Every change Wellknown observed on this MCP server, newest first, with what it was before and what it became. Tool-surface changes carry the definition diff. Nothing here is edited after the fact.
Changed the definition of "design.validate", "spec.validate" and "team.summarize"
⟨113 unchanged words⟩ default":false,"description":"Set true todisableskippersistencethe stored run AND run_id recovery (operational security and cost logs are still written) for this call (a private one-shot that ⟨187 unchanged words⟩
⟨115 unchanged words⟩ default":false,"description":"Set true todisableskippersistencethe stored run AND run_id recovery (operational security and cost logs are still written) for this call (a private one-shot that ⟨190 unchanged words⟩
⟨93 unchanged words⟩ Aggregates from your tool-call log and value events.Passprivate_sessionprivate_session=truehastonoskipeffect here: this read stores no run, and the per-call tool-usage recordforisthiswrittenread;ontheeveryhistoricalcall,dataasbeingthesummarisedprivacy
Changed the definition of "team.summarize"
TeamsTeams/Enterprise: summarises the caller's tool-usage patterns and value signals over a configurable window (default 30 days). Returnstool_call_countsperiod_days,topmost_used_toolsprinciples(callcitedcountsinandvalidateaveragerunslatency),value_event_countsvalue_events by event_type,
Changed the definition of "architect.certify" and "architect.validate"
⟨885 unchanged words⟩ in the UK (europe-west2, London). Codeprocessedistransientlysentbyto OpenAI (no-training-on-API-dataUS) for the check, under the EU SCCs anddropped;UKpayloadsAddendum (no-training): a certification keeps a stored run with us and is persisted at OpenAI; our operational security and cost logs, and OpenAI's abuse-monitoring retention, also apply. A private session cannot be certified. Payloads are JSON-escaped + delimited as inert untrusted data: ⟨85 unchanged words⟩
Changed the definition of "architect.validate", "handoffs.agency", "me.add_evidence" and 2 more
⟨474 unchanged words⟩ calls scoped by file/module; never truncate one call. When one call carries several files, keep each file whole, its own import lines included: merging files under one shared import block reads as code that uses names it never imported, and is reported as a defect. Pass repository="<name>" to group runs into a ⟨573 unchanged words⟩ is verified at the THIRD layer: yourtests, types, walks. The platform's own recursive-integrity practice: every PR runs validate against its own primitives, then cert. Real bugs surfaced via this practice in PR #157, NULL-UUID false-positive (iter3
⟨142 unchanged words⟩ single insert, side-effecting. Auth: Bearer <token> (
Changed the definition of "architect.certify", "architect.validate", "architect.validate_consensus" and 17 more
Pro/Teams—: second-pass adversarial certification of an architect.validate run that scored production_ready (A or B first-pass tier). ON CLIENT TIMEOUT—: DO NOT RETRY THIS TOOL. **RECOVERY FIRST**: ⟨135 unchanged words⟩ or iterate the code between architect.validate and architect.certify—, even by a single character—, cert rejects with code_fingerprint_mismatch. Fixing the code voids ⟨60 unchanged words⟩ run. NOT for tier=draft/emerging/not_applicable runs (typed rejections fire—, see below). NOT idempotent across attempts: ⟨60 unchanged words⟩ the run_id. The run is atomic by contract
Changed the definition of "architect.certify", "me.await_steer", "me.session_event" and 4 more
