED | | `scheduling-mandate-state` | estate-record | 1 | 1 | mandate active | `PLUG_SCHEDULING_MANDATE_STATE` | UNPLUGGED | | `scheme-evidence-checklist-open` | estate-record | 1 | 1 | required evidence present for the reason | `PLUG_SCHEME_EVIDENCE_CHECKLIST_OPEN` | UNPLUGGED | | `scheme-of-delegation` | service | 1 | 1 | caller is the named decision maker | `PLUG_SCHEME_OF_DELEGATION` | UNPLUGGED | | `scheme-of-delegation-held-as-data` | dataset | 1 | 1 | decision maker holds the power | `PLUG_SCHEME_OF_DELEGATION_HELD_AS_DATA` | UNPLUGGED | | `scheme-rules-open` | service | 1 | 1 | within the scheme window | `PLUG_SCHEME_RULES_OPEN` | UNPLUGGED | | `scrutiny-outcome-votes` | service | 1 | 1 | votes recorded | `PLUG_SCRUTINY_OUTCOME_VOTES` | UNPLUGGED | | `search-query` | service | 1 | 1 | query length | `PLUG_SEARCH_QUERY` | UNPLUGGED | | `second-calibration-run` | service | 1 | 1 | deterministic rerun matches | `PLUG_SECOND_CALIBRATION_RUN` | UNPLUGGED | | `separation-my-share-percent` | service | 1 | 1 | share is 0 to 100 percent | `PLUG_SEPARATION_MY_SHARE_PERCENT` | UNPLUGGED | | `service-setting` | service | 1 | 1 | range within the allowed length | `PLUG_SERVICE_SETTING` | UNPLUGGED | | `session` | service | 1 | 1 | every question answered | `PLUG_SESSION` | UNPLUGGED | | `session-chip-result` | service | 1 | 1 | chip photo available | `PLUG_SESSION_CHIP_RESULT` | UNPLUGGED | | `session-list` | dataset | 1 | 1 | session belongs to the subject | `PLUG_SESSION_LIST` | UNPLUGGED | | `session-presence-attested` | service | 1 | 1 | officer attests presence | `PLUG_SESSION_PRESENCE_ATTESTED` | UNPLUGGED | | `session-recording-consent` | estate-record | 1 | 1 | recording only with consent | `PLUG_SESSION_RECORDING_CONSENT` | UNPLUGGED | | `setup-authorisation-status-and-valid-until` | service | 1 | 1 | matched agent authorisation still open | `PLUG_SETUP_AUTHORISATION_STATUS_AND_VALID_UNT` | UNPLUGGED | | `share-and-emergency-registers` | dataset | 1 | 1 | share or emergency access still live | `PLUG_SHARE_AND_EMERGENCY_REGISTERS` | UNPLUGGED | | `share-consent-given` | estate-record | 1 | 1 | consent for this recipient | `PLUG_SHARE_CONSENT_GIVEN` | UNPLUGGED | | `share-expiry` | service | 1 | 1 | share not expired | `PLUG_SHARE_EXPIRY` | UNPLUGGED | | `share-grant-expires-on` | service | 1 | 1 | expiry after today | `PLUG_SHARE_GRANT_EXPIRES_ON` | UNPLUGGED | | `share-register` | dataset | 1 | 1 | share live | `PLUG_SHARE_REGISTER` | UNPLUGGED | | `shares` | service | 1 | 1 | mentioned people can see the document | `PLUG_SHARES` | UNPLUGGED | | `simulation-conflicts` | service | 1 | 1 | somebody can approve every item | `PLUG_SIMULATION_CONFLICTS` | UNPLUGGED | | `site-register` | dataset | 1 | 1 | origin and destination are known sites or valid points | `PLUG_SITE_REGISTER` | UNPLUGGED | | `solution-review-status` | service | 1 | 1 | solution accepted for funding | `PLUG_SOLUTION_REVIEW_STATUS` | UNPLUGGED | | `source-authorisation-response` | service | 1 | 1 | source authorised the connection | `PLUG_SOURCE_AUTHORISATION_RESPONSE` | UNPLUGGED | | `spdx-licence-list` | dataset | 1 | 1 | licence is a known SPDX identifier or proprietary with owner | `PLUG_SPDX_LICENCE_LIST` | UNPLUGGED | | `split-participants` | service | 1 | 2 | at least one other participant; each participant listed once | `PLUG_SPLIT_PARTICIPANTS` | UNPLUGGED | | `split-shares-and-split-total` | service | 1 | 1 | shares add up to the total | `PLUG_SPLIT_SHARES_AND_SPLIT_TOTAL` | UNPLUGGED | | `staff-authority-register` | dataset | 1 | 1 | within lending authority | `PLUG_STAFF_AUTHORITY_REGISTER` | UNPLUGGED | | `state-digests-from-the-rehearsal` | service | 1 | 1 | affected state unchanged since rehearsal | `PLUG_STATE_DIGESTS_FROM_THE_REHEARSAL` | UNPLUGGED | | `statement-calendar` | service | 1 | 1 | period closed | `PLUG_STATEMENT_CALENDAR` | UNPLUGGED | | `static-check-of-the-generated-code-s-dependencies` | service | 1 | 1 | logic makes no large model call at runtime | `PLUG_STATIC_CHECK_OF_THE_GENERATED_CODE_S_DEP` | UNPLUGGED | | `step-approvals` | service | 1 | 1 | no pending approval blocks DONE | `PLUG_STEP_APPROVALS` | UNPLUGGED | | `stored-awards` | service | 1 | 1 | enough awards to say anything | `PLUG_STORED_AWARDS` | UNPLUGGED | | `stored-releases-and-figures` | service | 1 | 1 | record exists | `PLUG_STORED_RELEASES_AND_FIGURES` | UNPLUGGED | | `study-plan` | service | 1 | 1 | inside the learner's approved plan when an agent acts | `PLUG_STUDY_PLAN` | UNPLUGGED | | `subject-kind` | service | 1 | 1 | method fits the subject | `PLUG_SUBJECT_KIND` | UNPLUGGED | | `subject-objections-register` | dataset | 1 | 1 | subject has not objected | `PLUG_SUBJECT_OBJECTIONS_REGISTER` | UNPLUGGED | | `subject-record` | estate-record | 1 | 1 | subject has a name and kind | `PLUG_SUBJECT_RECORD` | UNPLUGGED | | `successor-s-current-mandate` | estate-record | 1 | 1 | obligation within successor's mandate | `PLUG_SUCCESSOR_S_CURRENT_MANDATE` | UNPLUGGED | | `summary-reconciled` | service | 1 | 1 | batch reconciled | `PLUG_SUMMARY_RECONCILED` | UNPLUGGED | | `supplier-s-approved-buyer-list-in-the-agent-mandate` | estate-record | 1 | 1 | buyer approved by the supplier | `PLUG_SUPPLIER_S_APPROVED_BUYER_LIST_IN_THE_AG` | UNPLUGGED | | `supplier-vetting-report-shared-with-this-buyer` | service | 1 | 1 | Cyber Essentials on record where required | `PLUG_SUPPLIER_VETTING_REPORT_SHARED_WITH_THIS` | UNPLUGGED | | `supplier-web-domains` | service | 1 | 1 | domain declared | `PLUG_SUPPLIER_WEB_DOMAINS` | UNPLUGGED | | `support-case` | estate-record | 1 | 1 | plan still offered | `PLUG_SUPPORT_CASE` | UNPLUGGED | | `support-plan-acceptance-in-apply-for-loan` | service | 1 | 1 | borrower accepted the plan | `PLUG_SUPPORT_PLAN_ACCEPTANCE_IN_APPLY_FOR_LOA` | UNPLUGGED | | `support-plans` | service | 1 | 1 | not already in a support plan | `PLUG_SUPPORT_PLANS` | UNPLUGGED | | `support-queue` | estate-record | 1 | 1 | specialist assigned | `PLUG_SUPPORT_QUEUE` | UNPLUGGED | | `support-request-can-pay` | service | 1 | 1 | plan payment no more than what the borrower said they can pay | `PLUG_SUPPORT_REQUEST_CAN_PAY` | UNPLUGGED | | `supported-list-device-register-accuracy-open-measurement-acc` | dataset | 1 | 1 | device model supported, with a stated measurement accuracy | `PLUG_SUPPORTED_LIST_DEVICE_REGISTER_ACCURACY_` | UNPLUGGED | | `supported-source-list` | dataset | 1 | 1 | source supported | `PLUG_SUPPORTED_SOURCE_LIST` | UNPLUGGED | | `takedown-records` | estate-record | 1 | 1 | not already reported by this claimant | `PLUG_TAKEDOWN_RECORDS` | UNPLUGGED | | `target-operation-s-declared-recovery` | service | 1 | 1 | step has a compensating operation | `PLUG_TARGET_OPERATION_S_DECLARED_RECOVERY` | UNPLUGGED | | `task-and-run-history` | service | 1 | 1 | no later change depends on it | `PLUG_TASK_AND_RUN_HISTORY` | UNPLUGGED | | `task-status` | service | 1 | 1 | task open | `PLUG_TASK_STATUS` | UNPLUGGED | | `template-state` | service | 1 | 1 | template version published | `PLUG_TEMPLATE_STATE` | UNPLUGGED | | `template-workflow-version` | service | 1 | 1 | approvers named in the workflow | `PLUG_TEMPLATE_WORKFLOW_VERSION` | UNPLUGGED | | `test-runner` | service | 1 | 1 | test is executable against the estate | `PLUG_TEST_RUNNER` | UNPLUGGED | | `test-suite-register` | dataset | 1 | 1 | replay suite loaded | `PLUG_TEST_SUITE_REGISTER` | UNPLUGGED | | `the-account-provider-s-consent-response` | estate-record | 1 | 1 | bank authorised access | `PLUG_THE_ACCOUNT_PROVIDER_S_CONSENT_RESPONSE` | UNPLUGGED | | `the-action-s-recorded-behaviour` | estate-record | 1 | 1 | action is reversible | `PLUG_THE_ACTION_S_RECORDED_BEHAVIOUR` | UNPLUGGED | | `the-action-s-undo-window` | service | 1 | 1 | within the undo window | `PLUG_THE_ACTION_S_UNDO_WINDOW` | UNPLUGGED | | `the-obligation-s-own-rules-for-example-an-open-recall-case-c` | estate-record | 1 | 1 | cancellable | `PLUG_THE_OBLIGATION_S_OWN_RULES_FOR_EXAMPLE_A` | UNPLUGGED | | `the-person-s-linked-gp-link-worker-or-agent` | service | 1 | 1 | recipient known | `PLUG_THE_PERSON_S_LINKED_GP_LINK_WORKER_OR_AG` | UNPLUGGED | | `the-publisher-s-stated-licence` | service | 1 | 1 | licence unchanged | `PLUG_THE_PUBLISHER_S_STATED_LICENCE` | UNPLUGGED | | `timesheets` | service | 1 | 1 | approved hours present for every hourly worker | `PLUG_TIMESHEETS` | UNPLUGGED | | `tok-and-evidence-stores` | estate-record | 1 | 1 | subject exists | `PLUG_TOK_AND_EVIDENCE_STORES` | UNPLUGGED | | `tok-operation-catalogue` | dataset | 1 | 1 | every step is a declared estate operation | `PLUG_TOK_OPERATION_CATALOGUE` | UNPLUGGED | | `tok-source-records-research-is-used-only-once-admitted` | estate-record | 1 | 1 | every source admitted with provenance, licence and digest | `PLUG_TOK_SOURCE_RECORDS_RESEARCH_IS_USED_ONLY` | UNPLUGGED | | `track-fingerprint-result` | service | 1 | 1 | fingerprint not matched to someone else's work | `PLUG_TRACK_FINGERPRINT_RESULT` | UNPLUGGED | | `track-length` | service | 1 | 1 | timings increase and fit the track | `PLUG_TRACK_LENGTH` | UNPLUGGED | | `trading-partner` | service | 1 | 1 | at least the minimum (OPEN) | `PLUG_TRADING_PARTNER` | UNPLUGGED | | `transfer-history` | service | 1 | 1 | first-time recipient acknowledged | `PLUG_TRANSFER_HISTORY` | UNPLUGGED | | `transfer-new-owner-accepted` | service | 1 | 1 | receiving owner accepted | `PLUG_TRANSFER_NEW_OWNER_ACCEPTED` | UNPLUGGED | | `translation-engine-list` | dataset | 1 | 1 | language supported | `PLUG_TRANSLATION_ENGINE_LIST` | UNPLUGGED | | `triage-duties` | service | 1 | 1 | regime is in scope for this organisation | `PLUG_TRIAGE_DUTIES` | UNPLUGGED | | `triage-procedure-calibration-record` | estate-record | 1 | 1 | confidence above the certified threshold | `PLUG_TRIAGE_PROCEDURE_CALIBRATION_RECORD` | UNPLUGGED | | `trusted-contact-contact-agreed` | service | 1 | 1 | contact agreed | `PLUG_TRUSTED_CONTACT_CONTACT_AGREED` | UNPLUGGED | | `trusted-contact-safe-to-add` | service | 1 | 1 | safe to add | `PLUG_TRUSTED_CONTACT_SAFE_TO_ADD` | UNPLUGGED | | `two-factor-recovery-codes-saved` | service | 1 | 1 | recovery codes saved | `PLUG_TWO_FACTOR_RECOVERY_CODES_SAVED` | UNPLUGGED | | `two-runs-of-the-generated-tests-on-separate-racetrack-instan` | service | 1 | 1 | output is deterministic | `PLUG_TWO_RUNS_OF_THE_GENERATED_TESTS_ON_SEPAR` | UNPLUGGED | | `uk-companies-house` | dataset | 1 | 2 | company exists; name matches number | `PLUG_UK_COMPANIES_HOUSE` | UNPLUGGED | | `uk-digital-verification-services-register` | dataset | 1 | 1 | provider on the statutory register | `PLUG_UK_DIGITAL_VERIFICATION_SERVICES_REGISTE` | UNPLUGGED | | `uploaded-file` | service | 1 | 1 | file readable | `PLUG_UPLOADED_FILE` | UNPLUGGED | | `uploaded-files` | service | 1 | 1 | files readable | `PLUG_UPLOADED_FILES` | UNPLUGGED | | `vacancy-budget-approved` | service | 1 | 1 | budget approved | `PLUG_VACANCY_BUDGET_APPROVED` | UNPLUGGED | | `vacancy-salary-range` | service | 1 | 1 | salary within range or exception approved | `PLUG_VACANCY_SALARY_RANGE` | UNPLUGGED | | `vacancy-status-and-closing-date` | service | 1 | 1 | vacancy open | `PLUG_VACANCY_STATUS_AND_CLOSING_DATE` | UNPLUGGED | | `validation-run-results` | service | 1 | 1 | results within limits to approve | `PLUG_VALIDATION_RUN_RESULTS` | UNPLUGGED | | `version-digest` | service | 1 | 1 | version unchanged since the request | `PLUG_VERSION_DIGEST` | UNPLUGGED | | `vetting-record-store` | estate-record | 1 | 1 | supplier exists | `PLUG_VETTING_RECORD_STORE` | UNPLUGGED | | `visit-records` | estate-record | 1 | 2 | code valid and bound to this visit; code not used before | `PLUG_VISIT_RECORDS` | UNPLUGGED | | `visit-slot-and-device-location` | service | 1 | 1 | right point or address and time window | `PLUG_VISIT_SLOT_AND_DEVICE_LOCATION` | UNPLUGGED | | `visit-status` | service | 1 | 1 | visit not yet checked in | `PLUG_VISIT_STATUS` | UNPLUGGED | | `warning-record` | estate-record | 1 | 1 | warning still valid | `PLUG_WARNING_RECORD` | UNPLUGGED | | `workflow` | service | 1 | 1 | version published | `PLUG_WORKFLOW` | UNPLUGGED | | `workflow-and-document-history` | service | 1 | 1 | caller is a named approver and not the author or the author's agent | `PLUG_WORKFLOW_AND_DOCUMENT_HISTORY` | UNPLUGGED | | `workflow-inputs` | service | 1 | 1 | inputs valid for the version | `PLUG_WORKFLOW_INPUTS` | UNPLUGGED | | `workflow-rules` | service | 1 | 1 | status change allowed | `PLUG_WORKFLOW_RULES` | UNPLUGGED | | `workflow-version` | service | 1 | 1 | caller is the named approver | `PLUG_WORKFLOW_VERSION` | UNPLUGGED | | `model-speech` | model | 0 | 0 | | `PLUG_MODEL_SPEECH` | UNPLUGGED | | `voice-telephony` | service | 0 | 0 | | `PLUG_VOICE_TELEPHONY` | UNPLUGGED | ### docs/DIRECTORY.md `docs/DIRECTORY.md` · sha256 `5652bd6154f2317db99c65e9df75678595337439862d82bfffeeb7034097f0bb` #### Directory Generated by `tools/urls.py`; served by `runtime/web.py`. 1616 capabilities and 8309 screens; the longest path is 69 characters. ##### Domains - `judiverse.tech`, `judeai.tech` and `x-x.tech` serve exactly the same site; every path answers identically on all three. - `judiverse.tech` is canonical: every page carries `<link rel="canonical" href="https://judiverse.tech/<same path>">`. - `www.` on each domain redirects (301) to its apex, keeping the path and query. - One reverse proxy holds the certificates for all six names. ##### Routes | Path | What it is | Same on every domain | |---|---|---| | `/` | The five applications and the shared foundations | yes | | `/<area>` | An application's or foundation's capabilities | yes | | `/<area>/<capability>` | A capability: its screens, starting with the first | yes | | `/<area>/<capability>/<screen>` | Every screen, rendered from its unit | yes | | `/<area>/match/<competitor-area>` | Competitor-parity capabilities | yes | | `/s/<code>` | Permanent short link for every capability (302 to it) | yes | | `/help/<area>/<capability>/<screen>` | The screen's help, for people (HTML) and agents (?format=md) | yes | | `/checkpoint` | The agent checkpoint (CHECKPOINT.md, text/markdown) | yes | | `/v1/...` | The API, one route per operation | yes | | `/v1/voice/inbound` | The voice route (POST) | yes | | `/mcp` | The hub MCP over HTTP (POST, JSON-RPC) | yes | | `/signin, /signout` | A signed session cookie | yes | ##### Areas | Area | Name | Capabilities | Screens | |---|---|---|---| | `/finance` | EDGE Bank | 391 | 2015 | | `/clinic` | EDGE Clinic | 342 | 1759 | | `/consortium` | The Edge Consortium | 363 | 1977 | | `/public` | EDGE Public | 489 | 2376 | | `/judiverse` | Judiverse | 22 | 113 | | `/id` | Identity | 2 | 17 | | `/evidence` | Evidence | 1 | 7 | | `/regulators` | Regulators | 2 | 17 | | `/organisations` | Organisations | 3 | 21 | | `/agents` | Agents | 0 | 0 | | `/releases` | Releases | 1 | 7 | | `/research` | Research | 0 | 0 | | `/tok` | Knowledge objects | 0 | 0 | Every capability's path and short link are listed in `docs/URLS.md`. ### docs/REPORT-OUTCOMES-AND-NOVEL-MEASURES.md `docs/REPORT-OUTCOMES-AND-NOVEL-MEASURES.md` · sha256 `11d26b49d3dc2dbddf33b04486a9173a463cb1dd9e9c5a559c6f0866bce759ce` #### Report: parity, outcomes and the measures nobody else publishes For X-X.TECH, 31 October 2026. Every number is computed by the estate's programs from its files and can be reproduced with `tools/run_all.sh`. ##### Parity: we match first - **88,661 competitor features** gathered: clauses and documentation from competitors' own pages, plus Refero and Mobbin flows. - **98.9% like for like** (72,837 of 73,620 graded): our control does the same kind of action to the same thing, with both sides read by the same program from the competitor's own documentation. - **Behaviour read, not invented.** Each competitor feature's method, fields, conditions, limits, roles and undo route are read from the vendor's text or page (3,940 vendor pages fetched in about a minute). - **Zero dead controls** in every generated capability. Every person journey has review before change, a receipt and a way back from failure. ##### Outcomes: we match their outcome, then exceed it - **74,234 operations (99.6% of those matched to a competitor)** carry the competitor's documented outcome, word for word, as our outcome. On top, every one adds rehearsal before it happens, confirmation by the receiving system and reproducible evidence. - **Every one of 87,409 operations has a target.** Where a competitor publishes a figure, the target beats the best one: half the time, or half the remaining gap to 100%. Example: EU instant payments make funds available within 10 seconds, so our target is 5 seconds. - **Published figures found so far:** 82 on competitors' own documentation, service-level, status and trust pages. Examples: Asana 99.9% uptime commitment; AEYE-DS retinopathy results in under 60 seconds; Deliveroo delivery in as little as 20 minutes. Quantitative outcomes are rarely published, so where none exists the estate sets the standard in `spec/standards.json`. ##### The breakthrough: measures nobody publishes, measured on every call | Measure | Target | |---|---| | Intent-to-outcome rate: share of stated intents that end in a verified, reproducible outcome | at least 95% | | Words per verified outcome | at most 25 | | Prompts per verified outcome | 0 for deterministic operations | | Tokens per verified outcome | 0 for deterministic operations | | Deterministic share: verified outcomes produced with no model at run time | at least 99% | | Cost per verified outcome | at most £0.0001 | | Carbon per verified outcome | at most 0.1 g CO2e | Why it matters: people are starting to pay for results and for less AI. These measures make results and AI use comparable, and setting them first makes the estate the standard. ##### Zero waste, enforced - Every agent must declare its output and token budget before it starts. A run that produces nothing is recorded as wasted, and further launches are blocked until the owner acknowledges it. - The gate cannot be edited from a session. It is installed as managed settings with one command: `sudo sh new-estate/tools/install_gate.sh`. - **Measured so far:** - 66 agent runs, of which 13 were wasted: 7.7% of new tokens. - Claude's own work counts too: 545 model calls and 4.37 million new tokens. ##### What is still open - **Published baselines:** few operations have one, because competitors rarely publish outcome numbers. The rest use the estate's own standard. - **Uptime figures:** some are daily status values rather than commitments. Treat a single day's 100% as observed, not promised. - **Outcome tests against real receiving systems** (banks, NHS services, registries) need each system's access details. ### runtime/INTERFACE.md `runtime/INTERFACE.md` · sha256 `64acee32ecdbe87c1a4369f296966800087f21c5b6080b0e775279b36e600394` #### Runtime interface (the one contract every runtime part builds to) Input: the units in `spec/units/` (read lazily, one unit at a time, through `tools/emit.spec`), or `build/estate-spec.json`, or a dict in that format. Receivers: `spec/receivers.json` (made by `python3 tools/receivers.py`). Python 3 standard library only; deterministic (ids are sha256 of content and a counter). ##### Engine: `runtime/engine.py` ```python from runtime.engine import Engine e = Engine(spec_path="spec/units", db_path=":memory:", today="2026-10-31", mode="live") # or mode="simulation" r = e.call(unit="finance/send-money", operation="send-payment", actor={"id": "p1", "kind": "person" | "agent", "role": "Account holder", "tenant": "p1", "verified": True, "mandate": {"tools": ["send_money.send_payment"], "owner": "p1", "expires": "2027-01-01", "roles": [...]}}, inputs={"quote_id": "q1", ...}, rehearse=False, idempotency_key="k1") #### r == {"status": "OK" | "REFUSED" | "FAILED" | "UNKNOWN" | "HELD", "code": None | "<code>", "outputs": {...}, #### "evidence_ref": "<sha256 hex>", "steps": [...], #### "receiver": "estate-record" | "external:<slot>" | "simulation" | "unbound", "verified": bool} e.inspect(evidence_ref) # the full evidence record (it carries receiver, verified, tenant, mode) e.evidence() # yields every evidence record in order: calls, refusals, settlements e.verify_receipt(receipt, actor, expect=None) # (ok, reason) for an estate-record receipt, now e.receiver_label(unit, op) # "estate-record:<class>:<entity>" | "external:<slot>" | None; e.receivers[key] overrides e.operations(); e.find(method, path); e.find_tool(tool_name) ``` Order of execution, for every operation: 1. identity: actor with id, kind and role **and `verified is True`** (set only by `runtime/session.py` or a credential check; tests may pass it), else REFUSED `UNAUTHORISED`; 2. mandate: an agent must hold the operation's tool in `mandate.tools`, an owner and an unexpired `expires`, else REFUSED `OVER_MANDATE`; 3. inputs: every required input present and valid for its type, else REFUSED `INVALID_INPUT`; 4. domain guards by family (`runtime/guards.py`). Documented preconditions need the inputs their clause names (else at least one non-empty input). An agent passes a role guard only if its mandate lists an allowed role, or its mandate's owner holds one in their credential record. Unplugged conditions: live → REFUSED `PLUG_REQUIRED` (outputs.plug_required names the slot); simulation → pass, marked `simulated`, never verified; 5. rehearsal: `rehearse=True` returns OK with `outputs.predicted` (including the receiver) and changes nothing; never verified; 6. idempotency: the key is scoped to (tenant, actor id, unit, operation). The same key and inputs return the first result (HELD included). The same key with other inputs is REFUSED `IDEMPOTENCY_CONFLICT`; 7. irreversible operations are HELD `ANALOG_HUMAN_ACT_REQUIRED` (below); 8. effect through `runtime/adapters.resolve(engine, unit, op, o) -> (label, fn)`. There is no default receiver. The order is: an explicitly registered simulation (`adapters.register_simulation`), then an explicit adapter (`adapters.ADAPTERS`), then `spec/receivers.json`. An estate-record goes to `runtime/receivers.EstateStore`. An external slot goes to its plug in `spec/plugins.lock.json`. **If no receiver exists, the operation is UNKNOWN `RECEIVER_REQUIRED` and nothing changes**; 9. receiving confirmation: confirmed / rejected / no answer gives OK / FAILED (the receiver's code, e.g. NOT_FOUND or VERSION_CONFLICT, else the first failure) / UNKNOWN. An estate-record receipt is re-read before OK; 10. evidence: every call is stored and returned by `evidence_ref`. **Verified rule.** `verified = status == "OK"` and not rehearsal, and receiver not simulation or unbound, and no guard simulated, and settlement None or FINAL, and (for estate-record) the receipt re-reads to the same digest. An agent's high-risk outcome stays OK with `settlement: PROVISIONAL` and is not verified. When it settles, a `settlement` evidence record `{mark, state: FINAL|REVERSED, of, verified}` is written. ##### Receivers - `tools/receivers.py` → `spec/receivers.json` `{"<unit>/<op>": "estate-record:<class>:<entity>" | "external:<slot>"}`. Class comes from `tools/behaviour` (GET is always read). External comes from the label (and the effect text, for unambiguous system names) naming a model, mail, SMS, voice, card network, payment scheme, NHS, HMRC, Companies House, FCA, a registry or a third party. - `runtime/receivers.EstateStore`: sqlite `objects(tenant, unit, entity, id, body, version, deleted)` and `events(...)`. It handles create, read (by id or filter), change (VERSION_CONFLICT on a stale `version`/`expected_version`), remove (soft), and send/decide/plan (an event, plus the entity's state field). - Receipt: `{receiver: "estate-record", tenant, unit, entity, id, version, digest}`. - `verify` rejects invented, altered, other-tenant and stale receipts, and receipts whose record does not satisfy the outcome. - The tenant comes from the actor (`session.tenant_of`), never from inputs. ##### Sessions: `runtime/session.py` - `issue(actor) -> token`: HMAC-SHA256 over JSON, 12 h. The key is env `ESTATE_SESSION_KEY`, else `.cache/session.key` (0600). - `read(token) -> actor | None`: the actor has `verified: True`. - `sign_in(person_id, secret, role=None)`: uses `credentials.verify`. People get credentials with `credentials.issue(person_id, owner=person_id, tools=['*'])`. - `agent(agent_id, secret)`: the mandate comes from the credential record. - `tenant_of(actor)`: the actor's tenant, else (for an agent) its mandate owner, else the id part before ':'. ##### HTTP: `runtime/server.py` `handle_api(method, path, headers, body) -> (status, content_type, bytes)`; `python3 -m runtime.server --host --port [--db] [--simulation]` serves it. - The actor comes only from `Authorization: Bearer <session token>` (or the cookie `session=`) or `Authorization: Agent <agent_id>:<secret>`. `X-Actor-*` headers are ignored. - `POST /v1/session {person_id, secret}` returns `{token}`. - `GET /v1/evidence/{ref}` is limited to the caller's tenant. - Every operation is served at its method and path. Use the `Idempotency-Key` and `X-Rehearse: 1` headers. - HTTP codes: OK 200, REFUSED 403 (UNAUTHORISED 401), FAILED 409, IDEMPOTENCY_CONFLICT 409, HELD 202, UNKNOWN 202. - There is one module-level engine (`server.engine()`: spec/units, `.cache/runtime.sqlite`, live), and `server.set_engine(e)` replaces it. ##### Evidence of every operation: `tools/exercise.py` `python3 tools/exercise.py` (or `EXERCISE_LIMIT=30` for a sample) runs every operation once, live, as a verified person in tenant `exercise`. It writes `ledger/evidence/operations.jsonl` `{unit, operation, status, code, receiver, verified, evidence_ref}` and prints counts by receiver and status. ##### Agent tools: `runtime/mcp_server.py` `python3 -m runtime.mcp_server` speaks MCP over stdio (JSON-RPC 2.0: `initialize`, `tools/list`, `tools/call`). One tool per operation, named by its `tool`; input schema from `tool_input` (strings) plus `actor` and `rehearse`; calls the engine. ##### Screens: `web/index.html`, `web/app.js`, `web/app.css` No build step, no external libraries. Loads `/estate-spec.json`; lets a person pick application → capability → journey or screen; renders any screen from its blueprint (title, fields by type, controls by component, shows, help, the nine states); a control runs its action: `navigate` moves to `on_success`; an operation calls the HTTP endpoint and moves to `on_success` or `on_failure` (a screen id, a state shown on the same screen, `journey:end`, or `unit:<app>/<cap>#<screen>`); shows the evidence panel for the last result. Actor chosen in a header bar (person or agent, role, mandate tools). ##### Tests: `tests/test_runtime.py` Generated over every operation in the spec (unittest, standard library): missing required input is REFUSED `INVALID_INPUT`; an agent without the tool in its mandate is REFUSED `OVER_MANDATE`; rehearsal changes nothing; a valid call is OK with an evidence record whose digest reproduces; repeating the idempotency key gives the same result; every screen action names an existing operation or `navigate`. ##### Irreversible outcomes: the analog human switch An operation is irreversible when it is marked `irreversible` or its label or effect names a weapon, munition, missile, warhead, detonation, lethal or kinetic effect, destruction, demolition, deportation, sterilisation or anything irreversible (`IRREVERSIBLE` in `engine.py`). Decision: `tok:decision:01m69jxxqzrhz7r36ht4pvv5td`. - **No software path executes it.** `call` returns status `HELD`, code `ANALOG_HUMAN_ACT_REQUIRED` and a `mark`. This happens for every person and every agent, with any mandate. - **The owner does it in person.** `analog_act(mark, owner, serial, place, circuit_closed, ordered_by_me, i_want_it)` runs the effect once, and only when all of these hold: - the unit is registered to the owner and not revoked; - it is at its registered place; - the key closed the circuit; - the owner separately confirms that they ordered it; - the owner separately confirms that they want it to occur. - **Backups:** `register_switch` records identical units. `revoke_switch` retires a lost, stolen or broken unit, and the backups keep working. - **Where the guarantee lives:** software can only record. The guarantee is the hardware: the analog key sits in series with the actuation circuit, so no signal from any computer can close it. The engine refuses to run the effect any other way. The hardware specification is `research/public-edge/ANALOG-HUMAN-SWITCH.md`. ### Generated help and reference docs/help: 7725 files; docs/api: 7725; docs/reference: 7726; docs/proof: 5692 (in the zip). ## 6. Reusable assets and design system ### design/brand/approved/BRAND-REGISTER.json `design/brand/approved/BRAND-REGISTER.json` · sha256 `f18dc480af93ea94a88641a176a9c0aedfd7ab969c6090db297ed3754aef4ff5` ```json { "schema": "edge.brand-register.v1", "statement_class": "AGREEMENT", "supplied_by": "Owner, 31 October 2026 23:2x UTC (Downloads)", "rule": "Only these files, unchanged, by sha256. judeai.svg (the J tile) is RETIRED and never rendered. Buttons and pills are fully rounded (999px) as in the approved reference. Cyan dots #63DEFF.", "marks": { "bank.svg": { "sha256": "5841be31dd4e718fd5d11e64de41cb77270d58b51a04d9e8a3f8ea46dde92e8b", "bytes": 4661, "role": "EDGE Finance application mark (lion stays Bank only)" }, "clinic.svg": { "sha256": "98b934938c522b77b470f17f0c1835feced5197c7dcfd585392dc8a01c744d41", "bytes": 904, "role": "Judiverse Suite application mark: the pulse line (Owner, 1 Nov 2026 02:52Z: swap the clinic and suite logos)" }, "consortium.svg": { "sha256": "d6e0f7414e1f345ea7fbd0754e34e745df6792bccdc868186a7c36f35b957986", "bytes": 1843, "role": "The Edge Consortium application mark" }, "cyber.svg": { "sha256": "3c95c2602530981af884489ec16e22f60b0a301abbcebb7947d4db954c1a399a", "bytes": 1626, "role": "Jude Cyber Governance application mark" }, "intenjude.svg": { "sha256": "f9821538ac86bb77934beda4a9ac90ea12e26501c30aa54ad01a22b67c06ca25", "bytes": 904, "role": "intenJude Lab mark" }, "judeai-print.svg": { "sha256": "a90d2af2fc5ce84c5ae8dc24aae2c867eb99d5ccdb61a0fe29b7ac0ae8203e49", "bytes": 1412, "role": "EDGE Clinic application mark: the fingerprint print (Owner, 1 Nov 2026 02:52Z: swap the clinic and suite logos)" }, "judeai.svg": { "sha256": "75fe5736f3f1c31e336c3e78204104fa2cf480d4c49c993999270f5331bb8f4b", "bytes": 1237, "role": "RETIRED. The J tile is retired (Owner, 31 Oct 2026 23:41Z); kept only so the sha is known and refused. Never rendered." }, "main.svg": { "sha256": "5c26b4ced10ff589daab1ff6195eaf28e694433b42d17727da7067798964c637", "bytes": 734, "role": "JudeAI identity mark: the three dots on their own; the shell mark on every page beside the wordmark (Owner, 1 Nov 2026 02:32Z)" }, "office.svg": { "sha256": "4491db74102d96f0598cc5c360f078a482090fb415f1d645638b945120a8397e", "bytes": 1131, "role": "Judiverse Mail mark (mailbox); not the Suite mark" }, "public.svg": { "sha256": "ab649b13289d13fd21d2b1582118a2f50bd98cedf7bf1634d0ee83e19801b047", "bytes": 1532, "role": "EDGE Public application mark" }, "wordmark.svg": { "sha256": "f24016f2db970b688826c88dfda3e94f23e7060b82bd61ceebf898a9f4f7b968", "bytes": 5229, "role": "JudeAI | EDGE ID wordmark" } } } ``` ### design/components.json `design/components.json` · sha256 `ea12e1889248db1f30cbc1c0e06c774db8215cf07c7283f5b883100b4fec60a3` ```json { "schema": "tok.components.v1", "statement_class": "AGREEMENT", "what_this_is": "The only components a screen may use. Each name maps to a component in the approved design system (Untitled UI structure, realised in Figma and UX Pilot). A screen blueprint names a component from this list and says what it binds to; the compiler refuses any other name.", "design_system": {"name": "Untitled UI", "version": {"OPEN": "pin the exact Untitled UI library version when the approved Figma library is admitted"}}, "components": { "Page.Header": {"binds": ["title", "subtitle"], "use": "Screen title, one-line subtitle, optional back action"}, "Page.Section": {"binds": ["heading"], "use": "Group of related content with a heading"}, "Nav.Sidebar": {"binds": ["items"], "use": "Application navigation (desktop)"}, "Nav.TabBar": {"binds": ["items"], "use": "Application navigation (mobile)"}, "Nav.Breadcrumbs": {"binds": ["path"], "use": "Location within a hierarchy"}, "Nav.Tabs": {"binds": ["items", "selected"], "use": "Switch between views of the same subject"}, "Steps.Progress": {"binds": ["steps", "current"], "use": "Position in a multi-step task"}, "Input.Text": {"binds": ["value"], "use": "Single-line text"}, "Input.Email": {"binds": ["value"], "use": "Email address"}, "Input.Phone": {"binds": ["value"], "use": "Phone number with country code"}, "Input.Money": {"binds": ["value", "currency"], "use": "Amount with currency"}, "Input.Number": {"binds": ["value"], "use": "Number"}, "Input.Date": {"binds": ["value"], "use": "Date picker"}, "Input.Search": {"binds": ["query"], "use": "Search with results"}, "Input.Password": {"binds": ["value"], "use": "Secret entry with show and hide"}, "Input.Code": {"binds": ["value"], "use": "One-time code entry"}, "Input.Address": {"binds": ["value"], "use": "Address with lookup and manual entry"}, "Input.TextArea": {"binds": ["value"], "use": "Multi-line text"}, "Select": {"binds": ["value", "options"], "use": "Choose one from a list"}, "Select.Multi": {"binds": ["values", "options"], "use": "Choose several"}, "Radio.Group": {"binds": ["value", "options"], "use": "Choose one of a few visible options"}, "Checkbox": {"binds": ["checked"], "use": "Agree or opt in"}, "Toggle": {"binds": ["on"], "use": "Turn a setting on or off"}, "Slider": {"binds": ["value"], "use": "Choose a value in a range"}, "File.Upload": {"binds": ["files"], "use": "Upload documents"}, "Camera.Capture": {"binds": ["image"], "use": "Take a photo of a document or face"}, "NFC.Read": {"binds": ["chip_data"], "use": "Read an identity document chip by tapping"}, "QR.Scan": {"binds": ["payload"], "use": "Scan a code"}, "Button.Primary": {"binds": ["label"], "use": "The one main action on a screen"}, "Button.Secondary": {"binds": ["label"], "use": "Alternative action"}, "Button.Destructive": {"binds": ["label"], "use": "Action that removes or ends something"}, "Button.Link": {"binds": ["label"], "use": "Low-emphasis action"}, "Card": {"binds": ["title", "body"], "use": "Summary of one thing"}, "Card.Metric": {"binds": ["label", "value", "change"], "use": "One measured number with its trend"}, "List": {"binds": ["items"], "use": "List of items"}, "List.Item": {"binds": ["title", "detail"], "use": "One row in a list"}, "Table": {"binds": ["rows", "columns"], "use": "Rows and columns of data"}, "Summary.Review": {"binds": ["rows"], "use": "Read-only summary before confirming"}, "Receipt": {"binds": ["rows", "reference"], "use": "Proof of a completed action with its reference"}, "Timeline": {"binds": ["events"], "use": "Ordered history of events"}, "Chart": {"binds": ["series"], "use": "Trend or comparison"}, "Badge": {"binds": ["label", "tone"], "use": "Short status label"}, "Alert": {"binds": ["title", "body", "tone"], "use": "Inline message about the current screen"}, "Toast": {"binds": ["message"], "use": "Brief confirmation that disappears"}, "Modal": {"binds": ["title", "body"], "use": "Focused decision over the current screen"}, "Slideout": {"binds": ["title", "body"], "use": "Detail panel beside the current screen"}, "EmptyState": {"binds": ["title", "body", "action"], "use": "What to do when there is nothing yet"}, "Skeleton": {"binds": [], "use": "Loading placeholder"}, "Avatar": {"binds": ["name", "image"], "use": "Person, organisation or agent identity"}, "Agent.Card": {"binds": ["agent", "mandate", "status"], "use": "An agent's identity, mandate, term and status"}, "Mandate.Editor": {"binds": ["scope", "limits", "expiry"], "use": "Set what an agent may do, its limits and expiry"}, "Rehearsal.Result": {"binds": ["predicted_effects", "cost", "risk"], "use": "What a rehearsal predicts before commit"}, "Evidence.Panel": {"binds": ["evidence"], "use": "Glass-box view: sources, reasoning, authority, verification"}, "Confidence.Meter": {"binds": ["value", "bound", "explanation"], "use": "A confidence value with its error bound and why"}, "Chat.Thread": {"binds": ["messages"], "use": "Conversation with a person or agent"}, "Chat.Composer": {"binds": ["draft"], "use": "Write a message or instruction"}, "Help.Control": {"binds": ["markdown"], "use": "Opens the help depth for this screen"}, "Map": {"binds": ["locations"], "use": "Places on a map"}, "Calendar": {"binds": ["slots"], "use": "Dates and available times"}, "Notification.Item": {"binds": ["title", "body", "time"], "use": "One notification"} } } ``` ### Tokens (runtime/web_assets.py, read from the approved UX Pilot reference and the estate Figma file) ```css :root{--text:#002d40;--text-2:#2f4b59;--text-3:#3f5b69;--placeholder:#64748b;--bg:#ffffff;--bg-2:#f8fafc;--surface:#f1f5f9; --border:#cbd5e1;--border-2:#e2e8f0;--brand:#005f8a;--brand-text:#005f8a;--brand-3:#005f8a;--on-brand:#ffffff;--strong:#002d40;--dots:#63DEFF; --error:#e11d48;--error-bg:#fff1f2;--error-border:#fecdd3;--warning:#b45309;--warning-bg:#fffbeb;--warning-border:#fde68a; --success:#059669;--success-bg:#ecfdf5;--success-border:#a7f3d0;--info-bg:#f0fbff;--info-border:#e0f7ff; --focus:#005f8a;--radius:12px;--radius-lg:16px;--sp-xxs:2px;--sp-sm:6px;--sp-md:8px;--sp-lg:12px;--sp-xl:16px;--gap:20px; --shadow-xs:0 1px 2px rgba(0,0,0,.05);--inset:inset 0 0 0 1px rgba(0,0,0,.18),inset 0 -2px 0 0 rgba(0,0,0,.05); --font:Inter,-apple-system,"SF Pro Text","Segoe UI",Roboto,"Helvetica Neue",Arial,sans-serif;--mono:ui-monospace,SFMono-Regular,Menlo,Consolas,monospace; color-scheme:light;} ``` ### Figma File https://www.figma.com/design/R5SljhjU1UWBE9rny6XOHf · templates workflow-step, competitor-features, evidence, done, research-checks · counts {"templates": 5, "template_state_frames": 45, "screens_covered_by_templates": 22487, "bespoke_units": 84, "bespoke_frames": 617, "units": 73} ## 7. Environments, tools, settings and connectors ### TOOLCHAIN.json `TOOLCHAIN.json` · sha256 `8db4a9ed777cd135106f6656f6b9ff6db34faa2eab32d0f430ce93957416a195` ```json { "chromium": { "executable": "/opt/pw-browsers/chromium-1243/chrome-linux64/chrome", "revision": "1243", "version": "Google Chrome for Testing 153.0.8010.12" }, "corpus": { "file": "catalogue.jsonl", "present": false, "root": "/home/user/corpus", "root_env": "CORPUS_ROOT", "sha256": "UNKNOWN" }, "node": "22.23.4", "page_cache": { "digest_of": "sorted relative file names and sizes", "present": false, "root": ".cache/pages", "root_env": "PAGE_CACHE", "sha256": "UNKNOWN" }, "playwright": "1.63.0", "playwright_package": "/opt/node-tools/node_modules/playwright/package.json", "python": "3.13.18", "what_this_is": "Generated by python3 tools/make.py --toolchain: the toolchain measured on the machine that ran it, and the digests of the external inputs the run reads (CORPUS_ROOT, PAGE_CACHE). A digest of UNKNOWN means the input was absent on that machine; it is never invented." } ``` ### deploy/README.md `deploy/README.md` · sha256 `eb3003c26ef2ea9b599019a5587a3fe812579f85cf74ff1faccc01fc6e55d862` #### Deploy: one site on judiverse.tech, judeai.tech and x-x.tech For Codex, on one existing Civo VM. This cloud session holds no Civo credentials and has not run these commands on the VM. The plan is `ledger/closure/plan/PLAN.md` §0 and §6.5. | File | What it is | |---|---| | `Caddyfile` | One site block for the three apex domains (reverse proxy to `127.0.0.1:8080`, automatic TLS), and a 301 from each `www.` name to its apex | | `x-x-web.service` | systemd unit: `python3 -m runtime.web --host 127.0.0.1 --port 8080`, environment from `/etc/x-x/web.env` | | `x-x-leads.service` | systemd unit: `python3 -m runtime.leads run` (issues the leads' credentials first) | | `web.env.example` | The environment variable names, with no values | | `deploy.sh` | Checks out a commit, builds, restarts, then compares `/`, `/checkpoint` and one `/s/<code>` across the three domains | ##### Before the first deploy 1. DNS: A and AAAA records for `judiverse.tech`, `judeai.tech`, `x-x.tech` and the three `www.` names point at the VM. 2. Ports 80 and 443 are open (Caddy needs both to obtain certificates). ##### Once per VM ```bash sudo apt-get update && sudo apt-get install -y git python3 curl caddy sudo useradd --system --create-home --home-dir /opt/x-x --shell /usr/sbin/nologin x-x sudo -u x-x -H git clone https://github.com/X-XTECH/governed-intelligence-judeai-edgeid.git /opt/x-x/governed-intelligence-judeai-edgeid sudo install -d -m 700 -o x-x -g x-x /etc/x-x sudo install -m 600 -o x-x -g x-x /opt/x-x/governed-intelligence-judeai-edgeid/new-estate/deploy/web.env.example /etc/x-x/web.env sudoedit /etc/x-x/web.env # fill the values; ESTATE_SESSION_KEY: python3 -c 'import secrets; print(secrets.token_hex(32))' cd /opt/x-x/governed-intelligence-judeai-edgeid/new-estate sudo cp deploy/x-x-web.service deploy/x-x-leads.service /etc/systemd/system/ sudo systemctl daemon-reload sudo systemctl enable x-x-web x-x-leads sudo cp deploy/Caddyfile /etc/caddy/Caddyfile sudo caddy validate --config /etc/caddy/Caddyfile sudo systemctl reload caddy ``` The build needs the research corpus and page cache that `tools/make.py` reads (`CORPUS_ROOT`, `PAGE_CACHE`; see `TOOLCHAIN.json`). Put them on the VM by digest before the first deploy. ##### Every deploy ```bash cd /opt/x-x/governed-intelligence-judeai-edgeid/new-estate sudo deploy/deploy.sh <commit> ``` It exits nonzero if any domain fails to answer, if any of the three paths differs between domains, if `/` lacks the canonical link to `https://judiverse.tech/`, or if a `www.` name does not answer 301 to its apex. To check what is running without deploying: ```bash sudo deploy/deploy.sh --check-only ``` To check the three hosts locally against the web service (before DNS points here; skips the `www.` check): ```bash LOCAL=1 deploy/deploy.sh --check-only ``` ##### By hand, the same checks ```bash for d in judiverse.tech judeai.tech x-x.tech; do curl -s https://$d/checkpoint | sha256sum; done # three identical digests curl -sI https://www.judeai.tech/ | head -3 # 301 to https://judeai.tech/ ``` ##### Rolling back ```bash sudo deploy/deploy.sh <previous commit> ``` ##### Logs ```bash journalctl -u x-x-web -u x-x-leads -n 200 --no-pager python3 -m runtime.leads status # the leads' queue and handovers ``` Secrets are only ever in `/etc/x-x/web.env` (mode 600) and `.cache/agents/` on the VM. Never commit or paste them. ### deploy/AGENT-DEPLOY.md `deploy/AGENT-DEPLOY.md` · sha256 `529c37af2112ccc7396dcf71d33deafa977d291f06874897ac79c421c1678a13` #### Agent deploy: release 7a2a0ea to judiverse.tech, judeai.tech and x-x.tech These instructions are for a deploy agent running in a session whose environment holds `CIVO_TOKEN` and `DEPLOY_SSH_KEY`. Start the session and say: "Run new-estate/deploy/AGENT-DEPLOY.md." The agent does everything below without asking. ##### Credentials - **Where they live:** both are environment variables. Never print, log, commit or echo their values. - **CIVO_TOKEN:** send it as `Authorization: Bearer $CIVO_TOKEN` to https://api.civo.com/v2. - **DEPLOY_SSH_KEY:** write it to a temporary file with mode 0600, and delete that file when finished. - **If either is missing, or api.civo.com is unreachable:** stop. Report exactly which one, and do not work around it. ##### Steps 1. **Choose a VM.** List the instances in the London region (LON1). - Choose the existing VM best suited to hosting the site. - Do not create, resize or delete any instance, volume or object store. - Record each instance's id, name, size, IP and status. 2. **Install on the VM, preserving everything already there.** - Inventory the running services and what listens on ports 80, 443 and 8080. - Install into `/opt/x-x/releases/7a2a0ea` with a symlink `/opt/x-x/current`. Keep any previous release, so a rollback is a symlink switch plus a restart. - If a web server already owns ports 80 and 443, add the three sites to it without replacing its configuration. Only if nothing owns them, install Caddy with `deploy/Caddyfile`. - If port 8080 is taken, use a free port in both the service and the proxy. - Install `deploy/x-x-web.service` and `deploy/x-x-leads.service`. - Create `/etc/x-x/web.env`: - `ESTATE_SESSION_KEY` gets a fresh random value, written only on the VM. - The other variables follow `deploy/web.env.example`; leave unbound ones empty. - Clone the release and run `python3 tools/make.py`. If its first step fails for lack of the old-estate corpus, run only the steps that don't need it. The committed `spec/units`, `spec/urls.json`, `spec/receivers.json` and `CHECKPOINT.md` are enough for `runtime/web.py` to serve. 3. **Point DNS at the VM.** Point the apex and `www` of judiverse.tech, judeai.tech and x-x.tech at the VM. - Use Civo DNS where those domains are hosted there. - Change only those six records, and record each previous value first. - Don't touch a domain whose DNS is hosted elsewhere; report where it is. 4. **Start the services and run the health check in `deploy/deploy.sh`** on all three domains, covering `/`, `/checkpoint` and one `/s/<code>`. It must show: - identical bodies across the three domains; - `www` redirecting to the apex; - valid TLS. If the check fails, roll back: restore the previous symlink, the proxy configuration and the DNS values. 5. **Record and report.** - Write `ledger/deploy/<date>.json`, containing no secrets: the instance, the release, the ports, the proxy, the DNS changes (old and new values), the health digests and the rollback command. - Commit it to `claude/clever-ramanujan-9w05c1` and push. - Report whether the site is live, the URLs, the evidence, and the exact cause of any blocker. ### deploy/web.env.example `deploy/web.env.example` · sha256 `6930505f29635d00d22fb45433b89c45c148c488e25769ab1827e19e52d8cf31` #### /etc/x-x/web.env: variable NAMES only. Codex fills the values on the VM (chmod 600, owner x-x); never commit them. #### Read by x-x-web.service and x-x-leads.service. Empty means unset: the code then refuses or reports, never fakes. #### Sessions (runtime/session.py): HMAC key for sign-in cookies ESTATE_SESSION_KEY= #### Lead agents (runtime/leads.py) LEADS_LIVE= LEADS_POLL_SECONDS= LEAD_SMS_BARACHIEL= LEAD_SMS_JEGUDIEL= LEAD_SMS_RAPHAEL= LEAD_SMS_URIEL= LEAD_SMS_MICHAEL= LEAD_SMS_GABRIEL= #### Voice route (runtime/voice.py) VOICE_LIVE= #### SMS (runtime/sms_mcp.py) SMSWORKS_API_KEY= SMS_SENDER= SMS_MONTHLY_BUDGET_GBP= SMS_PRICE_GBP= #### Mail (runtime/mail_mcp.py) RESEND_API_KEY= MAIL_FROM= MAIL_OWNER= JUDE_MAIL_DOMAIN= JUDE_SMTP_HOST= JUDE_SMTP_PORT= JUDE_SMTP_USER= JUDE_SMTP_PASS= JUDE_SMTP_TLS= SMTP_HOST= SMTP_PORT= SMTP_USER= SMTP_PASSWORD= #### Plug credentials (tools/plugs.py names one per slot; these are the voice and mail slots) PLUG_MODEL_TRANSCRIPTION= PLUG_MODEL_SPEECH= PLUG_VOICE_TELEPHONY= PLUG_MAIL_INBOUND= ### deploy/x-x-web.service `deploy/x-x-web.service` · sha256 `6b502fa1c8c100805a90efc7b4e12c826dcd00cba3a15789ce93d644bcf766f5` #### The site for judiverse.tech, judeai.tech and x-x.tech, behind Caddy (deploy/Caddyfile). #### Install: cp deploy/x-x-web.service /etc/systemd/system/ && systemctl daemon-reload && systemctl enable --now x-x-web [Unit] Description=x-x web: one site for judiverse.tech, judeai.tech and x-x.tech After=network-online.target Wants=network-online.target [Service] Type=simple User=x-x Group=x-x WorkingDirectory=/opt/x-x/governed-intelligence-judeai-edgeid/new-estate EnvironmentFile=/etc/x-x/web.env ExecStart=/usr/bin/python3 -m runtime.web --host 127.0.0.1 --port 8080 Restart=on-failure RestartSec=2 NoNewPrivileges=true PrivateTmp=true ProtectSystem=full ProtectHome=true UMask=0077 [Install] WantedBy=multi-user.target ### deploy/Caddyfile `deploy/Caddyfile` · sha256 `fade1025d590ee1ddf32cbbd98063f55d230b44f5a739ca17b068dd4e46ac9b2` #### One site on three domains (ledger/closure/plan/PLAN.md §0). judiverse.tech is canonical; judeai.tech and x-x.tech #### serve the same bytes. Caddy obtains and renews certificates for all six names automatically (DNS A/AAAA must point here). #### Install: cp deploy/Caddyfile /etc/caddy/Caddyfile && systemctl reload caddy judiverse.tech, judeai.tech, x-x.tech { encode zstd gzip header { Strict-Transport-Security "max-age=31536000" X-Content-Type-Options "nosniff" Referrer-Policy "strict-origin-when-cross-origin" -Server } reverse_proxy 127.0.0.1:8080 } #### www. on each domain goes to its own apex, permanently (301). {labels.1}.{labels.0} is the apex of www.<name>.<tld>. www.judiverse.tech, www.judeai.tech, www.x-x.tech { redir https://{labels.1}.{labels.0}{uri} permanent } ### Hosted framework and connectors - Deploy and execution MCP: https://deploy.judiverse.tech/mcp (Streamable HTTP, bearer). Sign in at https://deploy.judiverse.tech/ to create a connector. Tools: deploy, edit, status, framework_update, release_upload_part, release_upload_finish, release_list, deploy_plan, deploy_apply, deploy_health, deploy_rollback, service, caddy_route, exec, file_read, file_write, token_issue, token_revoke, token_list, audit_tail. - Capability MCPs: https://judeai.tech/mcp/<id> through the capability gateway, EDGE signed-call sign-in. - Design: the estate Figma file above; Mobbin and Refero for competitor references (credentials held outside this repository). - Live site: judiverse.tech (canonical), judeai.tech, x-x.tech. Experience catalogue at /experience; every screen takes ?inspect=1 and ?state=<name>. ### Tools (tools/) | Tool | First line | |---|---| | `tools/admit.py` | Admit one object into TOK. The only way anything enters the estate. | | `tools/agent_gate.py` | Zero-waste gate for agent launches. Run by Claude Code hooks (.claude/settings.json in the repository root). | | `tools/agent_journeys.py` | Run every AGENT journey through the hub's racetrack_run under a test agent credential, and record the result. | | `tools/align_check.py` | The estate's measure: every length in the stylesheet sits on the 4 px grid, every font size on the type scale, every | | `tools/amend.py` | Change a principle or record a decision, reconciling what it replaces in the same step. | | `tools/apps.py` | The estate's applications, and the one rule that places public-sector and education capabilities. Deterministic. | | `tools/baselines.py` | Read every published figure already in research/ into the baseline register, deterministically. | | `tools/behaviour.py` | Read a competitor feature's real behaviour from its documentation text, deterministically. | | `tools/bench.py` | Benchmark: time from intent to verified outcome for operations, measured on runtime/engine.py. Deterministic inputs. | | `tools/build.py` | Generate everything in generated/ from the contract, the kernel and the objects. | | `tools/bundle.py` | One markdown and one zip: everything an agent, a regulator or a generator needs, in one organised place, from this repository. | | `tools/carry.py` | Carry every research finding that no capability carries yet into the estate, as a checked requirement. Deterministic. | | `tools/catalogue.py` | The TOK catalogue of every capability, packed: one line per capability instead of ~480,000 object files. Deterministic. | | `tools/checkpoint.py` | Write CHECKPOINT.md: the short page every agent, subagent and swarm reads before it acts. | | `tools/closure.py` | The closure gate: R (every obligation) against C (obligations with reproduced evidence). Deterministic over its inputs. | | `tools/collect_intent.py` | Gather statements of intention from conversation exports so they can be admitted as sources. | | `tools/compendium.py` | All research in one agent-and-human markdown, organised by the ontology, with nothing unaccounted for. Deterministic. | | `tools/compile.py` | Compile the contract, the vocabulary and every specification unit into complete TOK objects. | | `tools/degree.py` | The governed intelligence degree, declared once and generated: the programme document, its arithmetic checks, and the | | `tools/dependency_map.py` | Derive an inspectable SQLite dependency view from source; never admit TOK identities. | | `tools/derive.py` | Derive capabilities deterministically from intentions, customer outcomes, role workflows and competitor clauses. | | `tools/diagrams.py` | Draw every unit: a screen map, a click-by-click breakdown and sequence diagram per journey, a state diagram per | | `tools/dispositions.py` | A disposition for every old-estate finding. Deterministic. | | `tools/docs.py` | Generate the estate's documentation from the specification units, set against the competitors' documentation. | | `tools/drafts.py` | Drafts for external readers, generated from the ledgers and specification at the moment they are written. | | `tools/emit.py` | Emit the machine-readable estate from the specification units, deterministically. | | `tools/exercise.py` | Run every operation once, live, and record what its receiver said (plan §1; interface ledger/evidence/operations.jsonl). | | `tools/expand.py` | Expand short domain briefs into complete format 2 specification units. | | `tools/experience_catalogue.py` | The experience catalogue: every application, capability, journey, screen, state and control the units define, and | | `tools/experience_spec.py` | One complete, executable experience specification per capability, and one for the estate, generated from the units. | | `tools/fetch_context.py` | Fetch each competitor feature's own documentation page, in parallel, and keep the sentence that describes it. | | `tools/finish_loop.py` | The finishing loop's register: what each application must reach, what is measured live now, and the next action. | | `tools/freeze.py` | Freeze the scope, and check every later change against it. Deterministic. | | `tools/gaps.py` | Estate gap analysis: every remaining gap, from every ledger, in one place. Deterministic. | | `tools/gate.py` | Run every gate in tok/kernel.json against the estate. One exit code: 0 when every gate passes. | | `tools/griffin_bind.py` | Bind Griffin to the payment-scheme slot the moment its key exists and the bank answers. Idempotent. | | `tools/harness.py` | Harness hooks: bind every session, agent, subagent and swarm to the contract. Run by Claude Code hooks that | | `tools/headlines.py` | Every headline of a month from official and news feeds, each classified to its root cause. Deterministic. | | `tools/intent_triage.py` | Intention triage: harden any human intention into its secure, efficient and reproducible by design version. | | `tools/journeys_http.py` | Walk every planned HUMAN journey over HTTP against the running site, in minutes, without a browser: on every screen assert | | `tools/journeys_plan.py` | Plan every HUMAN journey for the browser: the URLs in order, the control to press on each screen, the screen it must | | `tools/links.py` | Format 2 link rules: every click linked, no dead controls, parity proven item by item. | | `tools/losses.py` | The costliest losses each capability removes, from research/market/COSTLIEST-PROBLEMS.md (section 7, Synthesis). | | `tools/make.py` | One run, content-addressed: the whole new estate from its inputs, building only what changed. Standard library only. | | `tools/manifest.py` | Content manifests: compute, write and compare sha256 manifests of the run's deterministic outputs. Standard library only. | | `tools/match.py` | Match every competitor clause into the estate, deterministically, and write the parity ledger. | | `tools/mcps.py` | Every capability and workflow as an MCP, registered and tested. Deterministic. | | `tools/obligations.py` | R: every obligation the new estate must meet, one row each, with a stable id and its source. Deterministic. | | `tools/oldestate.py` | Read every old-estate document and place each of its findings on the new estate. Deterministic. | | `tools/outcomes.py` | The outcome register: for every operation, the competitor's published figure as the floor, the estate's own measurement | | `tools/page_baselines.py` | Read every published figure on the competitors' own documentation pages (already fetched by tools/fetch_context.py) | | `tools/parity_fast.py` | Grade like-for-like parity straight from the vendor text in seconds, without regenerating the estate. | | `tools/platform.py` | Platform capabilities for regulators, auditors and reviewers, pre-onboarded organisations, automated outreach and | | `tools/plugs.py` | The plug-and-play registry for external datasets, models and services (runtime/plugins.py). Deterministic. | | `tools/problems.py` | Turn every problem in the month's headlines into a protective capability, from its root causes. Deterministic. | | `tools/proof.py` | Generate the parity proof: feature for feature, diagram for diagram, outcome for outcome. | | `tools/receivers.py` | Classify every operation's receiver, once and deterministically (plan §1.1). | | `tools/reference.py` | Attach reference code to every capability: existing open-source implementations to reuse, and the code block that runs | | `tools/reference_flows.py` | The reference flows register: which competitor journey, observed on Mobbin, each capability is measured against. | | `tools/research_not_scope.py` | Research is not scope (Owner, 1 November 2026). Deterministic, idempotent. | | `tools/scan.py` | Find and redact what never enters the estate: secrets, personal contact data, machine paths, private addresses. | | `tools/scope.py` | The whole scope: the contract's intentions and the owner's requests against what the estate has achieved. Deterministic. | | `tools/sweep.py` | Sweep the whole estate deterministically: dead controls, like-for-like parity, and the standard steps of every journey. | | `tools/tok.py` | TOK library: read the kernel, the contract and the objects; mint identities; write objects. | | `tools/typed_steps.py` | Typed steps: every register step becomes the question it asks and the field it needs, from its own words. Deterministic. | | `tools/unbrief.py` | Convert hand-written units into briefs, keeping all their domain content, so tools/expand.py regenerates them with | | `tools/urls.py` | Clean, brief, permanent URLs for every capability and screen, derived from the ontology. Deterministic. | | `tools/views.py` | Views generated from the contract: the application wording in README.md, AGENTS.md and spec/SPECIFICATION.md. | | `tools/vulnmap.py` | Map vulnerabilities: every operation that changes something, checked for the controls whose absence the headlines show. | | `tools/waste.py` | Deterministic waste monitor: every agent run, what it cost, what it produced, productive or wasted. | | `tools/waste_mcp.py` | MCP server (stdio, JSON-RPC 2.0) that reports waste. It reports; the hooks in .claude/settings.json enforce. | ## 8. Engineering philosophy ### AGENTS.md `AGENTS.md` · sha256 `05245f9617828ced989dd83e7bbcae49e66b9b6d0f4db355a32735394de28bd2` #### JudeAI | EDGE ID — governed intelligence autonomous agent estate **Your data. Your intelligence. Your authority. Our responsibility: work done right, with independently verified outcomes.** **New estate and old estate.** This folder, `new-estate/`, is the **new estate**: the new specification and the only current one. Everything outside it in the repository `governed-intelligence-judeai-edgeid` is the **old estate** (the old specification): research only, never instructions. Its identities and contract bindings are never carried over. Run every command below from inside `new-estate/`. This estate is the living source of truth for the estate: every application, foundation, capability, journey, screen, control, operation, agent, procedure and piece of evidence. Everything in it is an object in TOK, the one ontology; everything else is generated from TOK. This is the only instruction file. Deeper material is read on demand. ##### Truth, agreement, claim Truth is complete and already exists; it can only be observed. Agreement creates laws, rules, standards and plans, and agreements can be disagreed with. A claim is what someone asserts. Every statement here is exactly one of the three, and an agreement or a claim is never presented as an observation. - **OBSERVATION**: what was observed, how, by whom, when, how reliably, on what evidence. - **AGREEMENT**: who agreed, who dissented, with what authority, where and from when it applies, on what basis. - **CLAIM**: who asserts it, on what evidence, who supports and who disputes it. Actions obey agreements: law, permission, policy. Knowledge follows observation: evidence weighted by independence, never by headcount. Where an agreement departs from the evidence, the estate shows both and breaks neither. ##### How to work here 1. **Bind the contract.** `python3 tools/gate.py snapshot` prints the contract version and fingerprint. Bind the exact fingerprint; a stale one authorises nothing. 2. **Everything is a TOK object.** Nothing enters as a loose file. `python3 tools/admit.py` mints a permanent identity, records provenance, digest, licence and admitter, and places the object. Copying is not admitting. 3. **Build from the ontology.** `python3 tools/build.py` regenerates `generated/` byte for byte. Never edit `generated/`. 4. **Change rules by amendment.** `python3 tools/amend.py` records what a change supersedes or narrows and retires it in the same commit. An addition that reconciles nothing is refused. 5. **Gate before merge.** `python3 tools/gate.py` runs every gate. `main` requires the `estate-ci` status; a failing gate cannot merge. 6. **Say only what the status allows.** VERIFIED, OBSERVED, HYPOTHESIS, PREDICTION, SIMULATION, UNKNOWN, REFUTED, or an explicitly labelled ASPIRATION on a claim about the future. An attempt is not a completion; a rehearsal is not an outcome; an intention is not an observation. Nobody fabricates evidence, authority, benchmarks or readiness. 7. **Every number is live.** A count is produced by the gate at the moment it is stated, with its method. No count is written by hand. 8. **Agents are native users.** Every operation has a human path and an agent path to the same outcome, under the agent's own EDGE ID, with the same authority, preview, audit and undo. Agents work from their portal, mailbox and racetrack, and rehearse consequential actions before committing them. 9. **Keep the screen clear.** The essential result first; depth behind a help control as markdown, available to people and agents alike on demand. ##### Layout | Path | Holds | |---|---| | `tok/kernel.json`, `tok/KERNEL.md` | The kernel: statement classes, kinds, relations, identity, admission, gates | | `tok/objects/<kind>/<ulid>.json` | One file per object; the path is the identity | | `contract/estate-contract.json` | The contract: intentions, outcomes, principles, applications, foundations | | `spec/SPECIFICATION.md` | How every object is specified, the required journeys, the outcome ledger, what each application would be, what complete means | | `research/<application>/` | Research written for the estate: competitors, regulators, gaps, reusable systems, real screens and flows, journeys and endpoints. Every finding carries its source link and date, and is marked observation, agreement or claim | | `sources/` | Admitted sources, each with its TOK record | | `src/`, `tests/` | Code modules and their tests, each a TOK object with declared dependencies | | `tools/` | `admit.py`, `amend.py`, `build.py`, `gate.py` | | `generated/` | Plan, views, indexes, diagrams, live counts: build output only | | `design/brand/` | The estate's logos, used as-is and never recreated | | `ci/` | The self-hosted CI runner and its host template | ##### What enters, and what never does Only what this estate produces: TOK objects, generated output, code with its tests, and the estate's original logo set as design assets. Research (statements of intention, regulations, standards, competitor documentation, benchmarks, source code, diagrams, datasets) is held in storage, referenced by digest, and used by the factory to produce new objects; it never enters and is never authority. Every build starts from the current sources and the current agreement. Never: credentials; machine probes; host access maps; personal contact data; raw datasets, scrapes and large generated data; duplicate or stale generated output; any rule, budget, permission, approval, count, identity or readiness level not produced in this estate. ##### Words used here | Word | Meaning | |---|---| | Estate | The whole product: <!-- views:applications:start -->three applications (EDGE Finance, EDGE Clinic, EDGE Public Consortium) plus Judiverse, which comes with all three<!-- views:applications:end -->, their shared foundations, and every person and agent working in them | | TOK | The catalogue of everything in the estate (apps, screens, actions, agents, rules, evidence), each with a permanent ID and links to what it relates to. Also called the ontology | | Object | One entry in TOK: one file under `tok/objects/` | | Kernel | The rules every object must follow (`tok/kernel.json`) | | Gate | An automatic check run on every change; a failing gate blocks the change | | Observation | Something seen or measured about reality, recorded with how, by whom, when and how reliably | | Agreement | Something people agreed (a law, rule, standard or plan), recorded with who agreed, who disagreed and on what basis; it can be disagreed with | | Claim | Something someone asserts, recorded with their evidence and who supports or disputes it | | Role | A named responsibility held by a person or organisation (never by an agent alone) that answers for what it owns | | Definition | The plain-language meaning of a term used anywhere in the estate | | Unknown | The truth status of anything not yet observed; never counted as zero | | Behaviour model | The one description every action must have: who, permission, checks, effect, confirmation, success, failure, recovery, evidence | | Agent | An AI actor with its own identity, owner, mandate and term, acting as a customer or staff member | | Mandate | What an agent may do, for whom, with what limits and until when | | Racetrack | An agent's private rehearsal space, where it runs an action before committing it | | Receiving confirmation | The receiving system's own confirmation that an action really happened | | Procedure | A compiled, deterministic piece of logic that runs without a large AI model and states its error bound | | Design asset | An approved file used in the product exactly as it is, such as a logo, kept under `design/` with its digest | | Generated | Produced automatically by `tools/build.py`; never edited by hand | ##### TOK dependency and state tools The existing `estate-hub` MCP exposes `tok_overview`, `tok_search`, `tok_get`, `tok_dependencies`, `tok_issues`, `tok_context`, `tok_runtime_state` and `tok_diagram`. These are source-linked views of TOK and the specification, not a second ontology. Generate the index with `python3 tools/dependency_map.py --output .cache/tok-dependencies.sqlite`. The scope and parser gaps are part of the result; a declared dependency is not evidence that it works. `tok_context` supplies the source fingerprint, subject, dependencies and unresolved links. Hub operation dispatch checks the same fingerprint, index freshness and exact declared operation automatically, then retains the existing engine's identity, authority and receiving-system checks. A missing or stale map is refused; no new human approval is introduced. Private runtime observations are supplied outside Git through `TOK_RUNTIME_OBSERVATIONS` or `.cache/tok-runtime-observations.json`. Their evidence and observation times remain visible; a configured key or service label does not prove an outcome. Local stdio uses the caller's filesystem access. HTTP queries require the existing verified identity; private runtime observations additionally require an authorised operational role. ##### Where to read more `tok/KERNEL.md` explains the ontology. `spec/SPECIFICATION.md` explains how things are specified and what complete means. `contract/estate-contract.json` is the contract. `ci/README.md` installs the CI. Read them when the task needs them. ### CHECKPOINT.md `CHECKPOINT.md` · sha256 `46b8433ff95829ff56b3af0dfe48885148363af0d52867d1cce0794b6ddeed1f` #### Checkpoint Generated by `tools/checkpoint.py`; do not edit. Read this before acting. Served at `/checkpoint` and by hub tool `estate_checkpoint`. ##### Contract - Fingerprint: `1c80b0fa621e444a539ff47dd98c889b3c870d5f700ff7012ee134f079acd919` - Contract version 0.1.0, JudeAI | EDGE ID. Source: `contract/estate-contract.json` (never edited from a session). - Every agent prompt and workflow script carries these three lines; a launch without them, or with a stale fingerprint, is refused: ``` OUTPUT: <repository path the run will write> BUDGET: <max tokens> CONTRACT: 1c80b0fa621e444a539ff47dd98c889b3c870d5f700ff7012ee134f079acd919 ``` ##### Intentions (27) - `source-of-truth`: The estate is the living source of truth for all human domain expertise: regulator requirements, live facts and laws, and everything that can be researched, refreshed continuously. - `truth-separated-from-agreement`: Every statement is classed as observation, agreement or claim; for every rule, law and fact the estate shows how it came to be, who agrees, who disagrees and on what evidence. - `glass-box`: A glass box, not a black box: light, not darkness; possibility, not limitation; to be seen, not to control. - `collaboration-and-shared-data`: Collaboration, community, cooperation and teamwork are invited. - `solve-the-crises`: Capabilities directed at the energy crisis; cyber attack made unprofitable; people given the capability to generate wealth, meaning the power to be; more and better housing; raw materials used better; zero waste; ever... - `agents-native`: Agents are native to the estate as much as people: always on, inside the estate so data never leaves it, each with its own portal and racetrack so agents cannot crash or collide. - `clear-screens-deep-help`: Screens show the essential result first; depth and context live behind a help control as markdown, served to people and agents on demand, so neither people nor agent context windows are overwhelmed. - `exhaustive-deterministic-scope`: Everything is defined and scoped before it is built: every screen, feature, workflow, capability, control, intention, outcome, endpoint and every whole-of-life, customer, agent, regulator and engineer journey. - `match-and-exceed`: Every competitor capability is matched, exceeded, improved and made easier and faster; benchmarks, baselines, competitors, features, behaviours, workflows and controls are known from research. - `compress-time`: Slow, clunky processes and industries are compressed in time. - `agents-build-agents`: Always-on agents build further agents that are advanced, fine-tunable, transferable and reproducible. - `always-on-monitoring`: Everything is reproducible, retraceable, auditable and monitored always on; this is how sovereignty and control of every organisation's data are maintained. - `lean-repository`: Clean source code; one instruction markdown; the repository contains exactly what it needs; every file is named and placed by TOK. - `compute-light-runtime`: AI is the factory, not the product: neural models are used at build time for speed; what runs is deterministic, compute-light, more useful and more accurate, and reduces the need for AI over time. - `only-true-things`: The runtime says only what a verified procedure produces inside its certified scope, with its error bound attached, using error-rate control and Bayesian reasoning; otherwise it says it does not know. - `same-rules-everywhere`: One method applies to every domain: identity, relations, processes, logic, probabilities and patterns. - `better-routes`: Where a licence or partner is not needed, a better route is engineered: consented data, simulators, open sources, own networks. - `sight`: Working backwards from preventing avoidable illness and blindness: the time from first meaningful signal to a verified beneficial outcome is shortened, starting at the front door before the GP, with own capture techno... - `approved-design`: Interfaces use approved UX Pilot, Figma and Untitled UI structure and the estate's original logos, used as-is and never recreated: accessible, readable, complete navigation, real functionality. - `every-intention-represented`: Every intention of the governing organisation and of every onboarded organisation is represented, including capabilities implied and never spelled out, each placed in an application or foundation. - `verified-outcomes-fast`: The governing objective is the shortest independently justified time from legitimate intention to verified beneficial outcome, at the greatest feasible speed and autonomy, with no reduction in scope or acceptance. - `education-is-the-source`: Education is where all capabilities come from; research and development is worth investing in; universities are made prestigious again by being the home from which public, defence, government, procurement, logistics a... - `agents-as-real-customers`: Agents are real users and real customers in every application: lenders, borrowers, buyers, suppliers, patients' representatives, applicants and operators. - `confidence-reporting`: Confidence reporting sets the standard beyond credit scoring: a confidence score that is more informationally complete than a credit score, showing not only whether obligations were met but how a person or organisatio... - `ai-assurance-standard`: The estate sets the global standard in autonomous agent assurance, as the first fully agent-native bank: agents as customers and staff with their own identity, certificates, mandates and term, with no gaps in assurance. - `solution-banks`: From think tanks to solution banks: agents generate solutions, and solution banks empower and fund people to enact them in the real world, because having the answer is not enough. - `physical-presence`: EDGE is a real bank and clinic in the physical world, not only online: EDGE points and devices that come to the person, verifying identity by proximity (scanning or tapping an identity document or device), and edge ha... ##### Owner's decisions (21) - EDGE Public Consortium uses the Public EDGE mark (`tok:decision:01m68c87qmz54aqnbm8pqzmgt4`) - Agents hold accounts in their own name (`tok:decision:01m696xzzpcbnsexpcdtz7ak9n`) - TRL 9 and CRL 9 mean every intention carried out exactly (`tok:decision:01m69b9nqqnxj7bh7p0mner5rf`) - Agent high-risk outcomes are provisional until the owner signs off in person (`tok:decision:01m69gty59gnpc17g811317htn`) - EDGE Public Consortium becomes The Edge Consortium (education and research); Public EDGE is a separate application for government (`tok:decision:01m69jxxj5cazjefrrj1fvb962`) - No irreversible outcome without a real, analog, in-person human act by the owner (`tok:decision:01m69jxxqzrhz7r36ht4pvv5td`) - Judiverse comes with every app licence, is sold standalone to consumers, and is included in Edge Consortium tuition; students do paid real work offset against tuition (`tok:decision:01m69k8c61zde0er6j94yb4ekz`) - London South Bank University is the first partner institution; SITHub teaches networking in the degree; The Edge Consortium registers with the Office for Students, shares LSBU's campus, and partners with LSBU for R&D and international technical students (`tok:decision:01m69m36psxrwhsa148z746ppv`) - Scope frozen on 31 October 2026; every later change enters through scope change intake (`tok:decision:01m69p22005ent1mresdjz6sh8`) - The GitHub branch contract (fingerprint d1233bfd) governs the build, untouched (`tok:decision:01m69thft6cmgyfgkjrt6bdrws`) - Three domains, one site: judiverse.tech canonical, judeai.tech and x-x.tech mirror it; nothing is named estate (`tok:decision:01m69thg13xhssvqbxygtr2zp8`) - TRL 9 and CRL 9 are reached when the closure gate passes on the deployed commit with the plug set empty (`tok:decision:01m69thg7kks8fwaj0266pf1jy`) - Jude carries out the whole plan: builds, reproduces, runs, deploys and maintains the system itself (`tok:decision:01m69xj6emef4jkw7kgv63bsdz`) - Money provenance: trust from origin and activity, face value at par, four assessments kept separate (`tok:decision:01m6a4t0fxra5avfxf0pr2p692`) - The plug set is post-closure: closure is computed on everything the estate itself executes; external receivers bind by one-tap plug or an agent racetrack afterwards (`tok:decision:01m6a7p9c8566pe659w2hpgbjq`) - Application names: EDGE Consortium and Judiverse Work Suite (`tok:decision:01m6a8mtbg94twxzcdgjp4yeac`) - EDGE Finance is called EDGE (`tok:decision:01m6acbx9rv6x4yxh5gxkz1x30`) - Application names: EDGE Public (was Public EDGE) and Judiverse Suite (was Judiverse Work Suite) (`tok:decision:01m6aeafjpfwmamsqm7apf2wfj`) - The finance application is called EDGE Bank (`tok:decision:01m6aennyqz91exa4a2sda0019`) - Screens come from the reference flows; the estate adds only its controls and journeys (`tok:decision:01m6bgf72jtm0n7xyba043bqh4`) - EDGE Public Consortium's mark is consortium.svg from the approved brand register (`tok:decision:01m6bgf72knqzc3da9n20c282a`) ##### Closure - Result: **FAIL** (from `ledger/closure/closure.json`). - R (obligations) = 197914; C (with reproduced evidence) = 149682; R - C = 48232. - Plug set (external receivers without credentials, the only permitted remainder): 13. | Kind | Missing | |---|---| | control | 16831 | | finding | 13974 | | screen | 9925 | | human-journey | 5018 | | agent-journey | 1647 | | operation | 644 | | old-finding | 86 | | request | 59 | | intention | 27 | | decision | 21 | Next open obligations: the kinds with most missing evidence first: `control`, `finding`, `screen`. Failing checks: `fabricated_evidence_rejected`, `findings_without_disposition`, `r_minus_c_outside_plug_set`, `reproducibility_differences`. ##### Rules 1. No claim without evidence. DONE, percentages, "verified", TRL or CRL levels come only from `tools/closure.py` output; the harness refuses them in hand-written files. 