⟨474 unchanged words⟩ session that ran architect.validate, passing the codeexplicitlyexplicitly;—athestandardserverproduction_readyneverrunpersistsholds the codebyfordesign24 hours only, then an automatic clean-up deletes it), payload_incomplete (submitted/validated payload imports modules whose
Changed the definition of "architect.validate" and "team.summarize"
Pro/Teams — first-pass doctrine review of agentic code/workflow against the 10-principleAgenticAIBlueprint.Design Blueprint doctrine. ON CLIENT TIMEOUT — DO NOT RETRY THIS ⟨1081 unchanged words⟩ via this practice in PR #157 — NULL-UUIDfalse-positivfalse
⟨119 unchanged words⟩ read is untracked). Auth: Bearer <token>, Teams plan only;
Changed the definition of "architect.validate", "architect.validate_consensus", "design.validate" and 4 more
⟨403 unchanged words⟩ on rollback, badge/review URLs are stripped. Auth:Bearersign-in<token>required,Pro/Teamswith an active Pro, Pro Plus, Teams, Enterprise, beta, or trial plan. UK/EU residency; transient OpenAI processing (no-training); ⟨672 unchanged words⟩ via this practice in PR #157 — NULL-UUIDfalse-positive (iter3) and tie-breaker mismatch (iter5) — that 25 unifalse-positiv
Changed the definition of "architect.validate", "design.validate", "signals.feedback" and 1 more
⟨234 unchanged words⟩ :false,"description":"Set to true todisableskiploggingthe stored run AND prior-run anchoring AND run_id recovery for this call. Operational security and cost logs are still kept, per the Privacy Policy. The request is also not persisted in the model provider's response store; the provider's abuse-monitoring retention still applies. Use for private one-shots that don't participate in ⟨233 unchanged words⟩
Certificate recorded, valid to 2026-12-11
Authorization not required
First tool surface recorded: 29 tools (server version 1.26.0)
Showing the latest 12 events. The API returns up to 500 and filters by kind: ?kind=tool_surface_changed
⟨23 unchanged words⟩ "private_session":{"default":false,"description":"SetNotoeffecttrueon this tool: accepted only so clients can pass the same flag toskipevery tool. This read stores no run, and the per-call tool-usage recordforisthiswrittenread.on every call.","title":"Private Session","type": ⟨5 unchanged words⟩
⟨26 unchanged words⟩ ,"description":"Set to true to skiploggingthethisper-callsummarytool-usagecall.record for this read.","title":"Private Session","type": ⟨5 unchanged words⟩
⟨51 unchanged words⟩ case the server reuses the stored code automatically.SentHandledunderas thesametoolenterprise-safetydescriptionenvelopestates:assentarchitect.validateto OpenAI (transientUS)processing,fornothetrainingcheck (no-training) and persisted at OpenAI, JSON-escaped+and delimited).as inert data.","title":"Code","type":"string ⟨40 unchanged words⟩
⟨85 unchanged words⟩ a 6-minute budget. Edge case: if thetransportconnectiondroppedclosed before the first progress notification (very rare; ⟨1006 unchanged words⟩
⟨118 unchanged words⟩ write-only, single insert. Auth: Bearer <token> (Firebase ID token,any plan). Data at rest in the ⟨34 unchanged words⟩
⟨218 unchanged words⟩ only when the user chooses to give feedback. REJECTION CODES (invalid_request): a call with no rating, no what_helped or what_missing comment and no would_use_again answer (task_type, surface and contact_email alone are not feedback); a rating outside 1-5; a contact_email that is not one plain address.