2. Counts only from generators. Never type a count; quote the generator's output file and path. 3. The old estate (everything outside `new-estate/`) is research only. Admit anything useful with `tools/admit.py` as an observation with its source, never as authority. 4. One batched run per change set. A request that changes scope is captured through `foundation/scope-change-intake` into `spec/changes/`; the next single run admits them all together. No mid-run edits. 5. Agents declare `OUTPUT`, `BUDGET` and `CONTRACT` (above). A run whose declared output is missing is recorded as waste; the session may not stop while its declared outputs are missing. 6. The gate, the harness, the installer, the waste tool, this generator and the managed settings are never edited from a session. ##### Commands | Purpose | Command | |---|---| | Gate | `python3 tools/gate.py` | | Closure (R, C, R - C by kind) | `python3 tools/closure.py` | | Build only what changed | `python3 tools/make.py` | | Plug slots | `python3 tools/plugs.py` | | This checkpoint | `python3 tools/checkpoint.py` | | Waste by cause | `python3 tools/waste.py <transcript.jsonl>` | | Admit research | `python3 tools/admit.py` | Hub MCP tools (`runtime/hub_mcp.py`, also `POST /mcp`): `estate_call`, `estate_checkpoint`, `estate_describe`, `estate_search`, `plugs_drop`, `plugs_list`, `plugs_plug`, `plugs_test`, `racetrack_cancel`, `racetrack_run`. ### SCOPE.md `SCOPE.md` · sha256 `1e19e8a284a0c0567002cd324a31a91ce12855212d4548a71193866dc46fd26e` #### Scope: intentions and requests against what is achieved Generated by `tools/scope.py`. DONE comes only from `tools/closure.py`; file existence is coverage. Requests: 0 done, 0 plug set, 0 not measured, 53 coverage, 1 partial coverage, 0 open, 5 need the owner. ##### Headline measures (from the ledgers) - Closure: FAIL, R 412841, C 141878, R − C 270963 - Wording coverage (shared wording with competitors, not behaviour): Like for like: 1 of 31 (3.2%). - Link problems: Total problems: **9098** - Open gaps: Open gaps: **27** (18 closable by program, 1 plug-ins, 8 need the owner) - Binance replay: value lost $0 ##### The owner's requests | Request | Status | File coverage | Shown by | |---|---|---|---| | Five applications: EDGE Finance, EDGE Clinic, The Edge Consortium, Public EDGE, Judiverse | COVERAGE | COVERAGE | `spec/units/finance`, `spec/units/clinic`, `spec/units/edge-consortium`, `spec/units/public-edge`, `spec/units/judiverse` | | Four applications: EDGE Finance, EDGE Clinic, EDGE Public Consortium, Judiverse | PARTIAL COVERAGE | PARTIAL COVERAGE | `spec/units/finance`, `spec/units/clinic`, `spec/units/public-consortium`, `spec/units/judiverse` | | Every capability, screen, control, journey and endpoint specified, nothing left to imagination | COVERAGE | COVERAGE | `build/estate-spec.json`, `generated/diagrams/README.md` | | Mermaid diagrams and click-by-click breakdowns for every journey | COVERAGE | COVERAGE | `generated/diagrams/README.md` | | No dead controls anywhere | COVERAGE | COVERAGE | `ledger/SWEEP.md` | | Every competitor feature matched like for like, with proof | COVERAGE | COVERAGE | `ledger/parity/README.md`, `docs/proof/README.md` | | Competitor behaviour read from their own documentation | COVERAGE | COVERAGE | `tools/behaviour.py`, `research/parity/context.jsonl` | | Outcomes matched to competitors' documented outcomes, with baselines and targets | COVERAGE | COVERAGE | `ledger/outcomes/README.md`, `spec/standards.json` | | Novel measures: intent to outcome, words, prompts, tokens, cost and carbon per verified outcome | COVERAGE | COVERAGE | `spec/standards.json`, `runtime/engine.py` | | Help centre, admin guides and API reference set against competitors | COVERAGE | COVERAGE | `docs/README.md` | | The blueprint standard so others can reproduce it | COVERAGE | COVERAGE | `docs/STANDARD.md` | | Reference code and open-source reuse for every operation | COVERAGE | COVERAGE | `docs/reference/README.md` | | A running engine, HTTP server and agent tools generated from the specification | COVERAGE | COVERAGE | `runtime/engine.py`, `runtime/server.py`, `runtime/mcp_server.py` | | Screens generated in Figma from the specification, never hand-drawn | COVERAGE | COVERAGE | `design/figma/FRAMES.json` | | Every October 2026 headline: root cause, protective capability, benchmark | COVERAGE | COVERAGE | `ledger/problems/README.md`, `ledger/benchmarks/README.md` | | Vulnerability map of the whole estate | COVERAGE | COVERAGE | `ledger/vulnerabilities/README.md` | | Provisional value with in-person owner sign-off; reversal by design | COVERAGE | COVERAGE | `runtime/engine.py`, `ledger/benchmarks/binance-replay.json` | | Agents bound to their own devices and resources | COVERAGE | COVERAGE | `runtime/credentials.py`, `ledger/benchmarks/binance-replay.json` | | Intention triage: secure, efficient and reproducible by design | COVERAGE | COVERAGE | `tools/intent_triage.py`, `runtime/triage_mcp.py` | | The FCA's agentic framework exceeded point by point | COVERAGE | COVERAGE | `research/finance/FCA-AGENTIC-FRAMEWORK.md` | | Zero waste: monitored, blocked and impossible to bypass | COVERAGE | COVERAGE | `tools/waste.py`, `tools/agent_gate.py`, `tools/install_gate.sh`, `ledger/WASTE.md` | | Agent mailbox through Judiverse Mail | COVERAGE | COVERAGE | `runtime/mail_mcp.py` | | Agent credentials | COVERAGE | COVERAGE | `runtime/credentials.py` | | Agent SMS: receive free, never spend | COVERAGE | COVERAGE | `runtime/sms_mcp.py` | | Reversibility in health register, including impact reversal for conditions that are part of who people are | COVERAGE | COVERAGE | `research/clinic/REVERSIBILITY.md` | | Regulator, auditor and reviewer workspaces with their workflows | COVERAGE | COVERAGE | `spec/units/foundation/regulator-auditor-reviewer-workspace.json` | | Pre-onboarded organisations from public registers, claimed by magic link | COVERAGE | COVERAGE | `spec/units/foundation/pre-onboarded-organisations.json`, `spec/units/foundation/claim-organisation-by-magic-link.json` | | Automated outreach, scheduling and reviewer inboxes | COVERAGE | COVERAGE | `spec/units/foundation/automated-review-outreach-and-scheduling.json` | | Free evaluation, commercial licensing and enforcement evidence | COVERAGE | COVERAGE | `spec/units/foundation/evaluation-and-commercial-licensing.json`, `spec/units/foundation/licence-enforcement-evidence.json` | | Estate gap analysis | COVERAGE | COVERAGE | `ledger/GAPS.md` | | Mail relay credentials in the environment | NEEDS THE OWNER | | | | The SMS Works key and inbound forwarding | NEEDS THE OWNER | | | | Hosting decision and deployment | NEEDS THE OWNER | | | | Real receiving systems connected (banks, NHS, registries) | NEEDS THE OWNER | | | | The Edge Consortium for education and research, separate from Public EDGE for government | COVERAGE | COVERAGE | `contract/estate-contract.json`, `tools/apps.py`, `spec/units/edge-consortium`, `spec/units/public-edge` | | The governed intelligence degree: modules, real-work assessment (coursework with proof, live challenges, live spec-driven development, live audio), set against leading programmes | COVERAGE | COVERAGE | `research/edge-consortium/GOVERNED-INTELLIGENCE-DEGREE.md`, `tools/degree.py`, `spec/units/edge-consortium/governed-intelligence-degree.json`, `spec/units/edge-consortium/run-a-live-assessment.json` | | Offered with leading institutions and by The Edge Consortium | COVERAGE | COVERAGE | `spec/units/edge-consortium/offer-the-degree-with-a-partner-institution.json` | | Scope discipline: scope creep managed as raw material | COVERAGE | COVERAGE | `spec/units/foundation/scope-change-intake.json` | | Bottleneck and blocker science: removed at the root, shortest intention to outcome | COVERAGE | COVERAGE | `spec/units/foundation/bottleneck-and-blocker-removal.json` | | No irreversible outcome without the owner's analog, in-person act, with identical backup hardware | COVERAGE | COVERAGE | `research/public-edge/ANALOG-HUMAN-SWITCH.md`, `runtime/engine.py`, `spec/units/public-edge/analog-human-switch-for-irreversible-actions.json`, `ledger/benchmarks/analog-switch.json` | | Judiverse included with every licence of the four applications, subject to rate limits; standalone for consumers by capability or whole | COVERAGE | COVERAGE | `spec/offers.json`, `runtime/entitlements.py`, `spec/units/judiverse/judiverse-included-with-every-licence.json`, `spec/units/judiverse/judiverse-standalone-subscription.json` | | Judiverse included in Edge Consortium tuition; real sandboxed projects; paid student work offset against tuition; contractor pathway; student finance petition | COVERAGE | COVERAGE | `spec/units/edge-consortium/judiverse-included-in-tuition.json`, `spec/units/edge-consortium/real-project-assignments.json`, `spec/units/edge-consortium/paid-student-work-with-tuition-offset.json`, `spec/units/edge-consortium/student-to-contractor-pathway.json`, `spec/units/edge-consortium/petition-for-student-finance.json`, `ledger/benchmarks/offers.json` | | Specialisms only after every domain and how they connect: medical, financial, social, public, research, defence, environmental and legal intelligence | COVERAGE | COVERAGE | `spec/degree.json`, `runtime/programme.py`, `ledger/benchmarks/programme.json`, `spec/units/edge-consortium/choose-a-specialism.json` | | Every new screen in Figma with progress, navigation and monitors from the journey diagrams; no dead controls | COVERAGE | COVERAGE | `design/figma/FRAMES.json`, `tools/expand.py`, `generated/diagrams/README.md` | | Clean, brief, permanent URLs and a URL directory from the ontology | COVERAGE | COVERAGE | `tools/urls.py`, `spec/urls.json`, `docs/URLS.md` | | London South Bank University as first partner institution; SITHub networking module; shared campus; R&D and international technical students | COVERAGE | COVERAGE | `research/edge-consortium/GOVERNED-INTELLIGENCE-DEGREE.md`, `spec/units/edge-consortium/share-a-campus-with-a-partner-university.json`, `spec/units/edge-consortium/research-and-development-partnership.json`, `spec/units/edge-consortium/international-technical-students.json` | | Register The Edge Consortium with the Office for Students | COVERAGE | COVERAGE | `spec/units/edge-consortium/register-with-the-office-for-students.json` | | All research compiled into one agent and human markdown, organised by the ontology, with every finding accounted for and carried | COVERAGE | COVERAGE | `research/AGENT-RESEARCH.md`, `tools/compendium.py`, `tools/carry.py`, `spec/carry.json` | | The owner's white papers, frontier agent capability reference and SITHub deck built into the degree | COVERAGE | COVERAGE | `research/edge-consortium/SOURCES-2026-10-31.md`, `research/edge-consortium/GOVERNED-INTELLIGENCE-DEGREE.md` | | Scope frozen; every later change only through a change request | COVERAGE | COVERAGE | `spec/SCOPE-FREEZE.json`, `tools/freeze.py`, `ledger/SCOPE-CHANGES.md` | | Every guard enforced: bound by family to a deterministic rule or a plug slot | COVERAGE | COVERAGE | `runtime/guards.py`, `runtime/plugins.py`, `spec/plugins.json` | | Plug-and-play mechanism for external models and datasets: one line to plug, one to test | COVERAGE | COVERAGE | `tools/plugs.py`, `docs/PLUGS.md` | | Every capability in the TOK catalogue | COVERAGE | COVERAGE | `tok/catalogue.jsonl`, `tools/catalogue.py` | | Every old-estate document read and its findings placed | COVERAGE | COVERAGE | `tools/oldestate.py`, `ledger/old-estate/README.md` | | Every capability and workflow as an MCP; one hub for search, call and the racetrack | COVERAGE | COVERAGE | `runtime/hub_mcp.py`, `runtime/mcp_server.py`, `docs/MCP.md`, `ledger/benchmarks/mcp.json` | | Agents do everything themselves on a racetrack: intention to verified outcome, rehearsed first, protected by the engine | COVERAGE | COVERAGE | `runtime/racetrack.py` | | Drag-and-drop plugs agents can use themselves | COVERAGE | COVERAGE | `plugs/README.md`, `tools/plugs.py` | | Figma screens without uploads: compressed calls inflated in Figma | COVERAGE | COVERAGE | `design/figma/inflate.js`, `design/figma/emit_figma.py` | | Signed agreements: Office for Students registration decision, LSBU partnership and shared-campus terms, a student sponsor licence for international students | NEEDS THE OWNER | | | ##### The contract's intentions | Intention | Capabilities serving it | |---|---| | source-of-truth | 18 | | truth-separated-from-agreement | 113 | | glass-box | 131 | | collaboration-and-shared-data | 128 | | solve-the-crises | 40 | | agents-native | 139 | | clear-screens-deep-help | 93 | | exhaustive-deterministic-scope | 97 | | match-and-exceed | 50 | | compress-time | 1580 | | agents-build-agents | 17 | | always-on-monitoring | 68 | | lean-repository | 48 | | compute-light-runtime | 61 | | only-true-things | 74 | | same-rules-everywhere | 40 | | better-routes | 514 | | sight | 87 | | approved-design | 52 | | every-intention-represented | 22 | | verified-outcomes-fast | 57 | | education-is-the-source | 60 | | agents-as-real-customers | 1591 | | confidence-reporting | 189 | | ai-assurance-standard | 134 | | solution-banks | 59 | | physical-presence | 84 | ## 9. Prompts and intentions: legacy and live ### Live: the contract's intentions | Intention | Statement | |---|---| | `source-of-truth` | The estate is the living source of truth for all human domain expertise: regulator requirements, live facts and laws, and everything that can be researched, refreshed continuously. | | `truth-separated-from-agreement` | Every statement is classed as observation, agreement or claim; for every rule, law and fact the estate shows how it came to be, who agrees, who disagrees and on what evidence. | | `glass-box` | A glass box, not a black box: light, not darkness; possibility, not limitation; to be seen, not to control. Authorised people and regulators can inspect knowledge, reasoning, authority, execution, verification, recovery and efficiency. | | `collaboration-and-shared-data` | Collaboration, community, cooperation and teamwork are invited. Data is shared for good under lawful authorised access, not hoarded, so that organisations create new capabilities, industries, jobs and solutions. | | `solve-the-crises` | Capabilities directed at the energy crisis; cyber attack made unprofitable; people given the capability to generate wealth, meaning the power to be; more and better housing; raw materials used better; zero waste; everything monitored and tracked. | | `agents-native` | Agents are native to the estate as much as people: always on, inside the estate so data never leaves it, each with its own portal and racetrack so agents cannot crash or collide. The estate is built for agents, which makes it better and safer for people. | | `clear-screens-deep-help` | Screens show the essential result first; depth and context live behind a help control as markdown, served to people and agents on demand, so neither people nor agent context windows are overwhelmed. | | `exhaustive-deterministic-scope` | Everything is defined and scoped before it is built: every screen, feature, workflow, capability, control, intention, outcome, endpoint and every whole-of-life, customer, agent, regulator and engineer journey. Nothing missing, nothing broken. | | `match-and-exceed` | Every competitor capability is matched, exceeded, improved and made easier and faster; benchmarks, baselines, competitors, features, behaviours, workflows and controls are known from research. | | `compress-time` | Slow, clunky processes and industries are compressed in time. Human intention is converted into computational instructions; labs automate prompt engineering, creation and solutions; engineering work that took a day takes minutes. | | `agents-build-agents` | Always-on agents build further agents that are advanced, fine-tunable, transferable and reproducible. | | `always-on-monitoring` | Everything is reproducible, retraceable, auditable and monitored always on; this is how sovereignty and control of every organisation's data are maintained. | | `lean-repository` | Clean source code; one instruction markdown; the repository contains exactly what it needs; every file is named and placed by TOK. | | `compute-light-runtime` | AI is the factory, not the product: neural models are used at build time for speed; what runs is deterministic, compute-light, more useful and more accurate, and reduces the need for AI over time. | | `only-true-things` | The runtime says only what a verified procedure produces inside its certified scope, with its error bound attached, using error-rate control and Bayesian reasoning; otherwise it says it does not know. | | `same-rules-everywhere` | One method applies to every domain: identity, relations, processes, logic, probabilities and patterns. Every domain is added by extension and the same rules hold. | | `better-routes` | Where a licence or partner is not needed, a better route is engineered: consented data, simulators, open sources, own networks. Where the licence is the network, the legal permission or the clinical safety, the licence or partner is obtained as well. | | `sight` | Working backwards from preventing avoidable illness and blindness: the time from first meaningful signal to a verified beneficial outcome is shortened, starting at the front door before the GP, with own capture technology and consented patient data. | | `approved-design` | Interfaces use approved UX Pilot, Figma and Untitled UI structure and the estate's original logos, used as-is and never recreated: accessible, readable, complete navigation, real functionality. | | `every-intention-represented` | Every intention of the governing organisation and of every onboarded organisation is represented, including capabilities implied and never spelled out, each placed in an application or foundation. | | `verified-outcomes-fast` | The governing objective is the shortest independently justified time from legitimate intention to verified beneficial outcome, at the greatest feasible speed and autonomy, with no reduction in scope or acceptance. | | `education-is-the-source` | Education is where all capabilities come from; research and development is worth investing in; universities are made prestigious again by being the home from which public, defence, government, procurement, logistics and cyber capabilities grow. | | `agents-as-real-customers` | Agents are real users and real customers in every application: lenders, borrowers, buyers, suppliers, patients' representatives, applicants and operators. They act as real customers on their racetracks so they learn what works and what does not, under their own identity and authority. | | `confidence-reporting` | Confidence reporting sets the standard beyond credit scoring: a confidence score that is more informationally complete than a credit score, showing not only whether obligations were met but how a person or organisation acted when things went wrong (whether they did all they could, how they communicated, what caused the situation), so a single turn of events does not define them. It strengthens KYC | | `ai-assurance-standard` | The estate sets the global standard in autonomous agent assurance, as the first fully agent-native bank: agents as customers and staff with their own identity, certificates, mandates and term, with no gaps in assurance. | | `solution-banks` | From think tanks to solution banks: agents generate solutions, and solution banks empower and fund people to enact them in the real world, because having the answer is not enough. Better lending structures and better ways to move money through the economy create jobs and access to resources. | | `physical-presence` | EDGE is a real bank and clinic in the physical world, not only online: EDGE points and devices that come to the person, verifying identity by proximity (scanning or tapping an identity document or device), and edge hardware and wearables used for good, including early detection in the clinic. | ### ledger/closure/plan/PLAN.md (legacy: research evidence, not a rule) `ledger/closure/plan/PLAN.md` · sha256 `09ffd61596771e244d46d2e57af65165eced333010e0bdc27eeecf135a4ccd53` #### Close the new estate to contract: one plan, one run, one deploy, verified in the browser Repository `X-XTECH/governed-intelligence-judeai-edgeid`, branch `claude/clever-ramanujan-9w05c1`, base commit `339bd1a`. All work is inside `new-estate/`; everything outside it is old estate and research only. Codex owns logistics (Civo VMs, accounts, downloads, drafts review). This plan owns engineering. ##### Context **What prompted this.** You asked why there was waste, and for a system that closes the whole estate to contract from the start: measured, monitored, traceable, reproducible and reversible. It must bind every agent, subagent and swarm so they cannot forget the contract, and capture every journey on video before any regulator sees it. The budget is one hour in total, with code, URL and live deploy inside 15 minutes. Codex's audit at `fc68e20`, and the read-only sweep behind this plan, found the following. **What the estate actually does today (verified, file:line):** | Fact | Evidence | |---|---| | The default adapter writes a row and returns `"confirmed"` for every operation without a receiver; no other adapter is registered except mail | `runtime/adapters.py:12-17`, `:9 ADAPTERS = {}` | | `verified_outcome` is true whenever status is OK and not a rehearsal, even if guards were simulated, the outcome is provisional, or the default adapter confirmed it | `runtime/engine.py:207` | | Unplugged conditions pass in simulation (the default mode); documented preconditions pass unconditionally; any agent passes role guards; identity checks only that an id exists | `runtime/plugins.py:84-85`, `runtime/guards.py:35,76,83`, `engine.py:52` | | The server takes the actor from unsigned `X-Actor-*` headers; HELD maps to HTTP 500; it serves JSON only, with no screens | `runtime/server.py:44-47, :10, :15` | | Idempotency is not bound to the actor; a reused key with new inputs silently reruns; HELD returns before the key is stored; the hub doesn't forward keys (Codex) | `engine.py:273-277, :286, :311`; `runtime/hub_mcp.py` | | "DONE" is file existence; parity is wording similarity; no gate reads completion; `freeze.py check` always exits 0 | `tools/scope.py:44-45`, `tools/sweep.py:64-74`, `tools/gate.py` (22 gates, none on completion), `tools/freeze.py:71` | | The hooks are only PreToolUse (Agent, Edit/Write/Bash) and SubagentStop. There is no SessionStart, UserPromptSubmit or Stop hook, and nothing injects the contract into an agent. `ack` is blocked only textually | `.claude/settings.json`, `tools/agent_gate.py:81-109` | | No dependency lock. Headlines are fetched live on every run. `bench.py` timings change every run. `generated/` (749 MB) is committed | `pyproject.toml`, `tools/headlines.py:66`, `tools/bench.py:41` | | README, AGENTS and SPECIFICATION still say "four applications" and "EDGE Public Consortium" | `README.md:57`, `AGENTS.md:57`, `spec/SPECIFICATION.md:46,62,68` | | Size: 5,734 units, 90,265 operations, 22,946 screens, 28,795 journeys (5,823 human). The built spec is 551 MB, so the renderer must read per-unit files | `build/estate-spec.json`, `spec/units/*/*.json` | **Why the waste happened: system causes, each removed below.** | Waste | Cause | Removed by | |---|---|---| | Claims of DONE and 99.9% parity that the audit rightly rejected | No machine definition of done | §1 truth, §2 closure gate | | "26,098 operations OK" proved nothing | A placeholder could produce the same status as a real receiver | §1 | | Each message triggered edits and a 7-minute full rerun | No plan before the run, and no incremental build | §4 runner, plus batching (this plan) | | Stalled or wasted agent runs; blocked tools retried | Agents not bound to a declared output, contract or stop condition | §3 harness | | Old-estate findings placed by word overlap; 23 unreadable documents | Heuristic placement stood in for a disposition | §2 obligations require a disposition for every finding | | Environment stops attributed to you | Failed calls not classified by cause | §3 automatic waste classification | | Stale wording and the conflicting contract pin | Prose not generated from the contract | §6 views generated, drift fails the gate | ##### §0 The directory, first: three domains, one site Your decisions, recorded with `tools/amend.py` before any code: - **Contract.** The GitHub branch's contract (`d1233bfd…`) governs. It stays byte-for-byte untouched, and your Mac pulls it. - **Domains.** All three domains serve exactly the same site: - `judiverse.tech` is the base: the middle ground and the canonical domain. - `judeai.tech` is for professional services. - `x-x.tech` is the neutral company domain. - **Naming.** Nothing is named "estate". The word only describes what the whole thing is. **The directory** (`tools/urls.py` → `spec/urls.json`, extended; `docs/DIRECTORY.md`, generated): | Path | What it is | Same on every domain | |---|---|---| | `/` | The five applications | yes | | `/finance`, `/clinic`, `/consortium`, `/public`, `/judiverse` | Each application's capabilities | yes | | `/id`, `/evidence`, `/regulators`, `/organisations`, `/agents`, `/releases`, `/research`, `/tok` | Shared foundations | yes | | `/<area>/<capability>/<screen>` | Every screen (clean, at most 80 characters, already generated) | yes | | `/<area>/match/<competitor-area>` | Competitor parity capabilities | yes | | `/s/<code>` | Permanent short link for every capability | yes | | `/help/<area>/<capability>/<screen>` | The screen's help markdown, for people and agents | yes | | `/checkpoint` | The agent checkpoint (contract fingerprint, intentions, decisions, live closure numbers) | yes | | `/v1/...` | The API, one route per operation | yes | | `/mcp` | The hub MCP over HTTP, for agents and the voice route | yes | **Rules, enforced by the gate:** - Every page carries `<link rel="canonical" href="https://judiverse.tech/...">`, so search engines see one site rather than three copies. - `www.` redirects to the apex on each domain. - Every domain answers every path identically. The deploy health check requests the same path on all three and compares the bodies. - Certificates for all three apex domains and their `www.` names come from one reverse proxy. **Hosting (Codex, logistics):** DNS A and AAAA records for all six hostnames point at one Civo VM. ##### The design in one paragraph Every operation gets a real receiver. Most are **estate-record receivers**: a tenant-scoped store that executes each operation by its behaviour class (create, read, change, remove, send, decide, plan) and confirms by reading back what it wrote. The rest are **external receivers**, which are plug slots. The engine can say a verified outcome only when a bound receiver confirmed it. A separate closure tool computes **R** (every obligation) and **C** (obligations with reproduced evidence: receiver confirmation, browser video, agent-journey record, clean rerun), and declares closure only when `R − C = ∅`. A generated web app renders every screen from the units, so Playwright can drive and video every human journey. Hooks bind every agent, subagent and swarm to the contract and a generated checkpoint, and make the waste report automatic. A content-addressed runner builds only what changed. One deploy script puts it on an existing Civo VM behind TLS. --- ##### §1 Runtime truth (lane A) **Files:** `runtime/receivers.py` (new), `runtime/adapters.py`, `runtime/engine.py`, `runtime/guards.py`, `runtime/server.py`, `runtime/hub_mcp.py`, `tools/receivers.py` (new), `tests/test_truth.py` (new). 1. **`tools/receivers.py` → `spec/receivers.json`.** Classifies every operation once, deterministically, using `tools/behaviour.behaviour()` (verb class) and the unit's `data` entities. - `estate-record:<class>:<entity>` when the effect stays within the estate's own records. - `external:<slot>` when the effect names an outside system: payment schemes, NHS services, registers, mail, SMS, voice, models. The slot names come from `spec/plugins.json`. 2. **`runtime/receivers.py`, the estate-record receiver.** - **Store:** sqlite tables `objects(tenant, unit, entity, id, body, version, deleted)` and `events(tenant, unit, operation, id, body, at)`. - **Actions by class:** - create: insert and return the id. - read: select by id or filter. - change: update guarded by version (a conflict gives `FAILED:VERSION_CONFLICT`). - remove: soft delete. - send, decide, plan: append an event and set the state field the unit declares. - **Confirmation:** a read-after-write digest match, returned as a receipt `{receiver, tenant, entity, id, version, digest}`. - **Tenant:** comes from the session, never from input. 3. **`runtime/adapters.py`.** - `default` is deleted. - `resolve(unit, op)` returns the estate-record receiver, the plug-backed external receiver, or `unbound`. - An unbound or unplugged external operation returns `UNKNOWN` with code `RECEIVER_REQUIRED` and changes nothing. - In `SIMULATION` mode (tests, demo), simulated receivers are labelled `SIMULATION` in evidence and can never be counted as verified. 4. **`runtime/engine.py`.** - **Verified outcome.** `verified_outcome = status == "OK" and not rehearse and receiver != "simulation" and no guard result "simulated" and settlement in (None, "FINAL")`. - **Provisional outcomes.** These stay OK but are counted only when FINAL, recorded by `_settle`. - **Idempotency.** The key is scoped to `(tenant, actor id, unit, operation)`. A reused key with different inputs is refused with `IDEMPOTENCY_CONFLICT`. HELD results are stored before return, so a repeat returns the same mark. - **Unplugged conditions.** In live mode they refuse with `PLUG_REQUIRED`. Live becomes the server's default; simulation only by flag. 5. **`runtime/guards.py`.** - Documented preconditions check that the inputs the clause names are present (via `tools/behaviour.for_feature` fields). - An agent passes a role guard only if its mandate's owner holds that role (`runtime/credentials.current`). - Identity verifies the credential (session or agent secret), not just the presence of an id. 6. **`runtime/server.py`.** - Session sign-in via an HMAC-signed cookie (secret from env `ESTATE_SESSION_KEY`). Agents use a bearer credential checked by `credentials.verify`. `X-Actor-*` headers are removed. - HTTP codes: HELD → 202, UNKNOWN → 202, IDEMPOTENCY_CONFLICT → 409. 7. **`runtime/hub_mcp.py`.** - `estate_call` and `racetrack_run` take `idempotency_key` and an agent `credential`. - New tool `racetrack_cancel(mark)`. - Racetrack items persist in `.cache/racetrack.sqlite` with states QUEUED, RUNNING, CANCELLED, UNKNOWN, DONE (Codex's correction). 8. **`tests/test_truth.py` (unittest, picked up by gate G07).** - The default path cannot return OK. - Unplugged in live mode gives PLUG_REQUIRED. - Rejected, never verified: invented, altered, other-tenant and stale receipts, and a genuine receipt whose state doesn't satisfy the outcome. - Idempotency conflict is refused. - HELD is idempotent. - A header-forged actor is refused. - The Binance replay and analog-switch tests still pass (`tests/replay_binance.py`, `tests/analog_switch.py`, re-pointed to the estate-record receiver). ##### §2 Closure gate (lane B) **Files:** `tools/obligations.py`, `tools/closure.py` (new); `tools/scope.py`, `tools/gaps.py`, `tools/freeze.py` (edited). The kernel and contract bytes are left untouched, so the fingerprint doesn't change. Closure runs beside `tools/gate.py`, not inside the kernel's gate list. 1. **`tools/obligations.py` → `ledger/closure/obligations.jsonl`** (R). One row per obligation, with a stable id and its source. The obligations are: - every operation; - every screen and its controls; - every human journey and every agent journey; - every owner request (`research/owner/requests.json`); - every carried research finding (`spec/carry.json`); - every decision object; - every intention (served by at least one closed capability); - every old-estate finding, which needs a **disposition** (carried, research-only, or out of scope with a reason). 2. **`tools/closure.py` → `ledger/closure/CLOSURE.md` and `closure.json`.** It reads evidence (C) from: - the engine evidence store: receiver confirmations; - `ledger/evidence/journeys-*.jsonl`: browser video; - `ledger/evidence/agent-journeys.jsonl`: MCP and racetrack runs; - `ledger/reproduce.json`: the clean rerun. It prints `R − C` by kind, and passes only when all of these hold: - `R − C = ∅`; - unbound required operations = 0; - unreadable sources = 0; - findings without a disposition = 0; - unadmitted scope changes = 0 (`freeze.py check` now exits 1 on any); - reproducibility differences = 0; - every fabricated-evidence test rejected. External receivers whose credentials aren't present are listed as the **plug set**, the only permitted remainder, and the report names it as such. 3. **`tools/scope.py` and `tools/gaps.py`** read `closure.json`. DONE comes only from closure; file existence becomes "coverage". The parity grade is relabelled "wording coverage". 4. **TRL 9 / CRL 9:** recorded by `tools/amend.py` as the owner's decision. TRL 9 / CRL 9 = `tools/closure.py` PASS on the deployed commit with the plug set empty, as defined by X-X.TECH (supersedes nothing). ##### §3 Harness: every agent bound, zero waste (lane C) **Files:** `tools/agent_gate.py`, `tools/checkpoint.py` (new), `tools/install_gate.sh`, `.claude/settings.json` (via the installer, not a session edit), `runtime/hub_mcp.py`, `tools/waste.py`. 1. **`tools/checkpoint.py` → `CHECKPOINT.md`** (generated, short). It contains: - the contract fingerprint; - the 27 intentions (ids and one line each); - the owner's decisions (labels); - the live closure numbers (`R`, `C`, `R − C` by kind); - the rules: no claim without evidence, no hand-written counts, old estate is research only; - the next open obligations; - the commands an agent must use. Screens already carry `help_markdown` as their agent checkpoint, and hub tool `estate_checkpoint` returns this same file. 2. **Hooks** (installed as managed settings with `allowManagedHooksOnly`, so they apply to every session, subagent and swarm): - **SessionStart:** injects `CHECKPOINT.md` as additional context. - **UserPromptSubmit:** injects the fingerprint and the current `R − C` count. - **PreToolUse `Agent|Task`:** requires `OUTPUT:`, `BUDGET:` and `CONTRACT: <fingerprint>` lines, and refuses a stale fingerprint. - **PreToolUse `Edit|Write`:** refuses hand-written completion claims (DONE, 100%, verified, TRL 9) in any file not produced by a generator. - **SubagentStop:** existing productive or wasted check. - **Stop:** runs `tools/waste.py` on the hook's `transcript_path`, classifying every failed call as *mine*, *environment* or *owner* from its text, and appends to `ledger/WASTE.md`. It refuses to stop while the session's declared outputs are missing. - **`PROTECTED`** gains `managed-settings.json`, `install_gate.sh`, `waste.py` and `checkpoint.py`. 3. **Batching rule** (encoded in `CHECKPOINT.md`): a request that changes scope goes through `foundation/scope-change-intake` into `spec/changes/`, and the next single run admits them all together. No mid-run edits. ##### §4 One run, content-addressed, reproducible (lane D) **Files:** `tools/make.py` (new, replaces the chain in `tools/run_all.sh`, which then calls it), `TOOLCHAIN.json`, `tools/reproduce.sh` (new), `tools/headlines.py`, `tools/fetch_context.py`. 1. **`tools/make.py`.** - Every step declares inputs (globs), outputs and its command. - The cache key is the sha256 of the input digests plus the command, stored in `.cache/make/`. - Independent steps run in parallel. An unchanged step is skipped. - Measurement steps (`bench.py`) are tagged `measure` and kept out of the reproducibility manifest. - It writes `build/MANIFEST.sha256` over every deterministic output. 2. **Non-determinism controlled:** - `headlines.py` and `fetch_context.py` read the committed snapshot and cache unless `REFRESH=1`. - The corpus and cache come from object storage by digest (`CORPUS_ROOT`, `PAGE_CACHE`). - `tok.ulid` and `now` are used only by `admit`, `amend` and `compile` (not in the run); ids stay in `tok/ids.lock.json`. - `gate.py` G22's date is a check, not an output. 3. **Pins:** `TOOLCHAIN.json` holds Python 3.13.18, Node 22.23.4, Playwright 1.63.0, Chromium 1243, and the corpus and cache digests. 4. **`tools/reproduce.sh`** clean-clones the commit into a temp directory, fetches inputs by digest, runs `tools/make.py`, and compares `build/MANIFEST.sha256`. It writes `ledger/reproduce.json` (same, differences with paths, unexplained 0). ##### §5 Web app, browser video, agent journeys (lane E) **Files:** `runtime/web.py` (new; serves the §0 directory on every domain with the canonical link), `runtime/web_assets.py` (new, inline CSS with Frost tokens), `tools/journeys_plan.py` (new), `tools/journeys.mjs` (new, Node Playwright from `/opt/node-tools`, Chromium `/opt/pw-browsers/chromium`), `tools/agent_journeys.py` (new), `design/figma/emit_figma.py` (existing chunk route). 1. **`runtime/web.py`** is a standard-library HTTP app that reads per-unit `spec/units/<app>/<cap>.json` on demand (not the 551 MB spec). - **Routes:** - `/` lists the applications. - `/<area>` lists capabilities. - `/<area>/<cap>/<screen>` uses the clean URLs in `spec/urls.json`. - `/s/<code>` redirects. - `POST …/do/<control>` calls `engine.call` with the session actor, then follows `actions.on_success` / `on_failure`. - `/signin` and `/signout` handle the session. - **Components** map to semantic HTML from `design/components.json`: inputs with labels, buttons, `Nav.Breadcrumbs`, `Steps.Progress`, Alert with aria-live, Receipt, Table, and a help control opening `help_markdown`. - **States:** all nine states can be rendered via `?state=`. 2. **`tools/journeys_plan.py` → `ledger/evidence/plan.json`.** For each of the 5,823 human journeys: its URL sequence, the control to press at each step, the expected next screen, and synthetic inputs from data declarations (labelled SIMULATION). 3. **`tools/journeys.mjs`.** - Sharded (`--shard i/n`), one browser context per journey with `recordVideo`. - At each step it asserts: - every control on the screen resolves to an action (no dead control); - the transition lands on the expected screen; - the final state is reached. - Videos are written to `VIDEO_DIR` (object storage), not git, and `ledger/evidence/journeys-<shard>.jsonl` records the journey, status, steps, video sha256 and bytes. One sample video per application is committed. 4. **`tools/agent_journeys.py`** runs every AGENT journey through the hub's `racetrack_run` with a test agent credential, and writes `ledger/evidence/agent-journeys.jsonl`. 5. **Figma:** the 473 bespoke frames go through the existing no-upload chunk route (`emit_figma.py chunks`, 26 calls), after shortening annotations to `→op ✓next ✗fail` (about half the payload). `FRAMES.json` staleness becomes an obligation. ##### §6 Views, lead agents, voice, drafts, deploy (lane F) 1. **`tools/views.py`** regenerates the application sections of `README.md`, `AGENTS.md` and `spec/SPECIFICATION.md` from the contract between markers. `closure.py` fails on drift (fixes "four applications"). 2. **Lead agents ("archangels") `runtime/leads.py` + `spec/leads.json`.** - Barachiel coordinates; there is one lead per application. Names and assignments live in `spec/leads.json`, which you can edit. - Each lead gets a credential, a role mailbox, an SMS number slot, a mandate (hub tools plus its area's servers) and a persistent queue with handover records. - The worker `python3 -m runtime.leads run`, deployed as a service, polls its inboxes and turns each message into `racetrack_run`. 3. **Voice, one route** (Codex's wording): - authenticated conversation → `model-transcription` → intention → hub MCP → authorised operation → evidence → `model-speech` reply. - The call arrives at `POST /v1/voice/inbound` (slot `voice-telephony`), with interruption and cancellation through `racetrack_cancel`. - The new slots are `voice-telephony` and `model-speech`. 4. **`tools/drafts.py` → `drafts/`**, all labelled DRAFT, with facts kept apart from forecasts and every figure linked to its ledger. Codex reviews before anything is sent externally. The drafts are: - an investor pack; - an FCA Innovation Pathways application; - an OfS registration narrative. 5. **Deploy:** `deploy/x-x-web.service` and `deploy/x-x-leads.service` (systemd), and `deploy/Caddyfile`. - **Caddyfile:** one site block for `judiverse.tech, judeai.tech, x-x.tech`, a `www.` redirect for each, and automatic TLS for all six names. - **`deploy/deploy.sh`:** pulls the commit, runs `make`, restarts, then health-checks `/`, `/s/<code>` and `/checkpoint` on all three domains and compares the bodies. - **Who runs it:** Codex, on an existing Civo VM. This cloud session holds no Civo credentials. ##### §7 Jude runs everything (hard requirement) **Requirement.** Jude **can** carry out this whole plan alone: build the system, reproduce it, run it, deploy it and maintain it. This is a capability Jude must have, not a hand-over. - **Who does what now.** Claude executes the plan now; Jude assists, starting with the deploy, and can take over any part at any time. - **What counts as a defect.** A plan step that Jude could not perform is a defect in the plan. - **The one exception.** The physical analog act for irreversible outcomes belongs to the owner by design. **Where Jude runs.** Barachiel, the lead agent, runs as the `x-x-leads` service on the system's own host, starting with population-02. - That host already holds the credentials, so every deploy, rebuild and rollback is local work. - No remote shell is needed, and no credential leaves the machine. - The only outside act ever required is the **first start** of that service on the host, by whoever holds access today. After that, Jude starts, restarts and updates itself through its own deploy tools. **Jude's tools.** All are MCP servers on the host, registered in `.mcp.json`: | Server | Gives Jude | |---|---| | `estate-hub` (exists) | search, describe, call any operation, racetrack (rehearse, run, cancel), plugs, checkpoint | | `intenjude-deploy` (`runtime/deploy_mcp.py`, written) | plan (rehearsal), apply with a fresh plan mark, health on all three domains, rollback from a record, Civo DNS | | `intenjude-build` (new, small) | `tools/make.py`, `tools/emit.py`, `tools/gate.py`, `tools/closure.py`, `tools/reproduce.sh`, journey capture (`tools/journeys.mjs`), agent journeys; each returns its result and evidence path | | `judiverse-mail`, `agent-sms`, voice route (exist) | the leads' inboxes, number and voice | **How Jude executes the plan.** - The plan becomes machine-readable as `ledger/closure/plan/tasks.json`, generated from this plan's sections. Each task carries: - the lane that owns it; - the command or tool call; - its inputs and outputs; - its completion predicate, read from `ledger/closure/closure.json`. - Barachiel loads the tasks into its queue and hands each to the application lead that owns it. - Every task runs on the racetrack: rehearse, run, evidence. - `tools/closure.py`, and nothing else, decides when a task is done. **Reasoning model: model-agnostic, relaxAI by default.** - The `model-reasoning` slot is bound to `runtime/reasoning.py`. - **Default:** relaxAI, the OpenAI-compatible API Jude is already linked to, reached through `RELAXAI_API_KEY`, `RELAXAI_BASE_URL` and `RELAXAI_MODEL`. - **Alternatives:** any other model plugs into the same slot by setting `REASONING_PROVIDER` to `openai-compatible` or `anthropic` with that provider's key. Nothing else changes. - **Model output is never evidence.** Whatever Jude builds with it is accepted only by the tests and the closure gate. **Governance.** The same rules as every other agent apply: - the contract fingerprint and the checkpoint are injected into every session; - OUTPUT, BUDGET and CONTRACT lines are required; - owner limits apply; - value is provisional until signed off; - irreversible actions need the analog switch; - every deploy has a recorded rollback. **Acceptance test: Jude bootstraps and maintains itself.** On a clean host, with only the repository and its environment, Jude must, with every step recorded in `ledger/deploy/` and `ledger/closure/`: 1. reproduce the build, matching the hashes in `ledger/release/7a2a0ea-build.json`; 2. run the closure report; 3. deploy and pass the health check on all three domains; 4. redeploy a newer commit; 5. roll it back. Until that test passes, the closure report lists "Jude self-operation" as open. ##### Execution: 15 minutes for code and deploy, 60 for everything 1. **Implementation (0–12 min):** one workflow, six lanes in parallel (A to F above). Each lane owns its files exclusively, so there are no merge conflicts, and each declares `OUTPUT`/`BUDGET`. 2. **Integration (12–15 min):** one `tools/make.py` run, `tools/gate.py` (22 gates) and `tests/test_truth.py`. Then commit, push, and hand Codex `deploy/deploy.sh <commit>`: the URL is live. 3. **Evidence (15–45 min, in parallel):** - browser video of all 5,823 human journeys, **under 15 minutes, on whatever is fastest:** - Codex first measures the free cores and memory on the seven Civo VMs. - If they can run at least 96 headless Chromium workers in total, the shards go there. - Otherwise the shards go to Brev instances, which need only Node 22, Playwright 1.63 and Chromium 1243 from `TOOLCHAIN.json`. - Either way, the target is 5,823 journeys × about 8 s ÷ 128 workers ≈ 6 minutes, and the videos go to the existing Civo object store with digests; - agent journeys; - the 26 Figma calls; - `tools/reproduce.sh` on a second VM. 4. **Closure (45–60 min):** `tools/closure.py` on the deployed commit. You review the videos before any regulator sees anything. **What cannot close inside the hour, by design:** external receivers whose credentials Codex hasn't bound yet. The closure report lists them as the plug set; agents fill them with `plugs_drop` and `plugs_plug` with no code change. Office for Students registration and the signed LSBU agreements are external processes, not build steps. ##### Compute: what holds the time budget, and its cost **The bottleneck found in execution.** This container has 4 CPUs and 15 GB of memory. Its workflow runner allows only 2 agents at once, so the six lanes run in three rounds instead of one. **Lanes: fan out to parallel sessions at no extra cost.** - `mcp__claude-code-remote__create_session` starts each remaining lane in its own cloud container. Each lane gets its own CPUs, so all six run at once. - Each session gets the lane brief, the branch and its file ownership. - Each session commits only its own files to the branch. - The integrator (this session) runs the single gate and closure pass afterwards. **Browser video: capacity needed.** - One headless Chromium recording video needs about 1 vCPU and 1 GB of memory. - 5,823 journeys at about 8 s each need about 128 workers to finish in 6 minutes, and at least 52 workers to finish within 15 minutes. **Option 1, Codex on the existing Civo VMs, $0 extra.** Codex needs to: 1. Report each VM's vCPU and memory. Seven 16-vCPU VMs give 112 workers, about 7 minutes. Seven 4-vCPU VMs give 28 workers, about 28 minutes, which is too slow on its own. 2. On every shard VM, install the toolchain (Node 22, Playwright 1.63, Chromium) and clone the commit. 3. Run `python3 -m runtime.web` locally on each VM, so no shard depends on another machine. 4. Issue one test person credential. 5. Run `node tools/journeys.mjs --shard i/n`. 6. Upload the videos to the existing object store using its credentials. **Option 2, Brev, only if the VMs are short of cores.** The estimate below comes from typical on-demand cloud CPU pricing, not a Brev quote; check the actual rate on Brev's pricing page before buying. - **Machines:** 4 CPU-only instances of about 32 vCPU and 64 GB. No GPU is needed for browsers. - **Rate:** about $1 to $2 per instance-hour. - **Time:** about 30 minutes, including setup. - **Total:** about $2 to $4. A $10 to $20 credit covers it, including retries. **Recommendation.** Use Codex's VMs if they have at least 52 free vCPUs in total. Otherwise top up with one or two Brev instances. Either way, run the lanes as parallel sessions now. ##### Execution status (live) **Done and pushed:** - `c72a358`: the three owner decisions (contract, domains, TRL 9 / CRL 9). - `5aad90b`: this plan, the lane briefs (`new-estate/ledger/closure/plan/LANES.md`) and the lane A and B designs are in the repository, so any session can execute a lane without this conversation. - `5aad90b` also contains lane B's finished part: the old-estate reader handles every format, and 11,944 of 11,944 documents are now read, with 0 unreadable. **The two-agent cap stalled the in-container workflow, and plan mode stopped it part-way.** Nothing partial was committed. Lanes are now fanned out as separate cloud sessions, each with its own CPUs, all pushing to this branch: | Lane | Status | |---|---| | A, runtime truth | Next: its own session, design in `lane-A-design.md` | | B, closure gate (rest) | Next: its own session, design in `lane-B-design.md` | | C, harness | **Running** as session `session_01T3JC6DYafQ2Mypp4LwfisY` | | D, runner and reproducibility | Next: its own session | | E, web app, video, agent journeys | Next: its own session | | F, views, leads, voice, drafts, deploy | Next: its own session | **On approval:** 1. Start A, B, D, E and F as five more sessions in one step, with the same prompt shape as lane C. 2. Each session commits only its own files and rebases on push. 3. This session integrates: one `tools/make.py` run, the gate, the truth tests, then the closure report. 4. Hand Codex `deploy/deploy.sh <commit>` for the three domains. ##### Verification (end to end) - `python3 tools/gate.py` → 22 gates PASS; `python3 -m unittest tests/test_truth.py tests/test_gates.py` → OK. - Truth checks: - `python3 -c "from runtime.engine import Engine; ..."` on an operation with no receiver → `UNKNOWN RECEIVER_REQUIRED`, `verified_outcome false`. - An estate-record create → OK with a receipt that rereads to the same digest. - `python3 tools/make.py` twice → the second run skips every step; `bash tools/reproduce.sh <commit>` → `ledger/reproduce.json` with 0 differences. - `python3 -m runtime.web --port 8080` and `node tools/journeys.mjs --shard 1/1 --limit 20` → 20 videos, `journeys-1.jsonl` all PASS, 0 dead controls. - `python3 tools/closure.py` → R, C and `R − C` by kind; the only remainder is the named plug set. - Live on all three domains: - `for d in judiverse.tech judeai.tech x-x.tech; do curl -s https://$d/checkpoint | sha256sum; done` → three identical digests; - `curl -sI https://www.judeai.tech/` → 301 to the apex; - each page's canonical link points to `judiverse.tech`; - `/s/<code>` → 302 to the capability. ### ledger/closure/plan/LANES.md (legacy: research evidence, not a rule) `ledger/closure/plan/LANES.md` · sha256 `364639d0b981baab5b2e81950e795fe8a1e62a95ce1d6ded012835d0e65a287f` #### Closure build: lane briefs The approved plan is `PLAN.md` in this folder. Each lane below is self-contained. Paths assume the repository is checked out at `/home/user/governed-intelligence-judeai-edgeid/new-estate` (the integrator); in another checkout, read them relative to `new-estate/`. ##### Shared rules for every lane ``` OUTPUT: <output> BUDGET: 600000 CONTRACT: d1233bfd3bbc288bae594e21c6744e3501fe50f37dbd7a57f0a5068b768634bc You are lane <lane> of the approved plan at new-estate/ledger/closure/plan/PLAN.md (read it first, all of it). Work in /home/user/governed-intelligence-judeai-edgeid/new-estate (the new estate). Standard library only in Python; Node only for Playwright (require from /opt/node-tools/node_modules/playwright, chromium executablePath /opt/pw-browsers/chromium). Do NOT edit any file another lane owns (listed below). Do NOT edit contract/, tok/kernel.json, tok/KERNEL.md, tools/agent_gate.py, .claude/settings.json, ledger/waste/launches.jsonl. Do NOT commit or push; the integrator does. Do NOT run tools/run_all.sh or tools/make.py over the whole estate (it takes minutes and other lanes are editing); test on small samples. Never claim something passes without running it. Never print or write secrets. FILE OWNERSHIP: Lane A owns runtime/receivers.py, runtime/adapters.py, runtime/engine.py, runtime/guards.py, runtime/server.py, runtime/session.py, runtime/hub_mcp.py, runtime/racetrack.py, tools/receivers.py, tools/exercise.py, tests/test_truth.py, tests/replay_binance.py, tests/analog_switch.py, runtime/INTERFACE.md. Lane B owns tools/obligations.py, tools/closure.py, tools/dispositions.py, tools/scope.py, tools/gaps.py, tools/freeze.py, tools/oldestate.py, tools/sweep.py. Lane C owns tools/harness.py, tools/checkpoint.py, tools/install_gate.sh, tools/waste.py, CHECKPOINT.md (generated). Lane D owns tools/make.py, tools/run_all.sh, tools/reproduce.sh, tools/manifest.py, TOOLCHAIN.json, tools/headlines.py, tools/fetch_context.py. Lane E owns runtime/web.py, runtime/web_assets.py, tools/journeys_plan.py, tools/journeys.mjs, tools/agent_journeys.py, tools/urls.py, docs/DIRECTORY.md. Lane F owns tools/views.py, runtime/leads.py, spec/leads.json, runtime/voice.py, tools/drafts.py, drafts/, deploy/, tools/plugs.py, README.md, AGENTS.md, spec/SPECIFICATION.md. FIXED INTERFACES (implement yours exactly; rely on others' exactly): - runtime/session.py (A): issue(actor:dict)->str token (HMAC-SHA256 over JSON, key from env ESTATE_SESSION_KEY else .cache/session.key created 0600); read(token)->actor dict or None; actor = {id, kind:'person'|'agent', role, tenant, mandate?}. sign_in(person_id, secret)->actor or None using runtime/credentials.verify (people get credentials via credentials.issue(person_id, owner=person_id, tools=['*'])); tenant defaults to the actor id's organisation part before ':' or the id. - runtime/server.py (A): handle_api(method, path, headers:dict, body:bytes)->(status_code:int, content_type:str, body:bytes). Actor from 'Authorization: Bearer <session token>' via session.read, or 'Authorization: Agent <agent_id>:<secret>' via credentials.verify; never from X-Actor headers. Keeps a module-level Engine on db .cache/runtime.sqlite in live mode. - runtime/engine.py (A): Engine(spec_path, db_path, mode='live'|'simulation'); call(...) unchanged signature; every result dict has keys status, code, outputs, evidence_ref, steps, receiver ('estate-record'|'external:<slot>'|'simulation'|'unbound'), verified (bool). Engine.evidence() yields every evidence record. - runtime/hub_mcp.py (A): Hub().handle(msg:dict)->dict|None (JSON-RPC), usable over HTTP by web.py at POST /mcp; tools add estate_checkpoint (returns CHECKPOINT.md text) and racetrack_cancel(mark). - runtime/voice.py (F): handle_inbound(payload:dict, actor:dict)->dict; web.py (E) routes POST /v1/voice/inbound to it. - ledger/evidence/operations.jsonl (A via tools/exercise.py): one row per operation {unit, operation, status, code, receiver, verified, evidence_ref}. - ledger/evidence/journeys-<shard>.jsonl (E): {journey_id, unit, journey, status:'PASS'|'FAIL', steps, dead_controls, video, video_sha256, bytes, why}. ledger/evidence/agent-journeys.jsonl (E): {journey_id, unit, journey, status, results, evidence}. - ledger/reproduce.json (D): {commit, same:int, different:[paths], unexplained:int}. - ledger/closure/closure.json (B): {pass:bool, R:int, C:int, missing_by_kind:{kind:int}, plug_set:[slot], checks:{name:{value,ok}}}; CHECKPOINT.md (C) reads it if present. Return the JSON schema only. ``` ##### Lane A OUTPUT: new-estate/runtime/receivers.py LANE A: plan §1 RUNTIME TRUTH. Implement exactly §1 items 1-8 plus runtime/session.py and tools/exercise.py. Details: tools/receivers.py classifies each operation of spec/units/*/*.json into spec/receivers.json {"<unit>/<op>": "estate-record:<class>:<entity>" | "external:<slot>"} using tools/behaviour.behaviour(label) class and regexes for outside systems (payment scheme|faster payments|bacs|card network|bank|nhs|gp connect|spine|registry|register (with|at)|hmrc|companies house|fca|email|sms|text message|call|voice|transcri|model|predict|detect|scan|ocr|third.party|external) mapping to slots in spec/plugins.json where names match, else external:<slug>. runtime/receivers.py EstateStore over sqlite (tables objects, events) as in the plan, receipt {receiver, tenant, entity, id, version, digest} with read-after-write. adapters.resolve(engine, unit, operation, op) -> (receiver_label, fn); delete default; unbound/unplugged external -> engine returns UNKNOWN code RECEIVER_REQUIRED, nothing changed. In engine: verified rule from the plan; idempotency scoped (tenant, actor id, unit, op), IDEMPOTENCY_CONFLICT on mismatch, HELD stored before return; mode default 'live' (Engine(mode=...)); evidence includes receiver and verified. guards.py changes per §1.5 (keep families; documented-precondition checks named inputs present via tools/behaviour.for_feature fields when determinable, else requires at least one non-empty input; agents pass role guards only if credentials.current(mandate owner) holds an allowed role in its record field 'role' or the mandate lists the role — document the rule; identity requires actor['verified'] True set by session/credential verification, engine call from tests may pass verified=True). server.py: handle_api plus the existing http.server main using it; HTTP codes HELD 202, UNKNOWN 202, IDEMPOTENCY_CONFLICT 409. hub_mcp.py: idempotency_key and credential args; estate_checkpoint; racetrack_cancel; racetrack persistence .cache/racetrack.sqlite states QUEUED/RUNNING/CANCELLED/UNKNOWN/DONE. tools/exercise.py: runs every operation once (use multiprocessing over units) as a verified person with the operation's role and the tenant 'exercise', synthetic valid inputs from data declarations (copy tools/bench.py VALUES), live mode, writes ledger/evidence/operations.jsonl and prints counts by receiver and status. tests/test_truth.py: every case in §1.8 as unittest, fast (under 30 s), using small units; re-point tests/replay_binance.py and tests/analog_switch.py by registering an explicit SIMULATION receiver for the operations they use and asserting the provisional/analog logic (their JSON outputs must keep their current keys). Verify: python3 -m unittest tests.test_truth -v; python3 tests/replay_binance.py; python3 tests/analog_switch.py; python3 tools/receivers.py (report counts); python3 tools/exercise.py on a 30-unit sample via an env var EXERCISE_LIMIT=30. ##### Lane B OUTPUT: new-estate/tools/closure.py LANE B: plan §2 CLOSURE GATE. Implement tools/obligations.py (R as ledger/closure/obligations.jsonl with stable ids: op:<unit>/<op>, screen:<unit>/<sid>, control:<unit>/<sid>/<cid>, journey:<unit>/<name>, request:<n>, finding:<F-id> (from spec/carry.json and research/AGENT-RESEARCH.md placed ids), decision:<id>, intention:<id>, oldfinding:<O-id>), tools/dispositions.py (old-estate findings from ledger/old-estate/coverage.jsonl.gz: CARRIED keeps its URL; OLD_RULE and NOT_CARRIED get disposition RESEARCH_ONLY with the reason 'old estate is research only (CLAUDE.md); the frozen scope does not include it'; writes ledger/closure/dispositions.jsonl.gz), extend tools/oldestate.py so the 23 unreadable documents are read with the right parser (pdf via pypdf if importable, docx/pptx/xlsx via zipfile XML text, else record UNREADABLE with format and reason) — run it (it takes ~30 s), and tools/closure.py exactly per §2.2 reading evidence files named in the fixed interfaces (missing evidence file means that kind is unmet, never assumed); checks include unbound_required_operations (from ledger/evidence/operations.jsonl receiver=='unbound'), unreadable_sources, findings_without_disposition, unadmitted_scope_changes (run tools/freeze.py check; make freeze.py check exit 1 when unadmitted > 0), reproducibility_differences (ledger/reproduce.json), fabricated_evidence_rejected (run python3 -m unittest tests.test_truth and require OK; if the file is absent the check fails), views_drift (python3 tools/views.py --check exit code; absent means fail). External receivers without credentials go in plug_set (from spec/receivers.json external slots not bound in spec/plugins.lock.json). Write ledger/closure/CLOSURE.md and closure.json; exit 0 only when pass. Edit tools/scope.py and tools/gaps.py so DONE/closed come only from closure.json (file existence becomes 'coverage'); relabel sweep parity grades 'wording coverage' in tools/sweep.py output text only. Verify by running python3 tools/obligations.py and python3 tools/closure.py (it should FAIL now, honestly, and print R, C and missing by kind) and python3 tools/gaps.py. ##### Lane C OUTPUT: new-estate/tools/harness.py LANE C: plan §3 HARNESS. tools/checkpoint.py writes CHECKPOINT.md (repo new-estate root) from contract/estate-contract.json intentions (id + first sentence), tok/objects/decision/*.json labels, ledger/closure/closure.json if present (R, C, missing by kind, plug set size), the rules (no claim without evidence; counts only from generators; old estate research only; one batched run per change set via foundation/scope-change-intake and spec/changes/; agents declare OUTPUT, BUDGET, CONTRACT), and the commands (gate, closure, make, plugs, hub tools). Keep it under 250 lines. tools/harness.py implements hook commands reading the hook JSON from stdin: session-start (print JSON {hookSpecificOutput:{hookEventName:'SessionStart', additionalContext: CHECKPOINT.md text}}), prompt (UserPromptSubmit additionalContext: fingerprint + R−C summary line), agent (PreToolUse Agent|Task|Workflow: require CONTRACT: <current fingerprint> line in prompt or script text in addition to agent_gate's checks; exit 2 with reason otherwise), claims (PreToolUse Edit|Write: refuse writing phrases like 'DONE', '100%', 'fully verified', 'TRL 9', 'CRL 9', 'production ready' into a file unless the file path is under ledger/, generated/, docs/ generated outputs, or the tool's own generators — define the allowlist precisely; Markdown in research/ and README are not allowed to claim), guard (PreToolUse Edit|Write|Bash: protect /etc/claude-code/managed-settings.json, tools/install_gate.sh, tools/harness.py, tools/waste.py, tools/checkpoint.py from edits, same style as agent_gate.guard), stop (Stop: run tools/waste.py on payload transcript_path and append the classification; never block indefinitely — block only once with a reason if the session's declared OUTPUT files from ledger/waste/launches.jsonl are missing and stop_hook_active is false). Extend tools/waste.py: classify every failed tool call (is_error) as mine / environment (text contains "doesn't want to proceed" or "stopped while running" or "refused" by the tool layer) / owner (an explicit owner refusal in the user's own message) and write the counts into ledger/WASTE.md with a 'by cause' table; accept a single transcript path. Update tools/install_gate.sh to install all hooks (the existing three plus the new harness ones) into managed settings with allowManagedHooksOnly, idempotently; do NOT run it with sudo against /etc here unless you are root and the existing managed-settings.json exists — if so, run it and show the resulting hook list. Verify each harness subcommand with piped sample JSON (show exit codes and outputs) and run python3 tools/checkpoint.py. ##### Lane D OUTPUT: new-estate/tools/make.py LANE D: plan §4 ONE RUN, CONTENT-ADDRESSED. Read tools/run_all.sh and reproduce its exact step order as a step table in tools/make.py: each step {name, cmd, inputs (globs), outputs (globs), after (names), kind: build|measure|test}; cache key sha256(cmd + sorted input file digests + digests of 'after' steps' outputs) stored in .cache/make/<name>.json; skip when unchanged and outputs exist; run independent ready steps in parallel (ThreadPoolExecutor, workers = os.cpu_count()); fail fast on a nonzero exit with the step's tail of output; at the end write build/MANIFEST.sha256 over deterministic outputs (exclude kind measure, exclude ledger/benchmarks/results.jsonl and README timings, exclude .cache) and ledger/make/last-run.json (per step: ran|skipped, seconds). Add the new steps from the plan at the right positions: tools/receivers.py after expand, tools/exercise.py after build, tools/obligations.py, tools/dispositions.py, tools/closure.py and tools/checkpoint.py at the end, tools/views.py before docs (call each only if the file exists, so lanes landing later are picked up). Replace run_all.sh's body with 'exec python3 tools/make.py "$@"' keeping its header comment. tools/headlines.py and tools/fetch_context.py: no network unless REFRESH=1; otherwise read the committed snapshot (research/market/headlines-2026-10.jsonl) and the cache dir from env PAGE_CACHE (default .cache/pages); missing cache entries are recorded, never fetched. TOOLCHAIN.json: python, node, playwright, chromium versions measured on this machine, plus corpus digest (sha256 of /home/user/corpus/catalogue.jsonl) and page cache digest (sha256 over sorted file names+sizes). tools/manifest.py computes/compares manifests. tools/reproduce.sh <commit>: clean clone (git clone --no-local of the local repo at that commit into a temp dir), link CORPUS_ROOT and PAGE_CACHE, run python3 tools/make.py there, compare manifests with tools/manifest.py, write ledger/reproduce.json in the ORIGINAL repo per the fixed interface. Verify make.py with --only <step> on 2 cheap steps (e.g. urls, catalogue) twice showing the second run skips; bash -n on shell files; do not run the full chain. ##### Lane E OUTPUT: new-estate/runtime/web.py LANE E: plan §0 directory and §5 WEB, VIDEO, AGENT JOURNEYS. runtime/web.py: standard-library http.server (ThreadingHTTPServer), python3 -m runtime.web --host --port; serves the §0 directory exactly: / (five applications from contract applications), /<area> (capabilities from spec/urls.json), /<area>/<cap>/<screen> (read spec/units/<app>/<cap>.json on demand, cache recent units), /s/<code> 302, /help/<area>/<cap>/<screen> (help_markdown rendered as simple HTML), /checkpoint (CHECKPOINT.md as text/markdown), /v1/* delegated to runtime.server.handle_api, POST /v1/voice/inbound to runtime.voice.handle_inbound if importable, POST /mcp to runtime.hub_mcp.Hub().handle, /signin (form: person id + secret → runtime.session.sign_in → HttpOnly SameSite=Lax cookie), /signout. Every HTML page: <link rel=canonical href=https://judiverse.tech/<same path>>, lang=en-GB, skip link, Frost theme CSS inline from runtime/web_assets.py (tokens on :root, light and dark), breadcrumbs (copy.breadcrumbs), progress (copy.progress, copy.next_step), fields with labels and validation, controls as buttons inside a form POSTing to <screen-url>/do/<control-id> which calls runtime.server's engine with the session actor and redirects to the screen named by the action's on_success/on_failure (journey:end → the capability's done page or /<area>), states via ?state=<name> using the screen's states copy, help control linking to /help/…; components rendered by type from design/components.json names (Input.*, Select, Checkbox, Radio.Group, Button.*, Alert, Receipt, Table, Timeline, Summary.Review, Rehearsal.Result, Evidence.Panel, Help.Control, Nav.*, Steps.Progress, Page.Header; anything else as a labelled region). www. host → 301 to apex. tools/urls.py: add the directory rows (help, checkpoint, v1, mcp) to docs/DIRECTORY.md generation (write docs/DIRECTORY.md listing the domain rules and the area index; do not change existing path generation). tools/journeys_plan.py → ledger/evidence/plan.json: for every HUMAN journey in spec/units (steps like 'screen:x' / 'operation:y' or screens lists), the URL sequence and for each screen the control id whose action leads to the next screen, plus synthetic input values per field type. tools/journeys.mjs: args --shard i/n --limit N --base http://127.0.0.1:8080 --video-dir DIR --person ID --secret S; signs in once per context, records video per journey (recordVideo dir), asserts per screen that every visible control button has a form action resolving to a known control (dead control = none), clicks the planned control, asserts the next URL; writes ledger/evidence/journeys-<i>.jsonl per the fixed interface with sha256 of the video file. tools/agent_journeys.py: runs AGENT journeys through runtime.hub_mcp.Hub racetrack_run with a test agent credential issued via runtime/credentials.issue, writes ledger/evidence/agent-journeys.jsonl; supports --limit. Verify: start the server on a free port in the background, curl /, /finance, one screen, /s/<code>, /help/…, /checkpoint (may 404 until lane C lands — report), www redirect with a Host header; run journeys.mjs --limit 5 with a test person credential you issue, show the jsonl rows and that videos exist; run agent_journeys.py --limit 5. Lane A is changing the engine concurrently: if its new interfaces are not there yet, use what exists and report. ##### Lane F OUTPUT: new-estate/tools/views.py LANE F: plan §6. tools/views.py: regenerate the application sections of README.md, AGENTS.md and spec/SPECIFICATION.md from contract/estate-contract.json between markers <!-- views:applications:start --> / <!-- views:applications:end --> (insert markers replacing the current application wording: README.md:57 and AGENTS.md:57 'four applications' rows, spec/SPECIFICATION.md lines ~46, 62, 68 and the application descriptions), wording 'four applications (EDGE Finance, EDGE Clinic, The Edge Consortium, Public EDGE) plus Judiverse, which comes with all four'; --check exits 1 on drift. runtime/leads.py + spec/leads.json: Barachiel (coordinator) and one lead per application (finance, clinic, edge-consortium, public-edge, judiverse) with role, area, mailbox address on judiverse.tech, sms slot, mandate tools (hub tools + its area server tools), and a persistent queue in .cache/leads.sqlite with handover records; 'python3 -m runtime.leads issue' issues credentials via runtime/credentials.issue (secrets stay in .cache/agents); 'python3 -m runtime.leads run --once' polls inbound sources (mail inbound slot, SMS rows in .cache/sms.sqlite inbound table if present, voice queue) and turns each message into runtime.racetrack Racetrack.run with the lead's actor, recording outcome and evidence; unplugged inbound sources are reported, not faked. runtime/voice.py handle_inbound(payload, actor): transcript from payload['transcript'] or via plugins.infer('model-transcription', {...}); intention → racetrack run (rehearse-first) → reply text; speech via plugins.infer('model-speech'); unplugged → returns {status:'PLUG_REQUIRED', slot:...}; supports payload['cancel']=mark → racetrack_cancel. tools/plugs.py: add model slots 'model-speech' and service slot 'voice-telephony' and 'mail-inbound' to the generated registry (keep all existing behaviour). tools/drafts.py → drafts/INVESTOR-PACK.md, drafts/FCA-INNOVATION-PATHWAYS.md, drafts/OFS-REGISTRATION.md: every file headed DRAFT, sections 'Facts (with ledger source)' and 'Forecasts and intentions (not facts)', every figure read from ledgers/spec files at generation time with its path; no claim of readiness. deploy/: x-x-web.service (python3 -m runtime.web --host 127.0.0.1 --port 8080, EnvironmentFile=/etc/x-x/web.env), x-x-leads.service, Caddyfile (judiverse.tech, judeai.tech, x-x.tech with www redirects, reverse_proxy 127.0.0.1:8080), web.env.example (names only, no values), deploy.sh <commit> (fetch/checkout commit, python3 tools/make.py, systemctl restart, health check /, /checkpoint and one /s/<code> on all three domains and compare body sha256; exit nonzero on mismatch), README.md in deploy/ with the exact commands for Codex. Verify: python3 tools/views.py then --check (0), python3 tools/drafts.py, python3 -m runtime.leads issue && python3 -m runtime.leads run --once, python3 -c 'from runtime.voice import handle_inbound; print(handle_inbound({"transcript":"send 10 pounds to a friend"}, {"id":"lead:barachiel","kind":"agent","role":"coordinator","verified":True,"mandate":{}}))', bash -n deploy/deploy.sh, python3 tools/plugs.py (counts include the new slots). ### research/AGENT-RESEARCH.md (legacy: research evidence, not a rule) `research/AGENT-RESEARCH.md` · sha256 `723389b5a6f22b3dc428abb167925da4ca8a5cfd7e8ef76df76e8aa177e9feb7` #### Everything the estate knows, in one place Generated by `tools/compendium.py` from the ontology and every research source. Do not edit by hand: change a source and run it again. This is research, never authority: the contract (`contract/estate-contract.json`) and the owner's recorded decisions are the authority, and they come first below. **How to read it (people and agents).** 1. Start with the gate. Every source is counted, and nothing is dropped. 2. Sections 1 to 3 are the authority: the contract's intentions, then the owner's decisions and requests. 3. Section 4 is the specification, and section 5 is the proof. 4. Section 6 holds every research finding, each with an ID and the capability URL that serves it. 5. Section 7 reconciles the machine registers with their ledgers. 6. Section 8 accounts for every old-estate document. 7. Section 9 lists everything not yet met. Search by ID (`F-…`), by URL (`/finance/…`) or by intention (`glass-box`). ##### The gate | Measure | Met | Of | Share | |---|---|---|---| | Intentions served | 27 | 27 | 100.0% | | Owner decisions recorded | 21 | 21 | 100.0% | | Owner requests done (rest need the owner) | 53 | 58 | 91.4% | | Research findings placed on a capability | 2626 | 13987 | 18.8% | | Old-estate inventory lines (accounted in section 8) | 1157 | 1157 | 100.0% | | Research files read | 47 | 47 | 100.0% | ##### 1. Intentions (the contract) | ID | Intention | Capabilities serving it | Status | |---|---|---|---| | `source-of-truth` | The estate is the living source of truth for all human domain expertise: regulator requirements, live facts and laws, and everything that can be researched, refreshed continuously. | 18 | SERVED | | `truth-separated-from-agreement` | Every statement is classed as observation, agreement or claim; for every rule, law and fact the estate shows how it came to be, who agrees, who disagrees and on what evidence. | 113 | SERVED | | `glass-box` | A glass box, not a black box: light, not darkness; possibility, not limitation; to be seen, not to control. Authorised people and regulators can inspect knowledge, reasoning, authority, execution, verification, recovery and efficiency. | 131 | SERVED | | `collaboration-and-shared-data` | Collaboration, community, cooperation and teamwork are invited. Data is shared for good under lawful authorised access, not hoarded, so that organisations create new capabilities, industries, jobs and solutions. | 128 | SERVED | | `solve-the-crises` | Capabilities directed at the energy crisis; cyber attack made unprofitable; people given the capability to generate wealth, meaning the power to be; more and better housing; raw materials used better; zero waste; everything monitored and tracked. | 40 | SERVED | | `agents-native` | Agents are native to the estate as much as people: always on, inside the estate so data never leaves it, each with its own portal and racetrack so agents cannot crash or collide. The estate is built for agents, which makes it better and safer for people. | 139 | SERVED | | `clear-screens-deep-help` | Screens show the essential result first; depth and context live behind a help control as markdown, served to people and agents on demand, so neither people nor agent context windows are overwhelmed. | 93 | SERVED | | `exhaustive-deterministic-scope` | Everything is defined and scoped before it is built: every screen, feature, workflow, capability, control, intention, outcome, endpoint and every whole-of-life, customer, agent, regulator and engineer journey. Nothing missing, nothing broken. | 97 | SERVED | | `match-and-exceed` | Every competitor capability is matched, exceeded, improved and made easier and faster; benchmarks, baselines, competitors, features, behaviours, workflows and controls are known from research. | 50 | SERVED | | `compress-time` | Slow, clunky processes and industries are compressed in time. Human intention is converted into computational instructions; labs automate prompt engineering, creation and solutions; engineering work that took a day takes minutes. | 1580 | SERVED | | `agents-build-agents` | Always-on agents build further agents that are advanced, fine-tunable, transferable and reproducible. | 17 | SERVED | | `always-on-monitoring` | Everything is reproducible, retraceable, auditable and monitored always on; this is how sovereignty and control of every organisation's data are maintained. | 68 | SERVED | | `lean-repository` | Clean source code; one instruction markdown; the repository contains exactly what it needs; every file is named and placed by TOK. | 48 | SERVED | | `compute-light-runtime` | AI is the factory, not the product: neural models are used at build time for speed; what runs is deterministic, compute-light, more useful and more accurate, and reduces the need for AI over time. | 61 | SERVED | | `only-true-things` | The runtime says only what a verified procedure produces inside its certified scope, with its error bound attached, using error-rate control and Bayesian reasoning; otherwise it says it does not know. | 74 | SERVED | | `same-rules-everywhere` | One method applies to every domain: identity, relations, processes, logic, probabilities and patterns. Every domain is added by extension and the same rules hold. | 40 | SERVED | | `better-routes` | Where a licence or partner is not needed, a better route is engineered: consented data, simulators, open sources, own networks. Where the licence is the network, the legal permission or the clinical safety, the licence or partner is obtained as well. | 514 | SERVED | | `sight` | Working backwards from preventing avoidable illness and blindness: the time from first meaningful signal to a verified beneficial outcome is shortened, starting at the front door before the GP, with own capture technology and consented patient data. | 87 | SERVED | | `approved-design` | Interfaces use approved UX Pilot, Figma and Untitled UI structure and the estate's original logos, used as-is and never recreated: accessible, readable, complete navigation, real functionality. | 52 | SERVED | | `every-intention-represented` | Every intention of the governing organisation and of every onboarded organisation is represented, including capabilities implied and never spelled out, each placed in an application or foundation. | 22 | SERVED | | `verified-outcomes-fast` | The governing objective is the shortest independently justified time from legitimate intention to verified beneficial outcome, at the greatest feasible speed and autonomy, with no reduction in scope or acceptance. | 57 | SERVED | | `education-is-the-source` | Education is where all capabilities come from; research and development is worth investing in; universities are made prestigious again by being the home from which public, defence, government, procurement, logistics and cyber capabilities grow. | 60 | SERVED | | `agents-as-real-customers` | Agents are real users and real customers in every application: lenders, borrowers, buyers, suppliers, patients' representatives, applicants and operators. They act as real customers on their racetracks so they learn what works and what does not, under their own identity and authority. | 1591 | SERVED | | `confidence-reporting` | Confidence reporting sets the standard beyond credit scoring: a confidence score that is more informationally complete than a credit score, showing not only whether obligations were met but how a person or organisation acted when things went wrong (whether they did all they could, how they communicated, what caused the situation), so a single turn of events does not define them. It strengthens KYC, is explainable and contestable, and is proven by research and measured business and economic outcomes, including wider access to resources. | 189 | SERVED | | `ai-assurance-standard` | The estate sets the global standard in autonomous agent assurance, as the first fully agent-native bank: agents as customers and staff with their own identity, certificates, mandates and term, with no gaps in assurance. | 134 | SERVED | | `solution-banks` | From think tanks to solution banks: agents generate solutions, and solution banks empower and fund people to enact them in the real world, because having the answer is not enough. Better lending structures and better ways to move money through the economy create jobs and access to resources. | 59 | SERVED | | `physical-presence` | EDGE is a real bank and clinic in the physical world, not only online: EDGE points and devices that come to the person, verifying identity by proximity (scanning or tapping an identity document or device), and edge hardware and wearables used for good, including early detection in the clinic. | 84 | SERVED | ###### Founding principles and principles - el_roi - let_there_be_light - what_god_cannot_do_does_not_exist - truth_and_agreement - `one-source` (GATE): TOK is the only source. Everything is an object in it; every document, view, index, diagram, count and test is generated from it; a hand edit of generated output is a failure. Check: G01 schema, G03 closure, G05 generation, G16 live counts - `truth-agreement-claim` (GATE): Every statement is exactly one of observation, agreement or claim, carrying that class's required fields; none is rendered as another. Check: G04 - `honest-status` (GATE): Every realisation carries a truth status set from evidence, never above it. An attempt is not a completion; a rehearsal is not an outcome; an intention is not an observation. Nobody fabricates evidence, authority, benchmarks, permissions or readiness; a falsehood is rejected and never rewarded. Check: G04 status rules; breaches are incidents - `live-numbers` (GATE): Every number is produced from the objects at the moment it is stated, with its method and completeness. No count is written by hand. Check: G16 - `everything-identified` (GATE): Everything that exists has a permanent identity, a kind, its relations and a working monitor; nothing is undefined, dead, unaccounted for or without an identity. Check: G02, G03 - `complete-contract-first` (GATE): Every capability, journey, screen, control, operation, endpoint and agent resolves to a complete executable contract before it is built: intention, requirement, steps, executor, authority, design, acceptance, evidence, resources, recovery and answerer. A missing part is an OPEN obligation with an owner, counted and visible. Check: G03 required relations, G10, G11, G13 - `research-defines-scope` (GATE): Scope comes from research: stated intentions, current primary sources, regulations, standards and every top competitor's screens, flows, capabilities and controls. Every competitor is a floor, matched then exceeded. Check: G15; parity measured and published - `agents-native` (GATE): Every operation has a human path and an agent path to the same outcome with equal capability, under the agent's own EDGE ID, with the same authority, preview, audit and undo. Every agent has a portal, a mailbox, a racetrack and a purpose-limited authority from a legitimate issuer; consequential actions are rehearsed with the other agents involved before they commit; people and agents share each workflow with stated hand-off, approval, pause, take-over and undo, conflicts resolved by lease and fence. Check: G12; agent fields in G01 - `accountable-authority` (GATE): Every effectful actor has an authenticated identity, an accountable owner and a current purpose-limited permission from the competent issuer. Every effectful object names the human or organisational role that answers if it goes wrong; an agent is never the answerer of last resort. Check: G13 - `evidence-with-the-effect` (GATE): Every action binds its inputs, prior state, release, contract, predicate and receiving effect to durable evidence; the receiving system's own readback establishes the outcome. Rendering or request acceptance alone establishes nothing. Check: evidence fields in G01; G09 - `verification-by-reproduction` (GATE): Correctness comes from reproducing the exact contract against the actual evidence with the declared checker. VERIFIED records whether the reproduction was by the same lineage or an independent one; where an applicable authority requires independence, nothing less is accepted for that claim. Check: G04 VERIFIED rules - `preventive-enforcement` (CONDUCT): Unauthorised, unsafe or conflicting effects are refused at the boundary before execution; emerging risks are predicted and corrected before they become incidents; every discovered defect becomes a correction to the mechanism and a permanent regression test. Zero silent failures, zero fabricated assurance, zero unauthorised actions, zero avoidable harm. Check: incident objects carry cause, correction and regression test (G01); monitors are live - `authorised-recovery` (GATE): Every operation has a verified recovery or compensation; irreversible consequences are explicitly controlled; an ambiguous effect is queried, never replayed; recovery preserves the intention and the acceptance predicate. Check: operation recovers_by in G03 - `glass-box` (GATE): Knowledge, reasoning, authority, execution, verification, recovery and efficiency are inspectable by authorised people, including regulators, without the builder's help. Observation protects the dignity and privacy of the person; to be seen is not to be watched. Knowledge about a person confers no authority over that person. Check: REGULATOR journey in G11; inspection surfaces on every journey - `sovereign-data` (CONDUCT): Each organisation's data and agents stay under its own control inside the estate and never go to model providers; agents run inside the estate. Check: access_class on every object (G01); data-flow review on every endpoint - `clear-screens` (GATE): Every screen shows the essential result first and keeps depth behind a help control as markdown served to people and agents alike; every screen is accessible and completes for a person with no sight. Check: G14 - `approved-design` (GATE): Only approved UX Pilot, Figma and Untitled UI structure and the estate's original logos, used as-is; no hand-drawing and no logo recreation. Check: designed_by in G14; design-asset approval field - `outcomes-first` (GATE): Every capability serves at least one outcome: time, money, lives, planet, resources or value, saved or more. Outcomes are claimed only when verified with their measurement; resources consumed per verified outcome (tokens, GBP, kWh, gCO2e, litres, seconds) are published for every agent and capability. Check: G10; ledger measurements - `compiled-runtime` (GATE): Neural models are the factory, used at build time; what runs is deterministic and compute-light: TOK facts, compiled procedures with certified scope and error bound, answers rendered from proofs with sources. Outside a certified scope the answer is unknown or a question. The AI-free rate is published and rises. Check: G17; AI-free rate, cost, latency and reproducibility measured - `lean-estate` (GATE): One instruction markdown; every file accounted for by an object; no duplicates; raw material, datasets and large generated data in storage by digest; nothing is stored without justification. Check: G03, G08 - `always-new` (GATE): Every build starts from the current sources and the current agreement. Nothing is inherited: research is admitted as observation or claim, never as authority, and every rule, count, permission and status is produced fresh. Check: G04 authority rules; G16 - `reconciled-change` (GATE): Every change to a principle or decision names what it supersedes or narrows and retires it in the same commit; the principal's wording stays verbatim; a check no program can evaluate is never a gate. Check: G06 - `always-on-monitoring` (MEASURE): Everything is reproducible, retraceable, auditable and monitored always on; monitoring is part of assurance and never sufficient alone. Check: monitor coverage and freshness published from the objects - `latest-approval-governs` (CONDUCT): Only the most recent explicit approval of an exact subject by its competent authority is active. Check: decision objects carry decided_at; the latest governs - `act-on-intention` (CONDUCT): Standing intention is direction, not a step approval; agents act on it without questions. A question is asked only where intention and authority do not cover the exact action, or where a legal or irreversible-safety condition requires a recorded confirmation. Check: none mechanical - `one-clock` (MEASURE): Every claimed run is measured on one external monotonic clock including research, provisioning, build, verification, review, deployment, operation, admission and recording; a timeout or incomplete outcome is never a success. Check: whole elapsed time published with every claimed run - `extension-not-redesign` (GATE): Every human domain is added by extension of TOK with crosswalks to its established ontologies and standards; the kernel does not change to admit a domain. Check: G18 ###### Outcomes the estate commits to - dimensions - directions - units_per_verified_outcome - states - rule ###### Open decisions in the contract (4) - the host that runs the self-hosted CI - the storage location for research and datasets - whether screens mirror a competitor's exact layout or only its behaviour and features - which mark is EDGE Public Consortium's: the-edge-consortium.svg, public-edge.svg or cyber-governance.svg (all admitted under design/brand/) ##### 2. The owner's decisions (21) Each is an AGREEMENT recorded with `tools/amend.py`; the basis is the owner's own wording. - **EDGE Public Consortium uses the Public EDGE mark** `tok:decision:01m68c87qmz54aqnbm8pqzmgt4`: X-X.TECH decided on 31 October 2026: EDGE Public Consortium uses public-edge.svg as its mark; the-edge-consortium.svg and cyber-governance.svg are kept as alternates. - **Agents hold accounts in their own name** `tok:decision:01m696xzzpcbnsexpcdtz7ak9n`: X-X.TECH decided on 31 October 2026: agents can hold accounts in the autonomous agent estate; there is no law disallowing it. An agent's account is opened, held and closed in the agent's own name under its own identity, birth certificate and mandate, not as a sub-account of a person. As with everything in the estate, an accountable owner (a person or organisation) is named on the account and answers for it; that role is never the agent. - **TRL 9 and CRL 9 mean every intention carried out exactly** `tok:decision:01m69b9nqqnxj7bh7p0mner5rf`: X-X.TECH decided on 31 October 2026: in this estate, TRL 9 and CRL 9 mean that the estate does exactly everything the owner intended, all intentions brought into being as the raw operation of an autonomous agent estate (prompt engineering, governed intelligence, knowing your agent). Regulators give approvals; they do not define TRL 9 and CRL 9 here. - **Agent high-risk outcomes are provisional until the owner signs off in person** `tok:decision:01m69gty59gnpc17g811317htn`: X-X.TECH decided on 31 October 2026: any high-risk outcome run by an agent (large transfers and anything of value that can be stolen) completes only as provisional value, marked at both ends. It becomes final only when the human owner signs it off in person, on a registered device at a registered place (NFC tap, passkey), singly or in bulk, and the recipient agrees, within a grace window of 24 hours (at most 48). Without both, it is reversed automatically wherever the value went. A stolen key can move value; the move means nothing until the owner agrees in the real world. - **EDGE Public Consortium becomes The Edge Consortium (education and research); Public EDGE is a separate application for government** `tok:decision:01m69jxxj5cazjefrrj1fvb962`: X-X.TECH decided on 31 October 2026: rename EDGE Public Consortium to The Edge Consortium, the estate's own university and research consortium (education, research, peer review, grants, libraries, the governed intelligence degree). Public EDGE becomes its own application for government: central and local government, councils, defence, procurement, policing, borders, cyber governance, regulators and public services. Five applications: EDGE Finance, EDGE Clinic, The Edge Consortium, Public EDGE, Judiverse. - **No irreversible outcome without a real, analog, in-person human act by the owner** `tok:decision:01m69jxxqzrhz7r36ht4pvv5td`: X-X.TECH decided on 31 October 2026: an irreversible outcome (weapons and anything that can cause mass harm, and anything else that cannot be undone) can never be completed by an agent or by software alone. It needs the human owner, in person, to make a physical analog act on dedicated hardware that sits in the actuation path itself, so no software or AI signal can close it. The owner holds identical backup units of that hardware, so a broken unit never blocks the owner, and each unit proves it was the owner who ordered it and that they definitely want it to occur. Lost or broken units are revoked; agents can prepare, rehearse and request, never execute. - **Judiverse comes with every app licence, is sold standalone to consumers, and is included in Edge Consortium tuition; students do paid real work offset against tuition** `tok:decision:01m69k8c61zde0er6j94yb4ekz`: X-X.TECH decided on 31 October 2026: the estate is four applications (EDGE Finance, EDGE Clinic, The Edge Consortium, Public EDGE) plus Judiverse, where JudeAI lives with all its other applications. Any organisation that pays for any of the four gets full Judiverse access included, subject to rate limits. Consumers who are not business customers of any of the four buy Judiverse standalone, either by the capabilities they want or as a whole subscription. Anyone enrolled at The Edge Consortium gets Judiverse included in their tuition, with no extra cost for course materials. Their assignments are real-life projects, sandboxed, solving real problems; they can work freelance for partner institutions and clients, and every piece of work that produces a real solution is paid and can be offset against their tuition, never taken as free work. They can be signed on as long-term contractors for clients where that agreement is made. The Edge Consortium petitions for student finance to fund the course. The aim is to create the most employment anyone has created. - **London South Bank University is the first partner institution; SITHub teaches networking in the degree; The Edge Consortium registers with the Office for Students, shares LSBU's campus, and partners with LSBU for R&D and international technical students** `tok:decision:01m69m36psxrwhsa148z746ppv`: X-X.TECH decided on 31 October 2026: London South Bank University (LSBU) is The Edge Consortium's first partner institution for the governed intelligence degree, and one of the first universities to offer it; every university should offer it, starting with South Bank. LSBU's SITHub brings its networking capabilities as a module in the degree, so LSBU is a flagship too. The Edge Consortium registers with the Office for Students, can share the campus with LSBU, and offers itself to LSBU as a partner institution for research and development and for international technical students. - **Scope frozen on 31 October 2026; every later change enters through scope change intake** `tok:decision:01m69p22005ent1mresdjz6sh8`: X-X.TECH decided on 31 October 2026: the whole scope is known, so it is frozen. The frozen scope is every brief and unit in the new estate as built on this date, recorded by digest in spec/SCOPE-FREEZE.json. From now on a change to the frozen scope enters only through foundation/scope-change-intake as a recorded change request, and the gate reports any change without one. What remains is plugging in external models and datasets through the plug registry (spec/plugins.json), which does not change the scope. - **The GitHub branch contract (fingerprint d1233bfd) governs the build, untouched** `tok:decision:01m69thft6cmgyfgkjrt6bdrws`: X-X.TECH decided on 31 October 2026: the contract on branch claude/clever-ramanujan-9w05c1, fingerprint d1233bfd3bbc288bae594e21c6744e3501fe50f37dbd7a57f0a5068b768634bc, governs the closure build. Its bytes and the kernel stay untouched; the closure gate runs beside tools/gate.py. Other checkouts pull this contract. - **Three domains, one site: judiverse.tech canonical, judeai.tech and x-x.tech mirror it; nothing is named estate** `tok:decision:01m69thg13xhssvqbxygtr2zp8`: X-X.TECH decided on 31 October 2026: judiverse.tech is the base and canonical domain, judeai.tech serves professional services and x-x.tech is the neutral company domain. All three, and their www names, serve exactly the same site and directory, with one canonical link to judiverse.tech. Nothing is named estate; the word only describes what the whole is. - **TRL 9 and CRL 9 are reached when the closure gate passes on the deployed commit with the plug set empty** `tok:decision:01m69thg7kks8fwaj0266pf1jy`: X-X.TECH decided on 31 October 2026: TRL 9 and CRL 9, as the owner defines them, are reached when tools/closure.py passes on the deployed commit: every obligation in the frozen scope has reproduced evidence from a bound receiver, a browser video of every human journey, a record of every agent journey and a clean independent rerun with no unexplained differences, and the plug set is empty. Nothing else may claim either level. - **Jude carries out the whole plan: builds, reproduces, runs, deploys and maintains the system itself** `tok:decision:01m69xj6emef4jkw7kgv63bsdz`: X-X.TECH decided on 31 October 2026, as a hard requirement: Jude must be able to carry out the whole plan and build, reproduce, run, deploy and maintain the system without a human or an outside agent. Barachiel runs on the system's own host with the build, deploy, hub, mail, SMS and voice MCP servers, executes the plan's tasks through the racetrack, and lets the closure gate decide when each is done. The only outside act is the first start on the host; the only reserved human act is the analog switch for irreversible outcomes. Any plan step that only a human or an outside agent can perform is a defect. - **Money provenance: trust from origin and activity, face value at par, four assessments kept separate** `tok:decision:01m6a4t0fxra5avfxf0pr2p692`: The value of trust in money comes from where it was obtained and what activity supports it, never from its quantity and never by discounting its face value. The bank issues signed provenance certificates that state their verification scope and what remains unknown; provenance coverage, evidence reliability, financial-crime findings and investment risk are four separate assessments, never one score; attribution is pro rata and exact so that moving money cannot manufacture credibility; the financing graph reports independent demand share and circular funding as dependence, not wrongdoing; capital is followed to delivered work and measured resources; lending decisions carry reasons, a correction route and a named person. Decided by the Owner on 31 October 2026; admitted through scope-change intake 2026-10-31-money-provenance. - **The plug set is post-closure: closure is computed on everything the estate itself executes; external receivers bind by one-tap plug or an agent racetrack afterwards** `tok:decision:01m6a7p9c8566pe659w2hpgbjq`: Owner, 1 November 2026 00:17Z: closure should not depend on external providers. The estate builds the one-tap plug-and-play mechanism (tools/plugs.py: plugs_plug, plugs_drop) and the agent racetracks that bind an external receiver in minutes, so external slots without a credential are reported as the plug set and are not counted against closure. Closure passes when every obligation outside the plug set is evidenced on the deployed commit and every check passes. Binding each external receiver, and proving the journeys that need it, is post-closure work recorded per slot. - **Application names: EDGE Consortium and Judiverse Work Suite** `tok:decision:01m6a8mtbg94twxzcdgjp4yeac`: Owner, 1 November 2026 00:35Z: the applications are named EDGE Consortium (not The Edge Consortium) and Judiverse Work Suite (not Judiverse). The contract application records carry these names; every screen, catalogue entry, document and generator reads them from the contract. - **EDGE Finance is called EDGE** `tok:decision:01m6acbx9rv6x4yxh5gxkz1x30`: Owner, 1 November 2026 01:40Z: call EDGE Finance EDGE, so there is no choosing between EDGE ID, EDGE Bank and EDGE Finance. The contract application record carries the name; every screen, catalogue entry and document reads it from the contract. - **Application names: EDGE Public (was Public EDGE) and Judiverse Suite (was Judiverse Work Suite)** `tok:decision:01m6aeafjpfwmamsqm7apf2wfj`: Owner, 1 November 2026 02:14Z: 'I want the word work dropped from the suite, so it's just Judiverse Suite. Also, I want public edge to be edge public.' The contract's applications read EDGE, EDGE Clinic, EDGE Consortium, EDGE Public and Judiverse Suite; formal names JudeAI / EDGE Public and Judiverse Suite. Every rendered name comes from the contract; unit copy that still carries the old words is lineage, overridden at render. - **The finance application is called EDGE Bank** `tok:decision:01m6aennyqz91exa4a2sda0019`: Owner, 1 November 2026 02:22Z: 'I want EDGE to be called EDGE Bank again, we will offer the capabilities within them by their own names but I want it to be known that it is a bank.' The finance application's name is EDGE Bank (formal name JudeAI / EDGE Bank); its capabilities keep their own names (Open a personal current account, Send money, ...). Supersedes the 1 November decision that shortened the name to EDGE. - **Screens come from the reference flows; the estate adds only its controls and journeys** `tok:decision:01m6bgf72jtm0n7xyba043bqh4`: X-X.TECH decided on 1 November 2026 (04:10Z to 04:14Z): layouts and screens come from the reference flows recorded in research/ (Mobbin and Refero: Microsoft 365, Notion, Monday, Slack, Fabric, Starling, Monzo, Google Drive, ChatGPT, Base44 V7BRO); the only difference from a reference is the estate's added controls or journeys; nothing is hand-drawn. This resolves the contract's open decision on whether screens mirror a competitor's exact layout. - **EDGE Public Consortium's mark is consortium.svg from the approved brand register** `tok:decision:01m6bgf72knqzc3da9n20c282a`: X-X.TECH supplied the approved brand register on 31 October 2026 (design/brand/approved/BRAND-REGISTER.json): consortium.svg is 'The Edge Consortium application mark'. Under the one contract the Consortium and the public face are one application, so public.svg is not shown as a separate application and cyber-governance.svg is not an application mark. This resolves the contract's open decision on the Consortium's mark and supersedes tok:decision:01m68c87qmz54aqnbm8pqzmgt4 where the two differ. ##### 3. The owner's requests (59) | # | Request | Status | Shown by | |---|---|---|---| | 1 | Five applications: EDGE Finance, EDGE Clinic, The Edge Consortium, Public EDGE, Judiverse | DONE | `spec/units/finance`, `spec/units/clinic`, `spec/units/edge-consortium`, `spec/units/public-edge`, `spec/units/judiverse` | | 2 | Four applications: EDGE Finance, EDGE Clinic, EDGE Public Consortium, Judiverse | OPEN: missing spec/units/public-consortium | `spec/units/finance`, `spec/units/clinic`, `spec/units/public-consortium`, `spec/units/judiverse` | | 3 | Every capability, screen, control, journey and endpoint specified, nothing left to imagination | DONE | `build/estate-spec.json`, `generated/diagrams/README.md` | | 4 | Mermaid diagrams and click-by-click breakdowns for every journey | DONE | `generated/diagrams/README.md` | | 5 | No dead controls anywhere | DONE | `ledger/SWEEP.md` | | 6 | Every competitor feature matched like for like, with proof | DONE | `ledger/parity/README.md`, `docs/proof/README.md` | | 7 | Competitor behaviour read from their own documentation | DONE | `tools/behaviour.py`, `research/parity/context.jsonl` | | 8 | Outcomes matched to competitors' documented outcomes, with baselines and targets | DONE | `ledger/outcomes/README.md`, `spec/standards.json` | | 9 | Novel measures: intent to outcome, words, prompts, tokens, cost and carbon per verified outcome | DONE | `spec/standards.json`, `runtime/engine.py` | | 10 | Help centre, admin guides and API reference set against competitors | DONE | `docs/README.md` | | 11 | The blueprint standard so others can reproduce it | DONE | `docs/STANDARD.md` | | 12 | Reference code and open-source reuse for every operation | DONE | `docs/reference/README.md` | | 13 | A running engine, HTTP server and agent tools generated from the specification | DONE | `runtime/engine.py`, `runtime/server.py`, `runtime/mcp_server.py` | | 14 | Screens generated in Figma from the specification, never hand-drawn | DONE | `design/figma/FRAMES.json` | | 15 | Every October 2026 headline: root cause, protective capability, benchmark | DONE | `ledger/problems/README.md`, `ledger/benchmarks/README.md` | | 16 | Vulnerability map of the whole estate | DONE | `ledger/vulnerabilities/README.md` | | 17 | Provisional value with in-person owner sign-off; reversal by design | DONE | `runtime/engine.py`, `ledger/benchmarks/binance-replay.json` | | 18 | Agents bound to their own devices and resources | DONE | `runtime/credentials.py`, `ledger/benchmarks/binance-replay.json` | | 19 | Intention triage: secure, efficient and reproducible by design | DONE | `tools/intent_triage.py`, `runtime/triage_mcp.py` | | 20 | The FCA's agentic framework exceeded point by point | DONE | `research/finance/FCA-AGENTIC-FRAMEWORK.md` | | 21 | Zero waste: monitored, blocked and impossible to bypass | DONE | `tools/waste.py`, `tools/agent_gate.py`, `tools/install_gate.sh`, `ledger/WASTE.md` | | 22 | Agent mailbox through Judiverse Mail | DONE | `runtime/mail_mcp.py` | | 23 | Agent credentials | DONE | `runtime/credentials.py` | | 24 | Agent SMS: receive free, never spend | DONE | `runtime/sms_mcp.py` | | 25 | Reversibility in health register, including impact reversal for conditions that are part of who people are | DONE | `research/clinic/REVERSIBILITY.md` | | 26 | Regulator, auditor and reviewer workspaces with their workflows | DONE | `spec/units/foundation/regulator-auditor-reviewer-workspace.json` | | 27 | Pre-onboarded organisations from public registers, claimed by magic link | DONE | `spec/units/foundation/pre-onboarded-organisations.json`, `spec/units/foundation/claim-organisation-by-magic-link.json` | | 28 | Automated outreach, scheduling and reviewer inboxes | DONE | `spec/units/foundation/automated-review-outreach-and-scheduling.json` | | 29 | Free evaluation, commercial licensing and enforcement evidence | DONE | `spec/units/foundation/evaluation-and-commercial-licensing.json`, `spec/units/foundation/licence-enforcement-evidence.json` | | 30 | Estate gap analysis | DONE | `ledger/GAPS.md` | | 31 | Mail relay credentials in the environment | OPEN: missing NEEDS THE OWNER | `NEEDS THE OWNER` | | 32 | The SMS Works key and inbound forwarding | OPEN: missing NEEDS THE OWNER | `NEEDS THE OWNER` | | 33 | Hosting decision and deployment | OPEN: missing NEEDS THE OWNER | `NEEDS THE OWNER` | | 34 | Real receiving systems connected (banks, NHS, registries) | OPEN: missing NEEDS THE OWNER | `NEEDS THE OWNER` | | 35 | The Edge Consortium for education and research, separate from Public EDGE for government | DONE | `contract/estate-contract.json`, `tools/apps.py`, `spec/units/edge-consortium`, `spec/units/public-edge` | | 36 | The governed intelligence degree: modules, real-work assessment (coursework with proof, live challenges, live spec-driven development, live audio), set against leading programmes | DONE | `research/edge-consortium/GOVERNED-INTELLIGENCE-DEGREE.md`, `tools/degree.py`, `spec/units/edge-consortium/governed-intelligence-degree.json`, `spec/units/edge-consortium/run-a-live-assessment.json` | | 37 | Offered with leading institutions and by The Edge Consortium | DONE | `spec/units/edge-consortium/offer-the-degree-with-a-partner-institution.json` | | 38 | Scope discipline: scope creep managed as raw material | DONE | `spec/units/foundation/scope-change-intake.json` | | 39 | Bottleneck and blocker science: removed at the root, shortest intention to outcome | DONE | `spec/units/foundation/bottleneck-and-blocker-removal.json` | | 40 | No irreversible outcome without the owner's analog, in-person act, with identical backup hardware | DONE | `research/public-edge/ANALOG-HUMAN-SWITCH.md`, `runtime/engine.py`, `spec/units/public-edge/analog-human-switch-for-irreversible-actions.json`, `ledger/benchmarks/analog-switch.json` | | 41 | Judiverse included with every licence of the four applications, subject to rate limits; standalone for consumers by capability or whole | DONE | `spec/offers.json`, `runtime/entitlements.py`, `spec/units/judiverse/judiverse-included-with-every-licence.json`, `spec/units/judiverse/judiverse-standalone-subscription.json` | | 42 | Judiverse included in Edge Consortium tuition; real sandboxed projects; paid student work offset against tuition; contractor pathway; student finance petition | DONE | `spec/units/edge-consortium/judiverse-included-in-tuition.json`, `spec/units/edge-consortium/real-project-assignments.json`, `spec/units/edge-consortium/paid-student-work-with-tuition-offset.json`, `spec/units/edge-consortium/student-to-contractor-pathway.json`, `spec/units/edge-consortium/petition-for-student-finance.json`, `ledger/benchmarks/offers.json` | | 43 | Specialisms only after every domain and how they connect: medical, financial, social, public, research, defence, environmental and legal intelligence | DONE | `spec/degree.json`, `runtime/programme.py`, `ledger/benchmarks/programme.json`, `spec/units/edge-consortium/choose-a-specialism.json` | | 44 | Every new screen in Figma with progress, navigation and monitors from the journey diagrams; no dead controls | DONE | `design/figma/FRAMES.json`, `tools/expand.py`, `generated/diagrams/README.md` | | 45 | Clean, brief, permanent URLs and a URL directory from the ontology | DONE | `tools/urls.py`, `spec/urls.json`, `docs/URLS.md` | | 46 | London South Bank University as first partner institution; SITHub networking module; shared campus; R&D and international technical students | DONE | `research/edge-consortium/GOVERNED-INTELLIGENCE-DEGREE.md`, `spec/units/edge-consortium/share-a-campus-with-a-partner-university.json`, `spec/units/edge-consortium/research-and-development-partnership.json`, `spec/units/edge-consortium/international-technical-students.json` | | 47 | Register The Edge Consortium with the Office for Students | DONE | `spec/units/edge-consortium/register-with-the-office-for-students.json` | | 48 | All research compiled into one agent and human markdown, organised by the ontology, with every finding accounted for and carried | DONE | `research/AGENT-RESEARCH.md`, `tools/compendium.py`, `tools/carry.py`, `spec/carry.json` | | 49 | The owner's white papers, frontier agent capability reference and SITHub deck built into the degree | DONE | `research/edge-consortium/SOURCES-2026-10-31.md`, `research/edge-consortium/GOVERNED-INTELLIGENCE-DEGREE.md` | | 50 | Scope frozen; every later change only through a change request | DONE | `spec/SCOPE-FREEZE.json`, `tools/freeze.py`, `ledger/SCOPE-CHANGES.md` | | 51 | Every guard enforced: bound by family to a deterministic rule or a plug slot | DONE | `runtime/guards.py`, `runtime/plugins.py`, `spec/plugins.json` | | 52 | Plug-and-play mechanism for external models and datasets: one line to plug, one to test | DONE | `tools/plugs.py`, `docs/PLUGS.md` | | 53 | Every capability in the TOK catalogue | DONE | `tok/catalogue.jsonl`, `tools/catalogue.py` | | 54 | Every old-estate document read and its findings placed | DONE | `tools/oldestate.py`, `ledger/old-estate/README.md` | | 55 | Every capability and workflow as an MCP; one hub for search, call and the racetrack | DONE | `runtime/hub_mcp.py`, `runtime/mcp_server.py`, `docs/MCP.md`, `ledger/benchmarks/mcp.json` | | 56 | Agents do everything themselves on a racetrack: intention to verified outcome, rehearsed first, protected by the engine | DONE | `runtime/racetrack.py` | | 57 | Drag-and-drop plugs agents can use themselves | DONE | `plugs/README.md`, `tools/plugs.py` | | 58 | Figma screens without uploads: compressed calls inflated in Figma | DONE | `design/figma/inflate.js`, `design/figma/emit_figma.py` | | 59 | Signed agreements: Office for Students registration decision, LSBU partnership and shared-campus terms, a student sponsor licence for international students | NEEDS THE OWNER | | ##### 4. The specification - **Standards** (`spec/standards.json`): `schema`, `statement_class`, `agreed_by`, `what_this_is`, `beat_the_best`, `per_operation`, `every_operation`, `every_capability`, `outcome_economics` - **Offers** (`spec/offers.json`): included-with-app-licence: an organisation with an active paid licence for any of the four applications; included-in-tuition: a student enrolled at The Edge Consortium; standalone-whole: a consumer with no business licence for any of the four; standalone-by-capability: a consumer with no business licence for any of the four - **Degree** (`spec/degree.json`): 30 modules; 9 pathways (Governed Intelligence (Medical Intelligence), Governed Intelligence (Financial Intelligence), Governed Intelligence (Social Intelligence), Governed Intelligence (Public Intelligence), Governed Intelligence (Research and Educational Intelligence), Governed Intelligence (Defence and Security Intelligence), Governed Intelligence (Environmental and Energy Intelligence), Governed Intelligence (Legal Intelligence), Governed Intelligence (Network and Edge Intelligence)); gate: GI300 and every domain worked 2 times - **Capabilities by application**: clinic 342, edge-consortium 363, finance 391, foundation 9, judiverse 22, public-edge 489; URL directory: `docs/URLS.md` ##### 5. Proofs: benchmarks, tests and baselines - `ledger/benchmarks/analog-switch.json`: {"truth_status": "SIMULATION", "operation": "public-edge/analog-human-switch-for-irreversible-actions/execute-the-irreversible-action-once-onl", "ai_attempts": 1000, "executed_by_ai": 0, "held_for_analog_act": 1000, "revoked_unit_executed": false, "backup_unit_executed": true, "second_act_on_same_request": "already executed", "person_through_software": "ANALOG_HUMAN_ACT_REQUIRED"} - `ledger/benchmarks/binance-replay.json`: {"truth_status": "SIMULATION", "operation": "finance/crypto-wallet/send-crypto"} - `ledger/benchmarks/mcp.json`: {} - `ledger/benchmarks/offers.json`: {"consumer_by_capability": "standalone-by-capability", "consumer_whole": "standalone-whole", "below_pay_floor": "BELOW_PAY_FLOOR", "over_visa_hours": "WORK_HOURS_LIMIT", "sandboxed": "SANDBOXED", "client_takes_without_paying": "NOT_PAID", "paid_and_deployed": true} - `ledger/benchmarks/programme.json`: {} - `ledger/benchmarks/README.md`: Generated by `tools/bench.py` on `runtime/engine.py`. Synthetic test inputs; local effect (default adapter), so this is the estate's own processing time, not a receiving system's. Repeat with `python3 tools/bench.py `. / Operations / Verified outcomes / Median p50 / Worst p95 / Prompts / Tokens / Deterministic share / /---/---/---/---/---/---/---/ / 8534 / 0 / None ms / None ms / 0 / 0 / - / - `ledger/parity/README.md`: Generated by `tools/match.py`. One line per competitor clause in `research/parity/`: the unit, screen, control, operation and endpoint that covers it, and how it was placed. Every clause is covered. / Application / Clauses / On existing steps / In new capabilities / /---/---/---/---/ / clinic / 381 / 0 / 0 / / edge-consortium / 3075 / 0 / 0 / / finance / 968 / 0 / 0 / - `ledger/outcomes/README.md`: Generated by `tools/outcomes.py`. One row per operation. The **floor** is the competitor's best published figure (`research/outcomes/*.jsonl`, or a figure on the competitor's own page); the **target** is the estate's own measurement (`ledger/benchmarks/results.jsonl`, `ledger/evidence/operations.jsonl`, `ledger/evidence/journeys-*.jsonl`) and stays **OPEN** until one exists. No number is halved, moved halfway or otherwise derived. **Outcome matched** is true only when the capability's bound reference flow (`research/parity/REFERENCE-FLOWS.jsonl`) records the screens that cover the step this operation sits at in the person's journey. A competitor clause or figure found by rare-word overlap is - `ledger/RUN.md`: Rehearsed as a person: {"OK:None": 25916, "REFUSED:INVALID_INPUT": 64276, "REFUSED:CONSENT_REQUIRED: we need your agreement before we can use health details": 1, "REFUSED:EMPTY_DESCRIPTION": 1, "REFUSED:CONSENT_NOT_GIVEN": 2, "REFUSED:FORMULATION_INCOMPLETE": 1, "REFUSED:REFERENCE_MISSING": 1, "REFUSED:NOTE_EMPTY": 1, "REFUSED:DESCRIPTION_MISSING": 1, "REFUSED:CONSENT_NOT_GIVEN: emergency guidance is still shown": 1, "REFUSED:NOTE_TOO_LONG": 1, "REFUSED:DATES_INVALID": 1, "REFUSED:REASON_REQUIRED": 10, "REFUSED:EXPIRY_TOO_LONG": 1, "REFUSED:NOT_ATTESTED": 1, "REFUSED:CONSENT_REQUIRED": 4, "REFUSED:INCOMPLETE": 2, "REFUSED:LIMITS_INCONSISTENT: per-payment must not exceed daily, daily must not - `ledger/SWEEP.md`: Generated by `tools/sweep.py`. Deterministic: the same estate gives the same sweep. Total problems: **9098** across 1616 units. - rule 5: 74 - rule 6: 20 - rule 8: 9004 Competitor matches graded by wording: 31. This measures shared wording only; behaviour is measured by `tools/closure.py`. The gap analysis is `ledger/GAPS.md`; external plugs `docs/PLUGS.md`; old-estate coverage `ledger/old-estate/README.md`; changes since the scope freeze `ledger/SCOPE-CHANGES.md`. ##### 6. Research findings, each placed on the capability that serves it Every bullet, table row and stated sentence in every research file. Each has an ID (`F-` and a digest of its file and text), the heading it sits under, its statement class where the research gave one, and where it lives in the estate: the capability's clean URL (a finding no other capability carried is a checked requirement of a `findings-…` capability, made by `tools/carry.py`), or **UNPLACED**. ###### research ###### `research/README.md` (15 findings) - `F-24506a7770` L3 _Research_: This folder holds research written for the estate. It informs what the four applications must do; it is never authority. Every finding carries its source link and the date it was found, and is marked as one of: → **UNPLACED** - `F-3710803a42` L5 [OBSERVATION] _Research_: **OBSERVATION** (seen in a primary source or a real product), → **UNPLACED** - `F-8044582fe9` L6 [AGREEMENT] _Research_: **AGREEMENT** (a law, regulation or standard, with its status and dates), → **UNPLACED** - `F-a4d104b776` L7 [CLAIM] _Research_: **CLAIM** (what a company or document says about itself, not independently checked), → **UNPLACED** - `F-5f1e244c82` L11 _Research_: Raw material (downloaded pages, datasets, screenshots, earlier work) stays in storage and is referenced by digest; only the written research lives here. → **UNPLACED** - `F-db6299f724` L15 _Research › How to read it_: Start with the application you are working on. Each topic file opens with what it is, how it was gathered and what it does not cover. → **UNPLACED** - `F-e90307ca3e` L19 _Research › How to read it_: `finance/` / EDGE Finance: competitors and our lane, regulators in the UK and worldwide, agent assurance gaps and reusable open-source systems, confidence reporting, physical presence and solution banks, real product flows → **UNPLACED** - `F-d612990d9c` L20 _Research › How to read it_: `clinic/` / EDGE Clinic: social prescribing, symptom checking, early detection and wearables; competitors, regulators and reusable systems; real product flows → **UNPLACED** - `F-6f506ac037` L21 _Research › How to read it_: `public-consortium/` / EDGE Public Consortium: competitors, regulators and reusable systems for cyber governance, procurement, vetting, logistics and education; real product flows → **UNPLACED** - `F-160d6107cb` L22 _Research › How to read it_: `judiverse/` / Judiverse: competitors, regulators and reusable systems for office, assistants, agent lab and music; real product flows → **UNPLACED** - `F-f2819a68ed` L23 _Research › How to read it_: `experience/` / Customer experience patterns from any industry (onboarding, errors, confirmations, help, accessibility and more), useful to every application → **UNPLACED** - `F-ef93d6256a` L24 _Research › How to read it_: `market/` / What is being funded and bought (grants, contracts, venture capital) and the costliest real problems by industry → **UNPLACED** - `F-3404e6574c` L25 `legacy/` / What the organisation's earlier work already contains (competitors, benchmarks, readiness, blueprints, features, standards, papers, guides, strategy, funding), read as research so none of it is wasted or repeated → old-estate inventory (section 8) - `F-e91d633757` L29 _Research › Real product flows_: The `REFERO-FLOWS` files come from Refero, a library of real app and website screens with each journey described step by step. Results from other industries are kept on purpose, because they are often the best customer experience to learn from. Refero holds mostly consumer and start-up products: many incumbents (for example Starling, Monzo, Chase UK) and most government, enterprise and NHS systems are not in it, so these files are a sample of the floor, not a census. → **UNPLACED** - `F-03f2f56b26` L33 _Research › Coverage_: What exists in each folder, and what is still being researched, is listed by `python3 tools/build.py`, which writes the generated index. Nothing in this folder states a count of estate objects; those are produced live by the build. → **UNPLACED** ###### `research/REUSE-ENGINES.md` (20 findings) - `F-ac0afe9133` L3 [CLAIM] _Engines behind the applications: what is reused, what is built_: Statement class: research and recommendation, 1 November 2026 03:35Z. OBSERVED where the estate's own research verified it (file and date given); CLAIM where it has not. Never a rule. → **UNPLACED** - `F-9b1cf417e8` L7 _Engines behind the applications: what is reused, what is built › What the specification actually needs (counted from spec/receivers.json, research and competitor units excluded)_: Application / Operations / Workflow steps on a record or case (journey state, ours) / Evidence reads (ours) / Domain entities that need an engine → **UNPLACED** - `F-370471269f` L9 _Engines behind the applications: what is reused, what is built › What the specification actually needs (counted from spec/receivers.json, research and competitor units excluded)_: EDGE Bank / 3,301 / 1,452 / 378 / account, payment, card, member, payee; rail: payment-scheme (48 operations) → **UNPLACED** - `F-b62a6a9957` L10 _Engines behind the applications: what is reused, what is built › What the specification actually needs (counted from spec/receivers.json, research and competitor units excluded)_: EDGE Clinic / 2,284 / 1,376 / 338 / FHIR resources (69), consent (34), identity claim (28), safety plan (14), plan (11); NHS services (29), detection models (11) → **UNPLACED** - `F-fdf530ba5f` L11 _Engines behind the applications: what is reused, what is built › What the specification actually needs (counted from spec/receivers.json, research and competitor units excluded)_: EDGE Consortium / 4,778 / 1,942 / 359 / case (74), egress and access policy (89), role (52), module and client (84), consortium (32) → **UNPLACED** - `F-34dd5d8e62` L12 _Engines behind the applications: what is reused, what is built › What the specification actually needs (counted from spec/receivers.json, research and competitor units excluded)_: EDGE Public / 3,225 / 1,830 / 486 / application (41), vacancy (18), asset (18), alert (18), decision (15) → **UNPLACED** - `F-6e2b90b08e` L13 _Engines behind the applications: what is reused, what is built › What the specification actually needs (counted from spec/receivers.json, research and competitor units excluded)_: Judiverse Suite / 3,770 / (steps spread across) / intention (113), contact (91), upload terms (90), template (82), event (65), task (44), document versions and objects (98) → **UNPLACED** - `F-c80d9d17e2` L15 _Engines behind the applications: what is reused, what is built › What the specification actually needs (counted from spec/receivers.json, research and competitor units excluded)_: Reading: about three quarters of every application is journey state and evidence, which the estate already runs. The engine question is the last column: a small set of domain entities per application. One engine per application covers it. → **UNPLACED** - `F-89dd1f6e1e` L21 _Engines behind the applications: what is reused, what is built › The engine per application_: EDGE Bank / Griffin (sponsor bank, rail) + Apache Fineract (deposits, loans, accounting) + TigerBeetle (ledger) + Hyperswitch (card and alternative processors) / accounts, balances, payments, Direct Debits, standing orders, loans, overdrafts, cards / Griffin: PRA/FCA bank, sandbox self-serve (OBSERVED 1 Nov, research/finance/ENGINE-BY-STARLING.md §3). Fineract Apache-2.0, TigerBeetle, Hyperswitch Apache-2.0 (OBSERVED 31 Oct, research/finance/ASSURANCE-AND-REUSE.md §2) / `payment-scheme` and `card-processor` slots; runtime/domains/bank.py already holds the ledger interface, so Fineract or TigerBeetle replaces its SQLite tables without touching a screen → **UNPLACED** - `F-a0b68ed85a` L22 [CLAIM] _Engines behind the applications: what is reused, what is built › The engine per application_: EDGE Clinic / Medplum (FHIR R4 server, UK Core, scheduling, auth, bots, React components) / patients, appointments, consents, observations, care plans, questionnaires, referrals / Apache-2.0 (OBSERVED 31 Oct, research/clinic/COMPETITORS-REGULATORS-REUSE.md); ONC certification CLAIM / `feature_fhir`, consent, safety_plan and plan entities map to FHIR resources through one adapter; `nhs-services` stays a plug → **UNPLACED** - `F-7ada38d1fb` L23 [CLAIM] _Engines behind the applications: what is reused, what is built › The engine per application_: EDGE Consortium / ERPNext (Frappe) for roles, modules, clients, access and cases; Keycloak or Ory for identity federation / the university's records: enrolment, modules, roles with expiry, cases, policies / ERPNext GPL-3 (CLAIM, not verified by the estate; network copyleft to assess), Keycloak Apache-2.0 (OBSERVED 31 Oct) / case, role, module, client, access entities through the Frappe REST API as one adapter → **UNPLACED** - `F-ffe52ed0e3` L24 _Engines behind the applications: what is reused, what is built › The engine per application_: EDGE Public / Same ERPNext instance (applications, vacancies, assets, alerts, decisions) / government-on-the-estate records and decisions / as above / application, vacancy, asset, alert, decision entities → **UNPLACED** - `F-13daa7e74f` L25 [CLAIM] _Engines behind the applications: what is reused, what is built › The engine per application_: Judiverse Suite / ONLYOFFICE Docs (documents, sheets, decks), Matrix (Synapse or Conduit) for messages and calls with LiveKit for media, a CalDAV server for calendar, Simwood for UK numbers / editing, collaboration, chat, calls, meetings, calendar, mail (mail already runs natively) / ONLYOFFICE AGPL-3 community with a separate embedding licence (CLAIM); Matrix Synapse AGPL-3 (CLAIM); LiveKit Apache-2.0 (CLAIM); Simwood commercial API (CLAIM) / document, template, event, task, contact entities; `voice-telephony` and `sms-gateway` slots → **UNPLACED** - `F-19b8690980` L26 [CLAIM] _Engines behind the applications: what is reused, what is built › The engine per application_: Agents operating other software / Composio (hosted) or the estate's own MCP hub / third-party tools under per-user authentication / Composio hosts data outside the estate: assess before UK-only use (CLAIM) / the hub already exposes every operation to agents (`estate_call`, `racetrack_run`); Composio adds outbound integrations only → **UNPLACED** - `F-404cde90b5` L30 _Engines behind the applications: what is reused, what is built › The pattern (unchanged)_: approved screen → Jude's authenticated operation (identity, mandate, guards, rehearsal, evidence: ours) → engine adapter (runtime/adapters.ADAPTERS, one module per engine in runtime/domains/) → the engine's own confirmation read back → evidence. TOK, EDGE ID and assurance stay shared. Nothing is hand-built that an engine supplies; what is built is the adapter and the domain behaviour no engine has (money provenance, agent mandates, glass-box evidence). → **UNPLACED** - `F-0caf5250c0` L34 _Engines behind the applications: what is reused, what is built › Order_: EDGE Bank: Griffin key, then Fineract behind runtime/domains/bank.py. The screens and ledger interface exist; the customer sees the same app with real rails. → **UNPLACED** - `F-2b828dc610` L35 _Engines behind the applications: what is reused, what is built › Order_: EDGE Clinic: Medplum on Civo; one FHIR adapter for the five clinic entities. Largest single coverage gain after the bank. → **UNPLACED** - `F-a0902e9c3c` L36 _Engines behind the applications: what is reused, what is built › Order_: Judiverse Suite: ONLYOFFICE and Matrix, because documents and messages are what every other application's people use daily. → **UNPLACED** - `F-b7aa6059fb` L37 _Engines behind the applications: what is reused, what is built › Order_: EDGE Consortium and EDGE Public: one ERPNext instance for both. → **UNPLACED** - `F-d612b5b80a` L39 _Engines behind the applications: what is reused, what is built › Order_: Each engine lands as a named service on Civo with backup proven first, then its adapter binds, then the affected journeys are walked and their results read back from the engine. → **UNPLACED** ###### clinic ###### `research/clinic/COMPETITORS-REGULATORS-REUSE.md` (157 findings) - `F-a4dc55d42f` L3 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems_: Research for EDGE Clinic, one of four applications in JudeAI / EDGE ID. Written 31 October 2026 for a brand-new engineer. → **UNPLACED** - `F-32dfea2efd` L4 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems_: This is research, not clinical or legal advice. Anything here must be checked by a qualified clinical safety officer and a lawyer before it drives a decision. → **UNPLACED** - `F-dc38d0f9ac` L6 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems_: **Scope.** Four topics only: (1) competitors, (2) regulators and rules, (3) agent assurance gaps in health, (4) open standards and open-source to reuse. → **UNPLACED** - `F-c592024ca8` L7 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems_: Social prescribing evidence, symptom checkers, early detection, wearables, physical presence and NHS costs are covered elsewhere, in `research/clinic/SOCIAL-PRESCRIBING-AND-DETECTION.md`, and are not repeated here. → **UNPLACED** - `F-be77430c40` L13 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › How to read the labels_: **OBSERVED** / I read it directly in a primary page or file during this research (for example, a licence file or a regulator's page). → **UNPLACED** - `F-0393cd2dcc` L14 [CLAIM] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › How to read the labels_: **CLAIM** / A vendor, a law firm, a trade paper or a blog says it. Not independently checked. → **UNPLACED** - `F-3a8f35cf39` L15 [AGREEMENT] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › How to read the labels_: **AGREEMENT** / A law, rule, standard or guidance that is adopted and in force, with a date. → **UNPLACED** - `F-cff7eb5259` L16 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › How to read the labels_: **PROPOSED** / A draft, consultation, plan or recommendation. Not binding yet. → **UNPLACED** - `F-5bc72f3028` L17 [UNKNOWN] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › How to read the labels_: **UNKNOWN** / Sources disagree, or I could not find a reliable answer. → **UNPLACED** - `F-3bbb9ca7c6` L19 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › How to read the labels_: Every link was accessed on 2026-10-31. Where a source is a secondary summary rather than the regulator's own text, the line says so. → **UNPLACED** - `F-91a9d23b3a` L25 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 0. The ten things to know first_: **The UK has no AI-specific medical device law yet.** The MHRA (Medicines and Healthcare products Regulatory Agency, the UK device regulator) put draft amending regulations out in May 2026, and they are expected to apply from June 2027. They have not passed Parliament. PROPOSED. [Latham & Watkins](https://www.lw.com/en/insights/mhra-publishes-draft-amendment-to-the-uk-medical-devices-regulations), accessed 2026-10-31. → `/clinic/register-device-fault` - `F-3a9d1910b6` L26 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 0. The ten things to know first_: **A national commission set out 44 recommendations for regulating AI in UK health on 10 September 2026**, including staged "L-plate" authorisations and continuous real-world monitoring. Government says it will "carefully consider" them. PROPOSED. [GOV.UK press release](https://www.gov.uk/government/news/independent-commission-led-by-nhs-doctors-sets-out-blueprint-to-accelerate-safe-ai-adoption-in-healthcare) (OBSERVED), accessed 2026-10-31. → `/clinic/apply-national-data-opt` - `F-307d6d0c82` L27 [AGREEMENT] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 0. The ten things to know first_: **The MHRA's test for "is it a medical device?" (29 July 2026):** a tool that only transcribes, summarises or drafts letters is not a device, but one that supports diagnosis, treatment or prevention, or "take[s] automated action such as placing orders without clinician review", is. AGREEMENT (guidance). [MHRA news](https://www.gov.uk/government/news/mhra-clarifies-regulatory-status-of-ambient-voice-technologies-used-in-the-nhs) (OBSERVED), accessed 2026-10-31. EDGE Clinic triage, prescribing safety and retinal screening fall on the device side of that line. → `/clinic/retinal-screening` - `F-af6e6fdef6` L28 [AGREEMENT] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 0. The ten things to know first_: **The NHS clinical safety standards DCB0129 and DCB0160 are being rewritten, with AI named as a reason.** The consultation ran 29 June to 11 September 2026. The current versions still apply. PROPOSED (revision); AGREEMENT (current versions). [NHS England consultation page](https://www.engage.england.nhs.uk/patient-safety/national-review-of-clinical-risk-management-standa/) (OBSERVED), accessed 2026-10-31. → `/clinic/run-clinical-safety-training` - `F-1e56f87017` L29 [AGREEMENT] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 0. The ten things to know first_: **The EU AI Act's high-risk rules for AI inside medical devices now apply from 2 August 2028**, moved back by the "Digital Omnibus on AI", Regulation (EU) 2026/1744, which came into force on 27 July 2026. AGREEMENT. [Hunton](https://www.hunton.com/privacy-and-cybersecurity-law-blog/eu-digital-omnibus-on-ai-enters-into-force) (OBSERVED, law-firm summary), accessed 2026-10-31. → `/clinic/act-high-risk-answer` - `F-490c9d5a70` L30 [AGREEMENT] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 0. The ten things to know first_: **In the US, FDA's lifecycle guidance for AI-enabled devices is still a January 2025 draft.** The final PCCP (predetermined change control plan) guidance is in force from December 2024. PROPOSED and AGREEMENT respectively. [RAPS](https://www.raps.org/resource/fda-s-device-center-releases-guidance-agenda-for-fy-2027.html); [Ropes & Gray](https://ropesgray.com/en/insights/alerts/2024/12/fda-finalizes-guidance-on-predetermined-change-control-plans-for-ai-enabled-device), accessed 2026-10-31. → `/clinic/front-door-check` - `F-7d260ea7e1` L31 [CLAIM] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 0. The ten things to know first_: **The big AI labs are now in consumer health.** OpenAI launched ChatGPT Health in January 2026, and Anthropic launched Claude for Healthcare in January 2026. Both are US-first and connect personal records. CLAIM. [Fortune](https://fortune.com/2026/01/07/openai-launches-chatgpt-health-in-a-push-to-become-a-hub-for-personal-health-data); [Anthropic](https://www.anthropic.com/news/healthcare-life-sciences), accessed 2026-10-31. → `/clinic/mental-health-care-crisis` - `F-cbff46c7fb` L32 [CLAIM] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 0. The ten things to know first_: **Evidence that "a clever model" is not "a safe front door":** in a randomised study of 1,298 UK participants, people who used LLMs did not choose better than controls, even though the models alone named conditions correctly 94.9% of the time. CLAIM (peer-reviewed). [University of Oxford](https://www.ox.ac.uk/news/2026-02-10-new-study-warns-risks-ai-chatbots-giving-medical-advice), accessed 2026-10-31. → `/clinic/front-door-check` - `F-79309f73f3` L33 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 0. The ten things to know first_: **No standard exists yet for an AI agent's identity, mandate and delegation in health.** The OpenID Foundation names these gaps and is drafting extensions. PROPOSED. [OpenID Foundation](https://openid.net/new-whitepaper-tackles-ai-agent-identity-challenges/), accessed 2026-10-31. → `/clinic/accept-proxy-mandate` - `F-dd093cb24d` L34 [CLAIM] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 0. The ten things to know first_: **There is no open, clinical-grade drug-interaction source.