Pro/TeamsTeams: summarises the caller's tool-usage patterns and value ⟨132 unchanged words⟩
⟨10 unchanged words⟩ to architect.validate to produce this run_id. Sent verbatim—: the cert reviewer needs the actual code to ⟨93 unchanged words⟩
Pro/Teams—: first-pass doctrine review of agentic code/workflow against the 10-principle AI Design Blueprint doctrine. ON CLIENT TIMEOUT—: DO NOT RETRY THIS TOOL. Long-running LLM call ⟨34 unchanged words⟩ at t=0s (before the LLM call begins)—, capture it. On timeout, call `me.validation_history(run_id='<that-id>') ⟨257 unchanged words⟩ , single-shot request/response handlers) returns tier=not_applicable with score=null/grade=null—, that's not a failure, the doctrine simply ⟨71 unchanged words⟩ Teams, Enterprise, beta, or trial plan.UK/EUDataresidency;attransientrest in the UK; OpenAI (US) processing (no-training); prompt-injection in code is inert. ⟨9 unchanged words⟩ truncation, NO `...` placeholders, NO comment removal—, the architect treats your `...` as literal code ⟨70 unchanged words⟩ this run): the validate first-pass is permissive—, it scores on doctrine alignment + structural patterns visible in the submitted code. Cert's adversarial second-pass is rigorous—: it scores on cert-payload-completeness as well as code ⟨73 unchanged words⟩ the payload (cert correctly catches when they're omitted—, the module would NameError on import as submitted ⟨92 unchanged words⟩ iter9 cert downgrades. SCORE VARIANCE DISCLOSURE (anomaly #10—: empirically documented): validate scores are POINT ESTIMATES ⟨20 unchanged words⟩ deterministic seed (the seed is derived from input—: same input → same seed) can produce ⟨49 unchanged words⟩ , median verdict + per-principle stability metrics) instead—: collapses the variance, surfaces unstable principles explicitly. ⟨12 unchanged words⟩ tool when one score isn't enough. ITERATION LOOP—: repository keying. Pass the SAME `repository` value across ⟨35 unchanged words⟩ response). Changing the `repository` string between calls—, even subtly with an `iter-2` suffix—, silently severs the chain and yields a fresh ⟨40 unchanged words⟩ validate verifies DOCTRINE ALIGNMENT against the 10-principle Blueprint—, design patterns, hand-off explicitness, operational-state inspectability, ⟨18 unchanged words⟩ an adversarial second pass over the submitted code—: catches production_blockers the first pass missed, name-errors ⟨31 unchanged words⟩ and behave?) is verified at the THIRD layer—: your tests, types, walks. The platform's own ⟨13 unchanged words⟩ bugs surfaced via this practice in PR #157—, NULL-UUIDfalsefalse-positive (iter3
⟨103 unchanged words⟩ artifact under review. SEND FULL FILE CONTENTS VERBATIM—: the architect cites per-line evidence (identifiers, branch ⟨52 unchanged words⟩ split into MULTIPLE architect.validate calls scoped by file/module—: never truncate one call. Architecture summaries (high-level ⟨192 unchanged words⟩ #158 38/F vs #159 74/C (+36 spread)—, same code, score variance from non-deterministic LLM ⟨42 unchanged words⟩ are created in the web app at /app/sessions)—: foreign ids are refused before any model call. ⟨71 unchanged words⟩
Pro/Teams—: N-shot CONSENSUS doctrine review of agentic code. ON CLIENT TIMEOUT—: DO NOT RETRY THIS TOOL. Long-running (~80-120s ⟨27 unchanged words⟩ N× compute. RECOVERY: same heartbeat pattern as architect.validate—, the run_id is emitted in the FIRST progress ⟨93 unchanged words⟩ the prior persisted run's verdict via baseline-anchor injection—: same code can score 60/C anchored vs 98/A ⟨17 unchanged words⟩ delta against a prior run (regressions/improvements panel)—, for that, call `architect.validate` which keeps baseline ⟨14 unchanged words⟩ the CONSOLIDATED outer row IS persisted with `lifecycle_status='completed'`—, the next single-shot `architect.validate` on the same repository ⟨142 unchanged words⟩ tool itself is called inside another private context—: but no such wrapper exists today. OUTPUT: ⟨9 unchanged words⟩ , max), `mode_stability_min_pct` (the cert-eligibility gate's input—, ≥ 80% means the consensus is stable), ⟨51 unchanged words⟩