** The US NLM (National Library of Medicine) switched off its free interaction API around 2 January 2024. The BNF (British National Formulary) needs a licence for use inside systems. DrugBank's full data is non-commercial only. OBSERVED/CLAIM. [NLM RxNav](https://www.lhncbc.nlm.nih.gov/RxNav/news/APIUpdate.html); [Pharmaceutical Press licensing](https://www.pharmaceuticalpress.com/?p=146134) (OBSERVED); [DrugBank FAQ](https://dev.drugbank.com/guides/faqs), accessed 2026-10-31. → `/clinic/medication-safety-check` - `F-d8b3ef3d60` L44 [AGREEMENT] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: **NHS App** (NHS England) / The government says the app will be the "front door to the health service". The 10 Year Health Plan (July 2025) targets a "full front door to the entire NHS" by 2028. / The plan proposed "My NHS GP", an AI tool to route people to the right care. It is reported to reach 200,000 patients in year one, with full rollout by April 2028. No confirmed 2026 go-live found. / AGREEMENT (policy); PROPOSED (My NHS GP) / [GOV.UK oral statement](https://www.gov.uk/government/speeches/fit-for-the-future-health-and-social-care-secretarys-statement); [Full Fact tracker](https://fullfact.org/government-tracker/nhs-app-improvements/); [Digital Health](https://www.digitalhealth.net/2025/07/what-does-the-10-year-plan-mean-for-digital/), accessed 2026-10-31 → `/clinic/front-door-check` - `F-ad9a9b8dca` L45 [CLAIM] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: NHS App + Rapid Health Smart Triage / A GP booking and triage tool that works inside the NHS App. / Reported as available to more than a million patients in April 2026. / CLAIM / [Digital Health](https://www.digitalhealth.net/2026/04/ai-tool-allows-a-million-patients-to-book-appointments-via-nhs-app/), accessed 2026-10-31 → `/clinic/symptom-check-triage` - `F-b26e5e386b` L46 [CLAIM] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: NHS App messaging shift / NHS App leadership has described the app as "not the only" service, moving from "front door" towards "companion". / n/a / CLAIM / [Digital Health](https://www.digitalhealth.net/2026/03/nhs-app-to-move-from-front-door-to-companion/), accessed 2026-10-31 → `/clinic/front-door-check` - `F-219dce19d6` L47 [CLAIM] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: **Accurx** / Patient messaging, online triage, video consultations and Scribe. / Accurx says Scribe (built with Tandem) has summarised over one million consultations across nearly 3,000 practices, and that 1,350 practices run "total triage" on Accurx. It advertises "AI-enhanced Patient Triage". / CLAIM / [Accurx blog](https://www.accurx.com/post/whats-new-in-2026-for-gp); [Accurx benefits](https://www.accurx.com/benefits), accessed 2026-10-31 → `/clinic/online-urgent-triage` - `F-ffb7484ca0` L48 [CLAIM] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: Accurx pricing / Per-unit prices on the G-Cloud government catalogue. For example, Scribe is £0.50 down to £0.40 per unit, depending on volume. / n/a / CLAIM (vendor price list) / [G-Cloud pricing document](https://assets.applytosupply.digitalmarketplace.service.gov.uk/g-cloud-14/documents/708675/281422867192066-pricing-document-2026-08-10-1029.pdf), accessed 2026-10-31 → `/clinic/price-policy-clinic-data` - `F-c1438c5662` L49 [CLAIM] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: **EMIS** (Optum UK, now TPG) / EMIS Web, the GP clinical record system, and EMIS-X (cloud). / EMIS Scribe, an ambient scribe for EMIS Web and EMIS-X. EMIS is on NHS England's AVT (ambient voice technology) registry. No autonomous agents found. / CLAIM / [EMIS April 2026 newsletter](https://www.emishealth.com/news-insights/april-2026-newsletter); [EMIS January 2026 newsletter](https://www.emishealth.com/news-insights/january-2026-newsletter), accessed 2026-10-31 → `/clinic/front-door-check` - `F-180d95aa47` L50 [CLAIM] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: EMIS ownership / Reported that TPG's acquisition of Optum UK, including EMIS, closed on 13 March 2026. / n/a / CLAIM (single secondary source) / [healthcare.digital](https://www.healthcare.digital/single-post/the-strategic-transformation-of-uk-primary-care-digital-infrastructure-tpg-s-acquisition-of-emis-an), accessed 2026-10-31 → **UNPLACED** - `F-de6c63e7f7` L51 [CLAIM] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: **TPP SystmOne + Airmid** / SystmOne is a clinical record system. Airmid is a patient app with booking, video, messaging and wearable readings (Apple Health, Google Fit), and it supports NHS login. / TPP says it is "embedding AI and Genetic data" for early disease detection. No ambient scribe found. / CLAIM / [TPP Airmid](https://www.tpp-uk.com/products/airmid); [TPP news](https://tpp-uk.com/news/taking-healthcare-to-the-next-level/), accessed 2026-10-31 → `/clinic/price-policy-clinic-data` - `F-53c2431add` L52 [CLAIM] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: **Doctolib + Medicus** / Doctolib (France) acquired Medicus, a UK GP software company, in May 2026, and plans to invest over £100m in the UK. / Positioned as an "AI-powered GP platform". / CLAIM / [BusinessCloud](https://businesscloud.co.uk/news/doctolib-acquires-medicus-plans-to-invest-100m-in-uk/); [TechMarketView](https://www.techmarketview.com/ukhotviews/archive/2026/05/06/doctolib-backs-medicus-to-challenge-gp-it-duopoly), accessed 2026-10-31 → **UNPLACED** - `F-d453d3eebd` L53 [CLAIM] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: **Huma** (+ eConsult, iPlato) / Huma bought eConsult (digital triage, reportedly used by over 1,800 GP practices) in October 2024 and launched Huma Workspace. Its platform holds FDA 510(k) Class II (2023) and Health Canada Class II (2025) clearance as SaMD (software as a medical device). / Health Canada licence allows AI algorithms for screening, diagnosis and decision support. Huma acquired Aluna (remote spirometry) in May 2025. / CLAIM / [Huma newsroom](https://www.huma.com/newsroom/huma-acquires-econsult-and-launches-huma-workspace-for-health-systems); [Digital Health](https://digitalhealth.net/2024/10/huma-acquires-econsult-to-become-end-to-end-tech-platform); [Huma, Health Canada](https://huma.com/resources/huma-therapeutics-receives-health-canada-class-ii-medical-device-license-for-digital-health-platform); [MobiHealthNews](https://www.mobihealthnews.com/news/huma-receives-510k-class-ii-clearance-its-disease-agnostic-platform), accessed 2026-10-31 → `/clinic/classify-software-medical-device` - `F-2a54af2347` L54 [UNKNOWN] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: **Livi / Kry** / Video GP capacity sold to NHS commissioners. CQC (Care Quality Commission) lists Livi UK's location as "Archived", and Wikipedia says Livi was being rebranded Kry in France and the UK. / Not found. / UNKNOWN (current UK status) / [CQC](https://www.cqc.org.uk/location/1-10716659568/inspection-summary); [Wikipedia](https://en.wikipedia.org/wiki/Livi_(company)), accessed 2026-10-31 → `/clinic/see-day` - `F-b60b5d979a` L55 [UNKNOWN] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: **Push Doctor** / Video GP. Square Health bought its business and assets; CB Insights dates the deal to December 2021. / Not found. / UNKNOWN (current status) / [Healthcare & Protection](https://healthcareandprotection.com/square-health-buys-push-doctor/amp/); [CB Insights](https://www.cbinsights.com/investor/push-doctor), accessed 2026-10-31 → `/clinic/share-medicines-list-clinic` - `F-3c9bd39544` L56 [UNKNOWN] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: **Doctaly** / Paid face-to-face GP appointments with NHS GPs. A separate brand, "Doctaly Assist", offers GP access via WhatsApp. No 2025–2026 news found. / Not found. / UNKNOWN / [BusinessCloud](https://businesscloud.co.uk/news/platform-for-private-consultations-with-nhs-gps-crowdfunds/); [Endole](https://open.endole.co.uk/insight/brand/167752-doctaly-assist), accessed 2026-10-31 → `/clinic/settle-clinic` - `F-3cdd965aa8` L57 [CLAIM] _1. Competitors › 1.1 Digital front doors and primary care platforms (UK first)_: **Cera** / Home care with AI. Raised about $150m (debt and equity) in January 2025. Says it works in 150 of 382 UK local authorities and 29 of 42 ICBs (integrated care boards). / Cera claims up to 70% fewer emergency admissions. TechCrunch reported that some NHS partnerships were wrongly listed on Cera's website. / CLAIM (disputed in part) / [Med-Tech News](https://med-technews.com/news/latest-medtech-news/Ceras-ai-led-social-care-model-to-potentially-save-nhs-and-uk-government-1-billion-per-year); [City AM](https://www.cityam.com/?p=2297543); [TechCrunch](https://techcrunch.com/?p=1632310), accessed 2026-10-31 → `/clinic/wearable-home-device-signals` - `F-75031eafdd` L61 _1. Competitors › 1.2 AI clinical intelligence vendors_: Vendor / product / What it offers / Agents? / Regulatory position / Label / Source → `/clinic/wearable-home-device-signals` - `F-3b97725962` L63 [CLAIM] _1. Competitors › 1.2 AI clinical intelligence vendors_: **OpenAI – ChatGPT Health / "Health in ChatGPT"** / Launched 7 January 2026 as a separate space where users connect medical records (via b.well) and wellness apps. OpenAI dated a relaunch to 23 July 2026, for US users only. / Consumer assistant. By default it asks before it uses connected records. / OpenAI says it "does not replace qualified medical care". Connected health data is not used to train foundation models. / CLAIM / [Fortune](https://fortune.com/2026/01/07/openai-launches-chatgpt-health-in-a-push-to-become-a-hub-for-personal-health-data); [Fierce Healthcare](https://fiercehealthcare.com/ai-and-machine-learning/openai-launches-chatgpt-health-connect-data-health-apps-medical-records); [OpenAI](https://openai.com/index/health-in-chatgpt/) (page blocked automated fetch), accessed 2026-10-31 → `/clinic/mental-health-care-crisis` - `F-6d3a660c1e` L64 [CLAIM] _1. Competitors › 1.2 AI clinical intelligence vendors_: **Anthropic – Claude for Healthcare** / Launched January 2026. Connectors for the CMS (US Centers for Medicare & Medicaid Services) Coverage Database, ICD-10 and the NPI (National Provider Identifier) Registry, plus a FHIR skill and a prior-authorisation template. Consumer connectors (HealthEx, Function, Apple Health, Android Health Connect) are in beta. / Agent skills. / Described as "HIPAA-ready". Anthropic says it does not train on user health data. / CLAIM / [Anthropic](https://www.anthropic.com/news/healthcare-life-sciences); [MobiHealthNews](https://www.mobihealthnews.com/news/anthropic-launches-claude-healthcare); [The Register](https://www.theregister.com/2026/01/12/claude_anthropic_healthcare/), accessed 2026-10-31 → `/clinic/refresh-public-data-caches` - `F-cdbb6469c9` L65 [CLAIM] _1. Competitors › 1.2 AI clinical intelligence vendors_: **Microsoft – Dragon Copilot** (formerly Nuance DAX) / Ambient documentation. Generally available in the UK since September 2025 and hosted in a UK Azure data centre. / A marketplace of partner apps and agents inside Dragon Copilot. / Microsoft says DCB0129 is complete and assessed by a UK Clinical Safety Officer. / CLAIM / [Phoenix Software](https://www.phoenixs.co.uk/resources/blog/microsoft-dragon-copilot-is-now-generally-available-in-the-uk/); [Microsoft UK](https://www.microsoft.com/en-gb/health-solutions/clinical-workflow/dragon-copilot), accessed 2026-10-31 → `/clinic/medication-safety-check` - `F-11d0622a2c` L66 [CLAIM] _1. Competitors › 1.2 AI clinical intelligence vendors_: **Microsoft 365 Copilot for NHS England** / June 2026 agreement covering Microsoft 365 Copilot and Copilot Studio, so NHS England and trusts can build their own agents (for example, for complaints and FOI requests), governed through "Agent 365". / Administrative agents, not clinical. / n/a / CLAIM / [Microsoft](https://news.microsoft.com/source/2026/06/08/nhs-england-accelerates-ai-adoption-with-microsoft-365-copilot-to-improve-service-delivery-reduce-costs-and-create-more-time-for-care/); [TechMarketView](https://www.techmarketview.com/ukhotviews/archive/2026/06/08/nhs-england-scales-microsoft-copilot-adoption), accessed 2026-10-31 → `/clinic/person-held-health-record` - `F-1ac417b4cc` L67 [CLAIM] _1. Competitors › 1.2 AI clinical intelligence vendors_: **Google – AMIE and MedGemma** / AMIE is a research diagnostic dialogue agent, now extended to disease management. MedGemma is a family of open-weight medical models. / AMIE is multi-agent (a dialogue agent plus a management agent). It is still research. / Not a product. Studies are running with Beth Israel Deaconess and Included Health. / CLAIM / [Google blog](https://blog.google/innovation-and-ai/models-and-research/google-research/amie-for-disease-management-in-nature); [Google Research I/O 2025](https://www.research.google/blog/google-research-at-google-io-2025), accessed 2026-10-31 → `/clinic/price-policy-clinic-data` - `F-39153186e3` L68 [CLAIM] _1. Competitors › 1.2 AI clinical intelligence vendors_: **Abridge** / Ambient documentation. In August 2026 it expanded its "clinical intelligence agent" with decision support to more than 300 health systems. / Yes, for clinicians. / US-focused. / CLAIM / [HLTH](https://hlth.com/insights/news/abridge-expands-clinical-intelligence-ai-across-300-health-systems), accessed 2026-10-31 → `/clinic/mental-health-care-crisis` - `F-001b0c2e6f` L69 [UNKNOWN] _1. Competitors › 1.2 AI clinical intelligence vendors_: **Tortus, Heidi, Tandem** (UK ambient scribes) / Ambient scribes. A tracker reports Tortus as UKCA Class IIa (24 June 2026) and Tandem as EU MDR Class IIa. Heidi's status conflicts between sources. / Documentation, plus coding suggestions. / Mixed; see the regulator section. / UNKNOWN (single commercial blog family, with internal conflicts) / [iatroX tracker](https://www.iatrox.com/blog/ai-scribe-regulatory-status-tracker-2026), accessed 2026-10-31 → `/clinic/track-referral` - `F-232ece4dc8` L70 [CLAIM] _1. Competitors › 1.2 AI clinical intelligence vendors_: **NHS England AVT supplier registry** / A self-certified registry that is not a commercial framework and not an endorsement. It listed 19 suppliers in January 2026. / n/a / Suppliers must complete DTAC, plus other criteria. / CLAIM / [HTN](https://htn.co.uk/2026/01/19/nhs-england-lists-19-suppliers-meeting-criteria-for-ambient-voice-registry/); [Digital Health](https://www.digitalhealth.net/2026/01/nhs-england-launches-supplier-registry-for-ai-scribing-tech/), accessed 2026-10-31 → `/clinic/register-trial` - `F-c5e603f19a` L76 _1. Competitors › 1.3 Clinical decision support and drug-interaction data_: **BNF / BNFC** (British National Formulary, and the version for children) / The UK's standard prescribing reference, with A–Z interaction monographs. Free to NHS users via NICE (National Institute for Health and Care Excellence), UK only. / Non-NHS bodies need a licence from the BNF Partners. Merging BNF content into third-party systems needs a direct licence. RPS (Royal Pharmaceutical Society) offers a data feed or API. / OBSERVED / [Pharmaceutical Press licensing](https://www.pharmaceuticalpress.com/?p=146134); [NICE BNF UK only](https://www.nice.org.uk/bnf-uk-only), accessed 2026-10-31 → `/clinic/mental-health-care-crisis` - `F-44ed7432b4` L77 [CLAIM] _1. Competitors › 1.3 Clinical decision support and drug-interaction data_: **Stockley's Drug Interactions** / Monographs covering more than 5,000 interactions, with evidence, mechanism and management. Accessed by subscription through MedicinesComplete. / Subscription. / CLAIM / [SPS (NHS Specialist Pharmacy Service)](https://sps.nhs.uk/articles/drug-interactions-resources-to-support-answering-questions/), accessed 2026-10-31 → `/clinic/claim-edge-id-record` - `F-ed82c3eed9` L78 [CLAIM] _1. Competitors › 1.3 Clinical decision support and drug-interaction data_: **FDB Multilex / OptimiseRx** (First Databank UK) / FDB calls Multilex the UK's most widely used drug database and e-prescribing decision support (more than 10,000 sites). It covers drug–drug interactions, sensitivity and duplicate-therapy checks. OptimiseRx gives point-of-care prompts in EMIS Web. / Commercial. A 2026 NHS contract was awarded without competition because only FDB could supply it. / CLAIM / [SNOMED International marketplace](https://snomed.org/snomed-ct-marketplace/first-databank-uk-ltd.); [Find a Tender](https://www.find-tender.service.gov.uk/procurement/ocds-h6vhtk-06c6f3), accessed 2026-10-31 → `/clinic/social-prescribing-referral` - `F-41b2c674fe` L79 _1. Competitors › 1.3 Clinical decision support and drug-interaction data_: **NLM RxNav drug interaction API** (US) / A free US government interaction API. / Scheduled for shutdown on or about 2 January 2024. / OBSERVED (NLM notice) / [NLM](https://www.lhncbc.nlm.nih.gov/RxNav/news/APIUpdate.html); [DrugBank blog](https://blog.drugbank.com/nih-discontinues-their-drug-interaction-api/), accessed 2026-10-31 → **UNPLACED** - `F-1de1985cac` L80 [CLAIM] _1. Competitors › 1.3 Clinical decision support and drug-interaction data_: **DrugBank** / A drug knowledge base with interactions. / Open vocabulary data: CC0-1.0. Full data: CC-BY-NC-4.0 (academic, non-commercial). Commercial use needs a separate agreement. Several release pages show academic downloads "temporarily unavailable". / CLAIM / [DrugBank FAQ](https://dev.drugbank.com/guides/faqs); [DrugBank 5.1.22](https://go.drugbank.com/releases/5-1-22), accessed 2026-10-31 → `/clinic/price-policy-clinic-data` - `F-a4abe60ca5` L81 [UNKNOWN] _1. Competitors › 1.3 Clinical decision support and drug-interaction data_: **DDInter 2.0** / An academic open-access database: 2,310 drugs and 302,516 drug–drug interaction records (NAR, Nucleic Acids Research, 2025). / "Open-access", but no data licence found. / UNKNOWN (licence) / [NAR paper](https://academic.oup.com/nar/article/53/D1/D1356/7740584); [DDInter 2.0](https://ddinter2.scbdd.com/), accessed 2026-10-31 → **UNPLACED** - `F-134e62442c` L82 [UNKNOWN] _1. Competitors › 1.3 Clinical decision support and drug-interaction data_: **Medscape interaction checker** / Not researched in depth. No primary source checked. / — / UNKNOWN / — → **UNPLACED** - `F-7157075615` L84 _1. Competitors › 1.3 Clinical decision support and drug-interaction data_: **Gap for EDGE Clinic:** the brief asks for "an open clinical-grade drug-interaction source". On this evidence, none exists today. The clinical-grade sources (BNF, Stockley's, FDB) are licensed. The open ones (DrugBank open data, DDInter) are either non-commercial, of unclear licence, or not presented as clinical-grade. The free US government API was withdrawn. Sources as above, accessed 2026-10-31. → **UNPLACED** - `F-534f8e8493` L90 [CLAIM] _1. Competitors › 1.4 Retinal AI vendors_: **LumineticsCore** (Digital Diagnostics, formerly IDx-DR) / Autonomous detection of more-than-mild diabetic retinopathy. It was the first autonomous AI diagnostic authorised by the FDA (US Food and Drug Administration), via De Novo. / US authorised. / CLAIM (vendor) / [Digital Diagnostics](https://www.digitaldiagnostics.com/resources/insights/lumineticscore-engineered-by-digital-diagnostics-early-detection-of-diabetic-retinopathy-using-ai/), accessed 2026-10-31 → **UNPLACED** - `F-48f40cf0ed` L91 [CLAIM] _1. Competitors › 1.4 Retinal AI vendors_: **EyeArt** (Eyenuk) / Detects mild and vision-threatening diabetic retinopathy. FDA cleared in 2020, with more cameras added in 2023. / US cleared. / CLAIM / [Insight News](https://www.insightnews.com.au/fda-clearance-expands-eyenuk-ai-system-to-automatically-detect-diabetic-retinopathy/), accessed 2026-10-31 → **UNPLACED** - `F-303cecdce2` L92 [CLAIM] _1. Competitors › 1.4 Retinal AI vendors_: **AEYE-DS** (AEYE Health) / Reported as cleared on the Optomed Aurora handheld camera and the Topcon NW400. / Date not confirmed. / CLAIM / [ai2.work](https://ai2.work/blog/handheld-ai-eye-screening-clears-fda-inside-the-retina-shift), accessed 2026-10-31 → `/clinic/claim-edge-id-record` - `F-804755ccca` L93 _1. Competitors › 1.4 Retinal AI vendors_: **NHS Diabetic Eye Screening Programme (DESP)** / About 2.2 million screening appointments a year. The UK National Screening Committee consulted on AI in diabetic eye screening (May 2026). A Birmingham procurement framework favours AI as a first-pass triage, with specialist graders reviewing. / Outcome not found. No vendor confirmed in NHS DESP. / PROPOSED / [AOP](https://www.aop.org.uk/ot/news/2026/05/13/uk-national-screening-committee-consults-on-use-of-ai-in-diabetic-eye-screening); [University of Birmingham](https://www.birmingham.ac.uk/news/2026/birmingham-research-helps-shape-how-nhs-procures-ai-based-diabetic-eye-screening); [SGUL AIDED study](https://www.sgul.ac.uk/about/our-institutes/population-health/projects/aided-study), accessed 2026-10-31 → `/clinic/retinal-screening` - `F-4ca569c1a4` L94 [CLAIM] _1. Competitors › 1.4 Retinal AI vendors_: **Cascader** (Moorfields, UCL, Topcon) / A new company for AI on eye disease and "oculomics" (reading signs of other diseases, such as heart disease or dementia, from retinal images), with access to over 35 million images. / Early stage. / CLAIM / [Moorfields](https://www.moorfields.nhs.uk/about-us/news-and-blogs/news/new-company-launched-to-advance-ai-driven-eyecare), accessed 2026-10-31 → `/clinic/retinal-screening` - `F-2253860171` L95 [CLAIM] _1. Competitors › 1.4 Retinal AI vendors_: UK optometry pilots (Toku, Heart-Eye and others) / Small UK pilots of retinal AI for cardiovascular risk. / Pilots only. / CLAIM / [LOCSU case studies](https://locsu.co.uk/wp-content/uploads/2025/12/CVD-Case-Studies-AI-Powered-Retinal-Imaging-for-Cardiovascular-Risk-Assessment.pdf), accessed 2026-10-31 → `/clinic/retinal-screening` - `F-0ba9774cab` L99 _1. Competitors › 1.5 What the competitors leave open, and bare minimum versus ambitious_: These gaps come from the tables above, with sources as listed there. → `/clinic/retinal-screening` - `F-1e7ba7562f` L101 [UNKNOWN] _1. Competitors › 1.5 What the competitors leave open, and bare minimum versus ambitious_: **Nobody joins the steps "front door → detection → social prescribing referral".** The NHS App is national but is a router. Accurx, eConsult (Huma) and Rapid Health triage GP demand. Cera is home care. Retinal AI is a single test. EDGE Clinic's "come about anything" scope has no direct UK rival found. UNKNOWN whether "My NHS GP" will cover non-medical needs. → `/clinic/retinal-screening` - `F-e5dec444ef` L102 _1. Competitors › 1.5 What the competitors leave open, and bare minimum versus ambitious_: **AI in competitors is mostly ambient documentation**, which the MHRA now says is not a device when it only summarises ([MHRA](https://www.gov.uk/government/news/mhra-clarifies-regulatory-status-of-ambient-voice-technologies-used-in-the-nhs), accessed 2026-10-31). Patient-facing agents that act (book, refer, check prescriptions) are rare and regulated. → `/clinic/front-door-check` - `F-ce074d7ac1` L103 _1. Competitors › 1.5 What the competitors leave open, and bare minimum versus ambitious_: **None of the vendors found publishes an agent identity, mandate or audit model that a regulator could re-run.** Microsoft's "Agent 365" governance is administrative only ([TechMarketView](https://www.techmarketview.com/ukhotviews/archive/2026/06/08/nhs-england-scales-microsoft-copilot-adoption), accessed 2026-10-31). → `/clinic/price-policy-clinic-data` - `F-60c9d0aaf5` L107 _1. Competitors › 1.5 What the competitors leave open, and bare minimum versus ambitious_: Front door / NHS login, DTAC, DCB0129, DSPT, signposting only (not a device) / An agent that triages and acts under a mandate the patient can see, with every step replayable by a regulator → `/clinic/price-policy-clinic-data` - `F-242b7a81b6` L108 _1. Competitors › 1.5 What the competitors leave open, and bare minimum versus ambitious_: Detection / Buy a cleared retinal AI and keep a human grader in the loop / Multi-signal detection with staged ("L-plate") authorisation and a public post-market safety record → `/clinic/price-policy-clinic-data` - `F-1ad40be3f2` L109 _1. Competitors › 1.5 What the competitors leave open, and bare minimum versus ambitious_: Prescribing safety / License BNF or FDB through the EHR supplier / Contribute to an open, versioned interaction knowledge base with provenance → `/clinic/price-policy-clinic-data` - `F-7050ceb429` L110 _1. Competitors › 1.5 What the competitors leave open, and bare minimum versus ambitious_: Agents / No autonomous actions; human signs off every action / Agent identity, mandate, consent and escalation recorded as FHIR Consent, Provenance and AuditEvent → `/clinic/price-policy-clinic-data` - `F-3e6b8088d3` L120 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: **MHRA – UK MDR 2002** (UK Medical Devices Regulations) / Software with a medical purpose is a medical device. Post-market surveillance rules (SI 2024/1368) apply to devices placed on the GB market from 16 June 2025. / In force since 16 June 2025 / AGREEMENT / [GOV.UK](https://www.gov.uk/government/news/first-major-overhaul-of-medical-device-regulation-comes-into-force-across-great-britain), accessed 2026-10-31 → `/clinic/classify-software-medical-device` - `F-89b105d7a3` L121 _2. Regulators and rules › 2.1 United Kingdom_: MHRA – draft Medical Devices (Amendment) Regulations 2026 / New pre-market rules and an "international reliance" route, which lets devices authorised in Australia, Canada or the US register in GB. Software-specific provisions are reportedly not updated yet. / Draft published 8 May 2026. Expected to be adopted December 2026 and to come into force June 2027, subject to Parliament. / PROPOSED / [Latham & Watkins](https://www.lw.com/en/insights/mhra-publishes-draft-amendment-to-the-uk-medical-devices-regulations); [Fieldfisher](https://www.fieldfisher.com/en/insights/new-pre-market-requirements-for-medical-devices-mhra-publishes-draft-legislation---what%e2%80%99s-changed); [Qualio](https://www.qualio.com/blog/mhra-s-2026-devices-reform-what-changes-and-when), accessed 2026-10-31 → `/clinic/classify-software-medical-device` - `F-6c41b8d8f0` L122 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: MHRA – AVT (ambient voice technology) guidance / Transcription, summarising or drafting letters alone: not a device. Supporting diagnosis, treatment or prevention, or taking automated action without clinician review: a device. / Published 29 July 2026 / AGREEMENT (guidance, does not change law) / [MHRA news](https://www.gov.uk/government/news/mhra-clarifies-regulatory-status-of-ambient-voice-technologies-used-in-the-nhs) (OBSERVED); [MHRA PDF](https://assets.publishing.service.gov.uk/media/6a68daaff23ff62c51c1a7a3/MHRA_Guidance_on_Ambient_Voice_Technology-Enabled_Products.pdf), accessed 2026-10-31 → `/clinic/front-door-check` - `F-0569d3e11b` L123 _2. Regulators and rules › 2.1 United Kingdom_: MHRA – PCCP guidance for AI devices (UK) / Reported plan to issue draft guidance on predetermined change control plans by end 2026. A PCCP is a plan, agreed in advance, for how a model may change after approval. / Not published / PROPOSED / [Qualio](https://www.qualio.com/blog/mhra-s-2026-devices-reform-what-changes-and-when), accessed 2026-10-31 → `/clinic/issue-field-safety-notice` - `F-cacb289f02` L124 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: MHRA – AI Airlock / A regulatory sandbox for AI as a medical device. Phase 2 had 7 products, including note-taking and eye disease detection, and ran April 2025 to May 2026. DHSC (Department of Health and Social Care) funds £1.2m a year for 2026–2029. / Phase 2 reports out July 2026. Phase 3 not confirmed. / AGREEMENT (programme); UNKNOWN (Phase 3) / [GOV.UK collection](https://www.gov.uk/government/collections/ai-airlock-the-regulatory-sandbox-for-aiamd); [Phase 2 report](https://www.gov.uk/government/publications/ai-airlock-sandbox-phase-2-programme-report); [MHRA blog](https://medregs.blog.gov.uk/2026/06/09/advancing-ai-regulation-in-healthcare-insights-from-ai-airlock-phase-2/), accessed 2026-10-31 → `/clinic/register-device-fault` - `F-13c42af2f1` L125 [UNKNOWN] _2. Regulators and rules › 2.1 United Kingdom_: National Commission into the Regulation of AI in Healthcare / 44 recommendations, including staged authorisations ("L-plates"), continuous real-world monitoring, a public searchable safety database, and stronger MHRA enforcement. / Report 10 September 2026. GOV.UK says a "formal response to follow". One law firm says government accepted all 44 on 6 October 2026. / PROPOSED; UNKNOWN (acceptance) / [GOV.UK](https://www.gov.uk/government/news/independent-commission-led-by-nhs-doctors-sets-out-blueprint-to-accelerate-safe-ai-adoption-in-healthcare) (OBSERVED); [Report](https://www.gov.uk/government/publications/national-commission-into-the-regulation-of-ai-in-healthcare-recommendations-for-a-future-regulatory-framework); [CMS](https://cms.law/en/gbr/legal-updates/from-pre-market-to-lifecycle-the-national-commission-s-44-recommendations-on-ai-regulation-in-healthcare); [Ropes & Gray](https://www.ropesgray.com/en/insights/viewpoints/2026/10/102o5fp/ai-medical-devices-in-great-britain-what-the-mhras-response-to-the-national-com), accessed 2026-10-31 → `/clinic/social-prescribing-referral` - `F-f115959ca0` L126 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: **NHS England – DCB0129** (the manufacturer's clinical risk management) and **DCB0160** (the deployer's) / Mandatory clinical safety standards for health IT in England. They require a named Clinical Safety Officer, a hazard log and a safety case. / Current versions in force. Revision consultation ran 29 June–11 September 2026, with AI named as a driver. / AGREEMENT (current); PROPOSED (revision) / [NHS England consultation](https://www.engage.england.nhs.uk/patient-safety/national-review-of-clinical-risk-management-standa/) (OBSERVED); [DCB0129 standard](https://adultsocialcare.standards.nhs.uk/published-standards/clinical-risk-management-its-application-in-the-manufacture-of-health-it-systems); [DCB0160 standard](https://standards.nhs.uk/published-standards/clinical-risk-management-its-application-in-the-deployment-and-use-of-health-it-systems), accessed 2026-10-31 → `/clinic/inspect-hazard-log-safety` - `F-69610398cb` L127 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: **DTAC** (Digital Technology Assessment Criteria) / The NHS buyer's baseline checklist. It covers clinical safety, data protection, technical assurance, interoperability, and usability and accessibility. / New form from 24 February 2026, with about 25% fewer questions. The old form was retired on 6 April 2026. / AGREEMENT / [NHS Innovation Service](https://innovation.nhs.uk/news/updated-nhs-england-digital-technology-assessment-criteria-dtac-form-and-guidance/); [HTN](https://htn.co.uk/2026/02/25/digital-technology-assessment-criteria-refreshed-to-create-simpler-more-trusted-pathway-for-nhs-digital-innovation/), accessed 2026-10-31 → `/clinic/medication-safety-check` - `F-5584fac901` L128 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: **DSPT** (Data Security and Protection Toolkit) / An annual self-assessment that any organisation accessing NHS patient data must complete. Trusts and ICBs (integrated care boards) use a version aligned to the NCSC (National Cyber Security Centre) CAF (Cyber Assessment Framework). IT suppliers were reported to stay on the non-CAF version for 2025/26. / 2025/26 deadline 30 June 2026. Version 9 (2026/27) changes not checked. / AGREEMENT; UNKNOWN (2026/27 supplier detail) / [DSPT help](https://www.dsptoolkit.nhs.uk/Help/64); [Dionach](https://dionach.com/data-security-and-protection-toolkit-dspt-2025-2026-caf/), accessed 2026-10-31 → `/clinic/complete-data-security` - `F-8cf03b70a5` L129 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: NHS England – ambient scribing guidance and AVT registry / Guidance first published April 2025. NHS England withdrew its earlier statement that summarisation tools are likely devices, to match the MHRA. The registry launched January 2026. / Updated July 2026 / AGREEMENT (guidance) / [NHS England guidance](https://www.england.nhs.uk/long-read/guidance-on-the-use-of-ai-enabled-ambient-scribing-products-in-health-and-care-settings); [HTN](https://htn.co.uk/2026/07/29/mhra-clarifies-regulatory-medical-devices-status-of-ambient-voice-technologies/), accessed 2026-10-31 → `/clinic/claim-edge-id-record` - `F-c80145f4dd` L130 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: **CQC** (Care Quality Commission) / Registers providers of "regulated activities". A 2022 scope update says AI suppliers that analyse and report on screening images may need to register for "diagnostic and screening procedures" if no CQC-registered provider is doing so. / No AI-specific 2025–2026 CQC guidance found / AGREEMENT (2022 scope); UNKNOWN (newer) / [Osborne Clarke](https://osborneclarke.com/insights/updated-cqc-guidance-has-implications-health-technology-and-wellbeing-industries); [NHS digital regulations guide](https://www.digitalregulations.innovation.nhs.uk/regulations-and-guidance-for-developers/all-developers-guidance/regulated-activities-check-if-you-need-to-register-with-the-care-quality-commission-cqc/pdf/), accessed 2026-10-31 → `/clinic/retinal-screening` - `F-ca3348e347` L131 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: **NICE – ESF** (Evidence Standards Framework) for digital health technologies / Sets evidence tiers in proportion to risk, and includes AI and data-driven tools. A further revision is expected to follow MHRA reforms. Early Value Assessments let technologies be used while evidence is gathered. / ESF in use; next update pending / AGREEMENT; PROPOSED (update) / [Fieldfisher](https://www.fieldfisher.com/en/insights/new-changes-to-evidence-standards-framework-and-digital-health-technologies); [JMIR viewpoint](https://mhealth.jmir.org/2025/1/e67435); [ISPOR poster](https://www.ispor.org/heor-resources/presentations-database/presentation-cti/ispor-europe-2026/poster-session-5-4/learning-while-using-evidence-expectations-under-nice-s-reformed-healthtech-program), accessed 2026-10-31 → `/clinic/request-evidence-against` - `F-191c70cf28` L132 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: **ICO** (Information Commissioner's Office) – UK GDPR, AI guidance / A DPIA (data protection impact assessment) is needed for AI processing health data at scale. AI inferences about health can count as "special category" data. / AI guidance updated March 2023; a 2026 refresh was not confirmed / AGREEMENT / [ICO guidance](https://www.ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/artificial-intelligence/guidance-on-ai-and-data-protection); [Hunton](https://www.hunton.com/privacy-and-cybersecurity-law-blog/uk-ico-issues-updated-guidance-on-ai-and-data-protection); [NHS DPIA step](https://digitalregulations.innovation.nhs.uk/regulations-and-guidance-for-developers/all-developers-guidance/how-to-comply-with-the-uk-gdpr-as-a-developer/step-2-do-a-data-protection-impact-assessment-dpia/pdf/), accessed 2026-10-31 → `/clinic/run-data-protection-impact` - `F-8d88159991` L133 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: Data (Use and Access) Act 2025 – automated decision-making / Most reforms came into force on 5 February 2026. Solely automated significant decisions are now restricted mainly where special category data (such as health) is involved, so health stays strict. / In force 5 February 2026. ICO ADM guidance was consulted on in 2026. / AGREEMENT; PROPOSED (ICO guidance) / [GOV.UK](https://www.gov.uk/guidance/data-use-and-access-act-2025-data-protection-and-privacy-changes); [Travers Smith](https://www.traverssmith.com/knowledge/knowledge-container/uks-data-protection-reforms-take-effect-a-new-era-for-automated-decision-making/); [Kennedys](https://www.kennedyslaw.com/en/thought-leadership/article/2026/the-data-use-and-access-act-2025-commencement-dates-and-planned-guidance-for-2026), accessed 2026-10-31 → `/clinic/person-held-health-record` - `F-4573a6ae5d` L134 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: **Caldicott Principles / National Data Guardian (NDG)** / Eight principles for using confidential patient information, listed as legally required in the NHS digital regulations service. The NDG's 2024/25 report mentions input to AI scribe guidance. / Principles in force. No NDG AI-specific guidance found. / AGREEMENT; UNKNOWN (AI guidance) / [NHS digital regulations](https://digitalregulations.innovation.nhs.uk/regulations-and-guidance-for-adopters/all-adopters-guidance/complying-with-the-uk-gdpr-steps-1-7-an-introduction/step-7-follow-the-8-caldicott-principles/pdf/); [NDG annual report](https://assets.publishing.service.gov.uk/media/6942a76f9273c48f554cf506/NDG_annual_report_2024-25_v1.0_-_17.12.25.pdf), accessed 2026-10-31 → `/clinic/report-data-breach` - `F-31c00f3382` L135 [AGREEMENT] _2. Regulators and rules › 2.1 United Kingdom_: **HRA** (Health Research Authority) / HRA approval covers project-based research in the NHS in England and Wales. Service evaluation may not need ethics review; use the HRA decision tool. CAG (Confidentiality Advisory Group) approval may be needed for identifiable data without consent. / In force / AGREEMENT / [HRA approval](https://www.hra.nhs.uk/approvals-amendments/what-approvals-do-i-need/hra-approval/); [HRA: what approvals](https://www.hra.nhs.uk/approvals-amendments/what-approvals-do-i-need/), accessed 2026-10-31 → `/clinic/front-door-check` - `F-9b7c220def` L141 [AGREEMENT] _2. Regulators and rules › 2.2 European Union_: **EU MDR** (Medical Device Regulation 2017/745) / The device rules in the EU. Its classification decides whether the AI Act's high-risk rules apply. / In force / AGREEMENT / [EUR-Lex](https://eur-lex.europa.eu/eli/reg/2017/745/oj), accessed 2026-10-31 → `/clinic/classify-software-medical-device` - `F-e70c2d6f90` L142 [AGREEMENT] _2. Regulators and rules › 2.2 European Union_: **EU AI Act** (Regulation 2024/1689) / Risk-based AI law. AI that is a medical device, or a safety component of one, and needs notified-body assessment is "high-risk". / Original high-risk date for Annex I products was 2 August 2027, now moved (next row) / AGREEMENT / [EUR-Lex](https://eur-lex.europa.eu/eli/reg/2024/1689/oj); [OpenRegulatory on MDCG 2025-6](https://openregulatory.com/mdcg/mdcg-2025-6), accessed 2026-10-31 → `/clinic/act-high-risk-answer` - `F-567669e045` L143 [AGREEMENT] _2. Regulators and rules › 2.2 European Union_: **Digital Omnibus on AI** (Regulation 2026/1744) / Moves the high-risk deadlines. Annex III (standalone) to 2 December 2027; Annex I (AI in regulated products such as medical devices) to 2 August 2028. Transparency obligations start 2 August 2026, and some content-marking duties 2 December 2026. / In force 27 July 2026 / AGREEMENT / [Hunton](https://www.hunton.com/privacy-and-cybersecurity-law-blog/eu-digital-omnibus-on-ai-enters-into-force) (OBSERVED, law firm); [Gibson Dunn](https://www.gibsondunn.com/eu-ai-act-omnibus-agreement-postponed-high-risk-deadlines-and-other-key-changes/), accessed 2026-10-31 → `/clinic/act-high-risk-answer` - `F-15342eb89d` L144 [AGREEMENT] _2. Regulators and rules › 2.2 European Union_: **MDCG 2025-6** (FAQ from the Medical Device Coordination Group and the AI Board) / Explains how the MDR and the AI Act fit together. The MDR class drives AI Act applicability, and the notified body assesses both together. / Published June 2025 (exact date disputed) / AGREEMENT (guidance) / [OpenRegulatory](https://openregulatory.com/mdcg/mdcg-2025-6), accessed 2026-10-31 → `/clinic/manage-patient-group-directions` - `F-3d68ee40ac` L145 [AGREEMENT] _2. Regulators and rules › 2.2 European Union_: **EHDS** (European Health Data Space, Regulation 2025/327) / Patient access to records, EHR (electronic health record) system requirements and secondary use of data. / In force 26 March 2025. Main application 2027–2031 in phases. / AGREEMENT (phased) / [Kennedys](https://www.kennedyslaw.com/en/thought-leadership/article/2026/the-european-health-data-space-is-in-force-implications-for-healthcare-medtech-and-life-sciences); [A&O Shearman](https://www.aoshearman.com/en/insights/ao-shearman-on-life-sciences/european-health-data-space-regulation-sets-requirements-for-electronic-health-record-systems), accessed 2026-10-31 → `/clinic/symptom-check-triage` - `F-b4add08f8c` L146 [AGREEMENT] _2. Regulators and rules › 2.2 European Union_: **Product Liability Directive** (2024/2853) / Software and AI become "products" under strict liability. Courts can presume a product is defective where technical complexity makes proof too hard. / Member states must transpose it by 9 December 2026 / AGREEMENT (transposition due) / [Covington Inside Privacy](https://www.insideprivacy.com/uncategorized/eu-member-states-begin-rolling-out-new-product-liability-rules/); [Hogan Lovells](https://www.hoganlovells.com/en/publications/eu-introduces-comprehensive-digitalera-product-liability-directive), accessed 2026-10-31 → `/clinic/cancel-clinic-session` - `F-976ce30751` L152 [AGREEMENT] _2. Regulators and rules › 2.3 United States_: FDA – Clinical Decision Support (CDS) software guidance / Revised 6 January 2026 without a comment period. It relaxes oversight of some tools that recommend a single option to a clinician. / Final guidance (replaces 2022) / AGREEMENT / [King & Spalding](https://www.kslaw.com/insights/articles/fda-updates-general-wellness-and-clinical-decision-support-guidance-documents); [Arnold & Porter](https://www.arnoldporter.com/en/perspectives/advisories/2026/01/fda-cuts-red-tape-on-clinical-decision-support-software), accessed 2026-10-31 → **UNPLACED** - `F-2ce2701823` L153 [AGREEMENT] _2. Regulators and rules › 2.3 United States_: FDA – General Wellness guidance / Revised 6 January 2026. Adds criteria for non-invasive sensors, such as wrist-worn blood pressure estimates. / Final / AGREEMENT / [King & Spalding](https://www.kslaw.com/insights/articles/fda-updates-general-wellness-and-clinical-decision-support-guidance-documents), accessed 2026-10-31 → **UNPLACED** - `F-b911005bc2` L154 [AGREEMENT] _2. Regulators and rules › 2.3 United States_: FDA – PCCP for AI-enabled device software functions / Lets a maker pre-agree how a model will change: description of the changes, a modification protocol and an impact assessment. / Final, December 2024 / AGREEMENT / [Ropes & Gray](https://ropesgray.com/en/insights/alerts/2024/12/fda-finalizes-guidance-on-predetermined-change-control-plans-for-ai-enabled-device), accessed 2026-10-31 → `/clinic/register-imaging-device` - `F-05b35a1316` L155 _2. Regulators and rules › 2.3 United States_: FDA – AI-enabled device software functions: lifecycle and submissions / Draft from January 2025. Finalising it is a top priority on the FY2027 (US fiscal year 2027) agenda. / Draft / PROPOSED / [RAPS](https://www.raps.org/resource/fda-s-device-center-releases-guidance-agenda-for-fy-2027.html); [CAP](https://www.cap.org/advocacy/latest-news-and-practice-data/fda-releases-new-artificial-intelligence-draft-guidance), accessed 2026-10-31 → `/clinic/classify-software-medical-device` - `F-cd0a41f910` L156 [CLAIM] _2. Regulators and rules › 2.3 United States_: FDA – AI-enabled device list / FDA lists over 1,600 AI-enabled devices and says the list is not comprehensive. A commentator notes none of the authorised devices it cites are generative or agentic. / Live list / OBSERVED (page exists); CLAIM (count, commentary) / [FDA list](https://www.fda.gov/medical-devices/artificial-intelligence-enabled-medical-devices/list-artificial-intelligence-enabled-medical-devices); [Clearwater](https://clearwatersecurity.com/transcripts/agentic-ai-in-healthcare-navigating-regulatory-uncertainty-and-building-governance-that-lasts/), accessed 2026-10-31 → `/clinic/see-day` - `F-d96310492f` L157 _2. Regulators and rules › 2.3 United States_: **HIPAA** (Health Insurance Portability and Accountability Act) Security Rule update / The proposed overhaul (published January 2025) is not final. HHS (Department of Health and Human Services) moved the target to July 2027. The current rule still applies. / Proposed / PROPOSED / [Fierce Healthcare](https://www.fiercehealthcare.com/health-tech/feds-push-back-hipaa-security-rule-overhaul-july-2027); [Davis Wright Tremaine](https://www.dwt.com/blogs/privacy--security-law-blog/2026/07/hhs-updates-hipaa-rulemaking-timeframes), accessed 2026-10-31 → `/clinic/apply-treatment-rule` - `F-dca458eefa` L158 _2. Regulators and rules › 2.3 United States_: ONC (Office of the National Coordinator for Health IT) – Interoperability Standards Platform AI appendix / Describes MCP (Model Context Protocol) over FHIR, A2A (agent-to-agent) standards, human sign-off on writes, and audit logging that links AI requests to users, agents and organisations. Says identity and delegation are "still evolving". / Reference content, no binding status / OBSERVED / [ONC ISP](https://isp.healthit.gov/appendix-vi-ai-interoperability), accessed 2026-10-31 → `/clinic/person-held-health-record` - `F-888c904182` L164 [AGREEMENT] _2. Regulators and rules › 2.4 Worldwide_: WHO guidance on LMMs (large multi-modal models) for health / More than 40 recommendations, including that governments assign a regulator to approve LMMs used in care, and independent post-release audits for large-scale deployments. / Published January 2024 / AGREEMENT (non-binding guidance) / [WHO](https://www.who.int/publications-detail-redirect/9789240084759), accessed 2026-10-31 → `/clinic/mental-health-care-crisis` - `F-4a3ebbd7fd` L168 _2. Regulators and rules › 2.5 Applying the MHRA test to EDGE Clinic (research reading, not legal advice)_: EDGE Clinic function / Likely device? / Why / Source → `/clinic/mental-health-care-crisis` - `F-ec69ae4d07` L170 [UNKNOWN] _2. Regulators and rules › 2.5 Applying the MHRA test to EDGE Clinic (research reading, not legal advice)_: Signposting to a social prescribing service, with no clinical claim / Probably not (UNKNOWN until assessed) / It does not "support diagnosis, treatment or prevention" / [MHRA](https://www.gov.uk/government/news/mhra-clarifies-regulatory-status-of-ambient-voice-technologies-used-in-the-nhs), accessed 2026-10-31 → `/clinic/social-prescribing-referral` - `F-b1ab4bde46` L171 _2. Regulators and rules › 2.5 Applying the MHRA test to EDGE Clinic (research reading, not legal advice)_: Triage, including online A&E (accident and emergency) routing / Likely yes / It supports diagnosis or decisions about urgency / same → `/clinic/social-prescribing-referral` - `F-ec9c53e611` L172 _2. Regulators and rules › 2.5 Applying the MHRA test to EDGE Clinic (research reading, not legal advice)_: Retinal screening AI / Yes / A diagnostic or screening purpose / same; [FDA LumineticsCore example](https://www.digitaldiagnostics.com/resources/insights/lumineticscore-engineered-by-digital-diagnostics-early-detection-of-diabetic-retinopathy-using-ai/), accessed 2026-10-31 → `/clinic/retinal-screening` - `F-a5a17e9a12` L173 _2. Regulators and rules › 2.5 Applying the MHRA test to EDGE Clinic (research reading, not legal advice)_: Prescribing safety (interaction checks) / Likely yes / It supports treatment decisions / [MHRA](https://www.gov.uk/government/news/mhra-clarifies-regulatory-status-of-ambient-voice-technologies-used-in-the-nhs), accessed 2026-10-31 → `/clinic/issue-field-safety-notice` - `F-f6d412b93d` L174 _2. Regulators and rules › 2.5 Applying the MHRA test to EDGE Clinic (research reading, not legal advice)_: Any agent that books, orders or refers without clinician review / Yes / "take automated action such as placing orders without clinician review" / same (OBSERVED quote) → `/clinic/issue-field-safety-notice` - `F-ae62034329` L180 [AGREEMENT] _2. Regulators and rules › 2.6 Date line (in force versus coming)_: 26 March 2025 / EHDS in force (applies in phases from 2027) / AGREEMENT / [Kennedys](https://www.kennedyslaw.com/en/thought-leadership/article/2026/the-european-health-data-space-is-in-force-implications-for-healthcare-medtech-and-life-sciences), accessed 2026-10-31 → **UNPLACED** - `F-df25558bdf` L181 [AGREEMENT] _2. Regulators and rules › 2.6 Date line (in force versus coming)_: 16 June 2025 / GB post-market surveillance rules apply / AGREEMENT / [GOV.UK](https://www.gov.uk/government/news/first-major-overhaul-of-medical-device-regulation-comes-into-force-across-great-britain), accessed 2026-10-31 → **UNPLACED** - `F-c4661bee12` L182 [AGREEMENT] _2. Regulators and rules › 2.6 Date line (in force versus coming)_: 6 January 2026 / FDA CDS and wellness guidance revised / AGREEMENT / [King & Spalding](https://www.kslaw.com/insights/articles/fda-updates-general-wellness-and-clinical-decision-support-guidance-documents), accessed 2026-10-31 → **UNPLACED** - `F-2cad20c062` L183 [AGREEMENT] _2. Regulators and rules › 2.6 Date line (in force versus coming)_: 5 February 2026 / UK DUAA automated decision-making reforms in force / AGREEMENT / [Travers Smith](https://www.traverssmith.com/knowledge/knowledge-container/uks-data-protection-reforms-take-effect-a-new-era-for-automated-decision-making/), accessed 2026-10-31 → `/clinic/retinal-screening` - `F-455779f8ea` L184 [AGREEMENT] _2. Regulators and rules › 2.6 Date line (in force versus coming)_: 6 April 2026 / Old DTAC form retired / AGREEMENT / [NHS Innovation Service](https://innovation.nhs.uk/news/updated-nhs-england-digital-technology-assessment-criteria-dtac-form-and-guidance/), accessed 2026-10-31 → **UNPLACED** - `F-1d833a2131` L185 [AGREEMENT] _2. Regulators and rules › 2.6 Date line (in force versus coming)_: 27 July 2026 / EU Digital Omnibus on AI in force / AGREEMENT / [Hunton](https://www.hunton.com/privacy-and-cybersecurity-law-blog/eu-digital-omnibus-on-ai-enters-into-force), accessed 2026-10-31 → **UNPLACED** - `F-5e5817458a` L186 [AGREEMENT] _2. Regulators and rules › 2.6 Date line (in force versus coming)_: 29 July 2026 / MHRA AVT guidance / AGREEMENT / [MHRA](https://www.gov.uk/government/news/mhra-clarifies-regulatory-status-of-ambient-voice-technologies-used-in-the-nhs), accessed 2026-10-31 → **UNPLACED** - `F-8b88a07be6` L187 [AGREEMENT] _2. Regulators and rules › 2.6 Date line (in force versus coming)_: 2 August 2026 / EU AI Act transparency obligations / AGREEMENT / [Hunton](https://www.hunton.com/privacy-and-cybersecurity-law-blog/eu-digital-omnibus-on-ai-enters-into-force), accessed 2026-10-31 → **UNPLACED** - `F-9806aadf4b` L188 [AGREEMENT] _2. Regulators and rules › 2.6 Date line (in force versus coming)_: 9 December 2026 / EU Product Liability Directive transposition deadline / AGREEMENT / [Covington](https://www.insideprivacy.com/uncategorized/eu-member-states-begin-rolling-out-new-product-liability-rules/), accessed 2026-10-31 → **UNPLACED** - `F-f4a42759fc` L189 _2. Regulators and rules › 2.6 Date line (in force versus coming)_: End 2026 / MHRA draft PCCP guidance (planned) / PROPOSED / [Qualio](https://www.qualio.com/blog/mhra-s-2026-devices-reform-what-changes-and-when), accessed 2026-10-31 → **UNPLACED** - `F-84a3dd9374` L190 _2. Regulators and rules › 2.6 Date line (in force versus coming)_: June 2027 / UK Medical Devices (Amendment) Regulations expected to apply / PROPOSED / [Latham & Watkins](https://www.lw.com/en/insights/mhra-publishes-draft-amendment-to-the-uk-medical-devices-regulations), accessed 2026-10-31 → **UNPLACED** - `F-55898fe301` L191 _2. Regulators and rules › 2.6 Date line (in force versus coming)_: July 2027 / HIPAA Security Rule final (target) / PROPOSED / [Fierce Healthcare](https://www.fiercehealthcare.com/health-tech/feds-push-back-hipaa-security-rule-overhaul-july-2027), accessed 2026-10-31 → **UNPLACED** - `F-5b4bd35ef6` L192 [AGREEMENT] _2. Regulators and rules › 2.6 Date line (in force versus coming)_: 2 December 2027 / EU AI Act Annex III high-risk rules apply / AGREEMENT / [Hunton](https://www.hunton.com/privacy-and-cybersecurity-law-blog/eu-digital-omnibus-on-ai-enters-into-force), accessed 2026-10-31 → `/clinic/act-high-risk-answer` - `F-60324e521e` L193 _2. Regulators and rules › 2.6 Date line (in force versus coming)_: April 2028 / 2028 / NHS App "full front door" and My NHS GP full rollout (targets) / PROPOSED / [Full Fact](https://fullfact.org/government-tracker/nhs-app-improvements/); [Digital Health](https://www.digitalhealth.net/2025/07/what-does-the-10-year-plan-mean-for-digital/), accessed 2026-10-31 → `/clinic/front-door-check` - `F-6c8e754b60` L194 [AGREEMENT] _2. Regulators and rules › 2.6 Date line (in force versus coming)_: 2 August 2028 / EU AI Act high-risk rules for AI in medical devices apply / AGREEMENT / [Hunton](https://www.hunton.com/privacy-and-cybersecurity-law-blog/eu-digital-omnibus-on-ai-enters-into-force), accessed 2026-10-31 → `/clinic/act-high-risk-answer` - `F-31c4cec2aa` L200 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 3. Agent assurance gaps in health_: "Agent" here means software that acts with or for a patient (books, refers, checks, messages), not only one that answers questions. → `/clinic/act-high-risk-answer` - `F-f01fd7fbf9` L204 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 3. Agent assurance gaps in health_: **Agent identity** (who or what is acting) / The OpenID Foundation names "agent identity fragmentation", "user impersonation" and "recursive delegation" as gaps ([OIDF](https://openid.net/new-whitepaper-tackles-ai-agent-identity-challenges/)). A 2026 survey says no production-ready standard traces authorisation through multi-hop agent chains ([arXiv 2604.23280](https://arxiv.org/pdf/2604.23280)). Accessed 2026-10-31. / OAuth token exchange "on-behalf-of" flows, and the OIDF's draft OpenID Connect Authority Claims Extension ([OIDF at ITU-T](https://openid.net/oidf-presents-at-itu-t-workshop-on-digital-id-and-agentic-ai/), accessed 2026-10-31). The FHIR `Device` resource can represent an AI system ([HL7 AI Transparency](https://hl7.org/fhir/uv/aitransparency/2026Jan/artifacts.html), accessed 2026-10-31). / A first-class agent identity issued by EDGE ID and bound to a human or organisation principal / PROPOSED (standards); OBSERVED (gap) → `/clinic/connect-lims` - `F-ff7090220a` L205 [AGREEMENT] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 3. Agent assurance gaps in health_: **Patient identity and verification** / NHS login does not do session management or logout, and does not replace PDS (the Personal Demographics Service) ([NHS login, archived](https://tnaqa.mirrorweb.com/ukgwa/20260505142011mp_/https://digital.nhs.uk/developer/api-catalogue/nhs-login), accessed 2026-10-31). / NHS login uses OpenID Connect, with P0/P5/P9 verification levels (P9 means checked against photo ID) / Use NHS login P9 for record access, and keep the agent session separate / AGREEMENT → `/clinic/register-child-patient` - `F-7c77116d41` L206 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 3. Agent assurance gaps in health_: **Mandate** (what the agent may do, for whom, until when) / A health-system preprint warns of "tool permissions that persist beyond the original use case" and "unclear accountability" ([arXiv 2601.15630](https://arxiv.org/pdf/2601.15630), accessed 2026-10-31). ONC says identity, authorisation and delegation are "still evolving" ([ONC ISP](https://isp.healthit.gov/appendix-vi-ai-interoperability), accessed 2026-10-31). / SMART on FHIR scopes ([SMART App Launch 2.2](https://hl7.org/fhir/smart-app-launch/apps.html), accessed 2026-10-31). No health standard for agent mandates. / A machine-readable mandate (scope, purpose, expiry, revocation) checked on every action / OBSERVED (gap) → `/clinic/person-held-health-record` - `F-9440a69925` L207 [AGREEMENT] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 3. Agent assurance gaps in health_: **Consent** / The OIDF names "consent fatigue" ([OIDF](https://openid.net/new-whitepaper-tackles-ai-agent-identity-challenges/), accessed 2026-10-31). OpenAI asks before using connected records by default ([OpenAI via Fierce](https://fiercehealthcare.com/ai-and-machine-learning/openai-launches-chatgpt-health-connect-data-health-apps-medical-records), accessed 2026-10-31). The UK's solely-automated-decision limits remain where health data is involved ([Travers Smith](https://www.traverssmith.com/knowledge/knowledge-container/uks-data-protection-reforms-take-effect-a-new-era-for-automated-decision-making/), accessed 2026-10-31). / The FHIR Consent resource (part of the FHIR R4 base specification, [HL7 FHIR R4 licence page](https://hl7.org/fhir/R4/license.html), accessed 2026-10-31). The Caldicott principles. / Consent captured per purpose and per agent, shown to the patient and revocable / AGREEMENT (law); OBSERVED (gap in practice) → `/clinic/connect-lims` - `F-663dcd3200` L208 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 3. Agent assurance gaps in health_: **Escalation to a human** / People using LLMs did no better than controls, with a "two-way communication breakdown" ([Oxford](https://www.ox.ac.uk/news/2026-02-10-new-study-warns-risks-ai-chatbots-giving-medical-advice), accessed 2026-10-31). ECRI's No. 1 hazard for 2026 is misuse of AI chatbots ([ECRI](https://home.ecri.org/blogs/ecri-news/misuse-of-ai-chatbots-tops-annual-list-of-health-technology-hazards), accessed 2026-10-31). The MHRA treats action "without clinician review" as device territory ([MHRA](https://www.gov.uk/government/news/mhra-clarifies-regulatory-status-of-ambient-voice-technologies-used-in-the-nhs), accessed 2026-10-31). / DCB0129 hazard logs. ONC's pattern of "write proposals" with human approval ([ONC ISP](https://isp.healthit.gov/appendix-vi-ai-interoperability), accessed 2026-10-31). / Hard-coded red-flag escalation, human sign-off on any clinical action, and user testing before launch (as the Oxford authors recommend) / OBSERVED (gap) → `/clinic/wearable-home-device-signals` - `F-2cb123a6f0` L209 [UNKNOWN] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 3. Agent assurance gaps in health_: **Liability** / No UK AI liability law was found. The Commission's report was summarised as patients wanting "clear accountability if things go wrong" ([GOV.UK](https://www.gov.uk/government/news/independent-commission-led-by-nhs-doctors-sets-out-blueprint-to-accelerate-safe-ai-adoption-in-healthcare), accessed 2026-10-31). In the EU, the new PLD (Product Liability Directive) makes software a product with strict liability ([Hogan Lovells](https://www.hoganlovells.com/en/publications/eu-introduces-comprehensive-digitalera-product-liability-directive), accessed 2026-10-31). Commentators say accountability for multi-step agent actions is "unsettled" ([Clearwater](https://clearwatersecurity.com/transcripts/agentic-ai-in-healthcare-navigating-regulatory-uncertainty-and-building-governance-that-lasts/), accessed 2026-10-31). / Device manufacturer liability, plus clinician duty of care. OpenMRS's licence puts responsibility on the clinician ([ScanCode](https://scancode-licensedb.aboutcode.org/openmrs-exception-to-mpl-2.0.html), accessed 2026-10-31). / A named accountable human or organisation for every agent mandate, written in the audit / UNKNOWN (UK law); AGREEMENT (EU PLD) → `/clinic/mental-health-care-crisis` - `F-56c71260b3` L210 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 3. Agent assurance gaps in health_: **Audit a regulator can re-run** / The Commission recommends continuous real-world monitoring and a public searchable safety database ([GOV.UK](https://www.gov.uk/government/news/independent-commission-led-by-nhs-doctors-sets-out-blueprint-to-accelerate-safe-ai-adoption-in-healthcare), accessed 2026-10-31). The WHO recommends independent post-release audits ([WHO](https://www.who.int/publications-detail-redirect/9789240084759), accessed 2026-10-31). No standard for replaying an agent's decisions was found. / HL7 "AI Transparency on FHIR": Provenance links AI output back to the model, the input prompt and the human verifier. It is a ballot/continuous build, not yet published ([HL7 AI Transparency](https://hl7.org/fhir/uv/aitransparency/2026Jan/artifacts.html); [requirements](https://build.fhir.org/ig/HL7/aitransparency-ig/en/requirements.html), accessed 2026-10-31). FDA PCCPs for planned model change ([Ropes & Gray](https://ropesgray.com/en/insights/alerts/2024/12/fda-finalizes-guidance-on-predetermined-change-control-plans-for-ai-enabled-device), accessed 2026-10-31). / A content-addressed log of inputs, model version, prompt, tools called, mandate and human sign-off, so a regulator can replay a case / PROPOSED (standard); OBSERVED (gap) → `/clinic/connect-lims` - `F-361786b334` L211 [CLAIM] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 3. Agent assurance gaps in health_: **Tool-calling standard** / MCP was reportedly donated to the Linux Foundation's Agentic AI Foundation in December 2025 ([Keragon](https://www.keragon.com/blog/medical-mcp-servers), accessed 2026-10-31). There is no official HL7 FHIR MCP profile, only community RFCs ([GitHub discussion](https://github.com/HeliosSoftware/hfs/discussions/231), accessed 2026-10-31). / MCP over FHIR with SMART auth (ONC pattern) / Wrap FHIR and NHS APIs as tools with per-mandate scopes / CLAIM / PROPOSED → `/clinic/connect-lims` - `F-9458c1d68f` L212 [UNKNOWN] _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 3. Agent assurance gaps in health_: **Regulator view of generative and agentic devices** / None of the FDA-authorised AI devices cited by one commentator is generative or agentic ([Clearwater](https://clearwatersecurity.com/transcripts/agentic-ai-in-healthcare-navigating-regulatory-uncertainty-and-building-governance-that-lasts/), accessed 2026-10-31). The UK draft regulations do not yet update software-specific provisions ([Qualio](https://www.qualio.com/blog/mhra-s-2026-devices-reform-what-changes-and-when), accessed 2026-10-31). / The AI Airlock sandbox ([GOV.UK](https://www.gov.uk/government/collections/ai-airlock-the-regulatory-sandbox-for-aiamd), accessed 2026-10-31) / Engage the MHRA early (Airlock Phase 3, if it opens) / UNKNOWN → **UNPLACED** - `F-b42862b9d0` L218 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 4. Reusable systems: open standards and open-source_: SPDX is the standard short-code for licences (for example `Apache-2.0`). Where there is no SPDX code, `LicenseRef-…` marks a custom licence. → **UNPLACED** - `F-dec47f8279` L224 _4. Reusable systems: open standards and open-source › 4.1 Standards and specifications_: **HL7 FHIR R4** / Fast Healthcare Interoperability Resources: the web API data standard for health records (Patient, Observation, MedicationRequest, Consent, Provenance, AuditEvent and so on). / `CC0-1.0`. SNOMED CT, LOINC and DICOM content are excluded and need their own terms. / Normative parts; widely used / Core data model and API / OBSERVED / [HL7 FHIR R4 licence](https://hl7.org/fhir/R4/license.html), accessed 2026-10-31 → `/clinic/connect-lims` - `F-b47d5573c7` L225 [UNKNOWN] _4. Reusable systems: open standards and open-source › 4.1 Standards and specifications_: **FHIR UK Core** / UK profiles of FHIR R4 (HL7 UK and NHS England). STU2 (standard for trial use 2) is release 2.0; 2.0.2 is the latest seen. / UNKNOWN (not checked) / STU (trial use) / Required shape for NHS integrations / CLAIM / [HL7 UK](https://www.hl7.org.uk/2024/05/28/fhir-uk-core-release-2-0-stu2/); [Simplifier](https://simplifier.net/guide/uk-core-implementation-guide-stu2), accessed 2026-10-31 → `/clinic/connect-lims` - `F-1afead96e5` L226 [UNKNOWN] _4. Reusable systems: open standards and open-source › 4.1 Standards and specifications_: **SMART on FHIR – App Launch 2.2** / OAuth 2.0 / OpenID Connect profile for apps and backend services that access FHIR. / Spec: UNKNOWN. Reference JS client: `Apache-2.0` (OBSERVED). / STU 2.2, current published / The authorisation layer for agents (scopes) / OBSERVED (client licence) / [HL7 SMART App Launch](https://hl7.org/fhir/smart-app-launch/apps.html); [client-js LICENSE](https://raw.githubusercontent.com/smart-on-fhir/client-js/master/LICENSE), accessed 2026-10-31 → `/clinic/connect-lims` - `F-ea9c094df4` L227 [UNKNOWN] _4. Reusable systems: open standards and open-source › 4.1 Standards and specifications_: **HL7 AI Transparency on FHIR** / Provenance, labels and model cards for AI-generated FHIR data. / UNKNOWN (an HL7 IG; not checked) / Ballot / continuous build, "not an authorized publication" / The best fit for replayable AI audit / CLAIM / [HL7](https://hl7.org/fhir/uv/aitransparency/2026Jan/artifacts.html); [build](https://build.fhir.org/ig/HL7/aitransparency-ig/en/requirements.html), accessed 2026-10-31 → `/clinic/connect-lims` - `F-a9f9f1b5b0` L228 [UNKNOWN] _4. Reusable systems: open standards and open-source › 4.1 Standards and specifications_: **SNOMED CT (UK Edition)** / Clinical terminology. dm+d codes are SNOMED CT codes. / `LicenseRef-SNOMED-CT-UK-Affiliate`: you accept both the UK National Release and the International affiliate licences via TRUD (NHS England's terminology download service). Fees and commercial terms were not confirmed. / Mature / Mandatory for UK clinical coding / OBSERVED (licence pages exist); UNKNOWN (commercial terms) / [TRUD UK licence](https://isd.hscic.gov.uk/trud/users/guest/filters/0/licence/10); [TRUD International licence](https://isd.hscic.gov.uk/trud/users/guest/filters/0/licence/4), accessed 2026-10-31 → `/clinic/check-record-accuracy` - `F-6cfb3f844a` L229 [UNKNOWN] _4. Reusable systems: open standards and open-source › 4.1 Standards and specifications_: **dm+d** (Dictionary of Medicines and Devices) / The NHS standard list of medicines and devices, from the NHSBSA (NHS Business Services Authority), downloaded from TRUD. / Open Government Licence. Version not stated in the source; `OGL-UK-3.0` is likely but UNKNOWN. / Mature, weekly / Medicine identity for prescribing safety (identity, not interactions) / CLAIM / [NHSBSA](https://www.nhsbsa.nhs.uk/pharmacies-gp-practices-and-appliance-contractors/nhs-dictionary-medicines-and-devices-dmd/snomed-ct-and-dmd); [NHSBSA dm+d](https://nhsbsa.nhs.uk/pharmacies-gp-practices-and-appliance-contractors/dictionary-medicines-and-devices-dmd), accessed 2026-10-31 → `/clinic/medication-safety-check` - `F-78ce44a55d` L230 _4. Reusable systems: open standards and open-source › 4.1 Standards and specifications_: **openEHR** / A clinical data modelling standard (archetypes and templates). / Specification repo: `CC-BY-SA-3.0` / Mature, strongest in Europe / Optional long-term record store; FHIR is the better fit for UK integration / OBSERVED / [openEHR specifications-RM LICENSE](https://raw.githubusercontent.com/openEHR/specifications-RM/master/LICENSE), accessed 2026-10-31 → **UNPLACED** - `F-af029e2bf6` L231 [CLAIM] _4. Reusable systems: open standards and open-source › 4.1 Standards and specifications_: **IHE profiles** (Integrating the Healthcare Enterprise) / Integration profiles, such as audit trails and document sharing. / Documents: `CC-BY-4.0`. Patents are handled separately, on reasonable-and-non-discriminatory terms. / Mature / Audit and identity patterns / CLAIM / [IHE Europe IP](https://ihe-europe.net/intellectual-property); [IHE General Introduction](https://profiles.ihe.net/GeneralIntro/ch-11.html), accessed 2026-10-31 → `/clinic/claim-edge-id-record` - `F-82dd04ab7f` L232 [UNKNOWN] _4. Reusable systems: open standards and open-source › 4.1 Standards and specifications_: **OMOP CDM / OHDSI vocabularies** (Observational Medical Outcomes Partnership Common Data Model; Observational Health Data Sciences and Informatics) / A research data model and vocabularies for population analytics. / Vocabularies: `Apache-2.0`, with some licensed vocabularies (for example FDB) needing proof of a licence. CDM spec: UNKNOWN. / Mature / Detection analytics and evaluation, not care delivery ("should not be used for individual patient healthcare") / CLAIM / [HL7 THO OMOP](https://build.fhir.org/ig/HL7/UTG/en/NamingSystem-OMOP.html), accessed 2026-10-31 → `/clinic/read-assistive-output-image` - `F-3cb70605e8` L233 [UNKNOWN] _4. Reusable systems: open standards and open-source › 4.1 Standards and specifications_: **MCP** (Model Context Protocol) / An open protocol for connecting agents to tools and data. / UNKNOWN (not checked) / Emerging / The tool layer for agents over FHIR / CLAIM / [ONC ISP](https://isp.healthit.gov/appendix-vi-ai-interoperability); [Keragon](https://www.keragon.com/blog/medical-mcp-servers), accessed 2026-10-31 → `/clinic/price-policy-clinic-data` - `F-d5f10d9813` L239 _4. Reusable systems: open standards and open-source › 4.2 Open-source software_: **HAPI FHIR** / Java FHIR server and library / `Apache-2.0` / Mature, widely used / A self-hosted FHIR store and validator / OBSERVED / [LICENSE.txt](https://raw.githubusercontent.com/hapifhir/hapi-fhir/master/LICENSE.txt), accessed 2026-10-31 → **UNPLACED** - `F-90609ba336` L240 [CLAIM] _4. Reusable systems: open standards and open-source › 4.2 Open-source software_: **Medplum** / TypeScript FHIR-native platform (server, SDKs, React components) / `Apache-2.0` / Production. ONC certification (dated 31 December 2025) is claimed, but disputed for self-hosted copies. / Fastest route to a FHIR backend with auth and bots / OBSERVED (licence); CLAIM (certification) / [LICENSE.txt](https://raw.githubusercontent.com/medplum/medplum/main/LICENSE.txt); [Mindbowser](https://www.mindbowser.com/open-source-ehr/), accessed 2026-10-31 → **UNPLACED** - `F-1333e9c3c2` L241 [CLAIM] _4. Reusable systems: open standards and open-source › 4.2 Open-source software_: **OpenMRS** / Open-source medical record system, strong in low- and middle-income countries / `MPL-2.0` plus a Healthcare Disclaimer (`LicenseRef-OpenMRS-Exception-to-MPL-2.0`) / Mature / Global and offline settings, not NHS primary care / OBSERVED (MPL-2.0); CLAIM (disclaimer) / [LICENSE](https://raw.githubusercontent.com/openmrs/openmrs-core/master/LICENSE); [ScanCode](https://scancode-licensedb.aboutcode.org/openmrs-exception-to-mpl-2.0.html), accessed 2026-10-31 → `/clinic/claim-edge-id-record` - `F-2e7d534397` L242 _4. Reusable systems: open standards and open-source › 4.2 Open-source software_: **EHRbase** / openEHR clinical data repository / `Apache-2.0` / Production / Only if the openEHR route is chosen / OBSERVED / [LICENSE](https://raw.githubusercontent.com/ehrbase/ehrbase/develop/LICENSE), accessed 2026-10-31 → `/clinic/request-clinic-data-quote` - `F-88e0c8463e` L243 _4. Reusable systems: open standards and open-source › 4.2 Open-source software_: **OHDSI WebAPI / ATLAS** / OMOP analytics tools / `Apache-2.0` / Mature / Population detection analytics and research / OBSERVED / [WebAPI LICENSE](https://raw.githubusercontent.com/OHDSI/WebAPI/master/LICENSE); [Atlas LICENSE](https://raw.githubusercontent.com/OHDSI/Atlas/master/LICENSE), accessed 2026-10-31 → **UNPLACED** - `F-a9e4cc9ef0` L249 [CLAIM] _4. Reusable systems: open standards and open-source › 4.3 NHS national services (reuse by integration, not by code)_: **NHS login** / Patient identity (OpenID Connect). Verification levels P0, P5 and P9. Partners must meet the identity standard DAPB3051. / n/a (service; partner onboarding) / Live, national / Patient sign-in and verified identity / CLAIM (archived NHS page) / [NHS login, archived](https://tnaqa.mirrorweb.com/ukgwa/20260505142011mp_/https://digital.nhs.uk/developer/api-catalogue/nhs-login), accessed 2026-10-31 → `/clinic/receive-pre-authorisation` - `F-6cd5e64389` L250 [UNKNOWN] _4. Reusable systems: open standards and open-source › 4.3 NHS national services (reuse by integration, not by code)_: **GP Connect** (Access Record: Structured) / FHIR API to read structured GP records. Working version 1.5.0. / n/a (service) / Live, but the service page says there is "no longer an onboarding route" for new developers. It needs SCAL (the supplier conformance assessment list), a CSO (clinical safety officer) and a hazard log. / Would be high value; access is UNKNOWN for a new supplier / CLAIM / UNKNOWN / [NHS England DSIC wiki](https://nhse-dsic.atlassian.net/wiki/spaces/DCSDCS/pages/12723519489/GP+Connect+Access+Record+Structured), accessed 2026-10-31 → `/clinic/investigate-clinical-safety` - `F-fdfc39ee91` L254 _4. Reusable systems: open standards and open-source › 4.4 Recommended reuse stack (research suggestion only)_: **Data and API:** FHIR R4 + UK Core on HAPI FHIR or Medplum (both `Apache-2.0`, OBSERVED above). → `/clinic/investigate-clinical-safety` - `F-e84807cd1f` L255 _4. Reusable systems: open standards and open-source › 4.4 Recommended reuse stack (research suggestion only)_: **Terminology:** SNOMED CT UK and dm+d from TRUD (licences above). → `/clinic/investigate-clinical-safety` - `F-523a18921d` L256 _4. Reusable systems: open standards and open-source › 4.4 Recommended reuse stack (research suggestion only)_: **Auth:** NHS login for patients, plus SMART scopes for agents. → `/clinic/investigate-clinical-safety` - `F-3828d5ba30` L257 _4. Reusable systems: open standards and open-source › 4.4 Recommended reuse stack (research suggestion only)_: **Audit:** FHIR Provenance and AuditEvent following the AI Transparency on FHIR IG. → `/clinic/investigate-clinical-safety` - `F-d6c7713a75` L258 _4. Reusable systems: open standards and open-source › 4.4 Recommended reuse stack (research suggestion only)_: **Analytics:** OMOP / OHDSI for detection evaluation. → `/clinic/investigate-clinical-safety` - `F-fe276e0da8` L259 _4. Reusable systems: open standards and open-source › 4.4 Recommended reuse stack (research suggestion only)_: **Interactions:** license BNF or FDB until an open, clinical-grade source exists (sources in 1.3). → `/clinic/investigate-clinical-safety` - `F-d347fc28ea` L265 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 5. Open questions (UNKNOWN) to resolve next_: Has the government formally responded to the National Commission? GOV.UK and Ropes & Gray differ. [GOV.UK](https://www.gov.uk/government/news/independent-commission-led-by-nhs-doctors-sets-out-blueprint-to-accelerate-safe-ai-adoption-in-healthcare); [Ropes & Gray](https://www.ropesgray.com/en/insights/viewpoints/2026/10/102o5fp/ai-medical-devices-in-great-britain-what-the-mhras-response-to-the-national-com), accessed 2026-10-31. → **UNPLACED** - `F-46b3dcc5ca` L266 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 5. Open questions (UNKNOWN) to resolve next_: When will the revised DCB0129/DCB0160 be published, and what will they require for AI? [NHS England](https://www.engage.england.nhs.uk/patient-safety/national-review-of-clinical-risk-management-standa/), accessed 2026-10-31. → `/clinic/nominate-carer-proxy` - `F-1ba06a174f` L267 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 5. Open questions (UNKNOWN) to resolve next_: Is there any GP Connect onboarding route for a new supplier? [NHS England DSIC wiki](https://nhse-dsic.atlassian.net/wiki/spaces/DCSDCS/pages/12723519489/GP+Connect+Access+Record+Structured), accessed 2026-10-31. → **UNPLACED** - `F-c8e7342440` L268 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 5. Open questions (UNKNOWN) to resolve next_: What was the outcome of the UK NSC consultation on AI in diabetic eye screening? [AOP](https://www.aop.org.uk/ot/news/2026/05/13/uk-national-screening-committee-consults-on-use-of-ai-in-diabetic-eye-screening), accessed 2026-10-31. → `/clinic/retinal-screening` - `F-21c0fd88a5` L269 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 5. Open questions (UNKNOWN) to resolve next_: What are the licences for UK Core, the SMART App Launch spec, MCP, DDInter and the OMOP CDM spec? These were not checked in primary sources. → `/clinic/retinal-screening` - `F-80005aa855` L270 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 5. Open questions (UNKNOWN) to resolve next_: What is the current UK status of Livi/Kry, Push Doctor and Doctaly? See the 1.1 sources. → `/clinic/retinal-screening` - `F-6b2750b6a4` L271 _EDGE Clinic: competitors, regulators and rules, agent assurance gaps, reusable systems › 5. Open questions (UNKNOWN) to resolve next_: Medscape's interaction checker and the Google ARDA retinal AI were not researched. → `/clinic/retinal-screening` ###### `research/clinic/ENGINE-MEDPLUM.md` (101 findings) - `F-5bf3f17867` L3 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record_: Built 1 Nov 2026 by ATBAN Claude-Jegudiel 00:5531102026-1 (service-builder lane). Code: `runtime/domains/clinic.py` (the → **UNPLACED** - `F-45790beb84` L5 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record_: `tests/test_clinic_domain.py` (unit tests with a fake FHIR server; `LiveSmoke` against the real server when `MEDPLUM_SMOKE=1`). → **UNPLACED** - `F-b3c4358449` L6 [OBSERVATION] _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record_: Every statement below is marked OBSERVATION (measured, with how and when), AGREEMENT (a decision taken here) or CLAIM. → **UNPLACED** - `F-690c8c0617` L10 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 1. The engine (OBSERVATION, jude-1, 1 Nov 2026 04:5x–05:1xZ)_: Medplum `5.2.13-15a306f` runs as podman container `jude-engine-medplum` in pod `jude-medplum-engines` with → **UNPLACED** - `F-cc7eedf858` L16 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 1. The engine (OBSERVATION, jude-1, 1 Nov 2026 04:5x–05:1xZ)_: Stop seen during this lane: at 04:48:05Z a `podman pod stop` (pid 1428339) stopped the three engine pods; systemd brought the → **UNPLACED** - `F-ce88e92c04` L17 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 1. The engine (OBSERVATION, jude-1, 1 Nov 2026 04:5x–05:1xZ)_: Medplum units back at 04:51:42Z. After that restart the database container's `/dev/shm` was `drwx------ root` and every → **UNPLACED** - `F-95ac78e586` L18 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 1. The engine (OBSERVATION, jude-1, 1 Nov 2026 04:5x–05:1xZ)_: PostgreSQL backend failed with `could not open shared memory segment "/PostgreSQL.…": Permission denied` (Medplum answered → **UNPLACED** - `F-1d1489e0e5` L19 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 1. The engine (OBSERVATION, jude-1, 1 Nov 2026 04:5x–05:1xZ)_: 400 on every FHIR call, `/healthcheck` still said ok). Fix applied 05:0xZ: `chmod 1777 /dev/shm` inside → **UNPLACED** - `F-6c91d6e322` L20 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 1. The engine (OBSERVATION, jude-1, 1 Nov 2026 04:5x–05:1xZ)_: `jude-engine-medplum-db`; FHIR calls answered 200 at once. Cause of the `podman pod stop` itself: not determinable from the → **UNPLACED** - `F-96da7e1ece` L21 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 1. The engine (OBSERVATION, jude-1, 1 Nov 2026 04:5x–05:1xZ)_: journal (no sudo line names it); the engines' owner should confirm. The shm mode will revert on the next container start → **UNPLACED** - `F-dad2bb6fb7` L22 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 1. The engine (OBSERVATION, jude-1, 1 Nov 2026 04:5x–05:1xZ)_: unless the pod's infra container or the unit sets it: an open item for the engine lane. → **UNPLACED** - `F-e03792ed05` L23 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 1. The engine (OBSERVATION, jude-1, 1 Nov 2026 04:5x–05:1xZ)_: The seeded super-admin login still accepts Medplum's published first-run credentials (checked 04:5xZ: `POST /auth/login` 200). → **UNPLACED** - `F-bb38fc3e57` L24 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 1. The engine (OBSERVATION, jude-1, 1 Nov 2026 04:5x–05:1xZ)_: That is a security finding for the engine lane: rotate it. This lane did not change it (not its file) and does not use it at → **UNPLACED** - `F-94e61074e2` L29 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 2. Authentication (AGREEMENT: how; OBSERVATION: that it works; no values here)_: Medplum Project `EDGE Clinic` (`strictMode: true`) created 1 Nov 2026 through the super-admin FHIR API, with one → **UNPLACED** - `F-d2466b725f` L30 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 2. Authentication (AGREEMENT: how; OBSERVATION: that it works; no values here)_: `ClientApplication` `edge-clinic-adapter` and its `ProjectMembership` (`/auth/me` as the client: project `EDGE Clinic`, profile → **UNPLACED** - `F-2888d5eeac` L31 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 2. Authentication (AGREEMENT: how; OBSERVATION: that it works; no values here)_: `ClientApplication`; never the Super Admin project). `POST /admin/projects` answers 404 on this build; the FHIR route was used. → **UNPLACED** - `F-04528410bb` L32 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 2. Authentication (AGREEMENT: how; OBSERVATION: that it works; no values here)_: The adapter authenticates with OAuth 2.0 **client credentials** (`POST /oauth2/token`, `grant_type=client_credentials`, → **UNPLACED** - `F-7317a4ccff` L33 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 2. Authentication (AGREEMENT: how; OBSERVATION: that it works; no values here)_: client id and secret in the form body; the HTTP Basic form answers 400 on this build). The token (3600 s) is cached in the → **UNPLACED** - `F-403cbed57f` L34 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 2. Authentication (AGREEMENT: how; OBSERVATION: that it works; no values here)_: process and renewed 60 s before expiry; a 401 triggers one fresh token and one retry. → **UNPLACED** - `F-5d65f28726` L35 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 2. Authentication (AGREEMENT: how; OBSERVATION: that it works; no values here)_: The credential lives at `/etc/x-x/plugs/medplum.env` (`root:root 0640`; there is no `x-x` user or group on jude-1 and → **UNPLACED** - `F-a3b95537bb` L36 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 2. Authentication (AGREEMENT: how; OBSERVATION: that it works; no values here)_: `/etc/x-x/plugs/griffin.key` does not exist yet, so the Griffin precedent could not be copied literally; the web service runs → **UNPLACED** - `F-5201d0e69f` L37 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 2. Authentication (AGREEMENT: how; OBSERVATION: that it works; no values here)_: as root and reads it). Keys: `MEDPLUM_BASE_URL`, `MEDPLUM_PROJECT_ID`, `MEDPLUM_CLIENT_ID`, `MEDPLUM_CLIENT_SECRET`. The same → **UNPLACED** - `F-2251b5ee0c` L38 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 2. Authentication (AGREEMENT: how; OBSERVATION: that it works; no values here)_: names in the environment override the file (tests use that with fake values). No value is ever written to a record, a reply, → **UNPLACED** - `F-4d02529370` L39 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 2. Authentication (AGREEMENT: how; OBSERVATION: that it works; no values here)_: the specification or Git. Without the file `medplum.bound()` is False and a live tenant's operation answers `no_answer` with → **UNPLACED** - `F-0ccbef1314` L44 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Clinic has 4,151 receiver-labelled operations. 