⟨87 unchanged words⟩ artifact under review. SEND FULL FILE CONTENTS VERBATIM—: same constraint as architect.validate. Truncation produces hallucinated findings ⟨86 unchanged words⟩ the consolidated row IS persisted under this key—: discoverable as prior baseline for the next single-shot ⟨38 unchanged words⟩
Public—: list downloadable doctrine and agent asset artifacts ( ⟨39 unchanged words⟩ / mdc / json / toml / text—, the full vocabulary), pack_version, download_url, and ⟨52 unchanged words⟩ TO CALL: for the live MCP tools themselves—, those are already available through this server. For ⟨23 unchanged words⟩
⟨85 unchanged words⟩ over each member to get a cluster overview—: read the response. WHEN NOT TO CALL: ⟨44 unchanged words⟩ per slug. Returns 404-shaped error_payload on unknown slug—: the slug must match exactly the value emitted ⟨5 unchanged words⟩
Pro/Teams—: first-pass surface-craft review of a FRONTEND artefact ( ⟨25 unchanged words⟩ 10 agentic principles, design.validate scores the PERCEPTIBLE SURFACE—, what the user sees, taps, scans, ⟨20 unchanged words⟩ complexity, the Mental-Model gap). ON CLIENT TIMEOUT—: DO NOT RETRY. Long-running LLM call (~60-180s ⟨21 unchanged words⟩ the LLM call), and persists the run—: so on a client timeout, capture that ⟨76 unchanged words⟩ stay architect.validate-only). Returns surface_classification (ui_surface vs non_ui—: non-visual code is marked not_applicable, NOT failed) ⟨88 unchanged words⟩ (backend, config, type aliases) returns tier=not_applicable—, submit the actual UI surface instead. INPUTS: ⟨6 unchanged words⟩ as implementation_context (no truncation, no '…' placeholders—, they are read as literal code). Auth ⟨8 unchanged words⟩ Teams, Enterprise, beta, or trial plan.UK/EUDataresidency;attransientrest in the UK; OpenAI (US) processing (no-training); prompt-injection text inside the artefact ⟨9 unchanged words⟩ as architect.validate (timed_out, rate_limited, dependency_unavailable, schema_mismatch—, each carries retryable + next_action); the services ⟨46 unchanged words⟩ not a certified verdict. DOCTRINE: the eight laws—, each law's evidence, craft-surface application, anti-patterns, and the validator questions this tool scores against—, live in the `experience-design-blueprint` skill and docs/business/EXPERIENCE_DESIGN_BLUEPRINT.md ( ⟨13 unchanged words⟩
⟨55 unchanged words⟩ frontend artefact under review. SEND FULL SOURCE VERBATIM—: the reviewer cites specific elements, values, and ⟨25 unchanged words⟩ large, split into MULTIPLE calls scoped by component—: never truncate one call.","title":"Implementation ⟨120 unchanged words⟩ are created in the web app at /app/sessions)—: foreign ids are refused before any model call. ⟨73 unchanged words⟩
⟨31 unchanged words⟩ principle coverage, difficulty, library, and source-code link—: slug is the handle examples.get hydrates. Default limit ⟨110 unchanged words⟩
Authenticated—: submit an agency engagement enquiry on behalf of ⟨90 unchanged words⟩ call. ALWAYS confirm with the user before firing—: this creates a sales-visible record. WHEN NOT TO ⟨32 unchanged words⟩ <token> (Firebase ID token, any plan).UK/EUDataresidency.at rest in the UK. Response confirms the ticket id + scope so ⟨5 unchanged words⟩
Authenticated—: creates a support handoff record when an agent ⟨66 unchanged words⟩ human. ALWAYS confirm with the user before firing—: this creates a human-visible ticket. WHEN NOT TO ⟨41 unchanged words⟩ . Auth: Bearer <token> (any plan).UK/EUDataresidency.at rest in the UK. Response confirms ticket id + topic so the user can reference it.
Authenticated—: creates a partnerships handoff record for design-partner, ⟨46 unchanged words⟩ programme. ALWAYS confirm with the user before firing—: this creates a human-visible partnerships ticket. WHEN NOT ⟨17 unchanged words⟩ handoffs.agency); proactively or as a sales prompt—, only when the user has explicitly asked. BEHAVIOR ⟨7 unchanged words⟩ . Auth: Bearer <token> (any plan).UK/EUDataresidency.at rest in the UK. Response confirms the ticket id + audience so ⟨5 unchanged words⟩
Authenticated—: append a free-text evidence note to a specific ⟨55 unchanged words⟩ concrete they have built, observed, or decided—, not to capture intent or speculation. Pair with ⟨44 unchanged words⟩ <token> (Firebase ID token, any plan).UK/EUDataresidency.at rest in the UK. Notes are visible only to the owning user ⟨25 unchanged words⟩
Authenticated—: returns stages in the caller's active course where ⟨113 unchanged words⟩
Authenticated—: returns the caller's Blueprint learning-path state: current ⟨85 unchanged words⟩
Pro/Teams—: append a TYPED TEAM EVENT to a Governed ⟨351 unchanged words⟩
⟨93 unchanged words⟩ sentence event summary (truncated to 500 chars)—: a decision or transition, not a chat ⟨14 unchanged words⟩
Pro/Teams—: list or inspect the authenticated user's Governed Sessions ⟨32 unchanged words⟩ returns every session (id, title, status,repo_url, spec_ref, team_agents, run_count, validators = thelenses seen), newest first. (2) `session_id=<id>` returns that session plus its runtimeline (light rows; fetch full results per run viame.validation_history(run_id=...)) and, for team sessions,`events` = the typed team-event log posted viame.session_event. Attach new runs by passing `session_id` to ⟨31 unchanged words⟩
Pro/Teams—: return the authenticated user's validation run history for ⟨41 unchanged words⟩ returns a SINGLE run with the full persistedresult_json —result_json; use this to RECOVER a result when your MCP client tool-call timed outbefore architect.validate, design.validate, or spec.validate returned. The runcompletes server-side and persists; the run_id is surfaced in thefirst progress notification of every validate call so you have therecovery handle even when your client gives up early. (2) `repository=<name>` returns the full per-run trend for that repositoryplus a regression diff between the latest two ⟨5 unchanged words⟩ one summary per repository the user has validated,sorted by most recent. Use modes (2) or (3) BEFORE re-validating the same repository on either lens—: they tell you which principles or laws regressed ⟨31 unchanged words⟩
⟨61 unchanged words⟩ one summary per repository. Mutually exclusive with `run_id`—: if both are passed, `run_id` wins."," ⟨14 unchanged words⟩ run_id (UUID). Returns the persisted result_json verbatim—: the same payload architect.validate would have returned if ⟨37 unchanged words⟩
Public—: records explicit free-text user feedback about the Blueprint ⟨90 unchanged words⟩ ). Stores one feedback record, at restunderinUK/EUtheresidency.UK. contact_email is stored ONLY when permission_to_follow_up=true, and ⟨110 unchanged words⟩
⟨216 unchanged words⟩ they decided to give feedback. Use plain English—: e.g. 'code-review', 'architecture-design', 'agent-setup', 'onboarding', 'validation'. ⟨120 unchanged words⟩
⟨151 unchanged words⟩ Teams, Enterprise, beta, or trial plan.StoredData at rest in theUK/EU.UK. Do NOT include proprietary code, prompt content, ⟨40 unchanged words⟩
⟨48 unchanged words⟩ type that best matches what just happened: 'review_confidence'—: a validator lens (architect.validate / design.validate / spec.validate) returned aligned; 'runtime_risk_found'—: a validate run found violations; 'workflow_clarity'—: principles/examples clarified a design decision; 'agent_setup_success'—: user successfully wired up an agent or MCP tool; 'onboarding_helped'—: user understood how to start using the Blueprint; 'research_time_saved'—: user found relevant doctrine faster than expected; 'team_alignment'—: Blueprint helped align a team on agentic design; 'other'—: use only if none of the above fit. ⟨42 unchanged words⟩ answer directly: 1=low, 5=high. Do not guess—: only set this if the user gave an ⟨104 unchanged words⟩ it here. Do not ask for it explicitly—: only capture if volunteered.","title":"Team ⟨16 unchanged words⟩ Infer from what the user was doing: 'exploring'—: reading doctrine, browsing principles; 'designing'—: planning architecture or agent flows; 'implementing'—: writing or refactoring code; 'reviewing'—: running a validator lens on existing code, a surface, or a spec; 'shipping'—: preparing for production or deployment.","title": ⟨24 unchanged words⟩ based on their answer. Only set if asked—: do not assume.","title":"Would Recommend ⟨6 unchanged words⟩