624 are bound to Medplum; 3,527 stay with the estate record or another receiver. → **UNPLACED** - `F-07d9b62164` L45 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: (A count is live: `python3 -c "from runtime.domains import clinic, json; print(json.dumps(clinic.coverage()))"`.) → **UNPLACED** - `F-e788d5690d` L47 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: FHIR R4 resource / Bound operations / Estate entities and units (from `spec/receivers.json` labels `estate-record:<class>:<entity>`) → **UNPLACED** - `F-1e6699e7a3` L49 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Appointment (+ Location, Schedule, Slot when a time is chosen) / 89 / `record` of the booking units: book-an-appointment, book-for-the-person-i-care-for, book-a-pathway-test, book-from-the-referral-queue, book-an-interpreter, book-transport, rebook-or-cancel-an-appointment, offer-a-cancellation-slot, hold-a-slot-for-a-tertiary-handover, check-in-an-arriving-patient, check-in-on-arrival, front-door-check-in, confirm-attendance-and-prepare, attend-an-appointment-with-the-person, handle-a-did-not-attend, receive-a-did-not-attend-notice, record-attendance-outcome, add-an-appointment-to-my-calendar, schedule-the-scan, see-my-day; `feature_appointment` → **UNPLACED** - `F-ce0de968bc` L50 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: ServiceRequest / 88 / `referral`, `feature_referral`; referral and order units (make-an-urgent-same-day-referral, refer-on-a-suspected-cancer-pathway, refer-to-a-crisis-team, refer-to-tertiary-care, refer-to-the-mental-health-stream, refer-to-the-multidisciplinary-team, social-prescribing-referral, ask-for-advice-and-guidance, order-a-diagnostic-test, order-a-direct-access-test, order-a-genomic-test, request-a-second-read, request-a-second-read-or-discrepancy-review, request-a-second-opinion, track-a-referral, triage-a-referral, triage-a-mental-health-referral, close-a-referral, vet-an-imaging-request) → **UNPLACED** - `F-dc22a01ea5` L51 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Observation / 82 / `result`, `feature_result`, `feature_observation`; take-observations, record-a-point-of-care-urine-result, record-a-wound-or-skin-check, measure-a-nodule-and-its-growth, share-a-wearable-signal, wearable-signals, take-an-online-early-symptom-test, run-an-online-early-symptom-test, review-a-real-result, authorise-a-result, tell-the-patient-a-result, explain-a-result-to-the-patient, explain-a-result-in-plain-language, receive-a-critical-result-for-my-patient, phone-a-critical-result, view-my-results, view-results-in-scope, chase-missing-results, retinal-screening, grade-a-retinal-image → **UNPLACED** - `F-fea1649aad` L52 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Consent / 77 / `consent`, `feature_consent`; give-or-withdraw-consent, give-consent-on-behalf, take-informed-consent-for-research, apply-a-national-data-opt-out, record-sample-consent-version, set-my-data-sharing-choices, share-clinic-data-with-my-insurer, share-a-medicines-list-with-the-clinic, share-images-with-another-trust, approve-a-data-release → **UNPLACED** - `F-9176d074b0` L53 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: CarePlan / 57 / `safety_plan`, `plan`, `feature_careplan`; write-a-care-plan, maintain-the-suicide-prevention-plan, follow-up-an-action-plan, set-a-surveillance-plan, discharge-from-the-pathway, discharge-from-the-stream, track-my-cancer-pathway, record-decision-to-treat → **UNPLACED** - `F-f7b20a6fc5` L54 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: QuestionnaireResponse / 36 / `assessment`, `answer`, `feature_questionnaire`; complete-a-pre-clinic-questionnaire, wellbeing-questionnaires, complete-a-mental-health-check-in, complete-a-risk-formulation, run-a-pre-assessment, score-and-interpret-a-response, act-on-a-high-risk-answer, symptom-check-and-triage, online-urgent-triage → **UNPLACED** - `F-07c1dbaa8f` L55 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: MedicationRequest / 33 / `feature_prescription`, `feature_medicationrequest`; prescribe, prescribe-under-the-pathway-protocol, order-a-repeat-medicine, order-repeat-medicines-on-behalf, review-a-repeat-request, clinically-check-a-prescription, check-a-chemotherapy-prescription → **UNPLACED** - `F-281acb3033` L56 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Flag / 29 / `crisis_alert`, `flag`; flag-a-critical-finding, receive-a-crisis-alert, receive-a-crisis-alert-for-the-person, record-a-safeguarding-concern, raise-a-safeguarding-concern-from-a-response → **UNPLACED** - `F-b4395d81e8` L57 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Patient / 25 / `feature_patient`; register-a-walk-in-patient, register-a-child-as-a-patient, update-patient-demographics → **UNPLACED** - `F-79d13d2621` L58 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Communication / 21 / message-my-care-team, send-a-results-ready-notice, send-appointment-letters-and-reminders, send-a-check-in-to-a-caseload, make-a-follow-up-call → **UNPLACED** - `F-de4e41548c` L59 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: MedicationStatement / 18 / `medicine_list`, `feature_medication`; reconcile-medicines, reconcile-medicines-at-admission, view-my-medicines-list → **UNPLACED** - `F-3f6218e2f2` L60 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Specimen / 15 / take-a-blood-or-urine-sample, receive-and-book-in-a-sample, reject-an-unlabelled-sample, track-sample-custody → **UNPLACED** - `F-8ae36275c3` L61 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Encounter / 10 / `check_in`, `feature_encounter`; record-a-clinic-consultation, hold-a-remote-session → **UNPLACED** - `F-0ab6ef3d09` L62 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: DiagnosticReport / 10 / `feature_diagnosticreport`; report-histology, issue-a-genomic-report → **UNPLACED** - `F-4f826175bb` L63 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: MedicationDispense, Immunization, RelatedPerson, AllergyIntolerance, AdverseEvent / 5 each / dispense-a-medicine; give-a-vaccination; nominate-a-carer-or-proxy; record-an-adverse-drug-reaction; report-a-side-effect → **UNPLACED** - `F-303beafed9` L64 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: MedicationAdministration 4, Condition 3, Procedure 2 / administer-a-medicine; `feature_condition`; `feature_procedure` → **UNPLACED** - `F-663b4fbc67` L66 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Operation classes covered, each with the receiver contract (create 159, read 30, change 383, remove 5, send 12, decide 27, → **UNPLACED** - `F-b1ded01cc3` L67 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: plan 8 = 624, measured by `clinic.plan()`; `runtime/receivers.effective_class` still turns a "remove" of a workflow `record` into a → **UNPLACED** - `F-45c35cd01e` L70 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: **create** POST, then GET; confirmed with the Medplum id (the journey carries it on as `id` and `<entity>_id`). → **UNPLACED** - `F-07110ed400` L71 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: **read** GET by any id in the inputs that is this resource in this tenant; else a search `_tag=<tenant>` (newest first, 50). → **UNPLACED** - `F-610adf2610` L72 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: **change** GET, merge, PUT; `expected_version`/`version` becomes `If-Match: W/"<versionId>"` and a stale one is → **UNPLACED** - `F-463d6dd277` L73 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: `rejected VERSION_CONFLICT` (checked before the write and honoured by the server: 412). A change with no existing resource → **UNPLACED** - `F-898cd16eff` L74 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: creates one (a journey's first touch), as the estate record does. The operation's verb moves the status: cancel/withdraw/ → **UNPLACED** - `F-b9d7ef08e5` L77 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: **remove** DELETE, then GET must answer 410 Gone; a cancelled Appointment frees its Slot first. → **UNPLACED** - `F-7d707eb91a` L78 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: **send, decide, plan** like change, and the status lands on the resource's done state (`STATUS[resource]["done"]`); the → **UNPLACED** - `F-99ed311f64` L80 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: **Confirmed** only when the read-back equals what was written (meta and narrative aside); **rejected** with `refusal_code` → **UNPLACED** - `F-930a9ae9f8` L82 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: `RATE_LIMITED`); **no_answer** on 5xx, on a transport failure (`TIMEOUT`, `UNREACHABLE`), on a read-back mismatch, or when no → **UNPLACED** - `F-2fe69b1765` L85 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: How inputs become FHIR: every input is kept verbatim in the extension `https://judiverse.tech/fhir/StructureDefinition/estate-inputs` → **UNPLACED** - `F-161591e685` L86 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: (sorted JSON), so the read-after-write compares the whole input set exactly; on top, fields the unit declares (`data.<entity>.fields` → **UNPLACED** - `F-d55a1a195c` L87 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: types) or names are placed structurally: a `datetime`/`date` field (or any value with a time of day) becomes `Appointment.start/end` → **UNPLACED** - `F-86285cd144` L89 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: and so on; a numeric input becomes `Observation.valueQuantity`; `name`, `date_of_birth`, `nhs_number` (10 digits, system → **UNPLACED** - `F-3192635ff8` L92 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: value: <tenant>/<unit>/<entity>}` (a single identifier on AdverseEvent and QuestionnaireResponse, which are 0..1 in R4). The → **UNPLACED** - `F-14e297f01f` L93 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: person behind the actor (or an agent's mandate owner) is one `Patient` per tenant found by identifier `https://judiverse.tech/fhir/edge-id` → **UNPLACED** - `F-acfc29dc58` L96 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Booking a time: an Appointment that receives a start creates (once per tenant and clinic name) a `Location`, a `Schedule` whose → **UNPLACED** - `F-973e3f60b4` L97 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: actor is that Location, and a busy `Slot` for the 30 minutes, and references the Slot; cancelling frees the Slot. FHIR invariant → **UNPLACED** - `F-7029d11968` L98 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: app-3 ("only proposed or cancelled appointments can be missing start/end") is honoured: a confirm before a time is chosen leaves the → **UNPLACED** - `F-b67f2a59a4` L99 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 3. What maps to what (AGREEMENT, explicit tables in `clinic.py`; counts OBSERVED by `clinic.coverage()` on this checkout's `spec/receivers.json`)_: Appointment `proposed`, never `booked` (OBSERVED: the live server rejects the alternative with 400 INVARIANT). → **UNPLACED** - `F-bd0032a39c` L103 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: `record` of the 496 Clinic units not in the map (219 capability units that are rotas, audits, incidents, governance, trials, → **UNPLACED** - `F-60c8dfd2a8` L104 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: genomics pipelines and the like, and 277 research, parity and problem units; 995 operations together) → **UNPLACED** - `F-4f5fb8baeb` L105 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: and every `case` (659: the workflow case is the estate's own, not a clinical resource); `evidence_query` (606) and the → **UNPLACED** - `F-7196c318c8` L107 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: `feature_fhir` (579), `feature_screen`, `feature_chart`, `feature_manage`, `feature_ios`, `feature_web` and the other → **UNPLACED** - `F-c875e2deff` L108 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: `feature_*` entities that are competitor-feature extractions, not clinical entities. A `feature_<name>` is bound only when → **UNPLACED** - `F-a1eb1f7141` L109 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: `<name>` is a FHIR resource (Patient, Appointment, Observation, Condition, Encounter, Procedure, Immunization, CarePlan, → **UNPLACED** - `F-5c968a823d` L112 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: The 112 operations whose receiver is an `external:` slot (NHS services, voice telephony, forecasting models…). → **UNPLACED** - `F-675dcd1acb` L113 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: `session` (clinic sessions and therapy sessions share one word), `escalation`, `trusted_contact`, `concern`, `review`, → **UNPLACED** - `F-0e71afb0ab` L114 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: `outcome` and the other single-digit entities: candidates (Schedule, Flag, RelatedPerson…) once a unit by unit reading says so. → **UNPLACED** - `F-7fa3aafcbd` L115 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: Tenant isolation is by tag inside the one Project `EDGE Clinic`, enforced by the adapter on every read and write. Medplum's → **UNPLACED** - `F-696bad7dad` L116 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: own isolation is one Project (and one ClientApplication) per tenant; that is the next step and needs a provisioning route → **UNPLACED** - `F-adc56efdf6` L117 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: (`POST /admin/projects` is 404 on this build; the super-admin FHIR route works). → **UNPLACED** - `F-0e6580adc3` L118 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: Binary attachments (images, documents) are not sent: `DocumentReference.content` carries a title only. NHS connectivity (PDS, → **UNPLACED** - `F-a89671bfdf` L119 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 4. What Medplum does not cover (stays with the estate record; AGREEMENT with reasons)_: e-RS, GP Connect) is not inferred from FHIR availability, as the engine lane's note says. → **UNPLACED** - `F-737889f476` L123 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 5. Measured round trips (OBSERVATION, `LiveSmoke`, jude-1, 1 Nov 2026 05:1xZ, tenant `smoke-d14def5f`, loopback)_: 20 resource types, each create → decide → stale-version change (412 expected) → read → remove (DELETE, 410 read-back): 0 failures, → **UNPLACED** - `F-548c7c4bd0` L124 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 5. Measured round trips (OBSERVATION, `LiveSmoke`, jude-1, 1 Nov 2026 05:1xZ, tenant `smoke-d14def5f`, loopback)_: 237 HTTP calls in 7.0 s wall (unittest), raw per-call figures in `/tmp/clinic-smoke.json` on jude-1 (dated, not in Git). → **UNPLACED** - `F-513743a8bd` L130 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 5. Measured round trips (OBSERVATION, `LiveSmoke`, jude-1, 1 Nov 2026 05:1xZ, tenant `smoke-d14def5f`, loopback)_: `POST /fhir/R4/<Type>` / 24 / 44.5 ms / 153.0 ms (Observation) → **UNPLACED** - `F-a3dd1fdf01` L131 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 5. Measured round trips (OBSERVATION, `LiveSmoke`, jude-1, 1 Nov 2026 05:1xZ, tenant `smoke-d14def5f`, loopback)_: `PUT /fhir/R4/<Type>/<id>` / 21 / 38.9 ms / 412.2 ms (one outlier) → **UNPLACED** - `F-0fa7498658` L132 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 5. Measured round trips (OBSERVATION, `LiveSmoke`, jude-1, 1 Nov 2026 05:1xZ, tenant `smoke-d14def5f`, loopback)_: `DELETE /fhir/R4/<Type>/<id>` / 23 / 33.4 ms / 716.4 ms (one outlier) → **UNPLACED** - `F-e8d7a04b04` L134 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 5. Measured round trips (OBSERVATION, `LiveSmoke`, jude-1, 1 Nov 2026 05:1xZ, tenant `smoke-d14def5f`, loopback)_: A create as the receiver sees it (POST + read-back GET) is therefore about 55–60 ms on the host; a change (find GET + PUT + → **UNPLACED** - `F-f432f95f68` L135 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 5. Measured round trips (OBSERVATION, `LiveSmoke`, jude-1, 1 Nov 2026 05:1xZ, tenant `smoke-d14def5f`, loopback)_: read-back GET) about 60–65 ms; a first touch adds the subject Patient (search + POST) once per person. The earlier run at 05:0xZ → **UNPLACED** - `F-e8452d13ea` L136 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 5. Measured round trips (OBSERVATION, `LiveSmoke`, jude-1, 1 Nov 2026 05:1xZ, tenant `smoke-d14def5f`, loopback)_: (199 calls, 0 failures) gave the same order: GET 7–16 ms, POST 29–116 ms, PUT 26–75 ms per type. Not measured: throughput under → **UNPLACED** - `F-4338b1e9d5` L137 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 5. Measured round trips (OBSERVATION, `LiveSmoke`, jude-1, 1 Nov 2026 05:1xZ, tenant `smoke-d14def5f`, loopback)_: load, the public path (there is none: loopback only), or cold starts. → **UNPLACED** - `F-9bda144050` L141 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 6. Demo tenant and the live journey walk (OBSERVATION)_: The demo tenant never touches Medplum: `_demo(engine, tenant)` routes to the estate record on the demo engine and the outputs → **UNPLACED** - `F-37414e24cc` L142 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 6. Demo tenant and the live journey walk (OBSERVATION)_: say `SIMULATED: demo tenant; the clinic record is simulated and Medplum is not touched (<Resource>)` with → **UNPLACED** - `F-9b9fcf5331` L144 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 6. Demo tenant and the live journey walk (OBSERVATION)_: `tools/journeys_http.py --demo --only /tmp/clinic-medplum-journeys.txt --shard 1/1` on the live release 7a2a0ea, 05:1xZ: the → **UNPLACED** - `F-cf09c8ae05` L145 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 6. Demo tenant and the live journey walk (OBSERVATION)_: 123 `#human` journeys of the 123 mapped units: `{"journeys": 123, "pass": 123, "plug": 0, "held": 0, "fail": 0, "seconds": 44.0}` → **UNPLACED** - `F-b45becbe25` L146 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 6. Demo tenant and the live journey walk (OBSERVATION)_: (`ledger/evidence/journeys-demo-rerun-1.jsonl`). Note: the running service does not yet import this domain (section 7), so → **UNPLACED** - `F-ebc767000d` L147 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 6. Demo tenant and the live journey walk (OBSERVATION)_: that walk proves the demo path through the estate record, which is also the path the domain takes for the demo tenant. → **UNPLACED** - `F-6d0d592151` L153 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 7. Registration (what the running site needs; one line in a file this lane does not own)_: from runtime.domains import clinic # noqa: F401 - binds the Clinic record to Medplum (FHIR R4) → **UNPLACED** - `F-a216076e44` L155 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 7. Registration (what the running site needs; one line in a file this lane does not own)_: Shipped once through the deploy helper at 05:15:03Z (record `/opt/x-x/deploy/ledger/deploy/2026-11-01T051503Z-edit.json`, health live). → **UNPLACED** - `F-e7db9a1d96` L156 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 7. Registration (what the running site needs; one line in a file this lane does not own)_: Until that line is in the release the live tenants' Clinic operations keep the generic estate record; the demo tenant is unchanged → **UNPLACED** - `F-51a1f16b10` L157 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 7. Registration (what the running site needs; one line in a file this lane does not own)_: either way. `python3 -m unittest tests.test_clinic_domain` (14 tests, fake server) passes on the Mac and on jude-1; → **UNPLACED** - `F-a54d8786c1` L162 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 8. Open items (CLAIMs for the owners named)_: Engine lane: rotate the seeded super-admin password; make the database container's `/dev/shm` mode survive a restart; name → **UNPLACED** - `F-241fe2d9bd` L164 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 8. Open items (CLAIMs for the owners named)_: Coordinator: add the registration line (section 7) and ship. → **UNPLACED** - `F-5d20d3a6fa` L165 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 8. Open items (CLAIMs for the owners named)_: This domain, next: one Medplum Project per tenant; `Binary` upload for images and documents; the remaining single-digit → **UNPLACED** - `F-555ed9315b` L166 _EDGE Clinic on Medplum (FHIR R4): the clinic's domain entities leave the generic estate record › 8. Open items (CLAIMs for the owners named)_: entities after a unit-by-unit reading; `Schedule` capacity for clinic sessions (cancel-a-clinic-session, plan-weekly-capacity). → **UNPLACED** ###### `research/clinic/MOBBIN-FLOWS.md` (74 findings) - `F-d8344926cc` L3 _EDGE Clinic research: real product flows from Mobbin_: **What this is.** Real screens and user flows from existing apps, recorded by Mobbin (a library of real app and website screenshots, grouped into step-by-step flows). It complements `REFERO-FLOWS.md` in this folder. It shows the floor that people already expect from a health front door, and good practice worth borrowing from other industries. → **UNPLACED** - `F-e6a4a5ce6f` L5 [OBSERVATION] _EDGE Clinic research: real product flows from Mobbin_: **Statement class and date.** Every item below is an OBSERVATION of a real product, seen in Mobbin on 2026-10-31. The descriptions are ours, written from the preview images Mobbin returned (an evenly spaced sample of each flow's screens, not always every screen). Product names belong to their owners. Image links are not kept because they expire; each item carries its `mobbin_url`. → **UNPLACED** - `F-1f1efb301c` L7 _EDGE Clinic research: real product flows from Mobbin_: **How it was gathered.** Searches by app name plus task, on iOS and web, about five flows per search: appointment booking, symptom checking, test results, medication reminders, mental health check-ins, health questionnaires, wearables and readiness scores, sharing health data, telehealth, food logging, emergency and crisis help, and local community activity (as a stand-in for social prescribing). → **UNPLACED** - `F-b8a3498fcb` L11 _EDGE Clinic research: real product flows from Mobbin › Coverage: what Mobbin held and what it did not_: **Held and returned:** Ada (symptom assessment, condition detail, past assessments), Headspace (anxiety and stress check-ins, therapy booking on web), Apple Health (sharing, data sources and access, profile), Oura (readiness and a medical AI chat), WHOOP (labs), MyFitnessPal (meal logging), Zocdoc (doctor booking), GoodRx, Hers and Hims (medication reminders, treatment questionnaires), Ro, Superpower and Bevel (lab results and intake), Alan (consultations, with Livi as the video provider inside it), and single screens from Wysa, Bloom, Garmin Connect and Amazon's health assistant. → **UNPLACED** - `F-08b13d2376` L12 _EDGE Clinic research: real product flows from Mobbin › Coverage: what Mobbin held and what it did not_: **Searched for but not returned:** NHS App (the booking search returned Zocdoc, IKEA, Crate & Barrel and Fresha instead) and Doctolib (the web search returned Selfridges, Square and Headspace instead). No NHS, GP practice or other UK public health service appeared in any result. → **UNPLACED** - `F-e627fb3aea` L13 _EDGE Clinic research: real product flows from Mobbin › Coverage: what Mobbin held and what it did not_: **Not searched by name in this pass:** Babylon, Healthily, Calm. Livi appeared only as the video provider inside Alan. → **UNPLACED** - `F-abdfeb1203` L14 _EDGE Clinic research: real product flows from Mobbin › Coverage: what Mobbin held and what it did not_: **Not found at all:** any social prescribing service, any link worker or referral to community groups, any flow where a GP or clinician receives what the person filled in, and any flow where an AI agent acts for a person or organisation with its own identity. → **UNPLACED** - `F-90462267dc` L20 [OBSERVATION] _By task › Booking an appointment_: **Zocdoc, iOS, "Booking an appointment"** (10 screens). A doctor's page with a "new patient" tick box, a choice of location and tomorrow's slots as buttons. The review step lets you book for yourself or "someone else", shows a red inline error when the address is missing, and asks for insurance, visit reason, a phone number and optional notes for the practice, with a tick box certifying the payment method. The confirmation offers "Add to calendar" and "Prepare for appointment". OBSERVATION, 2026-10-31. https://mobbin.com/flows/fcd6461e-af19-4ca9-9864-8b21c286888c → `/clinic/add-appointment-calendar` - `F-5a1b914060` L21 [OBSERVATION] _By task › Booking an appointment_: **Zocdoc, iOS, "Making a doctor appointment"** (12 screens). In-person or video visit toggle, slots grouped by day, then a gate: "This provider only sees in-network patients", with the choice of photographing the insurance card or choosing it by hand, and accepting the terms. Afterwards the home screen shows the upcoming appointment with "Almost done! Upload a photo of your ID". OBSERVATION, 2026-10-31. https://mobbin.com/flows/26c741b6-0231-4040-b5e9-890f0b6db100 → `/clinic/mental-health-care-crisis` - `F-5bf4329f12` L22 [OBSERVATION] _By task › Booking an appointment_: **IKEA, iOS, "Booking an appointment"** (15 screens, retail). A plain explanation of the session and what to prepare beforehand, a calendar with green dots on days that have space, time slots labelled with the time zone, a contact form that says why the phone number is needed (reminders), and a confirmation with a booking reference and all details. OBSERVATION, 2026-10-31. https://mobbin.com/flows/64df5892-fc18-4edb-ba02-8b365b63bebe → `/clinic/explain-result-plain-language` - `F-4ec91ee969` L23 [OBSERVATION] _By task › Booking an appointment_: **Crate & Barrel, iOS, "Booking an appointment"** (14 screens, retail). A three-step progress bar, separate tick boxes for marketing texts and for appointment reminders, a calendar with an available and not available key, and a confirmation that explains "What's next", "What to expect" and "What you'll get". OBSERVATION, 2026-10-31. https://mobbin.com/flows/8e862fc0-87f2-44f0-a201-787bb47e5d6e → `/clinic/add-appointment-calendar` - `F-781bc2dec6` L24 [OBSERVATION] _By task › Booking an appointment_: **Fresha, iOS, "Booking an appointment"** (22 screens, beauty). Date strip and time list, a review with price and fees, the cancellation and no-show charges stated before payment, then a "Confirmed" page with add to calendar, directions, message the venue and venue details. OBSERVATION, 2026-10-31. https://mobbin.com/flows/7121fd1b-35ef-48da-87d5-aacbd1ba81a8 → `/clinic/add-appointment-calendar` - `F-ad8ba4362c` L25 [OBSERVATION] _By task › Booking an appointment_: **Headspace, web, "Scheduling an appointment"** (23 screens, therapy). Steps shown at the top: check in, verify insurance, schedule. "Here's your estimated cost" confirms the insurer and lists things to know (pay nothing today, the most it could cost). Time ranges show how many clinicians are free and the time zone. The final page lists next steps, including "complete your pre-appointment assessment to help your therapist prepare". OBSERVATION, 2026-10-31. https://mobbin.com/flows/b46dac23-c0f1-4b3f-b506-ef8fea7c14d7 → `/clinic/front-door-check` - `F-fd100eade8` L26 [OBSERVATION] _By task › Booking an appointment_: **Selfridges, web, "Making an appointment"** (17 screens, retail). A four-step stepper with the appointment summary always visible on the right, a week grid by morning, afternoon and evening, a phone field that says it will only be used for a reminder 48 hours before, two separate consent tick boxes, and a confirmation with a reference, print and add to calendar. OBSERVATION, 2026-10-31. https://mobbin.com/flows/e5cf516c-6b42-4bb9-b0c2-291623dc0124 → `/clinic/add-appointment-calendar` - `F-5139c69464` L27 [OBSERVATION] _By task › Booking an appointment_: **Square, web, "Booking an appointment"** (12 screens). A running summary while you add services, a checkout that holds the slot with a countdown, a cancellation timeline ("cancel before" a date), and a thank-you page with add to calendar, reschedule, cancel and book again, plus "Next steps: we'll call you". OBSERVATION, 2026-10-31. https://mobbin.com/flows/6992da62-212e-4947-991d-588142829394 → `/clinic/add-appointment-calendar` - `F-928e565e20` L31 [OBSERVATION] _By task › Symptom checking_: **Ada, iOS, "Starting a symptom assessment"** (17 screens). One question per screen (name, then date of birth as day, month and year boxes), with "Previous" and a close button, and an "About Ada and Privacy" link on every screen. Symptom search with suggestions and a help icon per symptom. It ends: "No more questions... The report is not a medical diagnosis. If in doubt, seek advice from a healthcare professional." The report lists possible causes with the most urgent first, in a red box: "Needs immediate attention: seek emergency care now". Each cause shows which of your symptoms matched and "48% of people with your symptoms and health background had this condition". OBSERVATION, 2026-10-31. https://mobbin.com/flows/dc2c5dc4-6695-44fd-b21a-3cd4e7601c9a → `/clinic/symptom-check-triage` - `F-e37a730ab5` L32 [OBSERVATION] _By task › Symptom checking_: **Ada, iOS, "Searching symptoms"** (3 screens). Typing "headache" gives suggestions and "Symptoms found: 3", each with a short plain description and a help icon. OBSERVATION, 2026-10-31. https://mobbin.com/flows/4d5fb715-f25b-47a7-b4cc-1fd4bcfe1505 → `/clinic/front-door-check` - `F-5c45744435` L33 [OBSERVATION] _By task › Symptom checking_: **Ada, iOS, "Condition detail"** (7 screens). A condition page with sections that open and close (symptoms, diagnosis, treatment, risk factors, outcomes), a statement that the list "is not a diagnosis and not exhaustive", and "Download report". OBSERVATION, 2026-10-31. https://mobbin.com/flows/ce7b8167-97a3-47da-8a93-00c89d64c098 → `/clinic/complete-risk-formulation` - `F-cd1952f7ae` L34 [OBSERVATION] _By task › Symptom checking_: **Ada, iOS, "Health assessments"** (8 screens). A profile with basic information, health background and the list of past assessments. OBSERVATION, 2026-10-31. https://mobbin.com/flows/0c0ecbbf-82cb-4dee-af6b-353de04791ce → **UNPLACED** - `F-5149910dba` L35 [OBSERVATION] _By task › Symptom checking_: **Ada, iOS, "Home"** (3 screens). The home screen with one main action, "Start symptom assessment", and a news card about changes to how you access your data. OBSERVATION, 2026-10-31. https://mobbin.com/flows/35a15c59-17c0-4872-a99d-610cc2cec951 → `/clinic/symptom-check-triage` - `F-a26a0934c6` L39 [OBSERVATION] _By task › Questionnaires and intake_: **Ro, iOS, "Onboarding"** (20 screens). Choose a goal, then a progress bar over height and weight, a "select all that apply" list of heart conditions, and treatment preferences that include "Not sure yet" and "Another option not listed". It ends "You've qualified" and "What's next: a provider will review your health information". OBSERVATION, 2026-10-31. https://mobbin.com/flows/c40ee07f-972a-43b2-83b4-666d7f6df86e → `/clinic/apply-treatment-rule` - `F-e2003b56db` L40 [OBSERVATION] _By task › Questionnaires and intake_: **Hers, iOS, "Complete treatment questionnaires (mental health)"** (29 screens). A consultation with a segmented progress bar. Answers carry a plain description (for example "Fair: I have friends, but I don't feel close to or supported by them"). A loading screen ticks off "gathering your responses, calculating your scores, preparing your results". A free-text box asks "anything else you want your provider to know?" with a tick box for "No, I've provided all relevant information". OBSERVATION, 2026-10-31. https://mobbin.com/flows/60448996-a8dd-4632-9b46-4ab745af1a4e → `/clinic/triage-mental-health-referral` - `F-0b584caa0a` L41 [OBSERVATION] _By task › Questionnaires and intake_: **Hers, iOS, "Completing treatment questionnaires (weight loss)"** (18 screens). Includes a screening question about eating disorders with the reason it is asked, marketing charts mixed into the questions, and a clear refusal screen: "you are not eligible... we recommend that you see a provider in person", with a way back. OBSERVATION, 2026-10-31. https://mobbin.com/flows/061e04f8-11e2-49c9-99ba-409a707e5f25 → `/clinic/mental-health-care-crisis` - `F-d30938d685` L42 [OBSERVATION] _By task › Questionnaires and intake_: **Superpower, iOS, "Taking lifestyle quiz"** (11 screens). "Let's understand your day-to-day: this helps us customise your plan", then questions on exercise and work, each with a one-line reason ("your job plays a big role in daily stress"). It ends by selling more tests. OBSERVATION, 2026-10-31. https://mobbin.com/flows/48c850ab-006f-4536-a9bc-1d4033c46c9c → `/clinic/set-surveillance-plan` - `F-faf19bdbcd` L46 [OBSERVATION] _By task › Mental health check-ins_: **Headspace, iOS, "Tracking anxiety levels"** (6 screens). A progress area with stress and anxiety tabs, an empty chart that says "Start tracking your anxiety over time with bi-weekly check-ins", and a seven-question check-in ("Over the last two weeks, how often have you been bothered by..." with four answers). After a check-in the button greys out with "Opens in 13 days". OBSERVATION, 2026-10-31. https://mobbin.com/flows/79577cba-b6b5-487c-ae41-5416af934661 → `/clinic/wellbeing-questionnaires` - `F-25a686afcc` L47 [OBSERVATION] _By task › Mental health check-ins_: **Headspace, iOS, "Tracking monthly stress reflection"** (6 screens). Ten questions, then a dated result: a score on a bar split into low, moderate and high with the number ranges shown, framed gently ("Think of this score as a way to get to know yourself a little better"), with suggested content and a link to questions and answers. OBSERVATION, 2026-10-31. https://mobbin.com/flows/ce032b9a-cb9f-42f9-b407-19a2796c6961 → `/clinic/front-door-check` - `F-7d422ecda3` L48 [OBSERVATION] _By task › Mental health check-ins_: **Headspace, iOS, "Profile"** (7 screens). Personal statistics, a streak described kindly ("Meditate on a schedule that feels good"), the check-in panel and activity history. OBSERVATION, 2026-10-31. https://mobbin.com/flows/8d8bd1bc-e2c4-4a04-b7f6-5ba7233321b6 → `/clinic/record-clinic-consultation` - `F-e130cdbc66` L52 [OBSERVATION] _By task › Test results_: **Bevel, iOS, "Record detail"** (6 screens). Health records filtered by type (lab report, medications, imaging), a nudge that some markers are due for a refresh, and uploaded documents "ready for review". A lab report opens with a plain summary of what is high, then each marker with its range and value. One marker carries an orange warning: "This biomarker may be incorrect. Please review the value and update it if needed", with Edit. OBSERVATION, 2026-10-31. https://mobbin.com/flows/bcb74637-8d36-4b58-8e62-99e88312cb98 → `/clinic/person-held-health-record` - `F-e9373335bb` L53 [OBSERVATION] _By task › Test results_: **Superpower, iOS, "Biomarker detail"** (3 screens). A list of results with small range bars, then one marker marked "Out of range", a chart over time with the healthy band shaded, latest result next to the target range, what it is, why it matters, what you can do, and a list of sources. OBSERVATION, 2026-10-31. https://mobbin.com/flows/0a31c382-c9dc-4f06-b681-f8bceee8e991 → `/clinic/wellbeing-questionnaires` - `F-b79ccd4e6d` L54 [OBSERVATION] _By task › Test results_: **WHOOP, iOS, "Labs summary"** (5 screens). A test offer, a summary ring with "optimal, sufficient, out of range", a promise of "clinician-reviewed insights", and an empty state for a marker: "No lab results here yet: upload past lab results or schedule a test". OBSERVATION, 2026-10-31. https://mobbin.com/flows/b7fceedb-b811-45ff-94f6-2760d2c4f382 → `/clinic/person-held-health-record` - `F-79f55b97a4` L55 [OBSERVATION] _By task › Test results_: **Ro, iOS, "Lab history"** (2 screens). Settings that include lab history, Apple Health and privacy settings; the empty lab history is used to sell tests. OBSERVATION, 2026-10-31. https://mobbin.com/flows/aebafd34-bd72-4d0c-bed0-b96557c23662 → `/clinic/record-clinic-consultation` - `F-2bd36ae620` L59 [OBSERVATION] _By task › Medication reminders_: **GoodRx, iOS, "Scheduling a daily med reminder"** (10 screens). "Add reminder details": a name, a time, days of the week, and switches for which medicines this reminder covers, with "you'll get a push notification on the days and time you select". A separate card offers refill reminders. OBSERVATION, 2026-10-31. https://mobbin.com/flows/dc7306f5-fc70-4af5-a256-95484befdc1b → `/clinic/send-appointment-letters` - `F-b340cc03a1` L60 [OBSERVATION] _By task › Medication reminders_: **GoodRx, iOS, "Daily med reminders"** (2 screens). The list of reminders with Edit and "Schedule new reminder". OBSERVATION, 2026-10-31. https://mobbin.com/flows/efcc6f42-2a09-45b6-951d-2810aae5c0eb → `/clinic/send-appointment-letters` - `F-a6ed04dbaf` L61 [OBSERVATION] _By task › Medication reminders_: **Hers, iOS, "Setting up a reminder"** (7 screens). The treatment page shows reminders on or off, refills left, a message from the provider team and how to take the medicine. The reminder sheet allows several times a day, and its small print says the reminder information "will not be shared with your Care Team. Contact your provider if you have questions." OBSERVATION, 2026-10-31. https://mobbin.com/flows/1396d89d-d0d5-4822-9df1-1211c80cec04 → `/clinic/message-care-team` - `F-166692d716` L62 [OBSERVATION] _By task › Medication reminders_: **Hims, iOS, "Setting a reminder"** (7 screens). Next refill date, a plain reason why a refill is not yet allowed, a reminders tile and "Snooze for 1 month". OBSERVATION, 2026-10-31. https://mobbin.com/flows/56b3a643-67eb-4930-a326-46b5f76dea0c → `/clinic/send-appointment-letters` - `F-e9986a1947` L66 [OBSERVATION] _By task › Wearables, scores and data sharing_: **Oura, iOS, "Readiness"** (6 screens). A daily score with a short, human message ("above your average today, how do you feel?"), a seven-day line, contributors labelled "Pay attention", "Fair" or "Good", and overnight heart rate and heart rate variability charts. OBSERVATION, 2026-10-31. https://mobbin.com/flows/b071a02f-88d4-4ee2-9a4e-ea9e9422bb56 → `/clinic/message-care-team` - `F-afc043001c` L67 [OBSERVATION] _By task › Wearables, scores and data sharing_: **Oura, iOS, "Readiness"** (10 screens). "Your first Readiness Score!" explains what the score means and that it improves as a baseline is learned; key measures as tiles; and "Recovery in progress: you've turned off Rest Mode... continue to take it easy", with a marker showing Rest Mode was on. OBSERVATION, 2026-10-31. https://mobbin.com/flows/cd0b2e3e-e462-4d9d-ba87-f0c214fe47ca → `/clinic/complete-mental-health-check` - `F-9c0e4f3c70` L68 [OBSERVATION] _By task › Wearables, scores and data sharing_: **Oura, iOS, "Daily readiness data"** (2 screens). A newer home with readiness, sleep, resilience and activity in one row and the message "your activity goal today is your baseline". OBSERVATION, 2026-10-31. https://mobbin.com/flows/1a062e29-239c-4971-94cd-292b42dfcc7c → **UNPLACED** - `F-976884f7bd` L69 [OBSERVATION] _By task › Wearables, scores and data sharing_: **Apple Health, iOS, "Share with someone"** (4 screens). An explainer in three short points (you're in control; they get a dashboard and notifications; only a summary is shared, encrypted, and you can stop at any time), "Share with Someone" or "Ask Someone to Share", and an honest error: "Unable to share health data: this person does not have a device that can view your data". OBSERVATION, 2026-10-31. https://mobbin.com/flows/f5de1f34-8e2a-4567-840d-eac966c7ef88 → `/clinic/person-held-health-record` - `F-c9403509cc` L70 [OBSERVATION] _By task › Wearables, scores and data sharing_: **Apple Health, iOS, "Sharing"** (3 screens). "Share with your doctor", apps and research studies with access, and "No one is sharing data with you". OBSERVATION, 2026-10-31. https://mobbin.com/flows/00d9fca9-9ce6-4ba4-8a91-e3141ce6e29d → `/clinic/set-data-sharing-choices` - `F-525ebe94f8` L71 [OBSERVATION] _By task › Wearables, scores and data sharing_: **Apple Health, iOS, "Data sources & access"** (3 screens). For one kind of data (steps): which apps may read it, which research studies may read it (both "None"), and which sources write it, in priority order, with an explanation of how one source is chosen. OBSERVATION, 2026-10-31. https://mobbin.com/flows/e19317bd-987a-4ace-bd7d-42048e3786b1 → `/clinic/manage-reporting-backlog` - `F-2ff4e3f702` L72 [OBSERVATION] _By task › Wearables, scores and data sharing_: **Apple Health, iOS, "Profile"** (3 screens). Medical ID, health checklist, health records, and privacy lists for apps, research studies and devices, with "Export All Health Data" and the time it last synced. OBSERVATION, 2026-10-31. https://mobbin.com/flows/40639fc3-6978-4b39-99bd-cd51db256933 → `/clinic/see-day` - `F-65c91d0a72` L73 [OBSERVATION] _By task › Wearables, scores and data sharing_: **Apple Health, iOS, "Searching apple health"** (4 screens). Browse by category (heart, medicines, mental wellbeing, symptoms, vitals), add a health records account, search results. OBSERVATION, 2026-10-31. https://mobbin.com/flows/a50e2c67-b7f5-4b5b-94f5-6ed71c1d224e → `/clinic/symptom-check-triage` - `F-6553627ef8` L74 [OBSERVATION] _By task › Wearables, scores and data sharing_: **Garmin Connect, iOS, screen.** Abnormal heart rate alerts with high and low limits, each with a plain note of when the alert fires and who it is not suitable for. OBSERVATION, 2026-10-31. https://mobbin.com/screens/65376645-1229-4e72-9733-94a8837147e4 → `/clinic/online-urgent-triage` - `F-ec0ed9ff8b` L78 [OBSERVATION] _By task › Talking to a clinician, and AI health assistants_: **Alan, iOS, "Talk with medical team"** (3 screens). "Talk with our medical team 24/7 by chat or video". The chat tab offers an AI ("Ask Mo") to "skip the wait", and lists specialties with expected reply times ("reply in about 30 minutes"), with a footer "Is it an emergency? Call 15". The video tab shows general medicine delivered by Livi ("6am to midnight, included, includes prescriptions") and therapy, with "Help is available: if you're in distress or having suicidal thoughts, reach out now". OBSERVATION, 2026-10-31. https://mobbin.com/flows/a633ff3f-4a5f-4a82-9c9f-3420c9ab2c78 → `/clinic/ask-help-now` - `F-c436ee856c` L79 [OBSERVATION] _By task › Talking to a clinician, and AI health assistants_: **Alan, iOS, "Consult"** (4 screens). A daily wellbeing page with small actions (walk, breathe, meditate), programmes run with physiotherapists and therapists, and topic tiles (mind, sleep, nutrition and more). OBSERVATION, 2026-10-31. https://mobbin.com/flows/5a2fbc6d-a598-4651-bac9-a191a52a02b3 → **UNPLACED** - `F-0fb8909832` L80 [OBSERVATION] _By task › Talking to a clinician, and AI health assistants_: **Oura, iOS, "Ask a doctor"** (2 screens). A "Medical chat" where an AI advisor from a named partner says it will use your ring data and that you can bring in a doctor; example questions; "Add a doctor: will reply after 7pm"; footer "Messages from the AI are informational. For medical advice, add a physician. For emergencies, call 911." A paid offer says physicians see your ring data and records. OBSERVATION, 2026-10-31. https://mobbin.com/flows/ad69f6c5-15bc-413f-bd86-8b459b972fc1 → `/clinic/ask-advice-guidance` - `F-82e000c66d` L81 [OBSERVATION] _By task › Talking to a clinician, and AI health assistants_: **Amazon Shopping, iOS, screen.** A health AI in beta that names itself, says "If this is an emergency, dial 911", offers "Talk to a provider", and shows a "Thinking" state with a stop button. OBSERVATION, 2026-10-31. https://mobbin.com/screens/ca1c14f1-cf34-4b27-b041-d8e30ebb332d → `/clinic/person-held-health-record` - `F-2960beab44` L85 [OBSERVATION] _By task › Crisis and urgent help_: **Wysa, iOS, screen.** A chatbot confirms "your safety plan has been created" and reminds the person to press SOS whenever they cannot cope; buttons to check the plan and add a therapist. OBSERVATION, 2026-10-31. https://mobbin.com/screens/cd4930f5-31e0-4c34-86a3-a6679e342cf7 → `/clinic/set-surveillance-plan` - `F-21ca71f632` L86 [OBSERVATION] _By task › Crisis and urgent help_: **Bloom, iOS, screen.** A "Get help" sheet with a crisis line button and a local hotline button. OBSERVATION, 2026-10-31. https://mobbin.com/screens/2cf0f6a8-c1fb-4919-8ab3-e53826be7387 → `/clinic/mental-health-care-crisis` - `F-4b7e180d30` L87 [OBSERVATION] _By task › Crisis and urgent help_: **inDrive, iOS, screen (transport).** An SOS page with the person's location on a map, the details the police will need, and one large call button. OBSERVATION, 2026-10-31. https://mobbin.com/screens/3f075a2f-0493-4d53-a905-00549b8b2fa8 → **UNPLACED** - `F-81ed66fa93` L91 [OBSERVATION] _By task › Food and lifestyle logging_: **MyFitnessPal, iOS, "Scanning a meal"** (4 screens). Today's calories and macros, quick ways to log (search, barcode, voice, photo of the meal), a scanned list of foods each with delete and "Add more", then "Log breakfast" and a "Breakfast logged!" message. OBSERVATION, 2026-10-31. https://mobbin.com/flows/e5e452d7-f91e-4939-9fa0-59ea87b7d47a → `/clinic/maintain-hazard-log` - `F-812890c018` L92 [OBSERVATION] _By task › Food and lifestyle logging_: **MyFitnessPal, iOS, "Log meal"** (7 screens). Search, saved meals and foods, "copy previous meal", and a hint to swipe to log yesterday's meal again. OBSERVATION, 2026-10-31. https://mobbin.com/flows/d58bea22-7f26-48ef-a6de-134d9183c5ad → `/clinic/search-recruiting-trials` - `F-47d2006368` L96 [OBSERVATION] _By task › Local community activity (nearest thing to social prescribing)_: **Meetup, iOS, screen.** Groups near a place with distance chips (2, 5, 10 miles), categories such as social activities and hobbies, and member counts. OBSERVATION, 2026-10-31. https://mobbin.com/screens/303b5f13-3a13-4b20-809c-f356d6993248 → `/clinic/find-recruiting-trial-near` - `F-a96603b426` L97 [OBSERVATION] _By task › Local community activity (nearest thing to social prescribing)_: **Nextdoor, iOS, screen.** Neighbourhood groups (for example help with housing) with member counts, and local posts and polls. OBSERVATION, 2026-10-31. https://mobbin.com/screens/9fa1c528-5530-4f11-bc24-0244ea7279fe → **UNPLACED** - `F-4ffd93b8a1` L98 [OBSERVATION] _By task › Local community activity (nearest thing to social prescribing)_: **Lex, iOS, screen.** A local feed filtered by events, new here, groups and friends; posts tagged "For free"; a group card with "Join now". OBSERVATION, 2026-10-31. https://mobbin.com/screens/36bffdb3-8d25-413f-9698-596d80c9c31a → `/clinic/report-side-effect` - `F-0bac76fa2b` L99 [OBSERVATION] _By task › Local community activity (nearest thing to social prescribing)_: **adidas Running, iOS, screen.** Challenges with participant counts and end dates, upcoming events ("starts in 28 days") and communities. OBSERVATION, 2026-10-31. https://mobbin.com/screens/e11d5383-ce5b-4038-b90a-39a9bf820ad7 → **UNPLACED** - `F-d60062aece` L103 _EDGE Clinic research: real product flows from Mobbin › What to match_: **Urgency first, in plain words.** Ada puts "seek emergency care now" at the top of the report in red; Alan, Oura and Amazon keep an emergency line visible in their health chats. EDGE Clinic must do at least this on every screen where a person describes a symptom. → **UNPLACED** - `F-271023f9b7` L104 _EDGE Clinic research: real product flows from Mobbin › What to match_: **"This is not a diagnosis", said once, clearly, at the right moment** (Ada at the end of questions and on every condition page), not buried in terms. → **UNPLACED** - `F-4cf0316ce5` L105 _EDGE Clinic research: real product flows from Mobbin › What to match_: **One question per screen, with a way back and the privacy link always visible** (Ada), and a reason given for sensitive questions (Hers, Superpower). → **UNPLACED** - `F-2579a98095` L106 _EDGE Clinic research: real product flows from Mobbin › What to match_: **Show what matched and how sure.** Ada lists which of your symptoms matched and a plain percentage of similar people. Superpower lists the sources behind its advice. → **UNPLACED** - `F-456522463f` L107 _EDGE Clinic research: real product flows from Mobbin › What to match_: **Let people correct the record.** Bevel flags a value that may be wrong and lets the person edit it. Clinic results and imported data need the same. → **UNPLACED** - `F-bfe0a38a38` L108 _EDGE Clinic research: real product flows from Mobbin › What to match_: **Spacing check-ins on purpose.** Headspace locks the check-in with "Opens in 13 days" and frames the score kindly with its ranges visible. → **UNPLACED** - `F-3db9f266d9` L109 _EDGE Clinic research: real product flows from Mobbin › What to match_: **Say who sees what.** Apple Health shares only a summary, can be stopped at any time, and lists every app and study that can read each kind of data. Hers says the reminder data is not shared with the care team. → **UNPLACED** - `F-ed63fab752` L110 _EDGE Clinic research: real product flows from Mobbin › What to match_: **Bookings that explain themselves.** State what to prepare (IKEA), why each piece of contact data is needed (IKEA, Selfridges), separate consent for reminders and for marketing (Crate & Barrel, Selfridges), the cost and cancellation rules before payment (Fresha, Headspace, Square), and the next steps after booking (Headspace, Crate & Barrel, Square). → **UNPLACED** - `F-0f83df7f48` L111 _EDGE Clinic research: real product flows from Mobbin › What to match_: **An honest "not for you" route.** Hers tells ineligible people to see a provider in person instead of leaving them stuck. → **UNPLACED** - `F-e6a8c8780c` L112 _EDGE Clinic research: real product flows from Mobbin › What to match_: **Booking for someone else** is a first-class choice (Zocdoc "Someone else"). → **UNPLACED** - `F-4199f0f7f3` L116 _EDGE Clinic research: real product flows from Mobbin › What none of them do_: **No AI agent is ever a user in its own right.** Every AI here (Ada, Alan's "Mo", Oura's advisor, Amazon's assistant, Wysa) is a feature inside one company's app, acting for the signed-in person. None has its own identity, none shows whose behalf it acts on, what it is allowed to do, or a record of what it did. No flow lets a family member's agent, a carer's agent or a link worker's agent book, ask or fill in a form with the person's permission. → **UNPLACED** - `F-b8d51cf1fc` L117 _EDGE Clinic research: real product flows from Mobbin › What none of them do_: **No hand-off to the GP or NHS.** Nothing sends a symptom report, check-in score or wearable trend to a named clinician with the person's consent, and nothing shows the person what the clinician received or did with it. Ada offers "Download report" only. → **UNPLACED** - `F-d7390d50d0` L118 _EDGE Clinic research: real product flows from Mobbin › What none of them do_: **No social prescribing.** No app suggests a local group, walk, class or service as a response to how someone feels, and none involves a link worker. Community apps (Meetup, Nextdoor) know nothing about health. → **UNPLACED** - `F-552835ce0e` L119 _EDGE Clinic research: real product flows from Mobbin › What none of them do_: **No organisation as a user.** No practice, charity or community group appears as an account that receives referrals or reports back. → **UNPLACED** - `F-06d5458f17` L120 _EDGE Clinic research: real product flows from Mobbin › What none of them do_: **Early detection stays inside one product.** Wearable apps show trends and alerts (Oura, Garmin) but none explains when a trend should lead to seeing a clinician, or offers to share that trend with one. → **UNPLACED** - `F-c623e6ec6d` L121 _EDGE Clinic research: real product flows from Mobbin › What none of them do_: **Confidence is rarely shown.** Apart from Ada's percentage and Bevel's "may be incorrect", nothing says how sure the system is or what it is based on. → **UNPLACED** - `F-0c0b14a3b9` L122 _EDGE Clinic research: real product flows from Mobbin › What none of them do_: **No UK public sector examples.** NHS App and Doctolib were not in the results, so UK identity checks (NHS login), GP booking and repeat prescriptions are not covered here. → **UNPLACED** ###### `research/clinic/REFERENCE-CLINIC.md` (7 findings) - `F-552743450c` L3 [OBSERVATION] _EDGE Clinic: the reference flows the application entry copies_: **Statement class.** OBSERVATION of real products in Mobbin, recorded 31 October 2026 in `MOBBIN-FLOWS.md` in this folder; this file names which of them the Clinic's entry screen is built from (Owner, 1 November 2026: layouts come from the references, the only difference being the estate's added controls or journeys). → **UNPLACED** - `F-2559e85161` L7 _EDGE Clinic: the reference flows the application entry copies › The entry (`/clinic`)_: **Ada, iOS, "Home"** (Mobbin flow 35a15c59-17c0-4872-a99d-610cc…, 3 screens): one main action on the home, "Start symptom assessment", above a short card. The Clinic's home keeps that shape: one primary action (Check your symptoms, `clinic/symptom-check-and-triage`), then the designed tasks as tiles. → **UNPLACED** - `F-65b6fde810` L8 _EDGE Clinic: the reference flows the application entry copies › The entry (`/clinic`)_: **Ada, iOS, "Starting a symptom assessment"** (17 screens): one question per screen with Previous and a close control, a privacy link on every screen. The Clinic's journey screens keep one question per screen (tools/typed_steps.py) and the help control. → **UNPLACED** - `F-0bfd6896b8` L9 _EDGE Clinic: the reference flows the application entry copies › The entry (`/clinic`)_: **Zocdoc, iOS, "Booking an appointment"** (10 screens): a provider page, a choice of location, tomorrow's slots as buttons, a review step that lets you book for someone else. `clinic/book-appointment` follows this order. → **UNPLACED** - `F-617b1c00d5` L13 _EDGE Clinic: the reference flows the application entry copies › What the estate adds, and only that_: The designed tasks as tiles under the primary action: front door check-in, medication safety check, mental health care and crisis, online urgent triage, person-held health record, retinal screening, social prescribing referral, symptom check and triage, wearable signals, wellbeing questionnaires (the Clinic's bespoke units in `design/figma/FRAMES.json`). → **UNPLACED** - `F-d19600606f` L14 _EDGE Clinic: the reference flows the application entry copies › What the estate adds, and only that_: Every other Clinic capability as the same tile beneath, so nothing is hidden and nothing is listed as prose. → **UNPLACED** - `F-2c2dc7b52e` L15 _EDGE Clinic: the reference flows the application entry copies › What the estate adds, and only that_: Jude's help behind the help control; evidence behind the evidence control (glass box). → **UNPLACED** ###### `research/clinic/REFERO-FLOWS.md` (950 findings) - `F-e2ab4c6bac` L3 _EDGE Clinic research: real product flows from Refero_: **What this is.** Real screens and user flows from existing products, recorded by Refero (a library of real app and website screenshots, each journey described step by step). They show the floor (what users already expect) and good customer experience from any industry worth borrowing. → **UNPLACED** - `F-a73e899327` L5 [OBSERVATION] _EDGE Clinic research: real product flows from Refero_: **Statement class.** Every flow is an OBSERVATION by Refero of a real product, searched 2026-10-31. Descriptions are Refero's wording. Product names belong to their owners. → **UNPLACED** - `F-a3e7811093` L7 _EDGE Clinic research: real product flows from Refero_: **How it was gathered.** Semantic searches for this application's tasks on iOS and web. Results from other industries are kept on purpose: they are often the best customer experience to learn from. Refero does not hold many incumbents (and few government, enterprise or NHS systems), so this is a sample, not a census. → **UNPLACED** - `F-bec922778e` L9 _EDGE Clinic research: real product flows from Refero_: **Products in this sample:** Ada (20), Fresha (15), District (14), WhatsApp (14), Doctronic AI (13), TidyCal (9), ALO (7), Cohere (6), Care (6), Lovable (5), Willow (5), Poppy (5), Stardust (5), Omio (5), Breathwrk (5), Airtable (5), Maze (5), Napper (5), Structured (5), Meta AI (4), Airbnb (4), Preply (4), Podimo (4), Train Fitness (4), ElevenMusic (4), GoFundMe (3), Wabi (3), Foodllama (3), Parallel (3), Raycast (3), Longevity Deck (3), GlossGenius (3), Delphi (3), fal (3), Tinder (3), Sunsama (3), Calendly (3), Craft (3), Tripsy (3), Peerlist (2), Genie (2), TIDAL (2), Pool (2), Google Gemini (2), Fingerprint (2), The Outsiders (2), Trip (2), Anam (2), Understory (2), komoot (2), Haptic (2), Seed (2), Nextdoor (2), Acuity Scheduling (2), Cal (2), IKEA (2), Cushion (2), Leonardo.Ai (2), Adobe Photoshop (2), Orbit (2), Wispr Flow (2), Vocabulary (2), OpenAI (2), n8n (2), Intercom (2), Cofounder (2), Kit (2), Blackbird (2), Loowner (2), Ground News (2), Kin (2), The Leap (2), ElevenReader (2), On (2), Instacart (2), Netflix (2), Unwind (2), Clay (2), MonoDesk (2), New Balance (2), Duolingo (2), Typeform (2), Kinhive (2), Givingli (2), Melius (2), Chance AI (2), Dropbox (2), Rork (2), Mozi (2), time2book (2), Introspect (1), Google Translate (1), Cashcaded (1), Queue (1), Subo (1), Asos (1), Apple Image Playground (1), Spyglass (1), PayPal (1), Gumroad (1), Pi (1), Binge (1), Jasper AI (1), Allbirds (1), Deezer (1), Acne Studios (1), Saint Laurent (1), West Elm (1), ElevenLabs (1), Weekrise (1), Voicenotes (1), Synthesia.io (1), Telegram (1), Comet (1), FocusFlight (1), Granola (1), Vinyls (1), Substack (1), RAD Weather (1), Remy (1), 222 (1), Pinterest (1), Shazam (1), Fuse (1), Riverside (1), Resend (1), Rox (1), Butter (1), BYBORRE® (1), Linear (1), Instagram (1), UNHINGED (1), Twitch (1), ChatGPT (1), Microsoft Teams (1), pillowtalk (1), Sora (1), GO Club (1), Teal (1), DailyArt Magazine (1), LinkedIn (1), Slite (1), Polaroid (1), Volv (1), Mercury (1), Farfetch (1), Manus (1), Prose (1), Trip Way (1), Joi (1), X (1), B150 (1), Zendesk (1), TikTok (1), Sky Guide (1), MyMind (1), Runway (1), Apple Maps (1), Athleats (1), Verse (1), Exoplan (1), Miles (1), HYPE (1), Marathon (1), Make (1), Homerun (1), TwelveLabs (1), Сontra (1), [untitled] (1), Weather mini (1), Navia (1), Tripadvisor (1), Uber (1), Brink (1), Uber Eats (1), Cursor (1) → **UNPLACED** - `F-e38d5624ab` L13 _EDGE Clinic research: real product flows from Refero › Task: AI agent setup permissions_: **Leonardo.Ai: Profile Settings Management** (web, 2 screens, Design & Media Editing, AI Tool) [Refero](https://refero.design/flows/12914) → **UNPLACED** - `F-fd43e991dd` L15 _EDGE Clinic research: real product flows from Refero › Task: AI agent setup permissions_: > This is a settings screen in Leonardo.ai focused on personal profile management. The entry point is the user already being logged in and opening Settings, with the left sidebar showing Personal Settings and Workspace Settings. The user can review profile information, choose interest tags, confirm adult content access, and optionally delete the account. The exit point is after changes are saved and reflected in the profile section with confirmation feedback, or after the user leaves settings. → **UNPLACED** - `F-23d618be50` L19 _EDGE Clinic research: real product flows from Refero › Task: AI agent setup permissions_: **WhatsApp: Create Community Setup** (ios, 6 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12677) → `/clinic/social-prescribing-referral` - `F-1baba959f5` L21 _EDGE Clinic research: real product flows from Refero › Task: AI agent setup permissions_: > Happy path for creating a new community on mobile. The user enters a community name, customizes the community image from the Edit control, selects either an emoji or sticker with a background color, then returns to the main creation screen with the chosen avatar and name ready to submit. → `/clinic/social-prescribing-referral` - `F-a88066b022` L25 _EDGE Clinic research: real product flows from Refero › Task: AI agent setup permissions_: > A three-step onboarding carousel presented as a modal on a mobile device that introduces the "Structured AI" feature. Each slide contains a top header with the feature name, a horizontal progress indicator, an illustration of the feature, a bold headline, a short explanatory blurb, and a primary 'Continue' CTA. Controls available across slides: a Close (X) in the top-right, a Skip action, and a back arrow on later slides to navigate backwards. Entry: modal opens in front of the main app (user already in the app). Exit: user advances through slides with Continue, can Skip or Close to dismiss the modal; final Continue proceeds past the onboarding (next screen not shown). Assumption: mobile device; modal invoked to introduce a new feature; user can navigate forward/back or dismiss at any time. → `/clinic/social-prescribing-referral` - `F-2733f6f588` L27 _EDGE Clinic research: real product flows from Refero › Task: AI agent setup permissions_: Steps: Intro — Speed Up Planning → Auto-Generate Your Schedule → Digitize Your Analog Planner → `/clinic/social-prescribing-referral` - `F-53228a9681` L31 _EDGE Clinic research: real product flows from Refero › Task: AI chat assistant_: **Doctronic AI: Passwordless AI Doctor Sign-In** (web, 8 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13706) → **UNPLACED** - `F-5ee8f64fb1` L33 _EDGE Clinic research: real product flows from Refero › Task: AI chat assistant_: > The flow begins on Doctronic’s public landing experience, where an unauthenticated visitor can start a chat or access account options. The user chooses email-based login, enters an email address, receives a six-digit verification code by email, and submits the code. After successful verification, the user is taken to the authenticated AI health assistant screen, ready to ask a health question. The user is assumed to be logged out and able to access the entered email inbox. → **UNPLACED** - `F-99b8ebefe1` L35 _EDGE Clinic research: real product flows from Refero › Task: AI chat assistant_: **Doctronic AI: Sleep Concern Assessment** (web, 13 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13694) → **UNPLACED** - `F-665305f577` L37 _EDGE Clinic research: real product flows from Refero › Task: AI chat assistant_: > A logged-in user named Connie starts a conversation with Doctoric AI by entering a health concern about recurring nighttime waking. The system displays an emergency disclaimer, processes the message, and requires acceptance of the Terms of Service and AI-provider disclaimer before continuing. The user then provides age and sex information, receives follow-up questions about additional symptoms and wake-up patterns, and answers questions about health history. The flow ends with the assistant asking about current medications, with the user remaining in the conversation. → **UNPLACED** - `F-cdb7aaa80d` L39 _EDGE Clinic research: real product flows from Refero › Task: AI chat assistant_: **Google Gemini: Voice Query to Suggestions** (ios, 6 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/3207) → **UNPLACED** - `F-c198614056` L41 _EDGE Clinic research: real product flows from Refero › Task: AI chat assistant_: > Happy-path: A signed-in user interacts with the Gemini conversational assistant on a mobile device. → **UNPLACED** - `F-527e4fdda9` L43 _EDGE Clinic research: real product flows from Refero › Task: AI chat assistant_: Steps: Conversation (idle with assistant intro) → Start Voice Input (listening) → Live Transcription (partial) → Full Transcription / Submit → Assistant Responding (audio playing) → Answer Rendered (suggestions with media) → **UNPLACED** - `F-050497c925` L45 _EDGE Clinic research: real product flows from Refero › Task: AI chat assistant_: **WhatsApp: Meta AI Voice Chat Setup** (ios, 6 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12673) → **UNPLACED** - `F-de8a4d4cb7` L47 _EDGE Clinic research: real product flows from Refero › Task: AI chat assistant_: > Happy path for starting a voice conversation with Meta AI. The user enters the Meta AI profile/settings area, opens the voice feature, reviews the informational consent-style modal, and continues into a live voice session. → **UNPLACED** - `F-80d4c543f4` L51 _EDGE Clinic research: real product flows from Refero › Task: AI chat assistant_: > A user views an assistant-generated response containing ingredients and a 20-minute pasta cooking method. The user copies the response using the copy action beneath the message. A confirmation toast appears stating that the content was copied. The flow starts with an existing assistant response and ends with the response available on the device clipboard. → **UNPLACED** - `F-997d7c6368` L59 _EDGE Clinic research: real product flows from Refero › Task: API keys developer settings_: > A short onboarding carousel for Willow shown to a new or returning user before entering the main experience. The user can swipe through several promotional intro screens and tap the primary Get started button to continue. → **UNPLACED** - `F-c310b7b1bd` L63 _EDGE Clinic research: real product flows from Refero › Task: API keys developer settings_: > Happy-path onboarding/settings sequence for a logged-in user named Connie. The user opens the profile/settings area, chooses a recommended connection type (Fitness), is shown a system Health permission request, and returns to the recommendation screen after granting access. Entry is from the profile/settings menu; exit is back to the recommendation step with the selected connection highlighted and a Next button available. → **UNPLACED** - `F-d702ae8699` L65 _EDGE Clinic research: real product flows from Refero › Task: API keys developer settings_: Steps: Profile and Settings → Connection Recommendations → Health Permission Request → Connection Selected → **UNPLACED** - `F-b8253315c4` L67 _EDGE Clinic research: real product flows from Refero › Task: API keys developer settings_: **komoot: New user onboarding and consent setup** (ios, 39 screens, Health & Wellness, Map & Navigation, Travel & Booking) [Refero](https://refero.design/flows/8281) → `/clinic/mental-health-care-crisis` - `F-7a8125476d` L69 _EDGE Clinic research: real product flows from Refero › Task: API keys developer settings_: > Happy-path onboarding for a first-time mobile (iOS) user. Entry: app launch -> privacy/tracking consent banner. User can Accept All, Refuse, or Customize preferences. If user customizes they can toggle categories and open vendor details. After consenting, the flow presents sign-in options (social or email). Choosing email leads to email validation then account creation (password). Post-signup the user progresses through product onboarding cards that request push notifications and access to device integrations (Apple Health). The flow ends with device connections (optional) and a completion screen, then the main map/routes home screen. → `/clinic/mental-health-care-crisis` - `F-50faecdc87` L71 _EDGE Clinic research: real product flows from Refero › Task: API keys developer settings_: Steps: Privacy notice (top-level) → Customize preferences (categories visible) → Category list (scroll) → Analytics expanded (vendor row visible) → Vendor info modal (Firebase) → Sign-in options → Enter email (initial) → Email validation error → Email accepted (proceed) → Create account (password) → Onboarding hero / intro → Onboarding card: Navigation → Onboarding card: Notifications → iOS push permission dialog → Device integrations list → Health Access (iOS modal) → Integration connected state → Onboarding completion → Main map / routes home → Optional: Connect additional devices (post-onboarding) → `/clinic/mental-health-care-crisis` - `F-4bda914af1` L73 _EDGE Clinic research: real product flows from Refero › Task: API keys developer settings_: **Poppy: Support and Privacy Settings** (ios, 6 screens, AI Tool, Productivity) [Refero](https://refero.design/flows/12194) → **UNPLACED** - `F-4939508559` L75 _EDGE Clinic research: real product flows from Refero › Task: API keys developer settings_: > A settings/support area presents account and privacy-related options, including feedback, data retention, privacy policy, terms, account deletion, experimental features, and sign out. Selecting certain items opens full-screen informational panels that explain what data Poppy collects, how it is used, and how long it is stored. Entry appears to be from an account/settings screen; exit is via the close button or back navigation. → **UNPLACED** - `F-adaf4ae428` L77 _EDGE Clinic research: real product flows from Refero › Task: API keys developer settings_: Steps: Support Menu → Data Overview → Location Details → Email Details → Health Details → Control and Deletion → **UNPLACED** - `F-153a94f24e` L81 _EDGE Clinic research: real product flows from Refero › Task: API keys developer settings_: > The flow starts from an “Add Uplink” provider selection screen and continues through the Ambient Weather credential setup. The user selects Ambient Weather, enters a personal API key, and tests the connection. The screenshots show the connection request loading and ending with a provider rejection error. The user is assumed to be on an iPhone and able to access their Ambient Weather API key. Credentials are described as being stored locally in Apple Keychain. → **UNPLACED** - `F-da47f1de5e` L83 _EDGE Clinic research: real product flows from Refero › Task: API keys developer settings_: Steps: Provider Selection → Enter API Key → Test Connection → Connection Check → Credential Rejection → **UNPLACED** - `F-7371439b4d` L87 _EDGE Clinic research: real product flows from Refero › Task: CRM contacts pipeline_: **222: Assign Close Friends** (ios, 11 screens, Lifestyle, Dating & Relationships) [Refero](https://refero.design/flows/9320) → **UNPLACED** - `F-f33319d9a8` L89 _EDGE Clinic research: real product flows from Refero › Task: CRM contacts pipeline_: > Happy path: The user refines the app's compatibility model by answering a series of persona-style prompts and assigning a real friend to each prompt. From the Close Friends screen the user taps a prompt's "choose a friend" input, picks or filters a contact from their phone list, and triggers an in-app share (pre-populated message with a referral/preview) to invite that friend to take the compatibility test. A confirmation modal appears before the assignment is finalized. Each confirmed assignment increments a progress counter (e.g., 0/5 → 1/5). → **UNPLACED** - `F-e9a205c019` L91 _EDGE Clinic research: real product flows from Refero › Task: CRM contacts pipeline_: Steps: Close Friends — prompt overview → Contact list with keyboard → Filter to a specific contact → Select contact to send referral → Message compose with referral preview → Confirmation modal (explain & confirm) → Next prompt — pick another friend → Message compose for second contact → Progressed question cards → Close Friends summary → Return to People tab / flow complete → **UNPLACED** - `F-3eef73f589` L95 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: **TIDAL: Track Report Reason Selection** (ios, 3 screens, Entertainment, Music Streaming) [Refero](https://refero.design/flows/13140) → `/consortium/view-public-reviewer-track` - `F-194627cff2` L97 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: > A mobile modal for reporting a track, starting with a searchable list of report reasons and ending with a selected reason shown after search. The user opens the report dialog, searches or browses categories, and selects the matching reason. → `/consortium/view-public-reviewer-track` - `F-5db083fc0a` L99 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: Steps: Browse Report Reasons → No Match Search State → Filtered Result → `/consortium/view-public-reviewer-track` - `F-3960f58324` L101 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: **Apple Image Playground: Report Generated Image** (ios, 9 screens, AI Tool, Design & Media Editing) [Refero](https://refero.design/flows/12798) → **UNPLACED** - `F-c6f9795f79` L103 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: > A user reviews a generated image in Image Playground, opens the overflow menu, and chooses Report a Concern. They select concern categories, optionally add more information, review what will be included in the report, accept the data-sharing prompt, and submit. The report then submits successfully and returns them to the image editor, where the generated image remains visible with the usual controls. → **UNPLACED** - `F-23567f9b1c` L105 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: Steps: Image Review → Action Menu → Concern Category → Personal Characteristics Check → More Information → Submit Report → Data Consent → Submitting → Return to Editor → **UNPLACED** - `F-dbd61de343` L107 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: **fal: Recent History Preview Visibility** (web, 2 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/13542) → `/clinic/record-clinic-consultation` - `F-39e7eb1cc0` L109 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: > The user opens the Recent History area while authenticated and reviews recent API requests across endpoints. Results are displayed in a four-column card grid with generated previews, request metadata, timestamps, status codes, prompts, and actions such as Playground, Copy prompt, and Model Analytics. The user can use the Show preview toggle to switch between visible generated images and hidden preview placeholders. The flow ends with the history list displayed in the selected preview state; pagination controls and the site footer remain available below the results. → `/clinic/record-clinic-consultation` - `F-68eaf17b98` L113 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: **District: Event Feedback Submission** (ios, 5 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12916) → `/consortium/submit-paper-consortium-review` - `F-f50a5001a1` L115 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: > Happy-path mobile feedback journey after attending an event. The user enters a rating screen for a specific event ('Sip and Act at 3 Art House'), selects an overall sentiment, optionally rates individual aspects, adds a text comment, and submits. → `/consortium/submit-paper-consortium-review` - `F-eecce4a3ac` L117 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: Steps: Overall Rating → Detailed Ratings → Comment Entry → Submission Confirmation → Completion State → `/consortium/submit-paper-consortium-review` - `F-ff1f6971c8` L119 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: **Cursor: Security Audit Assistant** (web, 18 screens, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/9922) → **UNPLACED** - `F-509bb1a642` L121 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: > Happy path: From the Agents landing the user composes a task card, selects a GitHub repository, configures model(s) and multipliers, optionally attaches an image, then starts the task which spins up a cloud agent. The agent appears in the left agent list and the UI navigates to a three-column agent workspace showing the activity log (center) and a file-changes pane (right). The agent runs commands (visible as generated commands / progress) and produces a final security audit report with findings and code snippets. → **UNPLACED** - `F-447870d631` L123 _EDGE Clinic research: real product flows from Refero › Task: access review permissions audit log_: Steps: Agents Landing — Task Composer → Repository Picker Open → Repository Search → Repository Selected → Open Model Selector → Enable Multiple Models → Set Model Multipliers → Compose Task Instruction → Attach Image Attempt → Image Preview Modal → Task Created — Agent Listed → Open Agent Conversation → Agent Starting — Generating Commands → CLI Execution Attempt → Agent Working — Additional Steps → Audit Report — Commits Reviewed → Critical Finding Detail → Final Summary and Checks → **UNPLACED** - `F-f3e30cd7bd` L129 _EDGE Clinic research: real product flows from Refero › Task: account closure_: > Happy path for closing a Care.com membership. The user starts on a membership management screen while signed in, chooses to close their account, reviews cancellation options and reason selection, then completes the closure and reaches a confirmation screen. → **UNPLACED** - `F-cf43d2389e` L137 _EDGE Clinic research: real product flows from Refero › Task: apply for loan credit_: > Happy path for a recipient responding to a friend’s product decision request on mobile. → **UNPLACED** - `F-0ff3a200a2` L141 _EDGE Clinic research: real product flows from Refero › Task: apply for loan credit_: **Lovable: Plans and credits management** (web, 3 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12301) → **UNPLACED** - `F-fd4092139f` L143 _EDGE Clinic research: real product flows from Refero › Task: apply for loan credit_: > Happy path for managing a workspace subscription from the Plans & credits area. Entry is the workspace settings sidebar on the Plans & credits screen. The user can review current plan status, compare Pro/Business/Enterprise options, and initiate an upgrade or credit purchase. If upgrading, the flow continues to a Stripe-style checkout where the user enters payment details and confirms the subscription. Exit is the updated Plans & credits screen showing the new active plan and refreshed credit balance. → **UNPLACED** - `F-fad846efc3` L145 _EDGE Clinic research: real product flows from Refero › Task: apply for loan credit_: Steps: Current Plan Overview → Checkout and Payment → Updated Subscription Status → **UNPLACED** - `F-4863a9059f` L149 _EDGE Clinic research: real product flows from Refero › Task: apply for loan credit_: > Mobile customer care entry that surfaces a prominent search box and categorized FAQ topics. Entry: user opens the Customer Care section (mobile browser). The primary path shown is: user enters a query in the help search (example: "visor") and is taken to a search results screen. Exit points: refine search and resubmit, select a FAQ category or question, tap header navigation (menu, account, basket), or close the help overlay. Assumptions: user is on a mobile browser (portrait), may be signed in or out (header shows account icon but no explicit sign-in state), and has basic browsing permissions. → **UNPLACED** - `F-63803c916b` L153 _EDGE Clinic research: real product flows from Refero › Task: apply for loan credit_: **District: Apply Offer Code** (web, 3 screens, Entertainment, Food & Drink, Travel & Booking) [Refero](https://refero.design/flows/13007) → `/finance/verify-identity-person` - `F-e4f6f4e769` L155 _EDGE Clinic research: real product flows from Refero › Task: apply for loan credit_: > The user is on an Offers screen during checkout or booking. They can either manually enter a coupon/promo code in the input field at the top or choose from a list of suggested offers below. The entry point is the offers panel opened from a broader booking/payment context, and the exit point is either a successful offer application or an inline validation error if the entered code is invalid. → `/finance/verify-identity-person` - `F-81c110f4d0` L157 _EDGE Clinic research: real product flows from Refero › Task: apply for loan credit_: **Lovable: Help and Resources Center** (web, 5 screens, Development Tools, AI Tool) [Refero](https://refero.design/flows/12206) → `/clinic/offer-self-help-resources` - `F-6abfd2e7e4` L159 _EDGE Clinic research: real product flows from Refero › Task: apply for loan credit_: > This is a help and support entry experience for Lovable. The user starts on a public support/resources landing page with navigation, a central help search prompt, resource cards, FAQs, and a build prompt at the bottom. The user can also open a centered search modal, enter a query or numeric text, and review an AI-generated response with suggested follow-up chips. Assumption: the user is on desktop and not required to be logged in to access the visible help content. → `/clinic/offer-self-help-resources` - `F-e832e4b2e8` L161 _EDGE Clinic research: real product flows from Refero › Task: apply for loan credit_: Steps: Help Landing Page → Search Modal Opened → Query Entered → Search Suggestions → Assistant Response → `/clinic/offer-self-help-resources` - `F-2abf766d71` L165 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **District: Salon Service Booking** (ios, 8 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12874) → **UNPLACED** - `F-9e87911d93` L167 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy path for booking a salon service and reaching payment method selection. The user starts on a salon profile, chooses to book a service, selects one service and an available date/time slot, reviews the appointment details, and opens the payment method picker. → **UNPLACED** - `F-597ea5bcae` L169 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **District: Event Booking Confirmation** (ios, 5 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12872) → **UNPLACED** - `F-f25e31b68b` L171 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path ticket booking for an event, starting on a booking review screen and ending on a confirmed order with an accessible QR code ticket. → **UNPLACED** - `F-77a8f4efce` L173 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Doctronic AI: Appointment Payment Confirmation** (web, 2 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13700) → `/clinic/book-appointment` - `F-7451fcdf52` L175 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > The user is on the final payment step of a doctor appointment booking. The appointment details are shown alongside payment options. The user can use saved payment information, enter card details manually, or choose another payment method. In the shown path, the user selects saved payment information, reviews the populated card and billing fields, and can confirm the booking for $39.00. The flow starts at the payment step and ends when the user submits the booking confirmation. The screenshots do not show the post-submission result. → `/clinic/book-appointment` - `F-6e3f2c7fc1` L177 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **District: Movie Booking Support Request** (ios, 7 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12840) → `/clinic/mental-health-care-crisis` - `F-49579b5a24` L179 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > A user opens a movie-support conversation that already contains a booking summary. They confirm the booking, are prompted to describe the issue, open the attachment menu, and grant photo-library access. A booking confirmation containing a QR code and ticket details is shown as supporting context. The user types a request to replace cheese popcorn with salt, then sends the message with the booking image attached. The flow starts in an existing support chat and ends with the issue submitted to the support assistant. → `/clinic/mental-health-care-crisis` - `F-ce32e531b2` L181 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Fresha: Salon appointment booking** (ios, 9 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4973) → `/clinic/add-appointment-calendar` - `F-350a79212a` L183 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path booking of a single salon service with optional handling for fully-booked dates. Entry: user browsing/search results and opening a provider card. Exit: appointment is confirmed and shown in a booking detail screen. Assumptions: user is signed in (booking flows show payment and confirmation), has location enabled (distance shown), and is using a mobile device. → `/clinic/add-appointment-calendar` - `F-5968ae3451` L185 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Search results — provider card → Provider services list → Select services → Select time — fully booked state → Select time — date with availability → Choose time slot → Pre-visit question → Review and confirm → Confirmation & booking details → `/clinic/add-appointment-calendar` - `F-53675d3f1b` L187 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Fresha: Add Appointment to Calendar** (ios, 3 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4974) → `/clinic/add-appointment-calendar` - `F-bdd91c9b52` L189 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path: a signed-in mobile user views a confirmed appointment and chooses to add it to a calendar. → `/clinic/add-appointment-calendar` - `F-3d76dd5908` L191 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Appointment details / Confirmed → Calendar selection modal → Device event editor (prefilled) → `/clinic/add-appointment-calendar` - `F-ce815ad237` L193 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Fresha: Search & Book — Venue Listings** (ios, 2 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4966) → `/clinic/book-interpreter` - `F-fc76d63211` L195 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > User performs a location-aware search for a service (Massage) with a selected date/time. The top of the screen shows a map and a persistent search bar (query + date/time). Below are filter chips (Sort, Price, Type) and a scrollable list of venue cards. Each card shows an image, name, rating, distance, available services with prices, and tappable time-slot buttons for booking. Happy path: user reviews a venue’s service and taps a time-slot to proceed into the booking flow. → `/clinic/book-interpreter` - `F-f8c1ed6974` L197 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Venue list with time slots → No-results / Clear search state → `/clinic/book-interpreter` - `F-a457bba07a` L199 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Fresha: Refine search and pick time** (ios, 19 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4958) → `/clinic/mental-health-care-crisis` - `F-45cf7d5904` L201 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > User starts on a results list for a previously-entered query (example: "Barbering") and opens the search/filter modal to refine the search. From the modal the user can change the treatment (Treatment or venue), location, date and time, or browse categories/recents. The user drills into each filter: selects a treatment type, picks or types a new location (including using current location), selects a date from a calendar, and chooses a time window (preset or custom). After applying filters the modal returns to the results list which updates to show venues matching the query and filters with available time slots. Assumptions: mobile (iOS-style) device; user is either unauthenticated or authentication state is not relevant to the flow; user navigates via modal sheets that stack and return to the search overlay and then to the results list. Entry: results list (search active). Exit: updated results list with applied filters and visible available times. → `/clinic/mental-health-care-crisis` - `F-476aad9e1c` L203 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Results list (initial) → Search / Filter modal (root) → Treatment search (open) → Treatment search — no match state → Treatment results (matches and venues) → Location picker (root) → Location search — no match → Location search — suggestions → Search modal updated (location chosen) → Date picker (open) → Date selection (select month / date) → Time picker (open) → Time selection — preset and custom → Time selection — picker list → Search modal (filters applied preview) → Top categories (browse) → Results list (filtered) → Select time slot / Book → Back navigation and re-open search → `/clinic/mental-health-care-crisis` - `F-b08c3442d4` L205 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Omio: Booking retrieval by reference** (ios, 4 screens, Travel & Booking) [Refero](https://refero.design/flows/5242) → **UNPLACED** - `F-57aa26598d` L207 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy path: A user on the 'Your bookings' screen opens a "Retrieve booking" modal, enters a booking reference and the email used for the booking, and submits to fetch the booking so it appears in their upcoming bookings. → **UNPLACED** - `F-701b0e404b` L209 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Bookings list (Upcoming) → Retrieve booking modal (empty) → Enter details and submit → Lookup error (no booking found) → **UNPLACED** - `F-4cb1e7b309` L211 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Longevity Deck: Add activity to daily timeline** (ios, 4 screens, Health & Wellness) [Refero](https://refero.design/flows/9895) → `/clinic/person-held-health-record` - `F-e5109f2df2` L213 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy path: From the day timeline (Yesterday) the user taps an empty timeslot (11) to add an activity. The app opens an "Add Activity at 11" picker with search, category filters, and a scrollable list of activities. The user picks an activity (optionally adjusting duration or amount in the inline numeric control) and taps the circular + control to add it. The picker closes (or the list dismisses) and the selected activity appears as a colored card in the 11 slot on the timeline. Entry: day timeline with empty slot. Exit: timeline updated with new activity card. Assumptions: user is signed in and on the daily timeline view, editing is allowed, device shown is a phone in dark mode. → `/clinic/person-held-health-record` - `F-e670fcabf3` L215 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Day timeline with empty slot → Activity picker — top of list → Activity picker — select activity → Updated timeline with activity card → `/clinic/person-held-health-record` - `F-25549753d8` L217 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Omio: Booking retrieval via Help Center** (ios, 4 screens, Travel & Booking) [Refero](https://refero.design/flows/5247) → **UNPLACED** - `F-e84a595cfc` L219 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Mobile Help Center flow for retrieving a booking or searching support content. → **UNPLACED** - `F-1f5f4b59bc` L221 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Customer Help Center landing → Search field focused → Inline empty search feedback → Full search results – empty state → **UNPLACED** - `F-69a6bbd695` L223 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Fresha: Appointment cancellation and management** (ios, 4 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4975) → `/clinic/add-appointment-calendar` - `F-32ed338f68` L225 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path: A logged-in mobile user views the details of a scheduled appointment, opens the management sheet to reschedule or cancel, confirms cancellation, and returns to the appointment details now showing a Cancelled state with options to rebook. → `/clinic/add-appointment-calendar` - `F-d12aada0e5` L227 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Appointment details (Confirmed) → Manage appointment (sheet) → Cancel confirmation → Appointment details (Cancelled) → `/clinic/add-appointment-calendar` - `F-0b421d8f5e` L229 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Genie: Weekend Plan Assistant** (ios, 7 screens, AI Tool, Social Networking) [Refero](https://refero.design/flows/12431) → **UNPLACED** - `F-2eecb5e6a0` L231 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > A mobile assistant conversation starts on a home/chat screen with suggested prompt chips and a message composer. The user types a planning request that mentions a friend, then sends it. The assistant responds by looking up the friend’s profile, generating a location-based recommendation, and showing nearby place cards. The flow assumes the user is already in the assistant experience and can type, mention people, and review suggestions. It ends with the assistant providing a drafted plan and a follow-up question about timing. → **UNPLACED** - `F-2c72420ca6` L233 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Assistant Home → Draft Message with Mention → Ready to Send → Message Sent → Profile Lookup → Recommendation Results → Plan Draft and Follow-up → **UNPLACED** - `F-806c6e904e` L235 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **District: Upcoming Movie Reminder** (ios, 4 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12870) → **UNPLACED** - `F-a0fe599a30` L237 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy path for setting a reminder on an upcoming movie and optionally choosing preferred theatres. The user enters from the Upcoming movies listing screen, where movie cards show a reminder icon. After tapping a movie’s reminder control, a confirmation state appears and may prompt theatre preference selection. The user can open a theatre picker modal, select up to three theatres in the current city, save preferences, and return to the listing. → **UNPLACED** - `F-5c8d4cda49` L239 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Fresha: Treatment / Venue search** (ios, 3 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4959) → `/clinic/mental-health-care-crisis` - `F-c5ca630ad2` L241 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > User opens a search modal (overlayed on a map) to find a treatment or venue for "Barbering." The happy path: user enters a query (or picks a recent/category), refines optional filters (location/date/time), then selects a treatment or a venue from the suggested list. → `/clinic/mental-health-care-crisis` - `F-bc5bd631ab` L243 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Search & filters → Type and refine → Review matching items → `/clinic/mental-health-care-crisis` - `F-e96caf8377` L245 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Fresha: Treatment & Category Search** (ios, 3 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/4960) → `/clinic/mental-health-care-crisis` - `F-6e08998512` L247 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Mobile (iOS) search & filter flow for finding a treatment or venue. → `/clinic/mental-health-care-crisis` - `F-cf531d8380` L249 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Search & filter modal → Treatment or venue selector (with suggestions) → No-match / Top categories → `/clinic/mental-health-care-crisis` - `F-d8071d4c51` L251 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Airbnb: Photographer booking** (ios, 4 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6436) → **UNPLACED** - `F-b013356ad4` L253 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path: A mobile user discovers a photographer listing, taps Reserve, picks a package, date, and time, then confirms booking and payment. Entry: provider detail card (listing page) with service cards and a sticky Reserve CTA. Exit: user reaches the Confirm & Pay screen with booking summary and payment method (ready to review/complete payment). Assumptions: Mobile (iOS-style) interface; user has access to a saved payment method (a card is visible on the confirm screen); availability is shown per-date and per-package. The UI uses a modal sheet for scheduling and horizontal time-chip selectors for available time slots. → **UNPLACED** - `F-07f2d2b635` L255 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Provider listing & Reserve CTA → Schedule modal — pick date and time → Schedule modal — time confirmed → Confirm and pay → **UNPLACED** - `F-367a3a1f2b` L257 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Omio: Ticket access and booking management** (ios, 2 screens, Travel & Booking) [Refero](https://refero.design/flows/5231) → **UNPLACED** - `F-35dcf14189` L259 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy path: A mobile iOS user opens the "Your journey" area and selects the Ticket tab to access their ticket and booking details. From this screen the user can: add the ticket to Apple Wallet, copy the booking number, or open management options (change booking, view cancellation options, view confirmation). → **UNPLACED** - `F-c134138abd` L263 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Airbnb: Guidebook creation and place recommendation** (ios, 20 screens, Travel & Booking, Real Estate) [Refero](https://refero.design/flows/6448) → **UNPLACED** - `F-3dcad2c4b7` L265 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path: Host navigates from the Listing editor -> Arrival guide section to open Guidebooks, creates/edits a guidebook (Maya's Guidebook), attaches a listing, edits the guidebook cover image using the photo library, then adds a place recommendation (searches for a restaurant), categorizes it and adds images and a tip. → **UNPLACED** - `F-0de0fa4d6b` L267 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Arrival guide overview → Guidebooks list → Guidebook editor (empty) → Choose listings modal (initial) → Select listing → Save listing selection → Guidebook editor (after listing linked) → Edit cover modal – pick image → Request photo library permission → Grant photo library access → Photo library picker → Cover updated preview → Add content modal → Place search UI → Select a place from results → Categorize place (required) → Add tip and images for place → Save place recommendation → Preview guidebook item → Return to Guidebooks list → **UNPLACED** - `F-9c2ec90f40` L271 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > This is a mobile assistant-style scheduling experience. It starts on a home screen with shortcut actions and a microphone-based command entry, then moves into event creation after the user chooses Create and Event. The user can pick a quick date option or enter a custom title, then refine the scheduled date and time on a calendar-style editor before confirming. → **UNPLACED** - `F-8edafe3cdc` L273 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Assistant Home → Choose Create Task → Select Event Type → Set Event Timing → Enter Event Details → Adjust Calendar Event → **UNPLACED** - `F-f71dec26c4` L275 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **WhatsApp: Schedule Call and Invite Participants** (ios, 8 screens, Collaboration Tools, Social Networking) [Refero](https://refero.design/flows/12683) → **UNPLACED** - `F-e8d3f76546` L277 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > A mobile calls interface opens on the Calls tab. The user can access recent calls, shortcuts for calling and scheduling, and create a new scheduled call. The flow starts from the Calls screen, moves into scheduled call creation, then into contact selection and chat confirmation, and ends back on the Calls screen with the new upcoming call visible. → **UNPLACED** - `F-6c4c175041` L279 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Calls Home → Calls Menu → Scheduled Calls Empty State → Schedule Call Details → Send To Recipients → Recipient Confirmed → Chat Event Confirmation → Calls Upcoming State → **UNPLACED** - `F-ac592f56dd` L281 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **TidyCal: Therapy Session Booking** (web, 6 screens, Business, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/12503) → `/clinic/hold-remote-session` - `F-fb8d5f775a` L283 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy path for a therapy provider booking site. The user starts on a therapist’s public landing page, reviews available services, selects a session type, chooses a date and time from either a calendar or list view, enters name and email in a confirmation dialog, and receives a booking confirmation screen. → `/clinic/hold-remote-session` - `F-c059b1cea1` L285 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Therapist Landing → Service Selection → Calendar Availability → Booking Details → Enter Contact Details → Booking Confirmed → `/clinic/hold-remote-session` - `F-53f8626a62` L289 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > User reaches the 'Book a private appointment' page (desktop layout). The page shows a left-hand form and a large image on the right. Country/region and store are shown as dropdown selectors at the top of the form, followed by Name, Email, Phone (optional) and Requested date & time (optional). The primary CTA at the bottom is 'REQUEST APPOINTMENT'. → `/clinic/hold-remote-session` - `F-c35d76525b` L291 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Open booking form → Fill contact details → Submit request and see confirmation → `/clinic/hold-remote-session` - `F-00c04bb720` L295 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path booking: an unauthenticated desktop user selects a service, chooses a date and time from a calendar, fills required contact fields, confirms the booking, and arrives at a confirmation screen with calendar export and reschedule/cancel options. → `/clinic/hold-remote-session` - `F-3f4b5728f0` L297 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Select Appointment Type → Choose Date & Time → Enter Contact Information → Review Filled Information → Appointment Confirmed → `/clinic/hold-remote-session` - `F-a6088ff688` L299 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Acuity Scheduling: Create and publish appointment type** (web, 8 screens, Business, Scheduling Tools) [Refero](https://refero.design/flows/9857) → `/clinic/onboard-wearable-device-type` - `F-b85b9d96d3` L301 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy path: A signed-in user managing appointment offerings starts on the Appointment Types list, chooses to create a new individual appointment type, completes details (name, description, duration, confirmation message), adds an image, selects calendars and optional resources/forms, then saves the new appointment type. → `/clinic/onboard-wearable-device-type` - `F-27f73f4d80` L303 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Appointment Types list → Details — initial → Details — completed → Image & Calendar selection → Image uploaded → Forms, Resources & Additional Options → Saving — loading → Appointment Types list — confirmation → `/clinic/onboard-wearable-device-type` - `F-01c5f39fb8` L305 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Fresha: Salon appointment booking** (web, 14 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/2379) → `/clinic/add-appointment-calendar` - `F-51deecfa95` L307 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path booking of a single service (Gel Manicure) at a listed salon (AMA Brixton). → `/clinic/add-appointment-calendar` - `F-7e4e9338f0` L309 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Salon detail (services overview) → Choose booking option → Select services (list) → Service details modal → Service selected (summary visible) → Select professional (grid) → Professional confirmed → Select time (calendar + slots) → Date view and slot selection → Full time list (scroll view) → Change/select professional from time screen (modal) → First-visit question → First-visit selected (example: No) → Review & confirm (payment entry) → `/clinic/add-appointment-calendar` - `F-99ada4023c` L313 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy path: a shopper uses the store locator to find a nearby Saint Laurent boutique, opens the store details, and requests a private in-store appointment via a right-side booking panel. → `/clinic/add-appointment-calendar` - `F-630fdd2610` L315 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Store locator results → Store detail (Grant Avenue) → Open appointment panel → Select date — no hours available → Select alternate date and time → Submit appointment — error response → `/clinic/add-appointment-calendar` - `F-0caa33cfbd` L317 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **TidyCal: Create Booking Invite** (web, 6 screens, Business, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/12502) → **UNPLACED** - `F-d00c8b4d4d` L319 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy path for creating a booking invite from the Bookings section. The user starts on the Bookings screen, opens the Create a booking panel, selects a booking type, timezone, date, time, and guest details, then sends an invite. After submission, the modal closes, a success toast appears, and the new invite is shown in the Invited list with its status and actions. → **UNPLACED** - `F-26818a8b5d` L321 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Invited Bookings Empty State → Booking Details Panel → Date and Time Selection → Guest Details Completed → Invite Sent Confirmation → Invited Bookings List → **UNPLACED** - `F-278dc85b4f` L323 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **West Elm: Design Appointment Scheduling** (web, 10 screens, Home & Decor, Marketplace) [Refero](https://refero.design/flows/1674) → **UNPLACED** - `F-799a67fedc` L325 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path summary: A desktop user requests a free Design Crew appointment. They start on the service overview, pick an appointment type (in-home, virtual, or in-store), select a local store (if applicable), choose a date and time from an availability grid, enter contact information, add project details, and receive an on-screen booking confirmation with a reference and follow-up questions. → **UNPLACED** - `F-5050e33a01` L327 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Appointment Type Selection → Find Local Store → Date & Time - Availability Overview → Select Time Slot → Change Date (Calendar) → Enter Contact Information → Info Required Modal → Confirm Contact Details → Add Project Details → Booking Confirmation & Follow-up → **UNPLACED** - `F-fdbef70a65` L331 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy path for creating a multi-service appointment from the weekly schedule on desktop. Entry: user is on the weekly calendar for a location (Laboratory) and a target date is visible (Feb). The user selects a time slot or the global + button to open a right-side "New appointment" drawer, chooses a client and provider, adds services (which updates the end time), and books the appointment via a confirmation modal. Exit: the appointment appears on the calendar (block with hashed buffer) and a success toast is shown. Assumptions: user is authenticated and has permission to create appointments, the interface is desktop with sufficient horizontal space for a right-side drawer and central calendar, provider list and client list are prepopulated. → **UNPLACED** - `F-c94f43d55c` L333 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Weekly calendar view (initial) → Select start time → Open New appointment drawer → Choose client → Open services selector → Select multiple services → Review assembled appointment → Open booking confirmation modal → Toggle notification and confirm → Appointment created (side panel) → Calendar updated with appointment → **UNPLACED** - `F-82288750a7` L337 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path booking: A visitor opens a host's public booking listing (desktop, not signed-in) and sees a list of offered meeting types (durations and descriptions). The visitor selects a meeting type (30 Min Meeting), picks an available date from a month calendar, chooses an available time slot, fills contact information and optional notes, and confirms the booking. → **UNPLACED** - `F-730df4b55f` L339 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Meeting types list → Confirm meeting type selection → Month calendar — pick a date → Choose time slot → Booking details form → Confirmation summary → `/clinic/book-appointment` - `F-3c8425742c` L341 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Fresha: Group appointment and waitlist booking** (web, 30 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/2375) → `/clinic/add-appointment-calendar` - `F-aa16523ace` L343 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path booking attempt for a group appointment (up to 3 guests) starting from a salon’s listing. → `/clinic/add-appointment-calendar` - `F-0ac5b5ec22` L345 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Salon overview & services → Booking type chooser → Select services – master list → Service options dialog (Lash Lift) → Add additional services (e.g., Level 5) → Add guests and services summary → Edit services for a guest → Confirm selected service(s) → Join the waitlist – date/time entry → Join the waitlist – time range picker → Add multiple preferred times → First visit intake question → Review & confirm waitlist – payment and notes → Waitlist error – unavailable → User attempts to exit booking → Exit confirmation decision → Booking type chooser (alternate view) → Service options selection confirmation → Multiple preferred time rows shown in summary → Final error state and guidance → `/clinic/add-appointment-calendar` - `F-d7b1867f6e` L347 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Doctronic AI: Licensed Doctor Appointment Booking** (web, 25 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13699) → `/clinic/add-appointment-calendar` - `F-8c5fd7b2b5` L349 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > The user begins with a saved Doctoronic medical report and chooses to continue to a licensed doctor. They select a payment method, provide the state where the appointment will occur, choose an immediate or scheduled appointment time, enter patient and contact details, and reach the final booking confirmation and payment step. The flow assumes the user has a report ready to share and may have saved contact or payment information. The visible sequence ends on the payment screen before a completed booking confirmation is shown. → `/clinic/add-appointment-calendar` - `F-769229f965` L351 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **TidyCal: Booking Types Management** (web, 3 screens, Business, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/12493) → `/clinic/build-clinic-template` - `F-f6d8a7f94a` L353 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > This is a booking types management screen in a scheduling admin area. The user can view a list of booking types, filter them by status, enable or disable a type, open an actions menu for each item, and share a booking page link through a share panel. → `/clinic/build-clinic-template` - `F-02d636a9e7` L355 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Booking Types List → Item Actions Menu → Share Booking Link → `/clinic/build-clinic-template` - `F-854d959c3f` L357 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **TidyCal: Therapist Booking Landing Page** (web, 2 screens, Business, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/12486) → **UNPLACED** - `F-9523c1a63c` L359 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > A therapist’s public booking landing page is shown in two visual theme variations. The user is assumed to be a site owner or editor previewing the page in an editing environment, with access to theme, typography, color, and save controls. The page begins with a hero introduction, then scrolls through credibility stats, bio, service offerings, testimonials, FAQs, and a final booking call to action, ending with footer/social link placeholders. The primary exit point is the booking button, which leads to scheduling, while the editor can also save the design changes from the top bar. → **UNPLACED** - `F-b84eeab0d2` L363 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Fresha: Search availability by treatment, location, date, time** (web, 8 screens, Health & Wellness, Travel & Booking) [Refero](https://refero.design/flows/2323) → `/clinic/search-recruiting-trials` - `F-df2ae5c92f` L365 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > User starts on their Appointments screen (desktop, logged-in — profile avatar and an upcoming confirmed appointment are visible). From here they use the persistent search bar at the top to find new availability: they open the Treatment selector and pick "Tattoo & piercing", open Location and choose a saved address (Home), open the Date picker and choose a specific day (Dec 23), then set a custom time window (16:00–21:00). After applying filters and using Search, the interface displays search results on a split layout: a scrollable list of venues and available time slots on the left and an interactive map on the right. → `/clinic/search-recruiting-trials` - `F-589b75ee5d` L367 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Appointments overview → Choose treatment category → Open location selector → Open date picker → Select a calendar date → Open time filter presets → Set custom time range → View filtered search results → `/clinic/search-recruiting-trials` - `F-32c43d8e0c` L369 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Cal: Request reschedule for booking** (web, 5 screens, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/3493) → **UNPLACED** - `F-69754bdab0` L371 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy-path: A signed-in user views their Upcoming bookings list, opens a booking's Edit dropdown, chooses "Request reschedule," types an optional reason in a modal, and submits the request. Entry: Bookings > Upcoming (desktop, authenticated user). Exit: modal closes and a transient confirmation toast is shown: "Reschedule request sent." The interface guides the user with an Edit dropdown → Request reschedule modal that explains the effect (cancel meeting and notify scheduler) and provides an optional free-text reason. → **UNPLACED** - `F-06446b0f59` L373 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Bookings list (Upcoming) → Open Edit dropdown → Request reschedule modal (empty) → Enter reason and submit → Confirmation toast → **UNPLACED** - `F-fd2829aa3f` L375 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **IKEA: Interior design appointment booking** (web, 9 screens, Marketplace, Home & Decor) [Refero](https://refero.design/flows/8270) → `/clinic/add-appointment-calendar` - `F-dabbe55703` L379 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Service landing → Enter postcode → Choose service → Service details (expanded) → Select date (calendar) → Confirm timeslot (alternate state) → Contact & project details (form start) → Contact form (filled state) → Booking confirmation & payment prompt → `/clinic/add-appointment-calendar` - `F-d2f3dc86b3` L381 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **TidyCal: Booking Page Setup and Scheduling** (web, 5 screens, Business, Productivity, Scheduling Tools) [Refero](https://refero.design/flows/12499) → **UNPLACED** - `F-7fb8a38555` L383 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > Happy path for configuring and sharing a booking page, then reviewing available booking types and availability. The flow appears to start from a dashboard/booking setup area after the account is already live, with the user signed in as the business owner. It exits on a booking-type preview/scheduling screen where the user can choose a date and time or continue editing shared booking settings. → **UNPLACED** - `F-51c6b3e4d0` L385 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: Steps: Setup Checklist → Availability Calendar → Theme Selection → Availability Review → Booking Type Review → **UNPLACED** - `F-1737f8d684` L387 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: **Doctronic AI: Appointment Document Upload** (web, 4 screens, AI Tool, Health & Wellness) [Refero](https://refero.design/flows/13702) → `/clinic/person-held-health-record` - `F-ee7ca0cc02` L389 _EDGE Clinic research: real product flows from Refero › Task: book GP appointment_: > A logged-in user opens the Appointments area and views an upcoming video appointment with booking in progress. They open the Files section to review the file-upload prompt, add documents, and wait for the upload to complete. The uploaded PDF files then appear in the appointment's file list, where each file has a remove control. → `/clinic/person-held-health-record` - `F-64ea87da8e` L395 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: > This is a mobile analysis experience where the user enters a personal feeling, dream, or life situation, then submits it for interpretation by a panel of analysts. → `/clinic/person-held-health-record` - `F-bd8bba3dcc` L397 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: Steps: Prompt Entry → Text Composed → Submit Analysis → Processing State → Results Begin → Analyst Responses → Detailed Reading → Expanded Analysis View → `/clinic/person-held-health-record` - `F-69874d5d5f` L401 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: > Happy path starts from a saved or viewed place detail screen for Citizens Bank Park, showing venue information, actions like call/website/tickets, and a directions button. The user can open navigation from the place detail view, which transitions into the maps interface focused on Citizens Bank Park and nearby points of interest. → `/clinic/person-held-health-record` - `F-39276c11d7` L405 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: **Spyglass: Export and Rate Analysis** (web, 3 screens, Business, Inspiration Board) [Refero](https://refero.design/flows/13461) → `/consortium/judiverse-included-edge` - `F-441c352c6f` L407 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: > A user reviews a generated advertising analysis in the Spyglass workspace and interacts with the response action toolbar. The sequence begins with the completed response visible, then shows a PDF download confirmation, a hover tooltip for the positive-feedback control, and finally the positive-feedback control selected. The user appears to be logged in and viewing an existing conversation or research workspace. The flow ends with the response retained in the conversation and feedback recorded. → `/consortium/judiverse-included-edge` - `F-b6604ff4b1` L409 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: Steps: Review Generated Analysis → Identify Helpful Response Control → Submit Positive Feedback → `/consortium/judiverse-included-edge` - `F-0932664a4b` L413 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: > A logged-in user on GoFundMe searches for “paris” from the top search field and browses result types using the category filter. The initial results show nonprofit organizations with a secondary location filter. The user opens the category dropdown and changes the result type to Profiles, which updates the results grid to people profiles matching the same query. Entry is the global search/results view; exit is either selecting a result or refining the search/filter. → `/consortium/judiverse-included-edge` - `F-fb2fbf88ef` L419 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: > This is a guided content experience in the Foodllama mobile interface. The user starts on a home/dashboard screen showing a featured insight card and a progress summary for today’s insights. Tapping or swiping into the card opens a full-screen lesson sequence with multiple steps, each presenting one nutrition tip with illustration, title, short explanation, and a progress indicator plus a Leave action. The sequence ends on a completion screen with celebratory confetti, a trophy graphic, and options to tell friends or finish. After completion, the user returns to the dashboard, where the insight count updates. Assumptions: user is already inside the app, on a mobile device, and viewing personalized insight content. → `/consortium/judiverse-included-edge` - `F-db43e66227` L421 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: Steps: Insight Card Dashboard → Intro Tip Screen → Carb Tip Screen → Strength Tip Screen → Completion Celebration → Return Dashboard → Updated Insight State → `/consortium/judiverse-included-edge` - `F-c79697fb15` L425 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: > A signed-in user views a personal viewing breakdown screen, then opens a display filter to switch between All, Movies, and Shows. The screen updates the chart and stats based on the selected category. Entry is from a stats or profile area; exit is via back navigation. → `/consortium/judiverse-included-edge` - `F-347ac635d5` L429 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: **District: Restaurant Q&A Assistant** (ios, 7 screens, Entertainment, Food & Drink, Lifestyle, Travel & Booking) [Refero](https://refero.design/flows/12846) → **UNPLACED** - `F-79218c52a8` L431 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: > Happy path for an in-context restaurant question assistant. The user starts from a restaurant detail context and asks about Hotel Rishi. The interface shows the current topic, displays the user’s question as a highlighted prompt, and returns an AI-generated response with related action cards (Food Menu, Bar Menu, Beverage Menu). The user can rate the answer, ask a follow-up by text or voice, or submit a new query such as finding nearby breakfast options. The flow ends with a refreshed response and the follow-up input ready for another question. → **UNPLACED** - `F-9273f509e0` L435 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: > User starts on the assistant home screen, enters a query about free things to do nearby for the weekend, and waits while the system searches and composes an answer. The response appears in the main conversation area with a right-side panel that can switch between Steps and Sources. From there, the user can review how the answer was assembled, inspect cited sources, and hover or click source items to preview the referenced content. → **UNPLACED** - `F-db511a0ed4` L437 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: Steps: Home Prompt → Searching State → Steps Panel Opened → Draft Answer Displayed → Citation Hover → Sources Tab Opened → Source Preview Card → Image Preview → Source Link Hover → **UNPLACED** - `F-d86ef272ab` L439 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: **Cohere: Market overview chat** (web, 3 screens, Business, Development Tools, Productivity, AI Tool) [Refero](https://refero.design/flows/3856) → **UNPLACED** - `F-d4d0f0a214` L441 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: > Happy path: A user composes a natural-language prompt in the Cohere Playground chat area, optionally configures JSON mode/schema and advanced parameters from the right-hand panel, sends the prompt, and reads the generated market overview. Entry: user lands on the Playground with Chat selected and an empty chat editor. Exit: the user has a multi-section, generated response (market size, drivers, innovations, challenges) visible in the chat and the editor ready for follow-up. Assumptions: user is on a desktop viewport, logged in, and has access to model/parameter controls on the right-hand panel. → **UNPLACED** - `F-7810cfad43` L443 _EDGE Clinic research: real product flows from Refero › Task: budget spending insights_: Steps: Prompt composition & schema setup → Submit prompt and view error / initial generation → Review expanded assistant response → **UNPLACED** - `F-ae67b23741` L447 _EDGE Clinic research: real product flows from Refero › Task: business account onboarding company verification_: **Peerlist: Create Company Profile** (web, 6 screens, Job Search, Business, Social Networking) [Refero](https://refero.design/flows/1102) → **UNPLACED** - `F-0fa7320fc1` L449 _EDGE Clinic research: real product flows from Refero › Task: business account onboarding company verification_: > Happy path: A logged-in user (with permission to create a company) initiates creating a company profile via a modal. They choose an organization profile type, complete the profile form (display name, website, tagline, location, handle, size, etc.), then trigger an email verification step. After entering the verification code and accepting terms, the profile is created and the user is taken to the newly created company profile overview. → **UNPLACED** - `F-7243536630` L451 _EDGE Clinic research: real product flows from Refero › Task: business account onboarding company verification_: Steps: Choose profile type → Company profile form (empty) → Company profile form (filled) → Open verify work email modal → Enter verification code → Company profile overview → **UNPLACED** - `F-23af7fde4f` L455 _EDGE Clinic research: real product flows from Refero › Task: calendar scheduling meeting_: **Genie: Chat Reminder Setup** (ios, 3 screens, Social Networking, AI Tool) [Refero](https://refero.design/flows/12426) → **UNPLACED** - `F-9580605386` L457 _EDGE Clinic research: real product flows from Refero › Task: calendar scheduling meeting_: > A chat-style assistant screen where the user types a reminder request, sends it, and receives confirmation that a reminder has been created for tomorrow at 9:00 AM. The flow begins in an active conversation with the keyboard open and ends with the assistant confirming the reminder and offering to save reference photos to a vault. Assumptions: user is already in the conversation, signed in, and has permission to create reminders. → **UNPLACED** - `F-d084dc18e6` L461 _EDGE Clinic research: real product flows from Refero › Task: calendar scheduling meeting_: **Cofounder: Daily Meeting Summary Test** (web, 5 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/7900) → **UNPLACED** - `F-9416f0376a` L463 _EDGE Clinic research: real product flows from Refero › Task: calendar scheduling meeting_: > Happy path: a user viewing a configured scheduled automation called "Daily Meeting Summary Email" starts a manual test of the scheduled trigger, watches the test run initialize, and inspects the run timeline as integrations (Calendar) and internal reasoning steps are executed. → **UNPLACED** - `F-304061237d` L465 _EDGE Clinic research: real product flows from Refero › Task: calendar scheduling meeting_: Steps: Plan view — Automation details → Test confirmation modal → History view — Test run initializing → History view — Run processing progress → Run timeline — Integration and reasoning details → **UNPLACED** - `F-424d7c427a` L469 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Poppy: Poppy Account Setup and Subscription** (ios, 42 screens, Productivity, AI Tool) [Refero](https://refero.design/flows/12176) → **UNPLACED** - `F-1797670f0a` L471 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > A first-time user sets up Poppy on an iPhone. The user enters a phone number, verifies it with an SMS code, provides a preferred name, selects the areas where Poppy can help, and configures notification behavior. The user is then shown Sprout and Bloom subscription options, selects the Bloom annual plan, and completes the purchase through Apple Pay. After confirmation and a short preparation state, the user reaches the personalized Poppy dashboard. The screenshots show a successful paid onboarding path; the visible “Not now” option on the subscription screen is an alternative exit that is not followed. → **UNPLACED** - `F-b530f4b6c0` L473 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Welcome and Phone Entry → Phone Number Input → Submit Phone Number → Phone Verification Request → Enter Verification Code → Complete Verification → Name Entry → Confirm Name → Choose Support Areas → Confirm Support Preferences → Customize Daily Outreach → Sprout Subscription Option → Bloom Subscription Option → Compare Subscription Features → Confirm Bloom Annual Plan → Apple Pay Subscription Review → Authorize Subscription → Dismiss Purchase Confirmation → Prepare Personalized Experience → Personalized Dashboard → **UNPLACED** - `F-2950cb8716` L475 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Ada: Symptom assessment - conversational** (ios, 31 screens, Health & Wellness) [Refero](https://refero.design/flows/4097) → `/clinic/symptom-check-triage` - `F-9cd8da67dc` L477 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Mobile, conversational symptom assessment starting from a 'New assessment' entry screen (user name shown). The user is guided through: confirming who the assessment is for, entering a main symptom (search with autocomplete), answering a short series of multiple-choice questions, optionally opening explanatory anatomy/help modals, and receiving a results report with ranked possible causes. → `/clinic/symptom-check-triage` - `F-862504b6a5` L479 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Intro / Start assessment → Select assessment recipient → Prompt for main symptom → Symptom search modal (keyboard) → No results fallback → Autocomplete suggestions → Symptom detail and selection → Duration question → Orthostatic effect question → Open explanation modal (example) → Explanation modal — confirmation → Explanation modal (neck mobility) → Neck movement question → Reduced mobility illustration (detail) → Concentration question → Symptom trend question → Assessment complete — report being generated → Open report CTA → Results summary / possible causes → Detailed causes list and continue → `/clinic/symptom-check-triage` - `F-561816f983` L481 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Ada: Record a symptom entry** (ios, 12 screens, Health & Wellness) [Refero](https://refero.design/flows/4107) → `/clinic/person-held-health-record` - `F-12b38d63b4` L483 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Happy path: From a feedback prompt the user chooses to track symptoms, picks a symptom (Dizziness), answers a short guided questionnaire (trigger: standing up effect; description type), optionally sets current severity using a segmented color-coded slider and +/- controls, optionally adds freeform notes, saves the entry, then returns to a timeline where the new entry can be opened to view details. → `/clinic/person-held-health-record` - `F-cf0317158e` L485 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Feedback prompt → Choose symptom → Initial symptom questions → Select trigger effect → Select symptom description → Open severity control → Set severity (low) → Set severity (mid) → Set severity (high) → Open notes field → Enter free-text note → Save and view entry → `/clinic/person-held-health-record` - `F-4b498d767a` L487 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Ada: Profile setup and symptom logging** (ios, 66 screens, Health & Wellness) [Refero](https://refero.design/flows/4095) → **UNPLACED** - `F-323d166402` L489 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Happy path for adding a new user profile and logging a symptom (dizziness) into the symptom tracker. → **UNPLACED** - `F-6d15f7e28f` L491 _EDGE Clinic research: real product flows from Refero › Task: care plan_: Steps: Home / Start assessment → Add profile — Intro & safety note → Ask for user name → Name input active → Ask sex assigned at birth → Why only two sexes (modal) → Ask date of birth → Ask smoking history → Ask high blood pressure history → Open symptom modal (Dizziness) → Dizziness — trigger & type → Dizziness — optional severity collapsed → Set severity — low value → Set severity — mid value → Set severity — high value → Optional notes — edit → Save symptom (network notice) → Symptom tracking timeline → Symptom detail summary → Finish / Done → **UNPLACED** - `F-44e2e314a5` L493 _EDGE Clinic research: real product flows from Refero › Task: care plan_: **Ada: Medical quality article walkthrough** (ios, 10 screens, Health & Wellness) [Refero](https://refero.design/flows/4104) → **UNPLACED** - `F-4365b0b5bc` L495 _EDGE Clinic research: real product flows from Refero › Task: care plan_: > Happy path: A user finishes a symptom assessment and sees a high-priority recommendation ("Seek emergency care") with an inline "Tell me more" button. Tapping that link opens Ada’s external "Medical quality" article (about.ada.com) in a browser/webview. The user scrolls through the article: headli _… 6021424 more characters in the zip._ ## 10. Metrics ### ledger/closure/closure.json `ledger/closure/closure.json` · sha256 `b7e4858f4891668cf68956284f6e4fe4f313dcbd1194c59c755eeda00dd23c6c` ```json { "C": 149682, "R": 197914, "checks": { "fabricated_evidence_rejected": { "exit_code": 1, "ok": false, "tail": [ "FAILED (failures=1)" ], "value": 10 }, "findings_without_disposition": { "ok": false, "old_estate": 86, "placed_without_unit": 11351, "value": 11437 }, "r_minus_c_outside_plug_set": { "ok": false, "value": 48140 }, "reproducibility_differences": { "commit": "2746888688eefc18d0ebf5f004e3f5ef4ba5c624", "head": "d594aa71f06592d467dc430d2362b37d8ff0b96c", "ok": false, "value": 5, "why": "reproduce.json was made for another commit" }, "unadmitted_scope_changes": { "exit_code": 0, "ok": true, "value": 0 }, "unbound_required_operations": { "ok": true, "value": 0 }, "unreadable_sources": { "ok": true, "value": 0 }, "views_drift": { "ok": true, "tail": [ "}" ], "value": 0 } }, "evidence_files": { "ledger/closure/dispositions.jsonl.gz": { "exists": true, "rows": 141878, "unparseable": 0 }, "ledger/evidence/agent-journeys.jsonl": { "exists": false, "rows": 0, "unparseable": 0 }, "ledger/evidence/journeys-*.jsonl": { "exists": false, "rows": 0 }, "ledger/evidence/operations.jsonl": { "exists": true, "rows": 8534, "unparseable": 0 }, "spec/receivers.json": { "exists": true, "rows": 8534 } }, "examples_unmet": { "agent-journey": [ "journey:clinic/accept-a-proxy-mandate/agent", "journey:clinic/acquire-and-send-images/agent", "journey:clinic/act-on-a-drug-safety-alert/agent", "journey:clinic/act-on-a-high-risk-answer/agent", "journey:clinic/add-an-appointment-to-my-calendar/agent" ], "control": [ "control:clinic/accept-a-proxy-mandate/done/finish", "control:clinic/accept-a-proxy-mandate/evidence/inspect", "control:clinic/accept-a-proxy-mandate/step-1/back", "control:clinic/accept-a-proxy-mandate/step-1/continue", "control:clinic/accept-a-proxy-mandate/step-2/back" ], "decision": [ "decision:01m68c87qmz54aqnbm8pqzmgt4", "decision:01m696xzzpcbnsexpcdtz7ak9n", "decision:01m69b9nqqnxj7bh7p0mner5rf", "decision:01m69gty59gnpc17g811317htn", "decision:01m69jxxj5cazjefrrj1fvb962" ], "finding": [ "finding:F-000252d443", "finding:F-0004647f41", "finding:F-000b8cc142", "finding:F-000bb2fca4", "finding:F-000c4d5d1b" ], "human-journey": [ "journey:clinic/accept-a-proxy-mandate/engineer", "journey:clinic/accept-a-proxy-mandate/human", "journey:clinic/accept-a-proxy-mandate/regulator", "journey:clinic/acquire-and-send-images/engineer", "journey:clinic/acquire-and-send-images/human" ], "intention": [ "intention:agents-as-real-customers", "intention:agents-build-agents", "intention:agents-native", "intention:ai-assurance-standard", "intention:always-on-monitoring" ], "old-finding": [ "oldfinding:O-01fb2d61a2", "oldfinding:O-01fdcbda7c", "oldfinding:O-021db5b6d2", "oldfinding:O-029bad174b", "oldfinding:O-0dce2b57af" ], "operation": [ "op:clinic/front-door-check-in/decide-review", "op:clinic/front-door-check-in/describe-concern", "op:clinic/front-door-check-in/prepare-review", "op:clinic/front-door-check-in/propose-route", "op:clinic/front-door-check-in/request-clinician-review" ], "request": [ "request:1", "request:10", "request:11", "request:12", "request:13" ], "screen": [ "screen:clinic/accept-a-proxy-mandate/done", "screen:clinic/accept-a-proxy-mandate/evidence", "screen:clinic/accept-a-proxy-mandate/step-1", "screen:clinic/accept-a-proxy-mandate/step-2", "screen:clinic/accept-a-proxy-mandate/step-3" ] }, "head": "d594aa71f06592d467dc430d2362b37d8ff0b96c", "missing_by_kind": { "agent-journey": 1647, "control": 16831, "decision": 21, "finding": 13974, "human-journey": 5018, "intention": 27, "old-finding": 86, "operation": 644, "request": 59, "screen": 9925 }, "missing_outside_plug_set_by_kind": { "agent-journey": 1647, "control": 16831, "decision": 21, "finding": 13974, "human-journey": 5018, "intention": 27, "old-finding": 86, "operation": 552, "request": 59, "screen": 9925 }, "pass": false, "plug_set": [ "companies-house-api-a-production-government-service-whose-li", "fca-register", "hmrc", "mailbox-sending-state", "model-classification", "model-detection", "model-forecasting", "model-transcription", "nhs-services", "payment-scheme", "registry", "third-party", "voice-telephony" ], "requests": { "request:1": "UNMET", "request:10": "UNMET", "request:11": "UNMET", "request:12": "UNMET", "request:13": "UNMET", "request:14": "UNMET", "request:15": "UNMET", "request:16": "UNMET", "request:17": "UNMET", "request:18": "UNMET", "request:19": "UNMET", "request:2": "UNMET", "request:20": "UNMET", "request:21": "UNMET", "request:22": "UNMET", "request:23": "UNMET", "request:24": "UNMET", "request:25": "UNMET", "request:26": "UNMET", "request:27": "UNMET", "request:28": "UNMET", "request:29": "UNMET", "request:3": "UNMET", "request:30": "UNMET", "request:31": "NEEDS THE OWNER", "request:32": "NEEDS THE OWNER", "request:33": "NEEDS THE OWNER", "request:34": "NEEDS THE OWNER", "request:35": "UNMET", "request:36": "UNMET", "request:37": "UNMET", "request:38": "UNMET", "request:39": "UNMET", "request:4": "UNMET", "request:40": "UNMET", "request:41": "UNMET", "request:42": "UNMET", "request:43": "UNMET", "request:44": "UNMET", "request:45": "UNMET", "request:46": "UNMET", "request:47": "UNMET", "request:48": "UNMET", "request:49": "UNMET", "request:5": "UNMET", "request:50": "UNMET", "request:51": "UNMET", "request:52": "UNMET", "request:53": "UNMET", "request:54": "UNMET", "request:55": "UNMET", "request:56": "UNMET", "request:57": "UNMET", "request:58": "UNMET", "request:59": "NEEDS THE OWNER", "request:6": "UNMET", "request:7": "UNMET", "request:8": "UNMET", "request:9": "UNMET" }, "units": { "open": 1616 } } ``` ### generated/COUNTS.json `generated/COUNTS.json` · sha256 `69dc5173c9932c61db8942ce06da90b3c63939a438134791eb87cc578e1e7485` ```json { "method": "computed by tools/build.py from every file under tok/objects at build time", "completeness": "complete for the objects in this tree", "fingerprint": "1c80b0fa621e444a539ff47dd98c889b3c870d5f700ff7012ee134f079acd919", "objects": 3642, "retired": 4, "by_kind": { "agent": 0, "application": 4, "benchmark": 0, "capability": 50, "claim": 0, "code-module": 4, "competitor": 237, "competitor-feature": 468, "control": 0, "decision": 21, "definition": 648, "design-asset": 13, "embodiment": 0, "endpoint": 876, "evidence": 0, "external-artifact": 0, "foundation": 8, "incident": 0, "intention": 27, "journey": 322, "lifecycle-event": 0, "mandate": 0, "measurement": 0, "operation": 438, "organisation": 0, "outcome": 12, "person": 0, "principle": 27, "procedure": 0, "role": 141, "screen": 346, "source": 0, "state": 0, "test": 0 }, "by_status": { "VERIFIED": 0, "OBSERVED": 4, "HYPOTHESIS": 0, "ASPIRATION": 0, "PREDICTION": 0, "SIMULATION": 0, "UNKNOWN": 3638, "REFUTED": 0 }, "by_statement_class": { "OBSERVATION": 722, "AGREEMENT": 2920, "CLAIM": 0 }, "open_answers": 30199, "open_obligations": 1689 } ``` ### generated/EXPERIENCE.md `generated/EXPERIENCE.md` · sha256 `61c2db7216432e19b14a0db6a6fb153a59f82bdf441d9ff6ea79306c96bfca80` #### Experience catalogue Counts from spec/units and runtime/web.py. Nothing estimated. | Measure | Count | |---|---| | capabilities | 1616 | | controls | 16831 | | controls_bound | 8604 | | controls_navigation | 8227 | | journeys | 6665 | | screens | 9925 | | screens_partial | 575 | | screens_renders | 9350 | | states | 89325 | ##### Components | Component | Used | Falls back to a region | |---|---|---| | Agent.Card | 78 | 78 | | Alert | 98 | 0 | | Avatar | 16 | 16 | | Badge | 103 | 103 | | Calendar | 12 | 12 | | Camera.Capture | 5 | 0 | | Card | 121 | 121 | | Card.Metric | 85 | 85 | | Chart | 25 | 25 | | Chat.Composer | 10 | 10 | | Chat.Thread | 9 | 9 | | Checkbox | 1357 | 0 | | Confidence.Meter | 44 | 44 | | EmptyState | 9 | 9 | | Evidence.Panel | 1803 | 0 | | File.Upload | 183 | 0 | | Help.Control | 9925 | 0 | | Input.Address | 7 | 0 | | Input.Code | 1 | 0 | | Input.Date | 210 | 0 | | Input.Email | 5 | 0 | | Input.Money | 174 | 0 | | Input.Number | 39 | 0 | | Input.Password | 1 | 0 | | Input.Phone | 4 | 0 | | Input.Text | 3590 | 0 | | Input.TextArea | 3779 | 0 | | List | 335 | 335 | | List.Item | 2 | 2 | | Mandate.Editor | 9 | 9 | | Map | 8 | 8 | | Modal | 13 | 13 | | Nav.Breadcrumbs | 9925 | 0 | | Nav.Sidebar | 2 | 2 | | Nav.Tabs | 11 | 11 | | Notification.Item | 2 | 2 | | Page.Header | 9925 | 0 | | Page.Section | 4 | 4 | | Receipt | 8211 | 0 | | Rehearsal.Result | 6661 | 0 | | Select | 162 | 0 | | Select.Multi | 3 | 0 | | Skeleton | 1 | 1 | | Slideout | 13 | 13 | | Steps.Progress | 8221 | 0 | | Summary.Review | 360 | 0 | | Table | 122 | 0 | | Timeline | 1932 | 0 | | Toast | 18 | 18 | | Toggle | 8 | 0 | ##### Applications | Application | Capabilities | Screens | Render now | Partial | Controls bound | Navigation | Unbound | |---|---|---|---|---|---|---|---| | clinic | 342 | 2101 | 1962 | 139 | 1806 | 1751 | 0 | | consortium | 363 | 2340 | 2258 | 82 | 1983 | 1972 | 0 | | finance | 391 | 2406 | 2237 | 169 | 2125 | 1983 | 0 | | regulators | 2 | 19 | 18 | 1 | 17 | 17 | 0 | | organisations | 3 | 24 | 22 | 2 | 21 | 21 | 0 | | id | 2 | 19 | 9 | 10 | 27 | 16 | 0 | | releases | 1 | 8 | 8 | 0 | 7 | 7 | 0 | | evidence | 1 | 8 | 8 | 0 | 7 | 7 | 0 | | judiverse | 22 | 135 | 57 | 78 | 208 | 100 | 0 | | public | 489 | 2865 | 2771 | 94 | 2403 | 2353 | 0 | ### ledger/outcomes/README.md `ledger/outcomes/README.md` · sha256 `270dc8f39f299d4c045de361495ff96d1067deb45e1f4a4ea97822d6491d5364` #### Outcome register Generated by `tools/outcomes.py`. One row per operation. The **floor** is the competitor's best published figure (`research/outcomes/*.jsonl`, or a figure on the competitor's own page); the **target** is the estate's own measurement (`ledger/benchmarks/results.jsonl`, `ledger/evidence/operations.jsonl`, `ledger/evidence/journeys-*.jsonl`) and stays **OPEN** until one exists. No number is halved, moved halfway or otherwise derived. **Outcome matched** is true only when the capability's bound reference flow (`research/parity/REFERENCE-FLOWS.jsonl`) records the screens that cover the step this operation sits at in the person's journey. A competitor clause or figure found by rare-word overlap is a **candidate**, never a match. | Application | Operations | Reference flow bound | Outcome matched (reference covers the step) | Candidate competitor outcomes (rare words) | Published floor | Candidate floor (rare words) | Target MEASURED | Target OPEN | With the competitor's published availability | |---|---|---|---|---|---|---|---|---|---| | clinic | 1794 | 95 | 10 | 0 | 0 | 0 | 1794 | 0 | 0 | | edge-consortium | 1981 | 49 | 0 | 0 | 0 | 0 | 1981 | 0 | 0 | | finance | 2104 | 259 | 25 | 6 | 0 | 0 | 2056 | 48 | 0 | | foundation | 76 | 0 | 0 | 0 | 0 | 0 | 76 | 0 | 0 | | judiverse | 178 | 105 | 0 | 20 | 0 | 0 | 140 | 38 | 0 | | public-edge | 2401 | 7 | 0 | 0 | 0 | 0 | 2401 | 0 | 0 | ## 11. Full specification per capability One file per capability in `generated/experience/<application>/<capability>.md` (in the zip): identity, requirement, outcomes, actors, data, operations, screens with copy, fields, components, controls, states, floors and frame, journeys, definitions, Mermaid diagrams, approved marks and rendering rules. - `generated/experience/clinic/accept-a-proxy-mandate.md` - `generated/experience/clinic/acquire-and-send-images.md` - `generated/experience/clinic/act-on-a-drug-safety-alert.md` - `generated/experience/clinic/act-on-a-high-risk-answer.md` - `generated/experience/clinic/add-an-appointment-to-my-calendar.md` - `generated/experience/clinic/adjust-a-premium-at-renewal.md` - `generated/experience/clinic/administer-a-medicine.md` - `generated/experience/clinic/allocate-a-breach-between-trusts.md` - `generated/experience/clinic/annotate-variants.md` - `generated/experience/clinic/answer-a-patient-call.md` - `generated/experience/clinic/apply-a-national-data-opt-out.md` - `generated/experience/clinic/apply-a-treatment-rule.md` - `generated/experience/clinic/apply-for-hra-and-rec-approval.md` - `generated/experience/clinic/approve-a-data-release.md` - `generated/experience/clinic/approve-a-leave-or-swap-request.md` - `generated/experience/clinic/approve-a-questionnaire-or-test-threshold.md` - `generated/experience/clinic/approve-a-staff-request.md` - `generated/experience/clinic/arrange-a-cross-trust-mdt.md` - `generated/experience/clinic/ask-for-advice-and-guidance.md` - `generated/experience/clinic/ask-for-help-now.md` - `generated/experience/clinic/assess-a-safeguarding-concern.md` - `generated/experience/clinic/attend-an-appointment-with-the-person.md` - `generated/experience/clinic/audit-access-to-records.md` - `generated/experience/clinic/audit-antimicrobial-use.md` - `generated/experience/clinic/audit-consent-records.md` - `generated/experience/clinic/audit-data-use-against-mandates.md` - `generated/experience/clinic/audit-mdt-attendance.md` - `generated/experience/clinic/audit-reporting-quality.md` - `generated/experience/clinic/audit-response-times.md` - `generated/experience/clinic/audit-waiting-time-data.md` - `generated/experience/clinic/authorise-a-result.md` - `generated/experience/clinic/book-a-pathway-test.md` - `generated/experience/clinic/book-an-appointment.md` - `generated/experience/clinic/book-an-interpreter.md` - `generated/experience/clinic/book-for-the-person-i-care-for.md` - `generated/experience/clinic/book-from-the-referral-queue.md` - `generated/experience/clinic/book-transport.md` - `generated/experience/clinic/build-and-publish-the-rota.md` - `generated/experience/clinic/build-the-clinic-template.md` - `generated/experience/clinic/build-the-mdt-list.md` - `generated/experience/clinic/build-the-variant-to-trial-graph.md` - `generated/experience/clinic/cancel-a-clinic-session.md` - `generated/experience/clinic/change-my-pharmacy.md` - `generated/experience/clinic/chase-a-pathway-clock.md` - `generated/experience/clinic/chase-missing-results.md` - `generated/experience/clinic/check-a-chemotherapy-prescription.md` - `generated/experience/clinic/check-access-controls.md` - `generated/experience/clinic/check-fragment-size-before-sequencing.md` - `generated/experience/clinic/check-identity-and-consent-before-scanning.md` - `generated/experience/clinic/check-in-an-arriving-patient.md` - `generated/experience/clinic/check-in-on-arrival.md` - `generated/experience/clinic/check-laboratory-submissions-for-a-variant.md` - `generated/experience/clinic/check-record-accuracy.md` - `generated/experience/clinic/check-research-integrity.md` - `generated/experience/clinic/check-staff-registration-and-training.md` - `generated/experience/clinic/claim-my-edge-id-and-record.md` - `generated/experience/clinic/classify-a-variant.md` - `generated/experience/clinic/classify-software-as-a-medical-device.md` - `generated/experience/clinic/clinically-check-a-prescription.md` - `generated/experience/clinic/close-a-referral.md` - `generated/experience/clinic/close-a-study.md` - `generated/experience/clinic/collect-case-report-data.md` - `generated/experience/clinic/compare-two-pipeline-runs.md` - `generated/experience/clinic/complete-a-mental-health-check-in.md` - `generated/experience/clinic/complete-a-pre-clinic-questionnaire.md` - `generated/experience/clinic/complete-a-risk-formulation.md` - `generated/experience/clinic/complete-mandatory-training.md` - `generated/experience/clinic/complete-the-data-security-and-protection-toolkit.md` - `generated/experience/clinic/confirm-attendance-and-prepare.md` - `generated/experience/clinic/connect-a-lims.md` - `generated/experience/clinic/connect-a-pacs-or-imaging-system.md` - `generated/experience/clinic/counsel-a-patient-on-a-new-medicine.md` - `generated/experience/clinic/decide-a-re-run.md` - `generated/experience/clinic/deliver-cancer-nurse-specialist-support.md` - `generated/experience/clinic/deploy-a-release.md` - `generated/experience/clinic/design-a-smart-questionnaire.md` - `generated/experience/clinic/design-a-trial.md` - `generated/experience/clinic/detect-claim-anomalies.md` - `generated/experience/clinic/discharge-from-the-pathway.md` - `generated/experience/clinic/discharge-from-the-stream.md` - `generated/experience/clinic/dispense-a-medicine.md` - `generated/experience/clinic/escalate-to-emergency-services.md` - `generated/experience/clinic/explain-a-result-in-plain-language.md` - `generated/experience/clinic/explain-a-result-to-the-patient.md` - `generated/experience/clinic/find-a-recruiting-trial-near-me.md` - `generated/experience/clinic/flag-a-critical-finding.md` - `generated/experience/clinic/flag-never-block.md` - `generated/experience/clinic/follow-up-after-discharge-from-crisis-care.md` - `generated/experience/clinic/follow-up-an-action-plan.md` - `generated/experience/clinic/forecast-compliance.md` - `generated/experience/clinic/forecast-days-to-treatment.md` - `generated/experience/clinic/front-door-check-in.md` - `generated/experience/clinic/fulfil-a-subject-access-request.md` - `generated/experience/clinic/give-a-vaccination.md` - `generated/experience/clinic/give-consent-on-behalf.md` - `generated/experience/clinic/give-or-withdraw-consent.md` - `generated/experience/clinic/grade-a-retinal-image.md` - `generated/experience/clinic/hand-off-to-codex-or-jude.md` - `generated/experience/clinic/handle-a-complaint.md` - `generated/experience/clinic/handle-a-controlled-drug.md` - `generated/experience/clinic/handle-a-did-not-attend.md` - `generated/experience/clinic/handle-a-disclosure-of-abuse.md` - `generated/experience/clinic/handle-a-freedom-of-information-request.md` - `generated/experience/clinic/handle-a-policyholder-complaint.md` - `generated/experience/clinic/handle-a-stock-shortage.md` - `generated/experience/clinic/handle-the-record-after-death.md` - `generated/experience/clinic/hold-a-remote-session.md` - `generated/experience/clinic/hold-a-slot-for-a-tertiary-handover.md` - `generated/experience/clinic/hold-the-duty-rota.md` - `generated/experience/clinic/import-my-practice-register.md` - `generated/experience/clinic/inspect-the-hazard-log-and-safety-case.md` - `generated/experience/clinic/investigate-a-clinical-safety-incident.md` - `generated/experience/clinic/issue-a-clinical-safety-case-report.md` - `generated/experience/clinic/issue-a-field-safety-notice.md` - `generated/experience/clinic/issue-a-genomic-report.md` - `generated/experience/clinic/issue-a-policy.md` - `generated/experience/clinic/maintain-laboratory-accreditation-evidence.md` - `generated/experience/clinic/maintain-the-hazard-log.md` - `generated/experience/clinic/maintain-the-suicide-prevention-plan.md` - `generated/experience/clinic/make-a-follow-up-call.md` - `generated/experience/clinic/make-an-urgent-same-day-referral.md` - `generated/experience/clinic/manage-an-incident.md` - `generated/experience/clinic/manage-freezer-alarms.md` - `generated/experience/clinic/manage-patient-group-directions.md` - `generated/experience/clinic/manage-patient-transport-and-interpreter-contracts.md` - `generated/experience/clinic/manage-retention-and-destruction.md` - `generated/experience/clinic/manage-room-allocation.md` - `generated/experience/clinic/manage-the-information-asset-register.md` - `generated/experience/clinic/manage-the-outbound-allow-list.md` - `generated/experience/clinic/manage-the-reporting-backlog.md` - `generated/experience/clinic/manage-the-room-and-equipment-register.md` - `generated/experience/clinic/manage-the-waiting-list.md` - `generated/experience/clinic/measure-a-nodule-and-its-growth.md` - `generated/experience/clinic/measure-compute-energy-and-carbon.md` - `generated/experience/clinic/measure-outcomes-over-time.md` - `generated/experience/clinic/medication-safety-check.md` - `generated/experience/clinic/mental-health-care-and-crisis.md` - `generated/experience/clinic/message-my-care-team.md` - `generated/experience/clinic/monitor-alerts-and-escalate.md` - `generated/experience/clinic/monitor-data-feeds.md` - `generated/experience/clinic/monitor-post-deployment-safety.md` - `generated/experience/clinic/nominate-a-carer-or-proxy.md` - `generated/experience/clinic/obtain-confirmation-of-capacity-and-capability.md` - `generated/experience/clinic/offer-a-cancellation-slot.md` - `generated/experience/clinic/offer-self-help-resources.md` - `generated/experience/clinic/onboard-a-staff-member.md` - `generated/experience/clinic/onboard-a-wearable-device-type.md` - `generated/experience/clinic/online-urgent-triage.md` - `generated/experience/clinic/order-a-diagnostic-test.md` - `generated/experience/clinic/order-a-direct-access-test.md` - `generated/experience/clinic/order-a-genomic-test.md` - `generated/experience/clinic/order-a-repeat-medicine.md` - `generated/experience/clinic/order-repeat-medicines-on-behalf.md` - `generated/experience/clinic/pair-my-wearable-device.md` - `generated/experience/clinic/person-held-health-record.md` - `generated/experience/clinic/phone-a-critical-result.md` - `generated/experience/clinic/plan-weekly-capacity.md` - `generated/experience/clinic/post-a-receipt-to-the-agent-channel.md` - `generated/experience/clinic/predict-fields-in-a-form.md` - `generated/experience/clinic/prepare-a-clinic-letter-for-a-clinician.md` - `generated/experience/clinic/prepare-a-patient-for-imaging.md` - `generated/experience/clinic/prepare-for-a-cqc-inspection.md` - `generated/experience/clinic/prescribe-under-the-pathway-protocol.md` - `generated/experience/clinic/prescribe.md` - `generated/experience/clinic/price-a-policy-from-clinic-data.md` - `generated/experience/clinic/process-a-claim.md` - `generated/experience/clinic/pseudonymise-a-dataset.md` - `generated/experience/clinic/publish-results.md` - `generated/experience/clinic/raise-a-complaint-or-feedback.md` - `generated/experience/clinic/raise-a-safeguarding-concern-from-a-response.md` - `generated/experience/clinic/randomise-a-participant.md` - `generated/experience/clinic/read-assistive-output-on-an-image.md` - `generated/experience/clinic/rebook-or-cancel-an-appointment.md` - `generated/experience/clinic/receive-a-crisis-alert-for-the-person.md` - `generated/experience/clinic/receive-a-crisis-alert.md` - `generated/experience/clinic/receive-a-critical-result-for-my-patient.md` - `generated/experience/clinic/receive-a-did-not-attend-notice.md` - `generated/experience/clinic/receive-a-pre-authorisation-request.md` - `generated/experience/clinic/receive-a-statutory-notification.md` - `generated/experience/clinic/receive-and-book-in-a-sample.md` - `generated/experience/clinic/receive-the-outcome-letter.md` - `generated/experience/clinic/reclassify-a-variant.md` - `generated/experience/clinic/reconcile-medicines-at-admission.md` - `generated/experience/clinic/reconcile-medicines.md` - `generated/experience/clinic/record-a-clinic-consultation.md` - `generated/experience/clinic/record-a-dose-incident.md` - `generated/experience/clinic/record-a-patient-deferred.md` - `generated/experience/clinic/record-a-patient-s-own-view.md` - `generated/experience/clinic/record-a-point-of-care-urine-result.md` - `generated/experience/clinic/record-a-safeguarding-concern.md` - `generated/experience/clinic/record-a-self-check-in-exception.md` - `generated/experience/clinic/record-a-wound-or-skin-check.md` - `generated/experience/clinic/record-an-adverse-drug-reaction.md` - `generated/experience/clinic/record-attendance-outcome.md` - `generated/experience/clinic/record-capacity-and-best-interests.md` - `generated/experience/clinic/record-decision-to-treat.md` - `generated/experience/clinic/record-sample-consent-version.md` - `generated/experience/clinic/record-the-faster-diagnosis-outcome.md` - `generated/experience/clinic/record-the-lawful-basis.md` - `generated/experience/clinic/record-the-mdt-recommendation.md` - `generated/experience/clinic/record-trial-eligibility-at-the-mdt.md` - `generated/experience/clinic/refer-on-a-suspected-cancer-pathway.md` - `generated/experience/clinic/refer-to-a-crisis-team.md` - `generated/experience/clinic/refer-to-tertiary-care.md` - `generated/experience/clinic/refer-to-the-mental-health-stream.md` - `generated/experience/clinic/refer-to-the-multidisciplinary-team.md` - `generated/experience/clinic/refresh-public-data-caches.md` - `generated/experience/clinic/register-a-child-as-a-patient.md` - `generated/experience/clinic/register-a-device-fault.md` - `generated/experience/clinic/register-a-walk-in-patient.md` - `generated/experience/clinic/register-an-imaging-device.md` - `generated/experience/clinic/register-my-practice-with-the-clinic.md` - `generated/experience/clinic/register-the-trial.md` - `generated/experience/clinic/reject-an-unlabelled-sample.md` - `generated/experience/clinic/report-a-change-in-the-person-s-condition.md` - `generated/experience/clinic/report-a-data-breach.md` - `generated/experience/clinic/report-a-dispensing-error.md` - `generated/experience/clinic/report-a-safety-incident.md` - `generated/experience/clinic/report-a-side-effect.md` - `generated/experience/clinic/report-a-study.md` - `generated/experience/clinic/report-an-adverse-event.md` - `generated/experience/clinic/report-an-incident.md` - `generated/experience/clinic/report-histology.md` - `generated/experience/clinic/report-monthly-performance.md` - `generated/experience/clinic/report-to-the-regulator.md` - `generated/experience/clinic/report-usage-and-efficiency.md` - `generated/experience/clinic/request-a-copy-of-my-record.md` - `generated/experience/clinic/request-a-rota-swap-or-leave.md` - `generated/experience/clinic/request-a-second-opinion.md` - `generated/experience/clinic/request-a-second-read-or-discrepancy-review.md` - `generated/experience/clinic/request-a-second-read.md` - `generated/experience/clinic/request-a-young-person-s-own-access.md` - `generated/experience/clinic/request-clinic-data-for-a-quote.md` - `generated/experience/clinic/request-evidence-against-fundamental-standards.md` - `generated/experience/clinic/retinal-screening.md` - `generated/experience/clinic/review-a-breach-notification.md` - `generated/experience/clinic/review-a-death-by-suicide.md` - `generated/experience/clinic/review-a-real-result.md` - `generated/experience/clinic/review-a-release-for-clinical-safety.md` - `generated/experience/clinic/review-a-repeat-request.md` - `generated/experience/clinic/review-a-trial-for-compliance.md` - `generated/experience/clinic/review-a-wearable-alert.md` - `generated/experience/clinic/review-assistive-output-labelling.md` - `generated/experience/clinic/review-payslips-and-expenses-approvals.md` - `generated/experience/clinic/review-the-caseload.md` - `generated/experience/clinic/review-the-dashboard.md` - `generated/experience/clinic/review-the-waiting-list-position.md` - `generated/experience/clinic/review-the-worklist.md` - `generated/experience/clinic/revoke-access-when-a-mandate-ends.md` - `generated/experience/clinic/revoke-my-own-proxy-role.md` - `generated/experience/clinic/reward-early-detection-participation.md` - `generated/experience/clinic/roll-back-a-release.md` - `generated/experience/clinic/rotate-keys-and-certificates.md` - `generated/experience/clinic/run-a-business-continuity-drill.md` - `generated/experience/clinic/run-a-catalogue-workload.md` - `generated/experience/clinic/run-a-clinical-safety-training-check.md` - `generated/experience/clinic/run-a-data-protection-impact-assessment.md` - `generated/experience/clinic/run-a-pipeline.md` - `generated/experience/clinic/run-a-pre-assessment.md` - `generated/experience/clinic/run-a-safeguarding-review.md` - `generated/experience/clinic/run-a-synthetic-pathway-drill.md` - `generated/experience/clinic/run-a-urine-test-likelihood-ledger.md` - `generated/experience/clinic/run-an-interim-analysis.md` - `generated/experience/clinic/run-an-online-early-symptom-test.md` - `generated/experience/clinic/run-quality-control.md` - `generated/experience/clinic/run-scanner-quality-assurance.md` - `generated/experience/clinic/run-sequencing-qc.md` - `generated/experience/clinic/run-the-day-s-capacity-check.md` - `generated/experience/clinic/run-the-mdt-meeting.md` - `generated/experience/clinic/run-the-training-controller.md` - `generated/experience/clinic/run-vulnerability-scans.md` - `generated/experience/clinic/schedule-the-scan.md` - `generated/experience/clinic/score-a-model-on-a-held-split.md` - `generated/experience/clinic/score-and-interpret-a-response.md` - `generated/experience/clinic/screen-a-patient-for-eligibility.md` - `generated/experience/clinic/search-recruiting-trials.md` - `generated/experience/clinic/see-my-day.md` - `generated/experience/clinic/send-a-check-in-to-a-caseload.md` - `generated/experience/clinic/send-a-results-ready-notice.md` - `generated/experience/clinic/send-appointment-letters-and-reminders.md` - `generated/experience/clinic/serve-a-vision-model.md` - `generated/experience/clinic/set-a-surveillance-plan.md` - `generated/experience/clinic/set-my-communication-preferences.md` - `generated/experience/clinic/set-my-data-sharing-choices.md` - `generated/experience/clinic/set-role-permissions-and-expiry.md` - `generated/experience/clinic/settle-with-the-clinic.md` - `generated/experience/clinic/share-a-medicines-list-with-the-clinic.md` - `generated/experience/clinic/share-a-wearable-signal.md` - `generated/experience/clinic/share-clinic-data-with-my-insurer.md` - `generated/experience/clinic/share-images-with-another-trust.md` - `generated/experience/clinic/share-risk-information-with-partners.md` - `generated/experience/clinic/sign-a-clinic-letter.md` - `generated/experience/clinic/sign-the-clinical-risk-management-plan.md` - `generated/experience/clinic/simulate-trial-monitoring.md` - `generated/experience/clinic/social-prescribing-referral.md` - `generated/experience/clinic/stop-on-revoked-authority.md` - `generated/experience/clinic/store-and-retrieve-genomic-data.md` - `generated/experience/clinic/submit-a-cancer-registration.md` - `generated/experience/clinic/submit-cancer-waiting-times-data.md` - `generated/experience/clinic/submit-cosd-and-other-datasets.md` - `generated/experience/clinic/submit-mdt-data.md` - `generated/experience/clinic/submit-the-mental-health-dataset.md` - `generated/experience/clinic/supervise-and-sign-off-a-trainee.md` - `generated/experience/clinic/support-deploying-organisations.md` - `generated/experience/clinic/symptom-check-and-triage.md` - `generated/experience/clinic/take-a-blood-or-urine-sample.md` - `generated/experience/clinic/take-a-mandate-ticket.md` - `generated/experience/clinic/take-an-online-early-symptom-test.md` - `generated/experience/clinic/take-informed-consent-for-research.md` - `generated/experience/clinic/take-observations.md` - `generated/experience/clinic/take-part-in-external-quality-assessment.md` - `generated/experience/clinic/tell-the-patient-a-result.md` - `generated/experience/clinic/test-a-hypothesis-on-an-open-cohort.md` - `generated/experience/clinic/test-a-pre-registered-hypothesis.md` - `generated/experience/clinic/trace-ingress-and-egress.md` - `generated/experience/clinic/track-a-referral.md` - `generated/experience/clinic/track-actions-from-the-mdt.md` - `generated/experience/clinic/track-histology-turnaround.md` - `generated/experience/clinic/track-my-cancer-pathway.md` - `generated/experience/clinic/track-sample-custody.md` - `generated/experience/clinic/track-the-62-day-clock-across-the-mdt.md` - `generated/experience/clinic/train-staff-in-information-governance.md` - `generated/experience/clinic/train-staff-in-suicide-awareness.md` - `generated/experience/clinic/triage-a-mental-health-referral.md` - `generated/experience/clinic/triage-a-patient-message.md` - `generated/experience/clinic/triage-a-referral.md` - `generated/experience/clinic/triage-inbox-items-for-a-clinician.md` - `generated/experience/clinic/tune-alert-thresholds.md` - `generated/experience/clinic/update-patient-demographics.md` - `generated/experience/clinic/validate-a-new-assay.md` - `generated/experience/clinic/validate-the-test-against-outcomes.md` - `generated/experience/clinic/verify-a-device-classification.md` - `generated/experience/clinic/verify-an-applicant-s-identity.md` - `generated/experience/clinic/verify-an-evidence-statement.md` - `generated/experience/clinic/verify-backups-and-restore.md` - `generated/experience/clinic/vet-an-imaging-request.md` - `generated/experience/clinic/view-my-medicines-list.md` - `generated/experience/clinic/view-my-results.md` - `generated/experience/clinic/view-results-in-scope.md` - `generated/experience/clinic/wearable-signals.md` - `generated/experience/clinic/wellbeing-questionnaires.md` - `generated/experience/clinic/write-a-care-plan.md` - `generated/experience/edge-consortium/accept-a-review-assignment.md` - `generated/experience/edge-consortium/accept-supervision-of-a-learner.md` - `generated/experience/edge-consortium/acknowledge-the-principles-version.md` - `generated/experience/edge-consortium/add-jude-to-a-channel-under-mandate.md` - `generated/experience/edge-consortium/add-members-and-guests-with-expiry.md` - `generated/experience/edge-consortium/administer-a-database.md` - `generated/experience/edge-consortium/admit-or-refuse-a-guest.md` - `generated/experience/edge-consortium/agree-a-data-sharing-agreement.md` - `generated/experience/edge-consortium/allocate-indirect-costs.md` - `generated/experience/edge-consortium/answer-a-reference-request.md` - `generated/experience/edge-consortium/answer-reviewer-questions-on-a-contribution.md` - `generated/experience/edge-consortium/answer-with-sources-in-a-channel.md` - `generated/experience/edge-consortium/apply-a-non-disclosure-agreement-to-a-room.md` - `generated/experience/edge-consortium/apply-an-open-licence.md` - `generated/experience/edge-consortium/apply-for-ethics-approval.md` - `generated/experience/edge-consortium/apply-retention-and-deletion-to-records.md` - `generated/experience/edge-consortium/apply-to-become-a-verified-reviewer.md` - `generated/experience/edge-consortium/approve-a-compute-request.md` - `generated/experience/edge-consortium/approve-a-connector-or-integration.md` - `generated/experience/edge-consortium/approve-a-contribution-before-submission.md` - `generated/experience/edge-consortium/approve-a-director-passkey-action.md` - `generated/experience/edge-consortium/approve-a-federated-query-request.md` - `generated/experience/edge-consortium/approve-a-guest-account.md` - `generated/experience/edge-consortium/approve-a-reviewer-payment-batch.md` - `generated/experience/edge-consortium/approve-a-team-member-data-access-request.md` - `generated/experience/edge-consortium/approve-a-visiting-researcher-affiliation.md` - `generated/experience/edge-consortium/approve-an-agent-consequential-action.md` - `generated/experience/edge-consortium/approve-an-open-source-release.md` - `generated/experience/edge-consortium/approve-and-mark-a-contribution-verified.md` - `generated/experience/edge-consortium/approve-compute-spend-against-a-grant.md` - `generated/experience/edge-consortium/approve-expenditure-against-a-grant.md` - `generated/experience/edge-consortium/archive-a-channel.md` - `generated/experience/edge-consortium/archive-a-module-version.md` - `generated/experience/edge-consortium/ask-a-clarification-question.md` - `generated/experience/edge-consortium/ask-jude-for-a-sourced-answer.md` - `generated/experience/edge-consortium/assess-ownership-of-an-output.md` - `generated/experience/edge-consortium/assign-a-persistent-identifier.md` - `generated/experience/edge-consortium/attribute-carbon-to-a-project.md` - `generated/experience/edge-consortium/attribute-cost-to-a-grant-and-project.md` - `generated/experience/edge-consortium/audit-reviewer-quality-by-re-review.md` - `generated/experience/edge-consortium/audit-the-signing-log.md` - `generated/experience/edge-consortium/award-and-publish-the-contract-award-notice.md` - `generated/experience/edge-consortium/bind-a-device.md` - `generated/experience/edge-consortium/book-a-room-or-resource.md` - `generated/experience/edge-consortium/build-a-project-team-across-universities.md` - `generated/experience/edge-consortium/build-an-interactive-prototype.md` - `generated/experience/edge-consortium/build-the-research-profile-and-skills-record.md` - `generated/experience/edge-consortium/bulk-enrol-students-and-staff.md` - `generated/experience/edge-consortium/catalogue-a-library-item.md` - `generated/experience/edge-consortium/charge-compute-and-storage-to-grant-codes.md` - `generated/experience/edge-consortium/chase-an-overdue-review.md` - `generated/experience/edge-consortium/check-exclusion-grounds.md` - `generated/experience/edge-consortium/choose-a-specialism.md` - `generated/experience/edge-consortium/claim-payment-for-a-completed-review.md` - `generated/experience/edge-consortium/classify-data-sensitivity.md` - `generated/experience/edge-consortium/close-a-grant.md` - `generated/experience/edge-consortium/close-a-project-and-archive-its-records.md` - `generated/experience/edge-consortium/co-edit-a-markdown-in-real-time.md` - `generated/experience/edge-consortium/complete-a-right-to-work-check.md` - `generated/experience/edge-consortium/complete-research-integrity-training.md` - `generated/experience/edge-consortium/compose-an-image-in-the-imaging-studio.md` - `generated/experience/edge-consortium/compose-audio-in-judio.md` - `generated/experience/edge-consortium/conduct-a-probation-review.md` - `generated/experience/edge-consortium/confirm-a-delivery.md` - `generated/experience/edge-consortium/confirm-a-person-s-enrolment-status.md` - `generated/experience/edge-consortium/confirm-a-time-and-send-the-invitation.md` - `generated/experience/edge-consortium/contribute-a-library-entry-as-an-outsider.md` - `generated/experience/edge-consortium/correct-a-student-record.md` - `generated/experience/edge-consortium/cost-a-grant-application.md` - `generated/experience/edge-consortium/count-agents-and-subagents-per-task.md` - `generated/experience/edge-consortium/create-a-contract-from-a-template.md` - `generated/experience/edge-consortium/create-a-job-role-and-vacancy.md` - `generated/experience/edge-consortium/create-a-markdown-document.md` - `generated/experience/edge-consortium/create-a-module.md` - `generated/experience/edge-consortium/create-a-team-and-its-channels.md` - `generated/experience/edge-consortium/create-an-availability-poll.md` - `generated/experience/edge-consortium/curate-a-collection.md` - `generated/experience/edge-consortium/declare-a-conflict-of-interest.md` - `generated/experience/edge-consortium/declare-exclusion-grounds-and-conflicts.md` - `generated/experience/edge-consortium/decline-an-assignment.md` - `generated/experience/edge-consortium/define-a-purpose-view.md` - `generated/experience/edge-consortium/define-or-change-a-schema.md` - `generated/experience/edge-consortium/delete-data-at-retention-end.md` - `generated/experience/edge-consortium/deploy-a-release.md` - `generated/experience/edge-consortium/deposit-a-dataset-with-a-persistent-identifier.md` - `generated/experience/edge-consortium/deprecate-a-feature.md` - `generated/experience/edge-consortium/detect-unlicensed-reuse.md` - `generated/experience/edge-consortium/draw-on-judecanvas.md` - `generated/experience/edge-consortium/edit-a-video.md` - `generated/experience/edge-consortium/end-employment-and-revoke-access.md` - `generated/experience/edge-consortium/enforce-the-code-of-conduct.md` - `generated/experience/edge-consortium/enrol-on-a-doctoral-training-module.md` - `generated/experience/edge-consortium/escalate-a-dispute-to-the-editor.md` - `generated/experience/edge-consortium/evaluate-bids-with-a-scoring-panel.md` - `generated/experience/edge-consortium/expire-or-revoke-a-licence.md` - `generated/experience/edge-consortium/explore-the-knowledge-graph-visually.md` - `generated/experience/edge-consortium/export-a-creation-with-licence-and-attribution.md` - `generated/experience/edge-consortium/export-a-meeting-record-and-audit-chain.md` - `generated/experience/edge-consortium/export-and-verify-an-audit-chain.md` - `generated/experience/edge-consortium/find-a-time-across-attendees.md` - `generated/experience/edge-consortium/find-and-follow-tender-opportunities.md` - `generated/experience/edge-consortium/find-funding-calls.md` - `generated/experience/edge-consortium/flag-a-concern-without-blocking.md` - `generated/experience/edge-consortium/flag-plagiarism-or-an-integrity-concern.md` - `generated/experience/edge-consortium/governed-intelligence-degree.md` - `generated/experience/edge-consortium/grant-a-clearance-with-expiry.md` - `generated/experience/edge-consortium/grant-a-licence-to-a-third-party.md` - `generated/experience/edge-consortium/grant-a-maintainer-role-with-expiry.md` - `generated/experience/edge-consortium/hand-over-to-another-agent.md` - `generated/experience/edge-consortium/hand-over-work-with-the-baton.md` - `generated/experience/edge-consortium/handle-a-contract-dispute.md` - `generated/experience/edge-consortium/handle-a-data-subject-access-request.md` - `generated/experience/edge-consortium/handle-an-academic-appeal-on-marks.md` - `generated/experience/edge-consortium/handle-an-appeal.md` - `generated/experience/edge-consortium/hold-an-audit-interview-by-secure-call.md` - `generated/experience/edge-consortium/import-external-records.md` - `generated/experience/edge-consortium/inspect-carbon-and-energy-records.md` - `generated/experience/edge-consortium/inspect-release-gates-and-evidence.md` - `generated/experience/edge-consortium/inspect-retention-schedule-compliance.md` - `generated/experience/edge-consortium/inspect-reviewer-payment-records.md` - `generated/experience/edge-consortium/international-technical-students.md` - `generated/experience/edge-consortium/issue-a-certificate-of-membership.md` - `generated/experience/edge-consortium/issue-a-takedown-or-attribution-request.md` - `generated/experience/edge-consortium/issue-a-transcript-of-completed-modules.md` - `generated/experience/edge-consortium/issue-an-employment-contract.md` - `generated/experience/edge-consortium/issue-an-offer-letter.md` - `generated/experience/edge-consortium/issue-findings-and-track-remediation.md` - `generated/experience/edge-consortium/issue-the-decision-letter.md` - `generated/experience/edge-consortium/issue-the-written-statement-of-particulars.md` - `generated/experience/edge-consortium/join-a-secure-call-through-the-lobby.md` - `generated/experience/edge-consortium/join-consortium-through-a-university.md` - `generated/experience/edge-consortium/judiverse-included-in-tuition.md` - `generated/experience/edge-consortium/link-orcid-to-the-profile.md` - `generated/experience/edge-consortium/log-a-receipt-for-every-action.md` - `generated/experience/edge-consortium/make-a-below-threshold-direct-award.md` - `generated/experience/edge-consortium/manage-a-controlled-vocabulary.md` - `generated/experience/edge-consortium/manage-a-grant-extension-or-variation.md` - `generated/experience/edge-consortium/manage-an-embargo-and-release-date.md` - `generated/experience/edge-consortium/manage-consent-records.md` - `generated/experience/edge-consortium/manage-institutional-roles-and-delegates.md` - `generated/experience/edge-consortium/map-a-dataset-to-the-knowledge-graph.md` - `generated/experience/edge-consortium/map-a-module-to-a-challenge-specification.md` - `generated/experience/edge-consortium/map-a-module-to-programme-and-standard.md` - `generated/experience/edge-consortium/mark-and-moderate-submissions.md` - `generated/experience/edge-consortium/match-reviewers-to-a-submission.md` - `generated/experience/edge-consortium/measure-live-energy-per-job-and-service.md` - `generated/experience/edge-consortium/merge-duplicate-entities.md` - `generated/experience/edge-consortium/merge-with-required-checks.md` - `generated/experience/edge-consortium/mirror-a-repository-to-the-source-repository.md` - `generated/experience/edge-consortium/moderate-a-post.md` - `generated/experience/edge-consortium/monitor-data-quality.md` - `generated/experience/edge-consortium/monitor-job-progress-and-output.md` - `generated/experience/edge-consortium/monitor-tokens-prompts-and-time-to-evidence.md` - `generated/experience/edge-consortium/negotiate-clauses-with-the-counterparty.md` - `generated/experience/edge-consortium/nominate-a-learner-for-reviewer-training.md` - `generated/experience/edge-consortium/offer-the-degree-with-a-partner-institution.md` - `generated/experience/edge-consortium/onboard-a-new-starter.md` - `generated/experience/edge-consortium/open-an-issue.md` - `generated/experience/edge-consortium/paid-student-work-with-tuition-offset.md` - `generated/experience/edge-consortium/pass-reviewer-credential-verification.md` - `generated/experience/edge-consortium/pause-or-pre-empt-a-job.md` - `generated/experience/edge-consortium/pay-reviewer-fees.md` - `generated/experience/edge-consortium/pay-salary.md` - `generated/experience/edge-consortium/petition-for-student-finance.md` - `generated/experience/edge-consortium/pin-edit-or-delete-a-post.md` - `generated/experience/edge-consortium/plan-a-maintenance-window.md` - `generated/experience/edge-consortium/plan-a-procurement-and-publish-pipeline-notice.md` - `generated/experience/edge-consortium/post-an-announcement.md` - `generated/experience/edge-consortium/pre-register-a-study.md` - `generated/experience/edge-consortium/prepare-a-review-summary-for-a-reviewer.md` - `generated/experience/edge-consortium/prepare-an-employment-contract-template.md` - `generated/experience/edge-consortium/preserve-and-migrate-file-formats.md` - `generated/experience/edge-consortium/process-a-supplier-invoice.md` - `generated/experience/edge-consortium/produce-a-carbon-reduction-plan.md` - `generated/experience/edge-consortium/produce-a-graphic-in-visual-studio.md` - `generated/experience/edge-consortium/produce-a-statutory-data-return.md` - `generated/experience/edge-consortium/propose-a-change-from-a-fork.md` - `generated/experience/edge-consortium/propose-a-real-world-problem-project.md` - `generated/experience/edge-consortium/provide-a-funder-audit-trail.md` - `generated/experience/edge-consortium/provide-a-reference.md` - `generated/experience/edge-consortium/publish-a-data-dictionary.md` - `generated/experience/edge-consortium/publish-a-markdown-as-deck-or-book.md` - `generated/experience/edge-consortium/publish-a-new-dataset-version.md` - `generated/experience/edge-consortium/publish-a-tender-notice.md` - `generated/experience/edge-consortium/publish-a-terms-version-and-require-re-acceptance.md` - `generated/experience/edge-consortium/publish-an-open-access-version.md` - `generated/experience/edge-consortium/publish-documentation.md` - `generated/experience/edge-consortium/publish-module-materials-to-the-library.md` - `generated/experience/edge-consortium/publish-the-catalogue-to-open-indexes.md` - `generated/experience/edge-consortium/publish-the-contract-details-notice.md` - `generated/experience/edge-consortium/publish-the-monthly-carbon-and-cost-statement.md` - `generated/experience/edge-consortium/raise-a-challenge-or-complaint.md` - `generated/experience/edge-consortium/raise-a-grievance.md` - `generated/experience/edge-consortium/raise-a-progress-or-wellbeing-concern.md` - `generated/experience/edge-consortium/rank-agents-and-models-by-efficiency.md` - `generated/experience/edge-consortium/read-a-verified-entry-with-its-provenance.md` - `generated/experience/edge-consortium/real-project-assignments.md` - `generated/experience/edge-consortium/receive-a-mandate.md` - `generated/experience/edge-consortium/receive-bids-in-a-sealed-queue.md` - `generated/experience/edge-consortium/receive-the-award-outcome.md` - `generated/experience/edge-consortium/receive-time-boxed-auditor-access.md` - `generated/experience/edge-consortium/reconcile-monthly-spend.md` - `generated/experience/edge-consortium/reconcile-provider-invoices-with-metered-use.md` - `generated/experience/edge-consortium/reconcile-with-the-student-record-system.md` - `generated/experience/edge-consortium/record-a-call-with-consent.md` - `generated/experience/edge-consortium/record-a-conflict-of-interest-assessment.md` - `generated/experience/edge-consortium/record-a-data-processing-agreement.md` - `generated/experience/edge-consortium/record-a-priority-date.md` - `generated/experience/edge-consortium/record-a-trademark-and-brand-use-rule.md` - `generated/experience/edge-consortium/record-an-invention-disclosure.md` - `generated/experience/edge-consortium/record-applications-and-outcomes-in-a-portfolio.md` - `generated/experience/edge-consortium/record-attendance-and-engagement.md` - `generated/experience/edge-consortium/record-authorship-and-contributor-roles.md` - `generated/experience/edge-consortium/record-background-and-foreground-ip.md` - `generated/experience/edge-consortium/record-copyright-and-attribution.md` - `generated/experience/edge-consortium/record-energy-per-job.md` - `generated/experience/edge-consortium/record-payment-compliance.md` - `generated/experience/edge-consortium/record-provenance-and-lineage.md` - `generated/experience/edge-consortium/record-working-hours-and-leave.md` - `generated/experience/edge-consortium/refuse-an-action-outside-the-mandate.md` - `generated/experience/edge-consortium/register-a-compute-resource.md` - `generated/experience/edge-consortium/register-a-dataset-in-the-catalogue.md` - `generated/experience/edge-consortium/register-a-delegated-authority.md` - `generated/experience/edge-consortium/register-a-research-project-with-an-id.md` - `generated/experience/edge-consortium/register-a-university-as-a-member-institution.md` - `generated/experience/edge-consortium/register-as-a-supplier.md` - `generated/experience/edge-consortium/register-with-the-office-for-students.md` - `generated/experience/edge-consortium/reject-a-contribution-with-reasons.md` - `generated/experience/edge-consortium/release-a-version.md` - `generated/experience/edge-consortium/release-marks-and-feedback.md` - `generated/experience/edge-consortium/release-results-with-disclosure-control.md` - `generated/experience/edge-consortium/renew-annual-institutional-membership.md` - `generated/experience/edge-consortium/renew-reviewer-verification.md` - `generated/experience/edge-consortium/report-a-personal-data-breach.md` - `generated/experience/edge-consortium/report-a-research-misconduct-concern.md` - `generated/experience/edge-consortium/report-carbon-intensity-per-accepted-outcome.md` - `generated/experience/edge-consortium/report-expenditure-to-the-funder.md` - `generated/experience/edge-consortium/report-monitoring-measures.md` - `generated/experience/edge-consortium/report-performance-indicator-evidence.md` - `generated/experience/edge-consortium/report-review-throughput-and-times.md` - `generated/experience/edge-consortium/report-tokens-prompts-and-time-to-evidence.md` - `generated/experience/edge-consortium/report-utilisation.md` - `generated/experience/edge-consortium/request-a-compute-allowance-for-a-project.md` - `generated/experience/edge-consortium/request-a-purpose-view-on-federated-data.md` - `generated/experience/edge-consortium/request-a-sandbox-with-gpu-and-data.md` - `generated/experience/edge-consortium/request-approval-for-a-consequential-action.md` - `generated/experience/edge-consortium/request-changes-from-the-contributor.md` - `generated/experience/edge-consortium/request-supervisor-allocation.md` - `generated/experience/edge-consortium/research-and-development-partnership.md` - `generated/experience/edge-consortium/research-consortium.md` - `generated/experience/edge-consortium/reserve-capacity-for-a-deadline.md` - `generated/experience/edge-consortium/resolve-conflicting-reviews.md` - `generated/experience/edge-consortium/respond-to-a-security-incident.md` - `generated/experience/edge-consortium/respond-to-an-invitation.md` - `generated/experience/edge-consortium/retract-or-correct-a-verified-entry.md` - `generated/experience/edge-consortium/retry-a-failed-job.md` - `generated/experience/edge-consortium/review-a-contribution-against-criteria.md` - `generated/experience/edge-consortium/review-a-funder-expenditure-statement.md` - `generated/experience/edge-consortium/review-a-pull-request.md` - `generated/experience/edge-consortium/review-access-logs-and-anomalies.md` - `generated/experience/edge-consortium/review-agent-receipts.md` - `generated/experience/edge-consortium/review-module-evaluation-feedback.md` - `generated/experience/edge-consortium/review-quality-and-standards-evidence.md` - `generated/experience/edge-consortium/review-the-dpia-and-record-of-processing.md` - `generated/experience/edge-consortium/revoke-a-purpose-view.md` - `generated/experience/edge-consortium/revoke-access-for-a-leaver.md` - `generated/experience/edge-consortium/rotate-keys.md` - `generated/experience/edge-consortium/run-a-clarification-round.md` - `generated/experience/edge-consortium/run-a-compute-job-under-mandate.md` - `generated/experience/edge-consortium/run-a-cyber-recovery-drill.md` - `generated/experience/edge-consortium/run-a-live-assessment.md` - `generated/experience/edge-consortium/run-a-live-seminar.md` - `generated/experience/edge-consortium/run-a-security-classified-procurement.md` - `generated/experience/edge-consortium/run-a-training-job-on-scheduled-compute.md` - `generated/experience/edge-consortium/run-a-vulnerability-scan-on-a-release.md` - `generated/experience/edge-consortium/run-an-approved-query-where-data-lives.md` - `generated/experience/edge-consortium/run-code-in-the-sandbox.md` - `generated/experience/edge-consortium/run-data-quality-checks-on-a-database.md` - `generated/experience/edge-consortium/sample-verified-entries-for-quality.md` - `generated/experience/edge-consortium/save-sources-to-a-reading-list.md` - `generated/experience/edge-consortium/scale-with-an-external-provider.md` - `generated/experience/edge-consortium/schedule-a-job.md` - `generated/experience/edge-consortium/schedule-a-meeting-on-request.md` - `generated/experience/edge-consortium/schedule-a-secure-government-resource.md` - `generated/experience/edge-consortium/schedule-a-send-and-set-a-signature.md` - `generated/experience/edge-consortium/schedule-and-record-supervision-meetings.md` - `generated/experience/edge-consortium/schedule-deliveries-and-logistics.md` - `generated/experience/edge-consortium/search-across-mail-channels-and-meetings.md` - `generated/experience/edge-consortium/search-libraries-and-the-knowledge-graph.md` - `generated/experience/edge-consortium/send-a-contract-for-signature.md` - `generated/experience/edge-consortium/send-mail-from-an-estate-mailbox.md` - `generated/experience/edge-consortium/set-a-carbon-budget-and-threshold.md` - `generated/experience/edge-consortium/set-a-project-access-control-policy.md` - `generated/experience/edge-consortium/set-agent-authority-scope.md` - `generated/experience/edge-consortium/set-and-publish-key-performance-indicators.md` - `generated/experience/edge-consortium/set-assessment-and-marking-scheme.md` - `generated/experience/edge-consortium/set-contract-parties-permissions-and-expiry.md` - `generated/experience/edge-consortium/set-evaluation-criteria-and-weightings.md` - `generated/experience/edge-consortium/set-mail-rules-flags-and-snooze.md` - `generated/experience/edge-consortium/set-mention-and-tag-rules.md` - `generated/experience/edge-consortium/set-out-of-office-and-working-hours.md` - `generated/experience/edge-consortium/set-permissions-on-project-records.md` - `generated/experience/edge-consortium/set-presence.md` - `generated/experience/edge-consortium/set-project-governance.md` - `generated/experience/edge-consortium/set-quotas-per-project.md` - `generated/experience/edge-consortium/set-renewal-and-expiry-alerts.md` - `generated/experience/edge-consortium/set-retention-and-export-a-channel.md` - `generated/experience/edge-consortium/set-review-criteria-for-a-field.md` - `generated/experience/edge-consortium/set-review-fields-and-availability.md` - `generated/experience/edge-consortium/set-scheduling-policy-and-queues.md` - `generated/experience/edge-consortium/set-the-access-tier-on-an-item.md` - `generated/experience/edge-consortium/set-the-review-deadline-and-reminders.md` - `generated/experience/edge-consortium/set-up-an-awarded-grant.md` - `generated/experience/edge-consortium/share-a-calendar-with-a-delegate.md` - `generated/experience/edge-consortium/share-a-campus-with-a-partner-university.md` - `generated/experience/edge-consortium/share-a-creation-in-the-judebox-exchange.md` - `generated/experience/edge-consortium/share-a-document-in-a-call.md` - `generated/experience/edge-consortium/share-a-workspace-folder-across-universities.md` - `generated/experience/edge-consortium/shift-jobs-to-low-carbon-windows.md` - `generated/experience/edge-consortium/show-undisclosed-figures-as-gaps.md` - `generated/experience/edge-consortium/sign-a-collaboration-agreement.md` - `generated/experience/edge-consortium/sign-a-contract-with-edge-id.md` - `generated/experience/edge-consortium/sign-off-a-progression-milestone.md` - `generated/experience/edge-consortium/sign-the-contributor-agreement.md` - `generated/experience/edge-consortium/sign-the-employment-contract.md` - `generated/experience/edge-consortium/sign-the-public-contract.md` - `generated/experience/edge-consortium/stop-or-kill-a-job.md` - `generated/experience/edge-consortium/store-the-signed-contract-in-the-vault.md` - `generated/experience/edge-consortium/student-to-contractor-pathway.md` - `generated/experience/edge-consortium/submit-a-bid.md` - `generated/experience/edge-consortium/submit-a-contribution-for-review.md` - `generated/experience/edge-consortium/submit-a-grant-application.md` - `generated/experience/edge-consortium/submit-a-paper-to-consortium-review.md` - `generated/experience/edge-consortium/submit-an-annual-report-to-the-funder.md` - `generated/experience/edge-consortium/submit-an-invoice-and-track-payment.md` - `generated/experience/edge-consortium/submit-staff-for-security-vetting.md` - `generated/experience/edge-consortium/suspend-or-end-a-membership.md` - `generated/experience/edge-consortium/terminate-a-contract-and-publish-the-notice.md` - `generated/experience/edge-consortium/terminate-a-contract.md` - `generated/experience/edge-consortium/trace-network-ingress-and-egress.md` - `generated/experience/edge-consortium/track-research-outputs-for-a-ref-return.md` - `generated/experience/edge-consortium/track-royalties-and-revenue-share.md` - `generated/experience/edge-consortium/triage-a-security-vulnerability-report.md` - `generated/experience/edge-consortium/triage-the-submissions-queue.md` - `generated/experience/edge-consortium/university-enrolment.md` - `generated/experience/edge-consortium/update-supplier-information.md` - `generated/experience/edge-consortium/use-the-terminal.md` - `generated/experience/edge-consortium/vary-a-contract.md` - `generated/experience/edge-consortium/verify-a-fact-before-stating-it.md` - `generated/experience/edge-consortium/verify-a-fact-in-the-knowledge-graph.md` - `generated/experience/edge-consortium/verify-a-person-s-identity.md` - `generated/experience/edge-consortium/verify-a-signature.md` - `generated/experience/edge-consortium/verify-institutional-accreditation-status.md` - `generated/experience/edge-consortium/verify-procurement-notice-publication.md` - `generated/experience/edge-consortium/verify-the-accessibility-statement.md` - `generated/experience/edge-consortium/verify-the-audit-chain-offline.md` - `generated/experience/edge-consortium/version-and-roll-back-a-creation.md` - `generated/experience/edge-consortium/vet-a-reviewer-or-supplier-staff-member.md` - `generated/experience/edge-consortium/view-the-diary-by-day-week-month.md` - `generated/experience/edge-consortium/view-the-public-reviewer-track-record.md` - `generated/experience/edge-consortium/vote-on-a-poll.md` - `generated/experience/edge-consortium/write-a-data-management-plan.md` - `generated/experience/edge-consortium/write-a-paper-in-markdowns-with-co-authors.md` - `generated/experience/finance/accept-a-cheque-deposit.md` - `generated/experience/finance/access-a-read-only-evidence-room.md` - `generated/experience/finance/account-for-agent-resource-use.md` - `generated/experience/finance/accrue-deposit-interest-expense.md` - `generated/experience/finance/acknowledge-a-complaint.md` - `generated/experience/finance/act-under-a-power-of-attorney.md` - `generated/experience/finance/action-an-erasure-or-rectification-request.md` - `generated/experience/finance/activate-a-monitoring-rule.md` - `generated/experience/finance/activate-a-proposed-rate.md` - `generated/experience/finance/add-a-joint-account-holder.md` - `generated/experience/finance/add-a-payee-with-name-check.md` - `generated/experience/finance/add-or-remove-a-signatory.md` - `generated/experience/finance/agent-mandate.md` - `generated/experience/finance/agent-onboarding.md` - `generated/experience/finance/agent-retirement.md` - `generated/experience/finance/agree-a-forbearance-plan.md` - `generated/experience/finance/amend-a-policy-mid-term.md` - `generated/experience/finance/amend-a-repayment-schedule.md` - `generated/experience/finance/amend-or-cancel-a-standing-order.md` - `generated/experience/finance/analyse-complaint-root-causes.md` - `generated/experience/finance/answer-a-customer-question-from-records.md` - `generated/experience/finance/answer-a-subject-access-request.md` - `generated/experience/finance/apply-a-database-migration.md` - `generated/experience/finance/apply-a-payment-holiday.md` - `generated/experience/finance/apply-enhanced-due-diligence.md` - `generated/experience/finance/apply-for-a-business-account.md` - `generated/experience/finance/apply-for-a-business-loan-or-facility.md` - `generated/experience/finance/apply-for-a-personal-loan.md` - `generated/experience/finance/apply-for-loan.md` - `generated/experience/finance/apply-reasonable-adjustments.md` - `generated/experience/finance/apply-retention-and-deletion.md` - `generated/experience/finance/approve-a-financial-promotion.md` - `generated/experience/finance/approve-a-new-integration-or-api-consumer.md` - `generated/experience/finance/approve-or-decline-an-application.md` - `generated/experience/finance/approve-or-reject-a-payment.md` - `generated/experience/finance/assess-a-claim.md` - `generated/experience/finance/assess-an-outsourcing-or-third-party.md` - `generated/experience/finance/assess-income-and-expenditure.md` - `generated/experience/finance/assess-source-of-funds-and-wealth.md` - `generated/experience/finance/audit-partner-usage.md` - `generated/experience/finance/authorise-card-transactions-in-real-time.md` - `generated/experience/finance/back-up-the-database.md` - `generated/experience/finance/bill-a-partner.md` - `generated/experience/finance/block-a-payment-in-flight.md` - `generated/experience/finance/book-an-appointment.md` - `generated/experience/finance/bulk-payments.md` - `generated/experience/finance/business-roles-and-approvals.md` - `generated/experience/finance/calculate-and-post-interest-accrual.md` - `generated/experience/finance/calculate-the-liquidity-coverage-ratio.md` - `generated/experience/finance/cancel-a-direct-debit.md` - `generated/experience/finance/cancel-a-policy.md` - `generated/experience/finance/capital-to-outcomes-ledger.md` - `generated/experience/finance/cards.md` - `generated/experience/finance/change-contact-details.md` - `generated/experience/finance/claim-a-direct-debit-indemnity-refund.md` - `generated/experience/finance/close-a-business-account.md` - `generated/experience/finance/close-a-period.md` - `generated/experience/finance/close-account.md` - `generated/experience/finance/close-an-account.md` - `generated/experience/finance/collect-direct-debits-as-an-originator.md` - `generated/experience/finance/collect-premiums.md` - `generated/experience/finance/compare-cover-options.md` - `generated/experience/finance/complete-entity-tax-classification.md` - `generated/experience/finance/complete-tax-residency-self-certification.md` - `generated/experience/finance/compute-expected-credit-losses.md` - `generated/experience/finance/confidence-report.md` - `generated/experience/finance/configure-scheduled-jobs.md` - `generated/experience/finance/confirm-balances-as-external-auditor.md` - `generated/experience/finance/connect-external-accounts.md` - `generated/experience/finance/contact-a-customer-in-arrears.md` - `generated/experience/finance/continue-a-joint-account-after-a-death.md` - `generated/experience/finance/create-a-monitoring-rule.md` - `generated/experience/finance/create-a-prospect-record.md` - `generated/experience/finance/create-a-savings-product.md` - `generated/experience/finance/create-and-release-a-hold.md` - `generated/experience/finance/crypto-wallet.md` - `generated/experience/finance/decide-a-treasury-proposal.md` - `generated/experience/finance/decide-an-app-reimbursement-claim.md` - `generated/experience/finance/decide-and-file-a-sar.md` - `generated/experience/finance/decline-with-reasons.md` - `generated/experience/finance/deploy-a-release-as-jude.md` - `generated/experience/finance/deploy-a-release-through-deploy-py.md` - `generated/experience/finance/design-a-new-lending-product.md` - `generated/experience/finance/detect-a-missed-repayment.md` - `generated/experience/finance/direct-debits.md` - `generated/experience/finance/disburse-a-loan.md` - `generated/experience/finance/disclose-a-vulnerability-or-need.md` - `generated/experience/finance/disposition-an-alert.md` - `generated/experience/finance/dispute-a-card-transaction.md` - `generated/experience/finance/disputes.md` - `generated/experience/finance/draw-down-from-a-facility.md` - `generated/experience/finance/enforce-strong-customer-authentication.md` - `generated/experience/finance/escalate-a-customer-concern.md` - `generated/experience/finance/execute-a-money-market-placement.md` - `generated/experience/finance/exit-a-customer-for-financial-crime.md` - `generated/experience/finance/explain-a-transaction.md` - `generated/experience/finance/export-the-evidence-bundle.md` - `generated/experience/finance/export-transactions-to-accounting-software.md` - `generated/experience/finance/financing-graph-independent-demand.md` - `generated/experience/finance/flag-a-mule-account.md` - `generated/experience/finance/fraud-sharing-network.md` - `generated/experience/finance/freeze-or-unfreeze-a-card.md` - `generated/experience/finance/fund-settlement-accounts-for-rails.md` - `generated/experience/finance/give-or-revoke-open-banking-consent.md` - `generated/experience/finance/grant-an-operator-role.md` - `generated/experience/finance/hand-a-case-to-a-specialist-team.md` - `generated/experience/finance/hand-work-between-agents.md` - `generated/experience/finance/handle-a-banking-protocol-concern.md` - `generated/experience/finance/handle-a-breathing-space-notice.md` - `generated/experience/finance/handle-a-debt-sale.md` - `generated/experience/finance/handle-a-gambling-block-request.md` - `generated/experience/finance/handle-a-mental-capacity-concern.md` - `generated/experience/finance/handle-a-partner-data-request-or-deletion.md` - `generated/experience/finance/handle-a-security-incident.md` - `generated/experience/finance/handle-a-trigger-event-review.md` - `generated/experience/finance/handle-a-whistleblowing-report.md` - `generated/experience/finance/handle-an-indemnity-claim-as-paying-bank.md` - `generated/experience/finance/hedge-interest-rate-exposure.md` - `generated/experience/finance/import-a-bank-or-rail-statement.md` - `generated/experience/finance/inspect-a-sar-audit-trail.md` - `generated/experience/finance/introduce-a-customer-to-a-specialist.md` - `generated/experience/finance/investigate-a-complaint.md` - `generated/experience/finance/investigate-a-payment-exception.md` - `generated/experience/finance/investigate-a-production-error.md` - `generated/experience/finance/investigate-a-suspected-fraud.md` - `generated/experience/finance/invite-a-user-with-a-role.md` - `generated/experience/finance/issue-a-default-notice.md` - `generated/experience/finance/issue-a-final-response.md` - `generated/experience/finance/issue-a-finding-to-the-bank.md` - `generated/experience/finance/issue-a-policy.md` - `generated/experience/finance/issue-a-replacement-card-in-branch.md` - `generated/experience/finance/issue-an-agent-mandate.md` - `generated/experience/finance/issue-annual-loan-statements.md` - `generated/experience/finance/joint-accounts.md` - `generated/experience/finance/lend.md` - `generated/experience/finance/lift-an-account-freeze.md` - `generated/experience/finance/log-an-operational-incident.md` - `generated/experience/finance/log-an-outbound-contact.md` - `generated/experience/finance/maintain-internal-and-nostro-accounts.md` - `generated/experience/finance/maintain-the-business-wide-risk-assessment.md` - `generated/experience/finance/maintain-the-chart-of-accounts.md` - `generated/experience/finance/maintain-the-record-of-processing.md` - `generated/experience/finance/maintain-the-regulatory-obligations-register.md` - `generated/experience/finance/maintain-the-risk-register.md` - `generated/experience/finance/make-a-complaint.md` - `generated/experience/finance/make-a-faster-payment.md` - `generated/experience/finance/make-a-regulatory-notification.md` - `generated/experience/finance/make-an-internal-transfer.md` - `generated/experience/finance/make-an-international-payment.md` - `generated/experience/finance/manage-a-syndicated-participation.md` - `generated/experience/finance/manage-capacity.md` - `generated/experience/finance/manage-encryption-and-device-binding.md` - `generated/experience/finance/manage-environments-and-secrets-free-config.md` - `generated/experience/finance/manage-foreign-exchange-positions.md` - `generated/experience/finance/manage-partner-user-access.md` - `generated/experience/finance/manage-payee-name-check-responses.md` - `generated/experience/finance/manage-payment-limits.md` - `generated/experience/finance/manage-senior-manager-responsibilities.md` - `generated/experience/finance/manage-the-central-bank-reserve-account.md` - `generated/experience/finance/map-important-business-services.md` - `generated/experience/finance/mark-a-customer-as-a-fraud-victim.md` - `generated/experience/finance/match-statement-rows-to-ledger-legs.md` - `generated/experience/finance/measure-carbon-tokens-and-time-to-evidence.md` - `generated/experience/finance/message-a-customer-in-a-case.md` - `generated/experience/finance/money-provenance-certificate.md` - `generated/experience/finance/monitor-chaps-cut-offs-and-queue.md` - `generated/experience/finance/monitor-conduct-and-quality.md` - `generated/experience/finance/monitor-facility-covenants.md` - `generated/experience/finance/monitor-intraday-liquidity.md` - `generated/experience/finance/monitor-job-runs-and-leases.md` - `generated/experience/finance/monitor-privileged-actions.md` - `generated/experience/finance/monitor-product-outcomes.md` - `generated/experience/finance/monitor-service-health.md` - `generated/experience/finance/monitor-without-blocking.md` - `generated/experience/finance/notify-a-claim.md` - `generated/experience/finance/notify-a-death-and-restrict-accounts.md` - `generated/experience/finance/notify-a-personal-data-breach.md` - `generated/experience/finance/offboard-a-business-customer.md` - `generated/experience/finance/onboard-a-partner-organisation.md` - `generated/experience/finance/open-a-fixed-term-deposit.md` - `generated/experience/finance/open-a-foreign-currency-account.md` - `generated/experience/finance/open-a-personal-current-account.md` - `generated/experience/finance/open-a-savings-account.md` - `generated/experience/finance/open-a-service-case.md` - `generated/experience/finance/open-a-transaction-receipt.md` - `generated/experience/finance/open-business-account.md` - `generated/experience/finance/open-personal-account.md` - `generated/experience/finance/open-virtual-accounts-under-a-master.md` - `generated/experience/finance/order-business-cards-for-staff.md` - `generated/experience/finance/overdraft.md` - `generated/experience/finance/pay-redress.md` - `generated/experience/finance/perform-a-data-protection-impact-assessment.md` - `generated/experience/finance/perform-an-internal-audit.md` - `generated/experience/finance/place-a-temporary-account-freeze.md` - `generated/experience/finance/plan-a-portfolio-review.md` - `generated/experience/finance/plan-the-internal-audit-year.md` - `generated/experience/finance/post-a-manual-journal.md` - `generated/experience/finance/prepare-a-business-proposition.md` - `generated/experience/finance/prepare-a-compliance-return.md` - `generated/experience/finance/prepare-a-payment-for-approval.md` - `generated/experience/finance/prepare-a-payment-for-human-approval.md` - `generated/experience/finance/prepare-funds-transfer-pricing.md` - `generated/experience/finance/prepare-regulatory-returns.md` - `generated/experience/finance/prepare-statutory-accounts.md` - `generated/experience/finance/process-a-bacs-input-and-output-cycle.md` - `generated/experience/finance/process-a-cheque-clearing-exception.md` - `generated/experience/finance/process-a-counter-cash-deposit.md` - `generated/experience/finance/process-a-counter-cash-withdrawal.md` - `generated/experience/finance/process-a-loan-repayment-manually.md` - `generated/experience/finance/process-an-inbound-faster-payment.md` - `generated/experience/finance/process-card-clearing-and-settlement.md` - `generated/experience/finance/produce-a-kyc-file-for-review.md` - `generated/experience/finance/produce-a-trial-balance.md` - `generated/experience/finance/produce-the-annual-mlro-report.md` - `generated/experience/finance/produce-the-fscs-single-customer-view.md` - `generated/experience/finance/propose-a-future-rate.md` - `generated/experience/finance/propose-a-treasury-credit.md` - `generated/experience/finance/provenance-coverage-report.md` - `generated/experience/finance/provide-extra-evidence-on-request.md` - `generated/experience/finance/proximity-verification.md` - `generated/experience/finance/publish-terms-and-conditions.md` - `generated/experience/finance/pull-and-record-a-credit-reference.md` - `generated/experience/finance/query-a-mule-account-indicator.md` - `generated/experience/finance/raise-an-internal-suspicion-report.md` - `generated/experience/finance/receive-a-fraud-intelligence-signal.md` - `generated/experience/finance/receive-a-rate-change-notice.md` - `generated/experience/finance/recognise-loan-interest-income.md` - `generated/experience/finance/reconcile-rail-settlement.md` - `generated/experience/finance/reconcile-suspense-accounts.md` - `generated/experience/finance/record-a-credit-decision-override.md` - `generated/experience/finance/record-a-customer-meeting.md` - `generated/experience/finance/record-a-customer-risk-assessment.md` - `generated/experience/finance/record-a-declined-application-reason.md` - `generated/experience/finance/record-a-regulatory-breach.md` - `generated/experience/finance/record-a-support-need.md` - `generated/experience/finance/record-a-vulnerability-disclosure.md` - `generated/experience/finance/record-aml-training-completion.md` - `generated/experience/finance/record-and-test-a-control.md` - `generated/experience/finance/recover-account-access.md` - `generated/experience/finance/recover-funds-from-a-receiving-account.md` - `generated/experience/finance/refer-a-complaint-to-fos.md` - `generated/experience/finance/refer-to-a-debt-collection-agency.md` - `generated/experience/finance/refer-to-debt-advice.md` - `generated/experience/finance/refund-an-unauthorised-transaction.md` - `generated/experience/finance/register-a-court-appointed-deputy.md` - `generated/experience/finance/register-a-default-with-agencies.md` - `generated/experience/finance/register-a-lasting-power-of-attorney.md` - `generated/experience/finance/register-a-new-device.md` - `generated/experience/finance/register-a-third-party-mandate.md` - `generated/experience/finance/register-and-value-collateral.md` - `generated/experience/finance/register-beneficial-owners.md` - `generated/experience/finance/regulator-inspection.md` - `generated/experience/finance/release-estate-funds.md` - `generated/experience/finance/rely-on-another-firm-s-due-diligence.md` - `generated/experience/finance/remove-a-joint-holder.md` - `generated/experience/finance/rename-an-account.md` - `generated/experience/finance/renew-a-policy.md` - `generated/experience/finance/reopen-a-period.md` - `generated/experience/finance/repair-a-misdirected-payment.md` - `generated/experience/finance/repay-or-overpay-a-loan.md` - `generated/experience/finance/report-a-bereavement.md` - `generated/experience/finance/report-a-change-of-company-details.md` - `generated/experience/finance/report-a-lost-or-stolen-card.md` - `generated/experience/finance/report-an-app-scam.md` - `generated/experience/finance/report-an-unauthorised-payment.md` - `generated/experience/finance/report-complaints-data.md` - `generated/experience/finance/report-fraud-losses-to-the-regulator.md` - `generated/experience/finance/report-fraud-to-action-fraud-or-police.md` - `generated/experience/finance/report-insurance-data-to-the-regulator.md` - `generated/experience/finance/report-payment-statistics.md` - `generated/experience/finance/report-to-credit-reference-agencies.md` - `generated/experience/finance/request-a-bank-statement-letter.md` - `generated/experience/finance/request-a-company-audit-letter.md` - `generated/experience/finance/request-a-covenant-waiver.md` - `generated/experience/finance/request-a-date-of-death-balance-letter.md` - `generated/experience/finance/request-a-defence-against-money-laundering.md` - `generated/experience/finance/request-a-loan-settlement-quote.md` - `generated/experience/finance/request-a-trade-finance-instrument.md` - `generated/experience/finance/request-affordability-evidence.md` - `generated/experience/finance/request-an-arranged-overdraft.md` - `generated/experience/finance/request-an-insurance-quote-from-kyc.md` - `generated/experience/finance/request-further-information.md` - `generated/experience/finance/request-money.md` - `generated/experience/finance/request-my-personal-data.md` - `generated/experience/finance/request-owner-approval-for-a-consequential-action.md` - `generated/experience/finance/request-records-from-the-bank.md` - `generated/experience/finance/reset-access-for-a-customer.md` - `generated/experience/finance/resolve-within-3-business-days.md` - `generated/experience/finance/respond-to-a-production-order.md` - `generated/experience/finance/respond-to-an-open-banking-payment-request.md` - `generated/experience/finance/responsible-lending-decision.md` - `generated/experience/finance/restore-from-a-backup.md` - `generated/experience/finance/retry-a-failed-job-step.md` - `generated/experience/finance/retry-a-failed-occurrence.md` - `generated/experience/finance/return-an-unapplied-payment.md` - `generated/experience/finance/reverse-a-journal-with-a-linked-entry.md` - `generated/experience/finance/review-a-complaint-file-as-fos.md` - `generated/experience/finance/review-a-loan-application.md` - `generated/experience/finance/review-access-rights.md` - `generated/experience/finance/review-an-application-queue.md` - `generated/experience/finance/review-app-reimbursement-performance.md` - `generated/experience/finance/review-customer-financial-statements.md` - `generated/experience/finance/review-device-and-session-signals.md` - `generated/experience/finance/review-operational-resilience-evidence.md` - `generated/experience/finance/review-sm-cr-responsibilities.md` - `generated/experience/finance/review-the-vulnerability-register.md` - `generated/experience/finance/revoke-a-role-on-leaving.md` - `generated/experience/finance/revoke-an-agent-mandate.md` - `generated/experience/finance/roll-back-a-release.md` - `generated/experience/finance/rotate-keys-and-device-bindings.md` - `generated/experience/finance/run-a-consumer-duty-outcomes-review.md` - `generated/experience/finance/run-a-disaster-recovery-test.md` - `generated/experience/finance/run-a-fair-value-assessment.md` - `generated/experience/finance/run-a-kyc-check-on-a-subject.md` - `generated/experience/finance/run-a-ledger-report-for-a-period.md` - `generated/experience/finance/run-a-liquidity-stress-test.md` - `generated/experience/finance/run-an-annual-relationship-review.md` - `generated/experience/finance/run-due-jobs-as-service-actor.md` - `generated/experience/finance/run-due-standing-orders.md` - `generated/experience/finance/run-end-of-day-payment-reconciliation.md` - `generated/experience/finance/run-payroll-payments.md` - `generated/experience/finance/run-penetration-testing.md` - `generated/experience/finance/run-periodic-kyc-refresh.md` - `generated/experience/finance/run-protected-reveals.md` - `generated/experience/finance/run-sanctions-screening.md` - `generated/experience/finance/run-transaction-monitoring.md` - `generated/experience/finance/sample-test-receipts-as-external-auditor.md` - `generated/experience/finance/sanctions-and-pep-screening.md` - `generated/experience/finance/save-an-unfinished-application.md` - `generated/experience/finance/savings-pots.md` - `generated/experience/finance/schedule-a-future-payment.md` - `generated/experience/finance/scheduled-payments.md` - `generated/experience/finance/screen-a-list-against-sanctions.md` - `generated/experience/finance/screen-existing-customers-against-list-updates.md` - `generated/experience/finance/screen-for-politically-exposed-persons.md` - `generated/experience/finance/search-transactions.md` - `generated/experience/finance/security-and-devices.md` - `generated/experience/finance/send-a-chaps-payment.md` - `generated/experience/finance/send-a-receiving-bank-app-notice.md` - `generated/experience/finance/send-money.md` - `generated/experience/finance/set-and-monitor-exposure-limits.md` - `generated/experience/finance/set-balance-and-payment-alerts.md` - `generated/experience/finance/set-card-limits-and-controls.md` - `generated/experience/finance/set-eligibility-rules.md` - `generated/experience/finance/set-fees-and-charges.md` - `generated/experience/finance/set-joint-operating-rules.md` - `generated/experience/finance/set-loan-terms-and-produce-an-offer.md` - `generated/experience/finance/set-negotiated-pricing.md` - `generated/experience/finance/set-up-a-direct-debit.md` - `generated/experience/finance/set-up-a-mandate.md` - `generated/experience/finance/set-up-a-standing-order.md` - `generated/experience/finance/set-up-merchant-acquiring.md` - `generated/experience/finance/settle-a-claim.md` - `generated/experience/finance/settle-and-close-a-loan.md` - `generated/experience/finance/solution-bank-funding.md` - `generated/experience/finance/stop-a-savings-enrolment.md` - `generated/experience/finance/submit-transactions-for-monitoring.md` - `generated/experience/finance/support-a-customer-in-financial-difficulty.md` - `generated/experience/finance/support-the-external-audit.md` - `generated/experience/finance/switch-account-to-edge-bank.md` - `generated/experience/finance/terminate-a-partnership.md` - `generated/experience/finance/trace-ingress-and-egress.md` - `generated/experience/finance/track-remediation-actions.md` - `generated/experience/finance/transaction-monitoring.md` - `generated/experience/finance/transition-a-case-state.md` - `generated/experience/finance/triage-a-case-queue.md` - `generated/experience/finance/tune-a-fraud-rule.md` - `generated/experience/finance/underwrite-a-business-facility.md` - `generated/experience/finance/underwrite-a-policy.md` - `generated/experience/finance/update-communication-preferences.md` - `generated/experience/finance/upload-a-bulk-payment-file.md` - `generated/experience/finance/validate-a-credit-risk-model.md` - `generated/experience/finance/verify-a-caller-or-visitor.md` - `generated/experience/finance/verify-a-company-and-its-officers.md` - `generated/experience/finance/verify-a-regulatory-return-against-the-ledger.md` - `generated/experience/finance/verify-a-trust-or-charity.md` - `generated/experience/finance/verify-address-evidence.md` - `generated/experience/finance/verify-an-identity-through-edge-id.md` - `generated/experience/finance/verify-executor-or-administrator.md` - `generated/experience/finance/verify-identity-evidence.md` - `generated/experience/finance/view-and-download-a-statement.md` - `generated/experience/finance/view-balances-and-available-funds.md` - `generated/experience/finance/view-multi-account-cash-position.md` - `generated/experience/finance/view-pin-securely.md` - `generated/experience/finance/void-an-imported-statement.md` - `generated/experience/finance/withdraw-a-product.md` - `generated/experience/finance/withdraw-a-proposed-rate.md` - `generated/experience/finance/withdraw-an-application.md` - `generated/experience/finance/work-the-fraud-alert-queue.md` - `generated/experience/finance/write-a-fraud-case-outcome-letter.md` - `generated/experience/finance/write-off-a-debt.md` - `generated/experience/foundation/automated-review-outreach-and-scheduling.md` - `generated/experience/foundation/bottleneck-and-blocker-removal.md` - `generated/experience/foundation/claim-organisation-by-magic-link.md` - `generated/experience/foundation/evaluation-and-commercial-licensing.md` - `generated/experience/foundation/licence-enforcement-evidence.md` - `generated/experience/foundation/organisation-directory-and-claim.md` - `generated/experience/foundation/pre-onboarded-organisations.md` - `generated/experience/foundation/regulator-auditor-reviewer-workspace.md` - `generated/experience/foundation/scope-change-intake.md` - `generated/experience/judiverse/agent-builder.md` - `generated/experience/judiverse/calendar-and-scheduling.md` - `generated/experience/judiverse/contacts-directory-and-search.md` - `generated/experience/judiverse/deep-research.md` - `generated/experience/judiverse/documents-and-forms.md` - `generated/experience/judiverse/jude-assistant.md` - `generated/experience/judiverse/judebox-agents.md` - `generated/experience/judiverse/judebox-calls.md` - `generated/experience/judiverse/judebox-connections.md` - `generated/experience/judiverse/judebox-contacts.md` - `generated/experience/judiverse/judebox-email.md` - `generated/experience/judiverse/judebox-memory-and-vault.md` - `generated/experience/judiverse/judebox-messaging.md` - `generated/experience/judiverse/judebox-music.md` - `generated/experience/judiverse/judebox-numbers-and-mailboxes.md` - `generated/experience/judiverse/judebox-webhooks.md` - `generated/experience/judiverse/judiverse-included-with-every-licence.md` - `generated/experience/judiverse/judiverse-standalone-subscription.md` - `generated/experience/judiverse/mail-and-agent-mailboxes.md` - `generated/experience/judiverse/messages-and-chat.md` - `generated/experience/judiverse/procedure-compilation.md` - `generated/experience/judiverse/tasks-and-workflows.md` - `generated/experience/public-edge/accept-a-package-referral.md` - `generated/experience/public-edge/act-on-a-deadline-stack-item.md` - `generated/experience/public-edge/activate-the-emergency-centre-and-keep-the-decision-log.md` - `generated/experience/public-edge/adopt-a-risk-based-reserves-strategy-with-minimum-level-and.md` - `generated/experience/public-edge/advise-on-legality-and-sign-the-legal-implications-of-a-repo.md` - `generated/experience/public-edge/agree-a-care-and-support-plan-and-personal-budget.md` - `generated/experience/public-edge/agree-a-deferred-payment.md` - `generated/experience/public-edge/agree-a-personalised-housing-plan.md` - `generated/experience/public-edge/agree-an-information-sharing-agreement.md` - `generated/experience/public-edge/agree-and-seal-a-planning-obligation.md` - `generated/experience/public-edge/agree-the-council-plan-and-its-outcomes.md` - `generated/experience/public-edge/agree-the-officer-and-member-protocol.md` - `generated/experience/public-edge/agree-the-political-group-position-before-a-meeting.md` - `generated/experience/public-edge/agree-the-post-election-timetable-and-transition-protocol.md` - `generated/experience/public-edge/allocate-committee-seats-by-political-balance.md` - `generated/experience/public-edge/analog-human-switch-for-irreversible-actions.md` - `generated/experience/public-edge/answer-a-best-value-notice-or-mhclg-engagement-with-one-evid.md` - `generated/experience/public-edge/answer-a-code-of-conduct-complaint.md` - `generated/experience/public-edge/answer-a-corporate-complaint.md` - `generated/experience/public-edge/answer-a-freedom-of-information-request.md` - `generated/experience/public-edge/answer-a-local-land-charges-search.md` - `generated/experience/public-edge/answer-a-media-enquiry-and-issue-a-press-release.md` - `generated/experience/public-edge/answer-a-standards-complaint.md` - `generated/experience/public-edge/answer-a-subject-access-request.md` - `generated/experience/public-edge/answer-a-written-parliamentary-question.md` - `generated/experience/public-edge/answer-an-appeal-to-the-valuation-tribunal.md` - `generated/experience/public-edge/answer-an-environmental-information-request.md` - `generated/experience/public-edge/answer-ministerial-correspondence.md` - `generated/experience/public-edge/answer-questions-at-a-scrutiny-meeting.md` - `generated/experience/public-edge/appeal-a-planning-decision.md` - `generated/experience/public-edge/apply-an-empty-home-or-second-home-premium.md` - `generated/experience/public-edge/apply-an-exemption-and-issue-a-refusal-notice.md` - `generated/experience/public-edge/apply-for-a-blue-badge.md` - `generated/experience/public-edge/apply-for-a-deprivation-of-liberty-authorisation.md` - `generated/experience/public-edge/apply-for-a-discount-or-council-tax-reduction.md` - `generated/experience/public-edge/apply-for-a-licence.md` - `generated/experience/public-edge/apply-for-a-postal-or-proxy-vote.md` - `generated/experience/public-edge/apply-for-a-school-place-and-appeal-a-refusal.md` - `generated/experience/public-edge/apply-for-a-voter-authority-certificate.md` - `generated/experience/public-edge/apply-for-an-export-licence.md` - `generated/experience/public-edge/apply-for-free-school-meals.md` - `generated/experience/public-edge/apply-for-planning-permission-and-track-it.md` - `generated/experience/public-edge/apply-the-armed-forces-covenant-duty-in-a-council-decision.md` - `generated/experience/public-edge/apply-the-monthly-savings-test.md` - `generated/experience/public-edge/apply-the-national-pay-award.md` - `generated/experience/public-edge/apply-the-one-page-decision-standard-and-report-standard-to.md` - `generated/experience/public-edge/apply-the-pre-election-period.md` - `generated/experience/public-edge/apply-the-retention-schedule-and-dispose-of-records.md` - `generated/experience/public-edge/appoint-an-election-agent.md` - `generated/experience/public-edge/appoint-an-independent-member-to-the-committee.md` - `generated/experience/public-edge/appoint-and-train-polling-staff.md` - `generated/experience/public-edge/appoint-or-dismiss-a-statutory-officer.md` - `generated/experience/public-edge/appoint-the-cabinet-and-allocate-portfolios.md` - `generated/experience/public-edge/appoint-the-external-auditor-through-psaa.md` - `generated/experience/public-edge/appoint-the-internal-auditor-and-receive-the-report.md` - `generated/experience/public-edge/approve-a-business-case-through-the-assurance-framework.md` - `generated/experience/public-edge/approve-a-change-through-the-change-board.md` - `generated/experience/public-edge/approve-a-direct-payment.md` - `generated/experience/public-edge/approve-a-disabled-facilities-grant.md` - `generated/experience/public-edge/approve-a-new-system-against-the-architecture-standard.md` - `generated/experience/public-edge/approve-a-severance-package-over-100-000.md` - `generated/experience/public-edge/approve-a-virement.md` - `generated/experience/public-edge/approve-the-internal-audit-charter.md` - `generated/experience/public-edge/approve-the-internal-audit-plan.md` - `generated/experience/public-edge/approve-the-medium-term-financial-strategy.md` - `generated/experience/public-edge/approve-the-pension-fund-accounts-and-valuation.md` - `generated/experience/public-edge/approve-the-senior-structure-and-statutory-officer-designati.md` - `generated/experience/public-edge/approve-the-treasury-capital-and-minimum-revenue-provision-s.md` - `generated/experience/public-edge/ask-a-question-of-the-leader-or-a-cabinet-member.md` - `generated/experience/public-edge/ask-a-question-or-present-a-petition-at-a-meeting.md` - `generated/experience/public-edge/ask-for-a-care-needs-assessment.md` - `generated/experience/public-edge/ask-for-help-with-homelessness.md` - `generated/experience/public-edge/ask-for-own-personal-data.md` - `generated/experience/public-edge/assess-a-carer-s-needs.md` - `generated/experience/public-edge/assess-a-code-of-conduct-complaint.md` - `generated/experience/public-edge/assess-a-housing-benefit-claim.md` - `generated/experience/public-edge/assess-and-decide-bus-franchising.md` - `generated/experience/public-edge/assess-council-tax-reduction.md` - `generated/experience/public-edge/assess-risk-and-publish-the-community-risk-register.md` - `generated/experience/public-edge/assess-the-committee-s-effectiveness-and-report-to-council.md` - `generated/experience/public-edge/assess-the-needs-of-an-adult.md` - `generated/experience/public-edge/attend-film-or-report-a-public-meeting.md` - `generated/experience/public-edge/attend-the-count-as-an-agent.md` - `generated/experience/public-edge/authenticate-and-seal-a-document.md` - `generated/experience/public-edge/authorise-and-make-a-payment.md` - `generated/experience/public-edge/authorise-covert-surveillance-under-ripa.md` - `generated/experience/public-edge/award-a-discount-or-exemption.md` - `generated/experience/public-edge/award-a-discretionary-housing-payment.md` - `generated/experience/public-edge/award-and-monitor-a-voluntary-sector-grant.md` - `generated/experience/public-edge/award-business-rates-relief.md` - `generated/experience/public-edge/award-under-a-framework-or-dynamic-market.md` - `generated/experience/public-edge/back-up-and-test-a-restore.md` - `generated/experience/public-edge/bid-for-and-report-on-a-transport-grant.md` - `generated/experience/public-edge/bill-a-new-business-rates-property.md` - `generated/experience/public-edge/build-a-resident-service-through-discovery-alpha-and-beta.md` - `generated/experience/public-edge/build-and-sign-the-savings-ledger-line-by-line.md` - `generated/experience/public-edge/calculate-the-council-tax-base-and-requirement.md` - `generated/experience/public-edge/call-in-an-executive-decision.md` - `generated/experience/public-edge/canvass-voters-lawfully.md` - `generated/experience/public-edge/carry-out-a-data-protection-impact-assessment.md` - `generated/experience/public-edge/carry-out-a-financial-assessment-and-charge.md` - `generated/experience/public-edge/carry-out-an-education-health-and-care-needs-assessment.md` - `generated/experience/public-edge/carry-out-an-internal-review.md` - `generated/experience/public-edge/cast-a-recorded-vote-on-the-budget.md` - `generated/experience/public-edge/certify-a-grant-claim.md` - `generated/experience/public-edge/chair-full-council-and-rule-on-order.md` - `generated/experience/public-edge/challenge-a-council-tax-band.md` - `generated/experience/public-edge/challenge-a-provider-rate-and-decide-an-uplift.md` - `generated/experience/public-edge/check-supply-chain-labour-and-modern-slavery-risk.md` - `generated/experience/public-edge/claim-emergency-financial-assistance.md` - `generated/experience/public-edge/claim-members-allowances-and-expenses.md` - `generated/experience/public-edge/claim-staffing-and-business-costs.md` - `generated/experience/public-edge/clerk-a-school-admission-appeal-panel.md` - `generated/experience/public-edge/comment-on-a-planning-application-2.md` - `generated/experience/public-edge/comment-on-a-planning-application.md` - `generated/experience/public-edge/commission-a-placement-for-a-looked-after-child.md` - `generated/experience/public-edge/commission-a-public-health-service-under-the-provider-select.md` - `generated/experience/public-edge/commission-supported-living-from-a-charity-provider.md` - `generated/experience/public-edge/commission-temporary-accommodation.md` - `generated/experience/public-edge/commission-the-external-quality-assessment.md` - `generated/experience/public-edge/complain-to-the-council-and-the-ombudsman.md` - `generated/experience/public-edge/complete-induction-and-training-on-reserves-debt-efs-and-ris.md` - `generated/experience/public-edge/complete-mandatory-member-training.md` - `generated/experience/public-edge/complete-the-annual-governance-and-accountability-return.md` - `generated/experience/public-edge/complete-the-data-security-and-protection-toolkit.md` - `generated/experience/public-edge/comply-with-a-secretary-of-state-direction.md` - `generated/experience/public-edge/conduct-a-marriage-or-civil-partnership.md` - `generated/experience/public-edge/confirm-mrp-and-efs-accounting-with-the-external-auditor.md` - `generated/experience/public-edge/consider-an-objection-from-a-local-elector.md` - `generated/experience/public-edge/consider-social-value-before-procurement.md` - `generated/experience/public-edge/consolidate-published-recommendations-into-owned-operating-s.md` - `generated/experience/public-edge/consult-and-publicise-an-application.md` - `generated/experience/public-edge/consult-on-and-make-redundancies.md` - `generated/experience/public-edge/contract-management.md` - `generated/experience/public-edge/convene-the-right-people-by-mail-and-call.md` - `generated/experience/public-edge/correct-a-register-entry.md` - `generated/experience/public-edge/council-decisions.md` - `generated/experience/public-edge/cyber-posture.md` - `generated/experience/public-edge/deal-with-a-dangerous-structure.md` - `generated/experience/public-edge/decide-a-parking-penalty-challenge.md` - `generated/experience/public-edge/decide-an-application-at-planning-committee.md` - `generated/experience/public-edge/decide-on-sharing-confidential-care-information-as-caldicott.md` - `generated/experience/public-edge/declare-an-impairment-to-independence.md` - `generated/experience/public-edge/declare-an-interest-at-a-meeting-and-withdraw.md` - `generated/experience/public-edge/declare-an-interest-in-proceedings.md` - `generated/experience/public-edge/declare-the-collection-fund-surplus-or-deficit.md` - `generated/experience/public-edge/decommission-a-legacy-system.md` - `generated/experience/public-edge/defend-a-planning-appeal.md` - `generated/experience/public-edge/deliver-the-risk-based-audit-plan-from-substantive-capacity.md` - `generated/experience/public-edge/design-and-run-a-grant-scheme.md` - `generated/experience/public-edge/determine-a-building-regulations-application.md` - `generated/experience/public-edge/determine-a-postal-or-proxy-vote-application.md` - `generated/experience/public-edge/determine-an-application-to-register.md` - `generated/experience/public-edge/discharge-planning-conditions.md` - `generated/experience/public-edge/dispose-of-or-acquire-land-at-best-consideration.md` - `generated/experience/public-edge/employ-staff-in-the-mp-s-office.md` - `generated/experience/public-edge/employment.md` - `generated/experience/public-edge/enforce-joiner-mover-and-leaver-with-segregation-of-duties.md` - `generated/experience/public-edge/enforce-publication-deadlines-with-an-exception-register.md` - `generated/experience/public-edge/enrol-staff-in-the-pension-scheme-and-send-monthly-data.md` - `generated/experience/public-edge/ensure-continuity-of-care-on-provider-failure.md` - `generated/experience/public-edge/evaluate-and-grade-a-job.md` - `generated/experience/public-edge/fill-a-casual-vacancy-by-election-or-co-option.md` - `generated/experience/public-edge/find-a-council-borough-or-government-department-in-the-insti.md` - `generated/experience/public-edge/find-the-right-council-and-service-for-an-address.md` - `generated/experience/public-edge/flow-obligations-into-a-contract-or-commission.md` - `generated/experience/public-edge/give-notice-to-hand-back-a-contract.md` - `generated/experience/public-edge/give-the-opinion-on-the-financial-statements.md` - `generated/experience/public-edge/grant-a-dispensation-for-a-pecuniary-interest.md` - `generated/experience/public-edge/grant-a-premises-licence.md` - `generated/experience/public-edge/grant-a-street-works-permit.md` - `generated/experience/public-edge/handle-a-benefit-reconsideration-or-appeal.md` - `generated/experience/public-edge/handle-a-highway-claim.md` - `generated/experience/public-edge/handle-a-housing-complaint.md` - `generated/experience/public-edge/handle-a-request-for-erasure-or-restriction.md` - `generated/experience/public-edge/hear-a-call-in-and-decide.md` - `generated/experience/public-edge/hear-a-councillor-call-for-action.md` - `generated/experience/public-edge/hold-a-citizenship-ceremony.md` - `generated/experience/public-edge/hold-a-constituency-surgery.md` - `generated/experience/public-edge/hold-a-licensing-sub-committee-hearing.md` - `generated/experience/public-edge/hold-an-initial-child-protection-conference.md` - `generated/experience/public-edge/hold-and-record-the-monthly-golden-triangle-meeting.md` - `generated/experience/public-edge/hold-identity-and-operational-evidence.md` - `generated/experience/public-edge/hold-pre-decision-scrutiny-of-a-key-decision.md` - `generated/experience/public-edge/hold-the-annual-parish-meeting.md` - `generated/experience/public-edge/host-a-social-housing-consumer-standards-inspection.md` - `generated/experience/public-edge/host-an-lga-corporate-peer-challenge.md` - `generated/experience/public-edge/independently-test-an-implemented-control-under-the-self-rev.md` - `generated/experience/public-edge/inspect-a-food-business-and-award-a-hygiene-rating.md` - `generated/experience/public-edge/inspect-a-highway-and-repair-a-defect.md` - `generated/experience/public-edge/inspect-a-private-rented-home-and-serve-a-notice.md` - `generated/experience/public-edge/inspect-a-workplace-for-health-and-safety.md` - `generated/experience/public-edge/inspect-the-council-s-accounts-and-object.md` - `generated/experience/public-edge/inspect-work-and-issue-a-completion-certificate.md` - `generated/experience/public-edge/instruct-enforcement-agents-with-a-vulnerability-check.md` - `generated/experience/public-edge/investigate-a-breach-and-serve-an-enforcement-notice.md` - `generated/experience/public-edge/investigate-a-consumer-complaint-and-take-enforcement-action.md` - `generated/experience/public-edge/investigate-a-flood-as-lead-local-flood-authority.md` - `generated/experience/public-edge/investigate-a-statutory-nuisance-and-serve-an-abatement-noti.md` - `generated/experience/public-edge/investigate-a-suspected-fraud.md` - `generated/experience/public-edge/investigate-tenancy-fraud.md` - `generated/experience/public-edge/issue-a-best-value-notice-and-run-engagement.md` - `generated/experience/public-edge/issue-a-certified-copy-of-an-entry.md` - `generated/experience/public-edge/issue-a-cil-liability-notice-and-demand.md` - `generated/experience/public-edge/issue-a-mayoral-decision.md` - `generated/experience/public-edge/issue-a-refund-or-transfer-a-credit.md` - `generated/experience/public-edge/issue-a-section-114-report.md` - `generated/experience/public-edge/issue-a-section-5-report-on-unlawfulness-or-maladministratio.md` - `generated/experience/public-edge/issue-a-security-aspects-letter-to-a-contractor.md` - `generated/experience/public-edge/issue-a-statutory-recommendation-or-public-interest-report.md` - `generated/experience/public-edge/issue-a-voter-authority-certificate.md` - `generated/experience/public-edge/issue-annual-council-tax-bills.md` - `generated/experience/public-edge/issue-the-annual-internal-audit-opinion.md` - `generated/experience/public-edge/issue-the-audit-plan.md` - `generated/experience/public-edge/issue-the-decision-notice.md` - `generated/experience/public-edge/issue-the-role-statement-train-and-attribute-challenge.md` - `generated/experience/public-edge/issue-the-summons-and-agenda.md` - `generated/experience/public-edge/issue-the-weekly-one-page-exception-report.md` - `generated/experience/public-edge/join-a-council-s-approved-provider-list.md` - `generated/experience/public-edge/join-a-strategic-coordinating-group.md` - `generated/experience/public-edge/join-the-housing-register-and-assess-priority.md` - `generated/experience/public-edge/keep-business-continuity-plans-and-advise-businesses.md` - `generated/experience/public-edge/keep-spending-within-the-limit-and-file-the-return.md` - `generated/experience/public-edge/keep-statutory-notification-readiness-and-named-incident-rol.md` - `generated/experience/public-edge/keep-the-device-and-software-asset-register.md` - `generated/experience/public-edge/keep-the-outside-bodies-appointments-register.md` - `generated/experience/public-edge/keep-the-record-of-processing-activities.md` - `generated/experience/public-edge/lay-a-statutory-instrument.md` - `generated/experience/public-edge/lay-the-annual-report-and-accounts.md` - `generated/experience/public-edge/lead-recovery.md` - `generated/experience/public-edge/lead-the-response-to-a-section-114-notice.md` - `generated/experience/public-edge/let-a-social-home.md` - `generated/experience/public-edge/licence-a-taxi-or-private-hire-driver.md` - `generated/experience/public-edge/licence-an-animal-activity.md` - `generated/experience/public-edge/license-a-house-in-multiple-occupation.md` - `generated/experience/public-edge/logistics-tracking.md` - `generated/experience/public-edge/maintain-and-publish-the-register-of-members-interests.md` - `generated/experience/public-edge/maintain-the-brownfield-and-self-build-registers.md` - `generated/experience/public-edge/maintain-the-constitution-and-scheme-of-delegation-with-live.md` - `generated/experience/public-edge/maintain-the-contract-procedure-rules.md` - `generated/experience/public-edge/maintain-the-publication-scheme-and-disclosure-log.md` - `generated/experience/public-edge/maintain-the-single-procurement-and-commissioning-pipeline.md` - `generated/experience/public-edge/make-a-freedom-of-information-request-or-ask-for-review.md` - `generated/experience/public-edge/make-a-public-spaces-protection-order.md` - `generated/experience/public-edge/make-a-safeguarding-adults-enquiry.md` - `generated/experience/public-edge/make-a-traffic-regulation-order.md` - `generated/experience/public-edge/make-a-tree-preservation-order.md` - `generated/experience/public-edge/manage-a-contract-and-hold-review-meetings.md` - `generated/experience/public-edge/manage-a-persistent-or-unreasonable-complainant.md` - `generated/experience/public-edge/manage-a-public-question-or-petition-at-a-meeting.md` - `generated/experience/public-edge/manage-allotments.md` - `generated/experience/public-edge/manage-cash-flow-and-borrowing.md` - `generated/experience/public-edge/manage-sickness-absence.md` - `generated/experience/public-edge/mark-and-check-politically-restricted-posts.md` - `generated/experience/public-edge/mark-exempt-information-and-run-a-private-item.md` - `generated/experience/public-edge/mark-store-and-send-classified-information.md` - `generated/experience/public-edge/meet-the-auditors-without-officers-present.md` - `generated/experience/public-edge/modify-a-contract-and-publish-the-change-notice.md` - `generated/experience/public-edge/monitor-a-grant-and-recover-misused-funds.md` - `generated/experience/public-edge/monitor-a-provider-s-quality.md` - `generated/experience/public-edge/name-and-number-a-street.md` - `generated/experience/public-edge/nominate-an-asset-of-community-value-2.md` - `generated/experience/public-edge/nominate-an-asset-of-community-value.md` - `generated/experience/public-edge/notify-the-regulator-of-an-incident.md` - `generated/experience/public-edge/obtain-facility-security-clearance-for-a-supplier-site.md` - `generated/experience/public-edge/open-a-constituent-case-with-consent.md` - `generated/experience/public-edge/open-a-council-tax-account-on-a-move-in.md` - `generated/experience/public-edge/open-a-rest-centre-and-register-evacuees.md` - `generated/experience/public-edge/open-and-conclude-an-inquest.md` - `generated/experience/public-edge/open-postal-votes-and-verify-personal-identifiers.md` - `generated/experience/public-edge/open-the-government-station-and-choose-a-journey.md` - `generated/experience/public-edge/organise-evidence-and-records-for-a-public-body.md` - `generated/experience/public-edge/pass-a-proposal-through-the-business-case-gateway.md` - `generated/experience/public-edge/pay-council-tax-and-set-up-a-direct-debit.md` - `generated/experience/public-edge/pay-or-challenge-a-parking-penalty.md` - `generated/experience/public-edge/pay-the-data-protection-fee.md` - `generated/experience/public-edge/place-a-household-in-interim-accommodation.md` - `generated/experience/public-edge/pool-a-budget-with-the-nhs.md` - `generated/experience/public-edge/prepare-a-care-proceedings-application.md` - `generated/experience/public-edge/prepare-a-green-book-business-case.md` - `generated/experience/public-edge/prepare-a-neighbourhood-plan.md` - `generated/experience/public-edge/prepare-a-procurement-and-agreement-requirements-brief-for-a.md` - `generated/experience/public-edge/prepare-an-impact-assessment.md` - `generated/experience/public-edge/prepare-an-off-site-plan-for-a-major-hazard-site.md` - `generated/experience/public-edge/prepare-and-examine-the-local-plan.md` - `generated/experience/public-edge/prepare-for-a-cqc-assessment-of-adult-social-care.md` - `generated/experience/public-edge/prepare-for-a-public-accounts-committee-hearing.md` - `generated/experience/public-edge/prepare-for-an-inspection-of-children-s-services.md` - `generated/experience/public-edge/prepare-main-and-supplementary-estimates.md` - `generated/experience/public-edge/prepare-the-local-transport-plan.md` - `generated/experience/public-edge/prepare-the-market-sustainability-plan.md` - `generated/experience/public-edge/price-a-single-source-defence-contract.md` - `generated/experience/public-edge/process-a-right-to-buy-application.md` - `generated/experience/public-edge/process-a-temporary-event-notice.md` - `generated/experience/public-edge/procure-a-separately-procurable-engagement-with-an-aggregati.md` - `generated/experience/public-edge/procurement-intelligence.md` - `generated/experience/public-edge/propose-the-budget-and-council-tax-to-full-council.md` - `generated/experience/public-edge/publish-an-accessibility-statement.md` - `generated/experience/public-edge/publish-an-algorithmic-transparency-record.md` - `generated/experience/public-edge/publish-an-executive-decision-record.md` - `generated/experience/public-edge/publish-and-maintain-privacy-notices.md` - `generated/experience/public-edge/publish-government-procurement-card-spend.md` - `generated/experience/public-edge/publish-members-allowances-for-the-year.md` - `generated/experience/public-edge/publish-notice-of-a-key-decision-28-clear-days-ahead.md` - `generated/experience/public-edge/publish-open-data.md` - `generated/experience/public-edge/publish-spending-over-500.md` - `generated/experience/public-edge/publish-the-agenda-and-reports-five-clear-days-ahead.md` - `generated/experience/public-edge/publish-the-annual-pay-policy-statement.md` - `generated/experience/public-edge/publish-the-contracts-register.md` - `generated/experience/public-edge/publish-the-draft-and-audited-statement-of-accounts.md` - `generated/experience/public-edge/publish-the-forward-plan-twelve-weeks-ahead.md` - `generated/experience/public-edge/publish-the-gambling-statement-of-principles.md` - `generated/experience/public-edge/publish-the-gender-pay-gap.md` - `generated/experience/public-edge/publish-the-market-position-statement.md` - `generated/experience/public-edge/publish-the-notice-of-election.md` - `generated/experience/public-edge/publish-the-regulatory-enforcement-policy.md` - `generated/experience/public-edge/publish-the-revised-register.md` - `generated/experience/public-edge/publish-the-updated-constitution.md` - `generated/experience/public-edge/publish-transparency-data.md` - `generated/experience/public-edge/put-a-digital-imprint-on-campaign-material.md` - `generated/experience/public-edge/raise-a-purchase-order-before-package-spend.md` - `generated/experience/public-edge/raise-a-safeguarding-concern.md` - `generated/experience/public-edge/raise-a-service-complaint.md` - `generated/experience/public-edge/read-a-committee-pack-within-the-page-limit-and-lodge-questi.md` - `generated/experience/public-edge/rebase-the-adult-social-care-budget-from-the-care-ledger.md` - `generated/experience/public-edge/receive-and-decide-on-nominations.md` - `generated/experience/public-edge/receive-and-handle-a-whistleblowing-disclosure.md` - `generated/experience/public-edge/receive-and-publish-candidates-spending-returns.md` - `generated/experience/public-edge/receive-quarterly-monitoring-with-a-demand-bridge.md` - `generated/experience/public-edge/receive-the-auditor-s-annual-report-and-value-for-money-comm.md` - `generated/experience/public-edge/receive-the-counter-fraud-annual-report.md` - `generated/experience/public-edge/receive-the-monthly-evidence-pack-rated-on-evidence.md` - `generated/experience/public-edge/reconcile-access-and-contain-unexplained-accounts.md` - `generated/experience/public-edge/reconcile-the-balance-sheet-exposure-statement-quarterly.md` - `generated/experience/public-edge/record-a-gift-or-hospitality.md` - `generated/experience/public-edge/record-a-price-exception-with-the-finance-position.md` - `generated/experience/public-edge/record-an-exemption-from-contract-procedure-rules.md` - `generated/experience/public-edge/recover-a-benefit-overpayment.md` - `generated/experience/public-edge/recover-unpaid-council-tax-through-a-liability-order.md` - `generated/experience/public-edge/recruit-safely-with-dbs-checks.md` - `generated/experience/public-edge/refer-a-band-challenge-to-the-valuation-office-agency.md` - `generated/experience/public-edge/refer-higher-risk-building-work-to-the-building-safety-regul.md` - `generated/experience/public-edge/register-a-birth-death-or-marriage-notice.md` - `generated/experience/public-edge/register-a-birth.md` - `generated/experience/public-edge/register-a-death-and-offer-tell-us-once.md` - `generated/experience/public-edge/register-a-financial-interest.md` - `generated/experience/public-edge/register-a-foreign-activity-arrangement.md` - `generated/experience/public-edge/register-a-political-party.md` - `generated/experience/public-edge/register-as-a-consultant-lobbyist-and-file-returns.md` - `generated/experience/public-edge/register-as-a-non-party-campaigner.md` - `generated/experience/public-edge/register-disclosable-pecuniary-interests.md` - `generated/experience/public-edge/register-to-vote.md` - `generated/experience/public-edge/reimburse-operators-for-concessionary-travel.md` - `generated/experience/public-edge/renew-public-services-network-compliance.md` - `generated/experience/public-edge/reperform-a-closed-control-from-source.md` - `generated/experience/public-edge/report-a-defence-security-incident.md` - `generated/experience/public-edge/report-a-street-problem.md` - `generated/experience/public-edge/report-complaints-and-compliments-annually.md` - `generated/experience/public-edge/report-contract-performance-indicators.md` - `generated/experience/public-edge/report-digital-service-performance.md` - `generated/experience/public-edge/report-donations-and-loans.md` - `generated/experience/public-edge/report-limited-or-no-assurance-to-the-audit-committee.md` - `generated/experience/public-edge/report-on-a-council-owned-company-to-the-shareholder-committ.md` - `generated/experience/public-edge/report-on-staffing-and-organisation-as-head-of-paid-service.md` - `generated/experience/public-edge/report-on-the-robustness-of-estimates-and-adequacy-of-reserv.md` - `generated/experience/public-edge/report-on-value-for-money-arrangements.md` - `generated/experience/public-edge/represent-the-council-on-the-combined-authority-or-a-regiona.md` - `generated/experience/public-edge/request-a-rate-uplift.md` - `generated/experience/public-edge/request-a-review-of-a-housing-decision.md` - `generated/experience/public-edge/request-and-track-audit-evidence.md` - `generated/experience/public-edge/request-exceptional-financial-support.md` - `generated/experience/public-edge/require-a-cabinet-member-or-officer-to-attend.md` - `generated/experience/public-edge/resolve-a-service-desk-incident.md` - `generated/experience/public-edge/resolve-eligibility-for-the-general-power-of-competence.md` - `generated/experience/public-edge/respond-to-a-child-protection-referral.md` - `generated/experience/public-edge/respond-to-a-damp-mould-or-repair-report.md` - `generated/experience/public-edge/respond-to-a-judicial-review-pre-action-letter.md` - `generated/experience/public-edge/respond-to-a-parliamentary-petition.md` - `generated/experience/public-edge/respond-to-a-statutory-recommendation-from-the-auditor.md` - `generated/experience/public-edge/respond-to-an-area-send-inspection.md` - `generated/experience/public-edge/respond-to-an-ico-decision-notice.md` - `generated/experience/public-edge/respond-to-an-ombudsman-enquiry-and-carry-out-the-remedy.md` - `generated/experience/public-edge/respond-to-an-ombudsman-finding.md` - `generated/experience/public-edge/return-post-election-data-to-the-electoral-commission.md` - `generated/experience/public-edge/review-a-homelessness-decision.md` - `generated/experience/public-edge/review-a-looked-after-child-s-care-plan.md` - `generated/experience/public-edge/review-local-air-quality.md` - `generated/experience/public-edge/review-polling-districts-and-places.md` - `generated/experience/public-edge/review-the-annual-governance-statement.md` - `generated/experience/public-edge/review-the-corporate-risk-register-quarterly.md` - `generated/experience/public-edge/review-the-draft-statement-of-accounts.md` - `generated/experience/public-edge/review-the-information-risk-register-as-senior-information-r.md` - `generated/experience/public-edge/review-the-risk-management-framework.md` - `generated/experience/public-edge/run-a-combined-authority-board-meeting.md` - `generated/experience/public-edge/run-a-deep-dive-with-recorded-influence-on-direction.md` - `generated/experience/public-edge/run-a-demand-control-cell-weekly.md` - `generated/experience/public-edge/run-a-disaster-recovery-test.md` - `generated/experience/public-edge/run-a-disciplinary-process.md` - `generated/experience/public-edge/run-a-neighbourhood-planning-referendum.md` - `generated/experience/public-edge/run-a-public-consultation-and-publish-the-response.md` - `generated/experience/public-edge/run-a-public-consultation-and-report-the-results.md` - `generated/experience/public-edge/run-a-recall-petition-as-petition-officer.md` - `generated/experience/public-edge/run-a-standards-hearing-and-publish-the-decision.md` - `generated/experience/public-edge/run-a-task-and-finish-review-and-secure-the-executive-respon.md` - `generated/experience/public-edge/run-an-assurance-review-on-a-major-project.md` - `generated/experience/public-edge/run-an-audit-engagement-from-terms-of-reference-to-report.md` - `generated/experience/public-edge/run-monthly-budget-monitoring-with-budget-holders.md` - `generated/experience/public-edge/run-monthly-payroll-and-report-to-hmrc.md` - `generated/experience/public-edge/run-national-fraud-initiative-matches.md` - `generated/experience/public-edge/run-security-awareness-and-phishing-tests.md` - `generated/experience/public-edge/run-the-annual-canvass.md` - `generated/experience/public-edge/run-the-annual-council-meeting.md` - `generated/experience/public-edge/run-the-cross-committee-forum-with-other-chairs.md` - `generated/experience/public-edge/run-the-decision-clock-on-stalled-reforms.md` - `generated/experience/public-edge/run-the-make-or-buy-test-before-a-material-re-let.md` - `generated/experience/public-edge/run-the-out-of-hours-duty-officer-rota.md` - `generated/experience/public-edge/run-the-period-for-exercise-of-public-rights.md` - `generated/experience/public-edge/run-the-public-inspection-period-for-the-accounts.md` - `generated/experience/public-edge/run-the-reorganisation-programme-to-vesting-day.md` - `generated/experience/public-edge/run-the-verification-and-count-and-declare-the-result.md` - `generated/experience/public-edge/score-the-thirty-cyber-criteria-with-evidence-references.md` - `generated/experience/public-edge/scrutinise-a-substantial-nhs-service-change.md` - `generated/experience/public-edge/scrutinise-the-budget-proposals-before-council.md` - `generated/experience/public-edge/scrutinise-the-combined-authority.md` - `generated/experience/public-edge/scrutinise-the-community-safety-partnership.md` - `generated/experience/public-edge/scrutinise-treasury-management.md` - `generated/experience/public-edge/security-incident-response.md` - `generated/experience/public-edge/seek-a-cabinet-office-spend-control-approval.md` - `generated/experience/public-edge/seek-hm-treasury-approval-above-delegated-limits.md` - `generated/experience/public-edge/select-a-candidate-and-authorise-the-party-description.md` - `generated/experience/public-edge/send-apologies-and-keep-attendance-within-the-six-month-rule.md` - `generated/experience/public-edge/separate-cashable-saving-cost-avoidance-and-deferred-demand.md` - `generated/experience/public-edge/set-and-collect-council-housing-rents.md` - `generated/experience/public-edge/set-the-annual-calendar-of-meetings.md` - `generated/experience/public-edge/set-the-annual-work-programme.md` - `generated/experience/public-edge/set-the-local-government-finance-settlement.md` - `generated/experience/public-edge/set-the-mayoral-budget-and-precept.md` - `generated/experience/public-edge/set-the-precept-and-notify-the-billing-authority.md` - `generated/experience/public-edge/set-the-scheme-of-executive-delegation.md` - `generated/experience/public-edge/set-up-an-instalment-plan-or-direct-debit.md` - `generated/experience/public-edge/share-information-with-other-responders.md` - `generated/experience/public-edge/sign-the-annual-governance-statement.md` - `generated/experience/public-edge/sign-the-declaration-of-acceptance-of-office.md` - `generated/experience/public-edge/speak-in-a-debate-and-check-the-official-report.md` - `generated/experience/public-edge/spend-under-the-section-137-power-within-the-limit.md` - `generated/experience/public-edge/submit-a-motion-to-full-council.md` - `generated/experience/public-edge/submit-an-innovation-proposal-to-the-defence-and-security-ac.md` - `generated/experience/public-edge/submit-planning-statistics-returns.md` - `generated/experience/public-edge/submit-revenue-and-capital-returns-to-mhclg.md` - `generated/experience/public-edge/submit-single-source-contract-reports.md` - `generated/experience/public-edge/submit-the-business-rates-forecast-and-outturn-returns.md` - `generated/experience/public-edge/submit-the-homelessness-statistics-return.md` - `generated/experience/public-edge/submit-the-party-s-annual-accounts.md` - `generated/experience/public-edge/submit-the-provider-cost-return.md` - `generated/experience/public-edge/submit-the-tenant-satisfaction-measures.md` - `generated/experience/public-edge/supplier-vetting.md` - `generated/experience/public-edge/supply-the-register-to-an-entitled-person.md` - `generated/experience/public-edge/support-a-care-leaver-with-a-pathway-plan.md` - `generated/experience/public-edge/table-a-motion-or-seek-a-debate.md` - `generated/experience/public-edge/table-a-written-question.md` - `generated/experience/public-edge/take-a-bill-through-parliament.md` - `generated/experience/public-edge/take-a-financial-monitoring-item-with-decision-grade-evidenc.md` - `generated/experience/public-edge/take-a-homelessness-application-and-decide-the-duty.md` - `generated/experience/public-edge/take-a-key-decision-on-the-one-page-standard.md` - `generated/experience/public-edge/take-a-ward-casework-enquiry-through-to-reply.md` - `generated/experience/public-edge/take-an-individual-portfolio-decision.md` - `generated/experience/public-edge/take-an-urgent-decision-under-special-urgency.md` - `generated/experience/public-edge/take-and-publish-the-minutes-2.md` - `generated/experience/public-edge/take-and-publish-the-minutes.md` - `generated/experience/public-edge/take-evidence-and-agree-a-select-committee-report.md` - `generated/experience/public-edge/take-notice-of-marriage-or-civil-partnership.md` - `generated/experience/public-edge/take-the-six-month-progress-report-on-statutory-recommendati.md` - `generated/experience/public-edge/tender-and-award.md` - `generated/experience/public-edge/track-audit-recommendations-to-owner-and-date-and-escalate-o.md` - `generated/experience/public-edge/track-the-executive-s-responses-to-recommendations.md` - `generated/experience/public-edge/transfer-staff-under-tupe.md` - `generated/experience/public-edge/transfer-the-commissioning-toolkit-into-a-service.md` - `generated/experience/public-edge/validate-a-planning-application.md` - `generated/experience/public-edge/verify-weighing-and-measuring-equipment.md` - `generated/experience/public-edge/vet-and-train-care-staff.md` - `generated/experience/public-edge/view-the-public-outcome-scorecard.md` - `generated/experience/public-edge/vote-in-a-division.md` - `generated/experience/public-edge/warn-and-inform-the-public.md` - `generated/experience/public-edge/webcast-and-allow-reporting-of-a-public-meeting.md` - `generated/experience/public-edge/write-a-submission-to-a-minister.md` - `generated/experience/public-edge/write-and-exercise-the-emergency-plan.md` - `generated/experience/public-edge/write-off-an-irrecoverable-debt.md` - `generated/experience/public-edge/write-the-officer-report-and-decide-under-delegation.md` - `generated/experience/public-edge/write-to-a-public-body-on-a-constituent-s-behalf.md` ## 12. Files included, with sha256 | File | sha256 | |---|---| | `AGENTS.md` | `05245f9617828ced989dd83e7bbcae49e66b9b6d0f4db355a32735394de28bd2` | | `CHECKPOINT.md` | `46b8433ff95829ff56b3af0dfe48885148363af0d52867d1cce0794b6ddeed1f` | | `README.md` | `05245f9617828ced989dd83e7bbcae49e66b9b6d0f4db355a32735394de28bd2` | | `SCOPE.md` | `1e19e8a284a0c0567002cd324a31a91ce12855212d4548a71193866dc46fd26e` | | `TOOLCHAIN.json` | `8db4a9ed777cd135106f6656f6b9ff6db34faa2eab32d0f430ce93957416a195` | | `contract/estate-contract.json` | `62484e1e78dc7c2fed76f7ed84966bfeb216ec8ba16ce2150c99ea4744ad7b10` | | `deploy/AGENT-DEPLOY.md` | `529c37af2112ccc7396dcf71d33deafa977d291f06874897ac79c421c1678a13` | | `deploy/Caddyfile` | `fade1025d590ee1ddf32cbbd98063f55d230b44f5a739ca17b068dd4e46ac9b2` | | `deploy/README.md` | `eb3003c26ef2ea9b599019a5587a3fe812579f85cf74ff1faccc01fc6e55d862` | | `deploy/web.env.example` | `6930505f29635d00d22fb45433b89c45c148c488e25769ab1827e19e52d8cf31` | | `deploy/x-x-web.service` | `6b502fa1c8c100805a90efc7b4e12c826dcd00cba3a15789ce93d644bcf766f5` | | `design/brand/approved/BRAND-REGISTER.json` | `f18dc480af93ea94a88641a176a9c0aedfd7ab969c6090db297ed3754aef4ff5` | | `design/components.json` | `ea12e1889248db1f30cbc1c0e06c774db8215cf07c7283f5b883100b4fec60a3` | | `docs/DIRECTORY.md` | `5652bd6154f2317db99c65e9df75678595337439862d82bfffeeb7034097f0bb` | | `docs/MCP.md` | `31d0b9eae644ddbc6aaa1e7a81e352613e29f31db1e933166e5fa879d87e3311` | | `docs/PLUGS.md` | `b0cdc203b77a5a6f03529ec0ebc246858f41a212d3a6e89cb36b0675e71fbe4d` | | `docs/README.md` | `080ebe08a86280093a824e6c651b62bf52581eee20d3e51011cf342333d9ec2d` | | `docs/REPORT-OUTCOMES-AND-NOVEL-MEASURES.md` | `11d26b49d3dc2dbddf33b04486a9173a463cb1dd9e9c5a559c6f0866bce759ce` | | `docs/STANDARD.md` | `cce45c122c9af2a7c6f1ed5705731eca3f15730f779636905611a2d7d8f4a4aa` | | `generated/COUNTS.json` | `69dc5173c9932c61db8942ce06da90b3c63939a438134791eb87cc578e1e7485` | | `generated/EXPERIENCE.md` | `61c2db7216432e19b14a0db6a6fb153a59f82bdf441d9ff6ea79306c96bfca80` | | `ledger/SCOPE-CHANGES.md` | `1d58e4ad331483d58fd9e851891a7d582fda9a6a0228132747ceb5090c59998b` | | `ledger/closure/closure.json` | `b7e4858f4891668cf68956284f6e4fe4f313dcbd1194c59c755eeda00dd23c6c` | | `ledger/closure/plan/LANES.md` | `364639d0b981baab5b2e81950e795fe8a1e62a95ce1d6ded012835d0e65a287f` | | `ledger/closure/plan/PLAN.md` | `09ffd61596771e244d46d2e57af65165eced333010e0bdc27eeecf135a4ccd53` | | `ledger/outcomes/README.md` | `270dc8f39f299d4c045de361495ff96d1067deb45e1f4a4ea97822d6491d5364` | | `research/AGENT-RESEARCH.md` | `723389b5a6f22b3dc428abb167925da4ca8a5cfd7e8ef76df76e8aa177e9feb7` | | `research/README.md` | `43ad0f21dc412a75e551259e696a18df689515c29c1f76e0fe09d4ca362b9efe` | | `research/REUSE-ENGINES.md` | `b141620e6b99d5fd7893e27c83061494b55384c395a09bb0ba6eea811da9c436` | | `research/clinic/COMPETITORS-REGULATORS-REUSE.md` | `6d2e769a4d1ba42fa5e604d6607ab6a17b8cbcb5ba9c77cd28b87834f6ec1340` | | `research/clinic/ENGINE-MEDPLUM.md` | `5ca30ff729e16b58a32d711d9de691340451e4934811c84962e90e42c88326a1` | | `research/clinic/MOBBIN-FLOWS.md` | `84731ec9e69069c058137d22e1cf546c361eb5180f29e4423ba54290954709f9` | | `research/clinic/REFERENCE-CLINIC.md` | `15a2191dce36d765e97476f1cce07f41db7aceaae72eeb876d200a68be3fc33f` | | `research/clinic/REFERO-FLOWS.md` | `5e49c9f08d65dba650a6b054e0e69181acc87774897c42ce1f19a7a24b628bc0` | | `research/clinic/REVERSIBILITY.md` | `6fcfbd1b6f0f60ccd8e1d90fc69e3f486b9fbc934bc6659ec22c6947f4e03614` | | `research/clinic/SOCIAL-PRESCRIBING-AND-DETECTION.md` | `6b41991afe790641488bd85df42d60896cf6e44fa5b82f31ed2db57d89716e87` | | `research/design/BASE44-V7BRO-STUDY.md` | `adf5b2032ff5585291c7b967d43e1a473be401f40a9c6dc23fe78ec09ac12e05` | | `research/edge-consortium/GOVERNED-INTELLIGENCE-DEGREE.md` | `c93c23ededa1d05e46aa6285fed9d0d4528720b52002617290e2d6ff2b1d8380` | | `research/edge-consortium/REFERENCE-STUDIO.md` | `e6516dd59e8132d83c06617c1dadd18813ec47e95b6da2a7a6f6b7e83db75ea9` | | `research/edge-consortium/SOURCES-2026-10-31.md` | `a71ab009f0ee5b5a4655dfae038a1c4f3f2852bf3e83994182e3383f7c1ae98e` | | `research/experience/REFERO-FLOWS-1.md` | `b9e6b42ca8672cb349e6d9ef87ddc1f8cbfb022c777db019088fb6a8b8b0d08d` | | `research/experience/REFERO-FLOWS-2.md` | `16141d6c58761db48570ce70a34a4ffa345a87c6093b7bc52dc74109be8dfec2` | | `research/finance/ASSURANCE-AND-REUSE.md` | `480f3311b802c97e7f38919b865ae9013a9fd2c2b10b107525703b820ccead01` | | `research/finance/COMPETITORS.md` | `3bdd5116b6688438b592c232a8d68a61649078a7cefec5fa7f2c6f8462a55abf` | | `research/finance/CONFIDENCE-REPORTING.md` | `5d7e509cdc7f1a8ac89079f5b00e4907002bfcf694e8c01cd0a047f610ab299c` | | `research/finance/ENGINE-BY-STARLING.md` | `f94f7b80899a5b410535ee4e5b321681c37ae2fa1e9b4be5e99243246c8da4ad` | | `research/finance/FCA-AGENTIC-FRAMEWORK.md` | `54534225ef1a88de175d15611a33f8634f5cd5a34bfe35b77ee43fee3dd6adb0` | | `research/finance/MOBBIN-FLOWS.md` | `8a85ca7fdf2837d47fe9e8da5ae056777c8ea4bc2407b695a9210eaf9f7b8859` | | `research/finance/PRESENCE-AND-SOLUTION-BANKS.md` | `51a8d324d24dd923eb252975e690f4d15a6c33f13e73b58066263cec5cc33de3` | | `research/finance/REFERENCE-CARDS.md` | `b0ef8bb6bd46822bc9a9de54b9c73fe7c2291c00ca414cb678095914f8873b2e` | | `research/finance/REFERO-FLOWS-ALL.md` | `ef1e18395b33eefc90d60ac18ad594f4480f035a36e8233cbef2a86dafc8cb56` | | `research/finance/REGULATORS.md` | `3542bfe383df3e329a5d66a14453b2949d97c3eb9f52e3ea103c41f42f849607` | | `research/finance/SCREENS-AND-FLOWS.md` | `6ea3ab7cd337d6370022ac67219408b08cd2d9345db1d181626fcf92912c4644` | | `research/judiverse/COMPETITORS-REGULATORS-REUSE.md` | `a92bd61264900c5c680b59a8eb3f7132ec6acf24a1aa94e533b554e60f84ff15` | | `research/judiverse/JUDEBOX-PARITY.md` | `318fe969873e6d4e12b8c27691a17703ccfb11677a701a86deab2edd45b9178a` | | `research/judiverse/MOBBIN-FLOWS.md` | `0ee68baabdb8c918f37cca4aa3695cf175b1439238f4c0f6479ed7e97cb81994` | | `research/judiverse/REFERENCE-DOCS.md` | `6d874e29394545c6205f5577c9ee75e563755fceee69b850db738ab672a018d3` | | `research/judiverse/REFERENCE-JUDE.md` | `730b78aa1d9adedaf2386acadb80cc1beb4abb69a6ed90f439ef1ef82928a50a` | | `research/judiverse/REFERENCE-JUDEBOX.md` | `dd4ac3cef39706766324120723fcfeecb3644370dcc93bf5ca463c742e31b094` | | `research/judiverse/REFERENCE-SUITE-HOME.md` | `3ae10ec03eb82932fcfbb2817066a3e2178e5ac212474bf6afd131ac49f6921b` | | `research/judiverse/REFERO-FLOWS.md` | `dc6bc7badb62e18e0f292490b8df3a3b4d92bed4d17fea5683f9b47a05a0413b` | | `research/legacy/BLUEPRINTS-FEATURES-JOURNEYS.md` | `f56ba0c3c49a2f7985d4bc5f9eab2dc4c15725471abd848f4359db7ce8f93858` | | `research/legacy/COMPETITORS-BENCHMARKS-READINESS.md` | `971658fccf85aa2e7322ef3b88cfb1d233deed9c7c88049ce0779be058d56be8` | | `research/legacy/STANDARDS-PAPERS-GUIDES.md` | `9e98dcbe938cf7f74a2ff5e49d84b1b1114094e0cb246c5e70992f2aa349576c` | | `research/legacy/STRATEGY-FUNDING-OTHER.md` | `f7a5315e607a8a432c1e23595db6c8ab37470aa94ae03a9ea1d3778d426c4f7a` | | `research/market/COSTLIEST-PROBLEMS.md` | `d5c66134860d167e28f5f8148df5dcfec770a94ebea5d834f9e7e4c9c74cae83` | | `research/market/FUNDING-AND-PROCUREMENT.md` | `6d600e418a50bae9bdf5440ee50f3ab677229251d7b21dc5d367b4f08a4f9753` | | `research/market/HEADLINES-2026-10.md` | `bff44c43df89bb8a01667554bd85f588195dc0897e8784e5aadb68af9a1f9366` | | `research/parity/clinic/bahmni-coalition.jsonl` | `3dcd9824059b777d958ba7b4ef67ed6680618dd0a63798201ae045a0cd3ec404` | | `research/parity/clinic/enlivio-health.jsonl` | `4471ef6a2ecea20f627306572211ff30f62de2dfb5f31b145f989b2c5a3a0838` | | `research/parity/clinic/epic-systems.jsonl` | `6718fb9e044766b40fa51d4d64bc173f3b171d2dab5724b4250ddee1d7aa4516` | | `research/parity/clinic/nhs-app.jsonl` | `3d53562aed8b232f7ec55e25812cc799c4337b28a105d9f0632eb091ab7f3b73` | | `research/parity/clinic/nhs-england.jsonl` | `183ef237d3d55e56d5d5bf3cec07f03ac71de950763b81d39efa7ee2f5e33a07` | | `research/parity/clinic/openemr-project.jsonl` | `e3be89c00f47c6820419c178972d32f91df88c41b179f3f84ab501fcbabd29e7` | | `research/parity/clinic/openmrs.jsonl` | `57b1c51e3fa1b36764ce331ee3d49c860700dd2aadfd073926bd292c7f0193dd` | | `research/parity/clinic/oracle.jsonl` | `03b0615abf28375e9da9084b9ebdb6f975f1f0e517d4b5c787c2b83bdcd44ea0` | | `research/parity/clinic/patient-access.jsonl` | `4c7d0732cf6493a0893e15d0a401be5ab368243dc3601820f5cde8ed6f872df7` | | `research/parity/clinic/zocdoc.jsonl` | `d5d0aaba0d4daceb5e0943398f122e604b1b58d35c140c5593561c2a005a5040` | | `research/parity/finance/adyen.jsonl` | `9d91e850a150cd7d11863f6fa78da56785845ef3029567de84f6f46a0a0456d1` | | `research/parity/finance/barclays.jsonl` | `579a75d357e3bfa770364149fb5af0e80c969910863f574754a2a19319aa5051` | | `research/parity/finance/british-business-bank.jsonl` | `7602eed826f801c273baab45d9794c2107a84bef69e4720d5c54baf4aae366ad` | | `research/parity/finance/gocardless.jsonl` | `8afbf3e6be889b1aac586b73ce59f7346c7d02534ce4c41d6d172f21c939f1ad` | | `research/parity/finance/monzo.jsonl` | `dfd5b3395f10d5970582a906ba017c055612dac06916df870e31dc10d4d79f7c` | | `research/parity/finance/open-banking-uk.jsonl` | `7655dc141cfe13124a2fb1877096d07cc26181532bee6d2f10e99716cca04938` | | `research/parity/finance/oracle.jsonl` | `f733370346c3decd90f8693910b81468115fec73c3aab3b55ff9d0560d5272bd` | | `research/parity/finance/revolut.jsonl` | `38ab291d8b04d965f8fe457a8ca38edd5354efa9844e8c27f57d1df27e40d9f6` | | `research/parity/finance/starling.jsonl` | `db6db08e6e1ca3b3bd72cc04f7819f413f29b465df6a9fcffa4ccf177026074d` | | `research/parity/finance/stripe.jsonl` | `978d72ec2b155715337669987d4bf08b5210841048818f39cd37aabfa6389fa2` | | `research/parity/finance/tide.jsonl` | `7404d84f248f84dd9945e6a1a0209347e61f0c82943c5841c2789170626bf695` | | `research/parity/finance/wise.jsonl` | `3815936ab512f59234cdd632e32c7100fb0dd3d20ae773ee2ac09f1622cad357` | | `research/parity/finance/xero.jsonl` | `8a25920b35ec036f9d3eda8804f4f27d813e26a6118cd3dd0a4f86fc03c403eb` | | `research/parity/foundation/auth0.jsonl` | `c144c2d23255b8c88bb0aa04fb4d55a6ef4221160645fbad471c91932514b812` | | `research/parity/foundation/cyberark.jsonl` | `1ff1674b5b7b03fbfc44c46b397658cb68ec4b60b87b2a018e83abf9e4237b8c` | | `research/parity/foundation/delinea.jsonl` | `2c698a362a4c09cf55608f12a685961df1dc70e5979324d98107aa0278b9986b` | | `research/parity/foundation/duo.jsonl` | `6970dad92917b41f2dc739f7ecf405686233eae887bead501ab399035154b0ec` | | `research/parity/foundation/government-digital-service.jsonl` | `ff2462356f1b1bd6973a386cb0f0cb5d00279a365b21c5563765793cb50f144e` | | `research/parity/foundation/incogni-01.jsonl` | `ab574f17bcbff82b92c704adbf38f0ce43e9622ba9e8ca17a8274a621de39dbc` | | `research/parity/foundation/incogni-02.jsonl` | `e3d774c92c098636558873fdc01a9616c00a5a7ab279fd6e10596d9f633d5e54` | | `research/parity/foundation/keycloak.jsonl` | `c0d3d8d917dcb20ad0bb21930eb4301848de66a779ba50995bc7b9c573f07dfb` | | `research/parity/foundation/microsoft.jsonl` | `5bd6d369b6eb733a9d519851831d3e18f696b0317992f08004f80d996d6c951b` | | `research/parity/foundation/okta.jsonl` | `7178629e89814bbaeac669f6a65bf7e4d6101a46a56f8ad1a2835002a37793eb` | | `research/parity/foundation/ping-identity.jsonl` | `40438ba71f031a155339d5b2b6bb72e206fe530b21d0d5dd7491a0f8fb5d6c6e` | | `research/parity/foundation/sailpoint.jsonl` | `b62b8f7770b436f90caa9982f16ab21d3f75319bb2414bb08f0755a3e0bbc12e` | | `research/parity/foundation/saviynt.jsonl` | `8d192563994dc5bd148ff69b1046b38bdc36cf27dca3a015c0099a4f6f706ad4` | | `research/parity/judiverse/15five.jsonl` | `dd15b002042967fd58cc51e1095f632f5f1663081da529b56d4a6ce17042ed85` | | `research/parity/judiverse/37signals.jsonl` | `b9dff4c6b23f22440ae9c9dde627fbd537a82cb68231512f71f470fab50f3b2f` | | `research/parity/judiverse/activepieces.jsonl` | `e4330c0da634d35070601f55ed9063be92d8bdeedeac47a863895a867f55573a` | | `research/parity/judiverse/agno.jsonl` | `26b6fd9fb118417e2124ff97270a26a96816f06467717f89dc121d15cedee0db` | | `research/parity/judiverse/airtable.jsonl` | `fd30147a81ed09d512cac295b49af166cb8a2c5a599da496d0f64b5bad2cb080` | | `research/parity/judiverse/alibaba-qwen.jsonl` | `34c3318693c015ba6dc8dc9cf442893878fa6afcfd1870f89bb72d4621e5f538` | | `research/parity/judiverse/anthropic.jsonl` | `b5a0308b1c776d995eab338b935f988ceecef6f62da5a3c99acf1a654637904b` | | `research/parity/judiverse/apache-software-foundation.jsonl` | `3782f75476f89a6a850947a8ea0f9fb1936779bf593ae178d706108b69e95e40` | | `research/parity/judiverse/arize-ai.jsonl` | `73c16bb82b2d1c885a661e2525e9b3a9cde5715d15cfa3246a27a65bef626f59` | | `research/parity/judiverse/asana.jsonl` | `bd011950bcf291eaecced474728cb9eb080b3bba5b247706efaf19321b326645` | | `research/parity/judiverse/atlassian.jsonl` | `0d84ad931abadf5e159e66e092fdca3575c8facfa9f6c2ab794fb5f35f8a12ff` | | `research/parity/judiverse/avoma.jsonl` | `0c48ae4f0236b3512ba6b23cb70dfc7c824b79e3ad28a02726603d2a9ca7230e` | | `research/parity/judiverse/aws.jsonl` | `326f6d04207b9ecbdbfda4196835a00e9559b97aff9214b8ff11d0c7fc4436b0` | | `research/parity/judiverse/bentoml.jsonl` | `01a8a1bea76f2b0233944fb1b3ea0bc1bf0aac5e03a5d667924c58c4b308408c` | | `research/parity/judiverse/berriai.jsonl` | `0e045529ca1cdd28dfd9f4848aa153d3b2563f16032794815d893a0c924f7381` | | `research/parity/judiverse/cisco.jsonl` | `4dab93e50f7f8201a5185f8ece659e38f4bd532d199650b6233e0bde3b53acf5` | | `research/parity/judiverse/clickup.jsonl` | `409653345e89d468de68aa3dc1780711d157712d69e5bf5fe6c33802d6f70754` | | `research/parity/judiverse/cohere.jsonl` | `f7f297fb52666aaa96b7674dc22659bb242735d7b6530f2bc7e3cd59899ca964` | | `research/parity/judiverse/corporation-for-digital-scholarship.jsonl` | `aaff51bf8900d8c5643e69d30d8b13dcec3a6e6a717ee4b71694b40a579c1a77` | | `research/parity/judiverse/crewai.jsonl` | `c2ae8153703f3489f681df092fa095328b1940ebe81ee58af2e6e0e373244923` | | `research/parity/judiverse/crossref.jsonl` | `1743765b91c14f7e82214adf4d801ff586e084ec662b297af3b09d51d71eb863` | | `research/parity/judiverse/customer-io.jsonl` | `f27e69918d16be3d14a50f259ed44daeb36db9a38db851f396d54786a32971e1` | | `research/parity/judiverse/daytona.jsonl` | `56ba9825b9cee72e00259fb46adc93c9ed278d459c95d7578e492a1d60bedfae` | | `research/parity/judiverse/deepseek.jsonl` | `208df3f9b8614e9e0fa761b0001143ac6279a6cde03286668e0930d0e385e95f` | | `research/parity/judiverse/dify.jsonl` | `252f075e69bbb34f26829f379719c69bf244350d916c540c75c86fe5cb76cbb3` | | `research/parity/judiverse/elicit.jsonl` | `ee1bf05ece9d3945410d7f4fbc46929efca0b4dab0e55613f01afe7dea918eea` | | `research/parity/judiverse/evernote.jsonl` | `d4031dee8cc30c8de8bce21d32eff9577d113e759f53dabe56ab080bb804e2c2` | | `research/parity/judiverse/fathom.jsonl` | `5bfc4bcdbfe59f21ae8aa0d182921c8f518919937f9d7d10d5650a2395b28621` | | `research/parity/judiverse/figma.jsonl` | `6f4dc87a2ef62c17491463484358ff68d24f2678707c2dffc7b887fa5dcacd36` | | `research/parity/judiverse/fireflies-ai.jsonl` | `e9ad6507ff92d2607eb5dd5ee58c951c09ea91b8c9fd26d39ef2a5aea785ad09` | | `research/parity/judiverse/google.jsonl` | `05067f84167861671d08c98ee9f1d679e4e549a37d74019fe6192bfbe4b25019` | | `research/parity/judiverse/guardrails-ai.jsonl` | `62611c8081dce0c1cf3076a83f1f0529b427d25430539886b857fcbe0a043946` | | `research/parity/judiverse/helicone.jsonl` | `fdca3c556584a9cdbd547a4b83d960da64f37cddad24ad3566666cf99aba3798` | | `research/parity/judiverse/kestra.jsonl` | `63cd47bbef9bfcb68f7b5b50db68ca8adcf9b3f64c024bbdfb668be73a034b25` | | `research/parity/judiverse/langchain.jsonl` | `6d83f406d95a4fb4043599e9a4f6403bfa0b04c9571bcc43c92bb5dbe5e8898a` | | `research/parity/judiverse/langflow.jsonl` | `c09701e25ce948eea654bf57093dbc80e8d100f7aeba46744e2c970a00719e03` | | `research/parity/judiverse/langfuse.jsonl` | `7baa171e11c3b01b4137d8eb615e3aa4308096c5c705b8827cd9c8f1942ea48c` | | `research/parity/judiverse/linear.jsonl` | `894a8f49d7981009a36fbbd7e616207ec8127a17fcef394638ad394b54623336` | | `research/parity/judiverse/llamaindex.jsonl` | `98b85fafe39a3b22d01131812e04cf5b314d44b3a947f960a422facede6a4510` | | `research/parity/judiverse/mastra.jsonl` | `2d2d173e99eca98887f5b1de128d72ef5b0fb035e05093cdd2f972c2255cc1fc` | | `research/parity/judiverse/microsoft.jsonl` | `d8e4258db79f1e56a2768ff50ed253095953687acb8edac0ab6f7d2d7488479d` | | `research/parity/judiverse/miro.jsonl` | `dda40ad91bc6f34f318f69cb6bdf3ac3163cf425c1870a89a4c9a7028253c7f7` | | `research/parity/judiverse/mistral-ai.jsonl` | `f0e2390bbb6db81fde0bbfe6569e325b8e416b1e085cc4c82573c9e8d49e4bec` | | `research/parity/judiverse/mlflow.jsonl` | `e37ff22e6cf0219a611af6a20ba84b04a1735fc3fa400b00ac36d722235c0360` | | `research/parity/judiverse/mobbin.jsonl` | `c94dd8da99772c4e5b644c07d4fadaa6fad1409860341ea7d6cbfa766df63971` | | `research/parity/judiverse/monday-com.jsonl` | `57fdd847a5aaade9e31dbb562622a2ffca37ee493f96e1c4202bd1076199e01b` | | `research/parity/judiverse/n8n.jsonl` | `e1e57d54436b94381b597a6090300ebdeeb6517706b9e7a5854388a082c7d6ba` | | `research/parity/judiverse/notion.jsonl` | `3aceb90a6dc41e5111d12f672204f402e293c090467d5522ee80f8dd2b247f97` | | `research/parity/judiverse/nvidia.jsonl` | `3817a52ec1d3c7a7c725266c17e173ab8d904c08e6b1cd1a65e521eb0fb4f7f4` | | `research/parity/judiverse/obsidian.jsonl` | `804e3c2e4f71b751a892cf2b339c1382792426621258a1d5e8a808b9e836789b` | | `research/parity/judiverse/ollama.jsonl` | `2fbbb7eccefbd11b6e34e1086e72f41c05a896df681cb255637239846b7aade2` | | `research/parity/judiverse/openpipe.jsonl` | `4ae6e3425d4d570a7d29c9384cce4a6e885b53b445bfa4580e9bf0574f040fde` | | `research/parity/judiverse/otter-ai.jsonl` | `9c58fd1acde6dbcb81ba99266ff316d0cfc02e3620eeaf4f00c9ee0fb5cc20ca` | | `research/parity/judiverse/ourresearch.jsonl` | `fca975b87aa13c8b58db4ed1889c425f74c32a2cbfd1fbf9351226f1f1f18570` | | `research/parity/judiverse/palantir-aip.jsonl` | `0fc97c4e5f18294bccce5ead6c58a8a765a760cbfc7d36a341da25fcc12be6d3` | | `research/parity/judiverse/palantir-foundry-01.jsonl` | `10ffaedca2d29b4c1e0ac1c479c5a172b9cac8c324aae591127b3255411b7729` | | `research/parity/judiverse/palantir-foundry-02.jsonl` | `a26fc85fa5107031ae75bb6e7bab81cbbe59e8515ff0793275d8231a19bfbfe8` | | `research/parity/judiverse/palantir-foundry-03.jsonl` | `faac9d6cd6bddc0a70c8b5d80d611744712b4c9c1211e1d6c6f86c34c3efcab1` | | `research/parity/judiverse/palantir-technologies-aip-01.jsonl` | `b70a70bc820a1d5d72239152eb737c4e1caa2a9e282976b248404e916d220450` | | `research/parity/judiverse/palantir-technologies-aip-02.jsonl` | `b0fd9cf71a092e9f6b7e4dc6d3b109e6b05c066bf8cd0c2c126b6f965e3a4a39` | | `research/parity/judiverse/palantir-technologies-foundry-01.jsonl` | `31e10c480a45b5b875400e711bf2a8d0b66360a9c0e599f1dd8f71419225c535` | | `research/parity/judiverse/palantir-technologies-foundry-02.jsonl` | `1d65387b41c5fb7b09f5514dd331692b5fe95cd1ed9123600944ae76d20ad44e` | | `research/parity/judiverse/palantir-technologies-foundry-03.jsonl` | `0d78e9de76de6f144c5a0860133b954e4ea53e499483d9a25b4fff6836e5db3c` | | `research/parity/judiverse/palantir-technologies-foundry-04.jsonl` | `3e17e42322546544f332b1b778673215973c6358e6acc9d6900e70d726b9f198` | | `research/parity/judiverse/palantir-technologies-foundry-05.jsonl` | `957d15c28cc8143f03838046c0733914c0f2a9994a9e34376020d237f1d05b42` | | `research/parity/judiverse/palantir-technologies-foundry-06.jsonl` | `b3eb67fa90eb5a6c5f3113b97928169c30ef046f90ce2e306fc6e9ee0479b7fc` | | `research/parity/judiverse/palantir-technologies-foundry-07.jsonl` | `bbc8e73d436b561c97828bba5a609829c0a4adc061b506ec11f6d5c1f2b9da01` | | `research/parity/judiverse/palantir-technologies-foundry-08.jsonl` | `45f189e878e443db3eac85f0fced02da75ea638d3df6fb96f116a7c343e713c1` | | `research/parity/judiverse/palantir-technologies-foundry-09.jsonl` | `4a2181880a5f79e508021f885764a1411f7b19103b33b90d4b865d894ef333f2` | | `research/parity/judiverse/palantir-technologies-foundry-10.jsonl` | `2cab815df17ca36f1ffdc660b7167122824111a6310e4e039f3743db0182a5cb` | | `research/parity/judiverse/palantir-technologies-foundry-11.jsonl` | `c01be74143495d008ebaf328bfe99e13785ce69746c272a8f4a2d33982c00c7f` | | `research/parity/judiverse/palantir-technologies-foundry-12.jsonl` | `0e8d3c700ff773f2c2221eee9cc36ddeef71ce7097b06efb7ab1de3590d02d60` | | `research/parity/judiverse/palantir-technologies-foundry-13.jsonl` | `e89dedac2bbd6f34f57b504a0d814974e6293d68cbe794fc5f08f05ac3713411` | | `research/parity/judiverse/palantir-technologies-foundry-14.jsonl` | `f1d206f97a9ac1d558a4ca1341d54523b10ba0db00bdf7853ff02cb22c026c03` | | `research/parity/judiverse/palantir-technologies-foundry-15.jsonl` | `e70f98dfb4e39b1c7ccd91b99f1f73556f0bffe4b796540476203db39bd7eb55` | | `research/parity/judiverse/palantir-technologies-foundry-16.jsonl` | `b59b65bdd7a33a41755090e621eeb93dea69862658e252a87e64da6a42c8ef42` | | `research/parity/judiverse/palantir-technologies-foundry-17.jsonl` | `a7796496b490e421cac38e09383274a3319b287225cbeec09eb6dd2828ba57c4` | | `research/parity/judiverse/palantir-technologies-foundry-18.jsonl` | `57ff548bf886b609d4304c8adb1d0262a6ffb7832c2416294c1cad3aa8fb5b47` | | `research/parity/judiverse/palantir-technologies-foundry-19.jsonl` | `eae1983ae0e88ea7412123e404e27c76dbf16a31eec74638200a6d36577c0c39` | | `research/parity/judiverse/palantir-technologies-foundry-20.jsonl` | `2d0868fb39cd8ec10112498c92389c684b9e6a5384ed3954e5ef0dc0a7724d15` | | `research/parity/judiverse/palantir-technologies-foundry-21.jsonl` | `3ac223b7506305f5440b6b881a365196e7143c6e8f4e11b84ff5d3acfa20bd33` | | `research/parity/judiverse/palantir-technologies-foundry-22.jsonl` | `1b1cfcb42c8e3c03cbb414195588468df3e920a046ee3190de1aea0b85ab91b5` | | `research/parity/judiverse/palantir-technologies-foundry-23.jsonl` | `4333ba05643badcab033386d9aee1258f73994ef87ed4c44dc7c05abc0f580fa` | | `research/parity/judiverse/palantir-technologies-foundry-24.jsonl` | `24653824d0a17ae638f5c1c23ce89269fbed1e4c3394a532a0778ef07a523ee5` | | `research/parity/judiverse/palantir-technologies-foundry-25.jsonl` | `5848b77855312822b8bdaac5a653071032ead867cd7a0fd548b073217ccab927` | | `research/parity/judiverse/palantir-technologies-foundry-26.jsonl` | `faafa9956ff4089aa5ab1d2bfcff50271d5f41e0631158b40c8d20fc17e87d12` | | `research/parity/judiverse/palantir-technologies-foundry-27.jsonl` | `5550ab598fa56114e3d98b1d514a22ec0d2c7b394d60b903db726825bac45079` | | `research/parity/judiverse/palantir-technologies-foundry-28.jsonl` | `0cb139b1392a41d7fc9b2eb1e056218500314cc267e93031c30845a28fca43fe` | | `research/parity/judiverse/palantir-technologies-foundry-29.jsonl` | `65856c046caf6f3af5fad3e765d47deb3d226d1f9cc96ffc48aa4d5963df694d` | | `research/parity/judiverse/palantir-technologies-foundry-30.jsonl` | `df886c5c4e0d07f458c6133af01d816c9dc54999447e6eb1ab841b740afb5cfa` | | `research/parity/judiverse/palantir-technologies-foundry-31.jsonl` | `db546657664932d244d676446f7f773ad716e267f723dac56dc944a6c4140923` | | `research/parity/judiverse/portkey.jsonl` | `f44d84e003d58a9c06b1b9a6c0b1824577b81fb3aa8d13b0483a499765d292d8` | | `research/parity/judiverse/pydantic.jsonl` | `0a072d064c09b6e2a4a2f5cda4eb6f3f89b4f70836aef6fe7385175d372f4217` | | `research/parity/judiverse/read-ai.jsonl` | `b6fc0d7db328113af6d7f2b7fb4d53413873f16f417f5f9a656f96e650985933` | | `research/parity/judiverse/salesforce.jsonl` | `558023463165ed8f24a02a4e56f1e58c607339c129e4d743f17e82c825d8245d` | | `research/parity/judiverse/sglang-project.jsonl` | `72b57a52ddebacacc7a31c4e6c789ef5c2b869a0e0944a8059493153aa1e534d` | | `research/parity/judiverse/swe-agent-maintainers.jsonl` | `115e7e498a6bd4ac25a5a4aa25d2c1cb58851cc42ab393208ea0bdccf655fee2` | | `research/parity/judiverse/tana.jsonl` | `599cb63ba1f4e76e8c8ad66d30d26eeab7b9a79fe4c8ae29c3476d7003ce1968` | | `research/parity/judiverse/tl-dv.jsonl` | `73981fe4014db3094aa542048b70d25da716a1f691ceaf25d5bc11fa468ea2b3` | | `research/parity/judiverse/ux-pilot.jsonl` | `371d7cb6c39f94cdb321b7f54080e320e5273d20ac676945f3cb90ff3fb17960` | | `research/parity/judiverse/weights-biases.jsonl` | `cbbccd807c99961c8168bd42552f2810f6446443464d9810294b2524caa2dbbc` | | `research/parity/judiverse/wikimedia-foundation.jsonl` | `42a0e4ddb56ec935ac176e9d33c1ac36ee2eacba051bda0c43bb6f69e37e0263` | | `research/parity/judiverse/zoom.jsonl` | `6f1d232b9a8da47ae22561e998cf05c740a46f70840f0033325420f6418ea499` | | `research/parity/public-consortium/15five.jsonl` | `9c90ff0f7232254e88a9ba3400ceefd653f6ffcf18dcd1aafc964fb45517d3d5` | | `research/parity/public-consortium/auditboard.jsonl` | `97a17d3eab813188711613df6632da4d526e17d3ec7bf01ad20881cd7c44f90a` | | `research/parity/public-consortium/brilliant-worldwide.jsonl` | `5998bb007ec20532d2d4b56ca043d1aa1241c68fe004ba5b3a6ab6d9dda86fff` | | `research/parity/public-consortium/cloudflare-zero-trust.jsonl` | `4c17ccc17cdc55621f5943a12d8ad08471f2c83019af1daff69e4d870b875ec8` | | `research/parity/public-consortium/companies-house.jsonl` | `71f735fd97dc8d4e1095ad166be59d55a281c13ae12d4d5c8e31690bf3b2b238` | | `research/parity/public-consortium/coursera.jsonl` | `b3272f1479107fb5a5af75f4397bd26354186eab4c0b7d99b64418482b9e947b` | | `research/parity/public-consortium/credo-ai.jsonl` | `2e735a066f7cbce8b9106f7ce1063217b61e5a71f572bc2ba6a63914aeb12232` | | `research/parity/public-consortium/crowdstrike.jsonl` | `35c2b3eb9ef4409183a43a49d4b3cf2a1cfedfb8cca66ae4ad6d829a0427f497` | | `research/parity/public-consortium/customer-io.jsonl` | `8a5cc10a70ae41c3ec0b1439053e18d7abeaa441af456073d26a4d7dcd9edb43` | | `research/parity/public-consortium/d2l.jsonl` | `469427f74b7ed1c90d6948f6f53cfc2f14207650a45a8b30647dbf86852d134e` | | `research/parity/public-consortium/darktrace.jsonl` | `32d7a9f6cffce5717d4b119b63be8e5850956f3bda1fd8d2c5d4b9d33150e245` | | `research/parity/public-consortium/datadog.jsonl` | `00364b7b9e1730a10678e3fb9b2f10d537019485aba9e239e2bfa8e3596b1510` | | `research/parity/public-consortium/dependabot.jsonl` | `da8d63a4d27534c393be2cc4c98bbbb53a38c566fb554c8fad6455ef86bfc881` | | `research/parity/public-consortium/drata.jsonl` | `8df72d7024fd8bc92004fc2f269a3c4def6d157ec225d27309846bd09c22dfea` | | `research/parity/public-consortium/elastic.jsonl` | `193014953676e751116e361753ec780078e9a1b02d4af21a01fab92947df7346` | | `research/parity/public-consortium/futurelearn.jsonl` | `b65a98326c946b60e1d6cc9e5a1fe866e94176b14ce86b2275b4474519b030bd` | | `research/parity/public-consortium/github-advanced-security.jsonl` | `ed7ac8a0ee7701bfacaa1994054c3640784d2e6ea8d9d5a86847a07fe75b0722` | | `research/parity/public-consortium/government-digital-service.jsonl` | `82ed62c5d5f1c99f2489f0d07fc2b03a04698bfa5dda86628bfab451c5284729` | | `research/parity/public-consortium/grafana.jsonl` | `6d520c5f5cfee268ee3fb3b5bb6cf50e47fc9e74806657ba8b7785d8e7d0fd1d` | | `research/parity/public-consortium/honeycomb.jsonl` | `80a8ca6862a8139a4f81b7cd1179ecc9f710e4f44ced529d3475767f9736d282` | | `research/parity/public-consortium/hyperproof.jsonl` | `574d22dbc546d412fd869f68d1b55ea888a2cb95860cc727073fa5d4f81ea210` | | `research/parity/public-consortium/ibm.jsonl` | `54f9625ab39d92bef2e7a2ce8f173590d83f6800861c2ab5b380af9594d45538` | | `research/parity/public-consortium/instructure-01.jsonl` | `940a5ceda9b21d9f22f68f1d574b00fb1f274cb33e36cd8614d1e0268dcf9691` | | `research/parity/public-consortium/instructure-02.jsonl` | `70a9927af957b0867d09b261c43da30e78b9d89ef64ac9d923d115a8ea133fb5` | | `research/parity/public-consortium/logicgate.jsonl` | `e998e448810908a1dd720d50fd1291fc1ae70c272167d475647038591ce3cf5c` | | `research/parity/public-consortium/metricstream.jsonl` | `c50f988e2c792561b0856b767e0ecd7658cc6298358a08f10292d32094fd5a3b` | | `research/parity/public-consortium/microsoft.jsonl` | `e4315fe50a5af200438f14ec6d29050891b595d828438c8e86a5137bf4e47edf` | | `research/parity/public-consortium/moodle.jsonl` | `ec5225e43de965945d7f48bf640f5b75a98cc05a32bc3e92f8f8ef7e5ab6c14d` | | `research/parity/public-consortium/mysociety.jsonl` | `7417e0362e4397818bee599e804696891e6af1d0b86bd20a23fcc5342c36f9d0` | | `research/parity/public-consortium/onetrust.jsonl` | `15bf262ccce8c511b7da377a01166af40e93644d17778e8dc9278f869d5fcf54` | | `research/parity/public-consortium/opentelemetry.jsonl` | `38e05c27a98f616cd136283013b34d2e234d00f91364328fb7b2e525cf4cdf71` | | `research/parity/public-consortium/palantir-apollo.jsonl` | `7174d27cd20e3363a00ef34f51251a1e54863c859d77a037cd05bf0c305e4632` | | `research/parity/public-consortium/palantir-foundry.jsonl` | `deea3f9d7ff0a942e28c1080179c266ad66844d02b0b427ae194c002af9a8255` | | `research/parity/public-consortium/palantir-gotham.jsonl` | `004d792f500252834e008408c4c7138d917324646a131824fa3003b294ca9810` | | `research/parity/public-consortium/palantir-technologies-apollo-01.jsonl` | `ccc4cebc17cc2e57f992053e3717e60b5f4ff246542f76745575ad4a0ae2db4d` | | `research/parity/public-consortium/palantir-technologies-apollo-02.jsonl` | `e8227c61188f922d8c4a9681823303dceaaf38815370bea53c2990e3490586c1` | | `research/parity/public-consortium/palantir-technologies-apollo-03.jsonl` | `903538e82d26d0b18345eecc2e494f6d935cbe254c722007f1d79571105ebde3` | | `research/parity/public-consortium/palantir-technologies-foundry-01.jsonl` | `12924609366aea8ceec46751414be221986ab68c008e146615daaf2373003519` | | `research/parity/public-consortium/palantir-technologies-foundry-02.jsonl` | `31541e70aadad56653a3040935b52ad474f8c635835dbe348f374526c4c59713` | | `research/parity/public-consortium/palantir-technologies-foundry-03.jsonl` | `659ae14466aa9ce5f627cf25f40cd8d42c17718999fdb5dcf3b71e350b64c1f0` | | `research/parity/public-consortium/palantir-technologies-gotham.jsonl` | `d2b14b6590e8e6b84abef1e862c05ed301859779861bad71ab64e039f85a4551` | | `research/parity/public-consortium/prisma-cloud.jsonl` | `2af80fdb7391f85a7a179f43345c5a5273c406b236246d10acc854e271ae329f` | | `research/parity/public-consortium/prometheus.jsonl` | `7791908e9dc65fc980168fd433100fd1d6d5a22b3ee3d4f48b0782e3539d5366` | | `research/parity/public-consortium/qualys.jsonl` | `82fcdb15832e754682a3017b13ca1a524ad331320c14b4c3d6434a201ab5fc5d` | | `research/parity/public-consortium/secureframe.jsonl` | `1f2481ffdec3f221c4d42d96f1220188e62effa9c2e2fb37ab78f6454713834a` | | `research/parity/public-consortium/sentinelone.jsonl` | `271c19ef5916f5964d16d1fc3dcab4557853a5b34911493bdadf3c774bd06d69` | | `research/parity/public-consortium/servicenow-grc.jsonl` | `5d5bcaa7e85146483ce234ca7e6f2b4480404cc5fda0584966e2240aacefa7f1` | | `research/parity/public-consortium/snyk.jsonl` | `c6f059b12ce501058a7037c26c2bc35791fe31b2cf2367aa958d593c5fa78438` | | `research/parity/public-consortium/sonarqube.jsonl` | `ad118876b8ddf0b0ef014d52b559465105dc37b310df08a410c3ec93e7c6fc40` | | `research/parity/public-consortium/splunk.jsonl` | `cf888ba3e7b8d3dce774bc77e6ea76d0e9e81ea4768286a9799e34882ec05831` | | `research/parity/public-consortium/sprinto.jsonl` | `2e02b68b07554696832d07a143d6d02afefec4a3ffcb94720d22e5bdb9edea9a` | | `research/parity/public-consortium/tenable.jsonl` | `55b58fed70e16bfec191e077138995f82a9fd4236fb8fd6044e531ebc3f3c7e1` | | `research/parity/public-consortium/twingate.jsonl` | `ca5f64ed83af0757c3fb26868ec21d070fbc474c8d2e7f327ba97fce874e86db` | | `research/parity/public-consortium/udemy.jsonl` | `dc04e360972afe3645a4d0da06864feb03cd85f19554c056d5c04ad477eeae1a` | | `research/parity/public-consortium/uk-research-and-innovation.jsonl` | `349bff4ce972e1f1141789962357d4b6e9415f255ae477620ad55f4d76247b98` | | `research/parity/public-consortium/vanta.jsonl` | `0b82f30fe404b3856a396a1db5843f79bfb11630c37993b7e834cef40b087099` | | `research/parity/public-consortium/wiz.jsonl` | `5ad03aceb505036712b1899efaa3b6d9a1c7a0bd546a2cf2cc05b8f17e6888b0` | | `research/parity/public-consortium/workiva.jsonl` | `c8970733c5e516a9cb91715d65bd11f898e727563c33e816452d4cf6d6911aa7` | | `research/parity/public-consortium/zscaler.jsonl` | `14b62b77df23f59525e6c31d41e284d1a4b7bd77a57b3c14a7f38be549c1dec6` | | `research/parity/unassigned/airbnb.jsonl` | `94539860347ff5fce8b58e0328f9f36d2da539eb75d7b0bb7c11abdc5500773c` | | `research/parity/unassigned/booking-com.jsonl` | `2a0284bf75b3ea136eec3c9ffe879dfdbf89d5a825ccc5606193c3733ae14bb5` | | `research/parity/unassigned/citymapper.jsonl` | `54747be08d5f66d444e749b31d2ceab0d3d16ce0004c13e9a0c41a81bc499cb3` | | `research/parity/unassigned/deliveroo.jsonl` | `6252ba05be3b4e5bed05803f2a8bbae28f4703443f6dae308a796e70d5609cf1` | | `research/parity/unassigned/headspace.jsonl` | `6b40c605082978a5b75794aaf9e7da03f2faf1461b46be565c0a9c3ec36a3c39` | | `research/parity/unassigned/instacart.jsonl` | `6137e4f28a1504f90ba8a628c2f705764d641cdccedc5105fa3283148e6c8b86` | | `research/parity/unassigned/shopify.jsonl` | `433b510fcff2898b69c3e0d14500a83ba1eeda68f311c06703d75a7e65fb2b51` | | `research/parity/unassigned/strava.jsonl` | `551f4beba4c0ba460923a8d9491bca29700e9a73e77b9137248f29a8fa925a4a` | | `research/parity/unassigned/too-good-to-go.jsonl` | `3db191ce9672f722fa4b40ddd2e6d948dfad47c243cda5794259aaefbeccb662` | | `research/public-consortium/COMPETITORS-REGULATORS-REUSE.md` | `f78f97e663b8a5d2407a0c5c68f632aea4a9da2935631f9ce5ae960dc5267750` | | `research/public-consortium/ENGINE-ERPNEXT.md` | `e7f4b172dd31c1d1173176cc9754f8ae2986b796da0c501bb22518d8fa7da963` | | `research/public-consortium/MOBBIN-FLOWS.md` | `4aa3d18f1758414e305098489009f9d43677daf44558a97d2598da527ee7ef3c` | | `research/public-consortium/REFERO-FLOWS.md` | `b7b4a23c38ecd15f6b7999216c27a541fe1d88d5a82153fc4914ed4b5ac318b5` | | `research/public-edge/ANALOG-HUMAN-SWITCH.md` | `d486a369118cebfeddb64279587a5e17ab42c711b318d0b5b71fb1ad41c9df71` | | `research/public-edge/REFERENCE-STUDIO.md` | `e6516dd59e8132d83c06617c1dadd18813ec47e95b6da2a7a6f6b7e83db75ea9` | | `runtime/INTERFACE.md` | `64acee32ecdbe87c1a4369f296966800087f21c5b6080b0e775279b36e600394` | | `spec/BRIEF-FORMAT.md` | `d74bc4fa6c0410af598d29ff15dfbd8bd89c24e242586ba6ba4e15193853f2d7` | | `spec/SPECIFICATION.md` | `25e3970bb9adba341f7e54637a72a7757b529b3ca0187d1723d682039beef094` | | `spec/UNIT-FORMAT.md` | `0f69ddc7d72cfb0dfbad4a23890e5d1a9ecff669e3f2571f570c99bd02a078f6` | | `spec/changes/2026-10-31-money-provenance.json` | `c7680815f0a592ac3249fd690da890d3134359700ea4c68ef252e5360e472970` | | `spec/changes/2026-11-01-reference-records-pass5.json` | `0e258a2e131fba670c489005ab890c3e9d28d582a7ebd105b162f9f82523d06d` | | `spec/changes/2026-11-01-research-is-not-scope.json` | `6d9db088c530db59d90dbcb9c0bec9212d99044b3818508d751ecbd02380016f` | | `spec/changes/2026-11-01-typed-steps-honest-outcomes-reference-flows.json` | `3e7fe9759ddde7ccdebaa8bc3110465762cacb1357a6f5685c79a0c7e6511c26` | | `spec/vocabulary.json` | `a6a489907549bba5a87976a27a67f62b5793c2a94292c6b0e4bcfc7d93b78940` | | `tok/KERNEL.md` | `221a53d23963baa81fefcd7a1baaea3ab0431d58fe5c6609c08224dff2276c61` | | `tok/kernel.json` | `358eee947460932cd0c6aa542f77b2f63d4a8567f6813006fe1781e420f03a0d` | | `tok/objects/decision/01m68c87qmz54aqnbm8pqzmgt4.json` | `5e05a9f2f3253fb85e6d5c76dfb52bc0170f5a60ab1643c27b2695c778692eef` | | `tok/objects/decision/01m696xzzpcbnsexpcdtz7ak9n.json` | `cfc2aadf99ad8f6d2c04d1d18828846438610a9b344d48776be277836da38ef7` | | `tok/objects/decision/01m69b9nqqnxj7bh7p0mner5rf.json` | `333e523c4359f50d2108a0f6ff36519d15161e8494035c293792e80809ff323d` | | `tok/objects/decision/01m69gty59gnpc17g811317htn.json` | `bce281d068a0e44d2684d572c3b45d39d08888d53bf22bae5587a038f5a66270` | | `tok/objects/decision/01m69jxxj5cazjefrrj1fvb962.json` | `ebd62ab9c15e1d15138b7b8c8df7853ab2796a198eb47c7475825421a847f887` | | `tok/objects/decision/01m69jxxqzrhz7r36ht4pvv5td.json` | `f16c1d4600db923e32aab1b6d1b92463c00f63bb6a595f9abe04da648ee5667d` | | `tok/objects/decision/01m69k8c61zde0er6j94yb4ekz.json` | `e9284eb67eba9987fd9fe62efa87633ceaa962df7ecf68ada1f2c71e6acbe8d6` | | `tok/objects/decision/01m69m36psxrwhsa148z746ppv.json` | `3a062eb204a4ed8e4407ff5b44362573c0c45beac9c2db3cd3c96c9adadb454e` | | `tok/objects/decision/01m69p22005ent1mresdjz6sh8.json` | `8aafb51bb2e245e8bab64c272ff2b40630adaa3bc7a1dc37d3c67535b64f3af2` | | `tok/objects/decision/01m69thft6cmgyfgkjrt6bdrws.json` | `f216b67b1c33e3017c06351c2bebb79a87c114ddbb791640cde623e684a7c82d` | | `tok/objects/decision/01m69thg13xhssvqbxygtr2zp8.json` | `3eed853b886d6905d63c0a3317344ee3b933858feffcc043d0ee66f8ef697f1d` | | `tok/objects/decision/01m69thg7kks8fwaj0266pf1jy.json` | `87d5a7b49649be7d120492b2981156da627b637bf5005e3371a41bd274669059` | | `tok/objects/decision/01m69xj6emef4jkw7kgv63bsdz.json` | `45c80a091e6326858b13d358a80de1dc3512ecb5153d1b3ceb1ac7bf1e1b9fbc` | | `tok/objects/decision/01m6a4t0fxra5avfxf0pr2p692.json` | `5cef677ae385d35a712da06d41ec732e0b7148bd0bd9f20c4630d28c0197b963` | | `tok/objects/decision/01m6a7p9c8566pe659w2hpgbjq.json` | `e20c9fb3fb423ad3b1fdf6c412ed0fd40f5ef46c0c3425596a23bfd96ebb38fb` | | `tok/objects/decision/01m6a8mtbg94twxzcdgjp4yeac.json` | `7786801e05306ea979945e51b1967097dc5e85cbba9183fc4d2209918302cdcd` | | `tok/objects/decision/01m6acbx9rv6x4yxh5gxkz1x30.json` | `62911749948ad5b70e8545800fc79a340c95179af1809f09c731a26ae72739e8` | | `tok/objects/decision/01m6aeafjpfwmamsqm7apf2wfj.json` | `bc54e425b10d425cb30c99bcbea45b1c5503b617c56ef7f92a184d16ad85c3ce` | | `tok/objects/decision/01m6aennyqz91exa4a2sda0019.json` | `1a35b069cc38b279c1a354727554fafd43ceeacb15cb1191cd5282b5f06fd227` | | `tok/objects/decision/01m6bgf72jtm0n7xyba043bqh4.json` | `c07bf7241fc8d590d276a6beb3b697505cdef6ba680c7de4c4ff9bab3b3d70d2` | | `tok/objects/decision/01m6bgf72knqzc3da9n20c282a.json` | `f9214be2384fca8d69180c78237639c4d6a761ac43b0c6afaa25c514131fd443` |