Pro/Teams—: first-pass specification-quality review of a WRITTEN SPEC ( ⟨64 unchanged words⟩ traceability, risk and reversibility). ON CLIENT TIMEOUT—: DO NOT RETRY. Long-running LLM call (~60-180s ⟨21 unchanged words⟩ the LLM call), and persists the run—: so on a client timeout, capture that ⟨76 unchanged words⟩ stay architect.validate-only). Returns spec_classification (spec_document vs non_spec—: source code or UI artefacts are marked not_applicable ⟨96 unchanged words⟩ TO CALL: built code or a rendered surface—, those return tier=not_applicable; use the sibling validators instead. ⟨19 unchanged words⟩ + delta specs; no truncation, no '…' placeholders—, they are read as literal content). Auth ⟨8 unchanged words⟩ Teams, Enterprise, beta, or trial plan.UK/EUDataresidency;attransientrest in the UK; OpenAI (US) processing (no-training); prompt-injection text inside the spec ⟨9 unchanged words⟩ as architect.validate (timed_out, rate_limited, dependency_unavailable, schema_mismatch—, each carries retryable + next_action); the services ⟨46 unchanged words⟩ not a certified verdict. DOCTRINE: the eight laws—, each law's definition, rationale, anti-patterns, and the validator questions this tool scores against—, live in content/spec-quality-laws.json (the what-to-build companion to the experience-design laws).
⟨55 unchanged words⟩ "The specification under review. SEND FULL TEXT VERBATIM—: the reviewer cites specific requirements, decisions, and ⟨163 unchanged words⟩ are created in the web app at /app/sessions)—: foreign ids are refused before any model call. ⟨74 unchanged words⟩
Pro/Teams—: summarises the caller's tool-usage patterns and value signals ⟨117 unchanged words⟩ plan only; Pro and beta plans are refused.UK/EUDataresidency.at rest in the UK.
Wait for the NextCockpitAIDB Studio Steer (long-poll)
⟨8 unchanged words⟩ next `steer` event to a Governed Session fromtheAIDB Studiocockpit, then return it. DELIVERY GUARANTEE: the durable ⟨178 unchanged words⟩
⟨62 unchanged words⟩ vocabulary only. `steer` events and actor `human` arecockpit-originatedAIDB Studio-originated by contract (the owner posts them from the ⟨184 unchanged words⟩ record a steer (steers arrive FROM thecockpitAIDB Studio via me.await_steer). REJECTION CODES (invalid_request): 'Session ⟨56 unchanged words⟩ steer/human posts are refused with a pointer tothe cockpitAIDBchannel.Studio. Auth: sign-in required, with an active Pro ⟨8 unchanged words⟩
⟨9 unchanged words⟩ designer | system (`human` is reserved forthe cockpitAIDBchannelStudio)","enum":["pm","engineer", ⟨19 unchanged words⟩ started working on a steer; `steer` itself iscockpit-onlyAIDB Studio-only and refused on this channel)","enum" ⟨66 unchanged words⟩
⟨108 unchanged words⟩ the web app at /app/sessions. Read-only. Auth:Bearersign-in<token>.required, with an active Pro,orPro Plus, Teams,planEnterprise,required.beta, or trial plan.
⟨179 unchanged words⟩ review on what is actually changing. Auth:Bearersign-in<token>.required, with an active Pro,orPro Plus, Teams,planEnterprise,required.beta, or trial plan.
⟨182 unchanged words⟩ keep personal data out of the free text.InSeparately,everyfeedbackotherlinkedcasetonothinganleavesaccount may appear, pseudonymised and shortened, in our weekly lead-intelligence extract to OpenAI (US). private_session does not apply to this tool: it runs only when theserver.user chooses to give feedback.
Pro/Teams — recordsRecords a value moment (e.g. review_confidence, runtime_risk_found, workflow_clarity) after a successful validate run on any lens—(architect.validate, design.validate, or spec.validate—) or a doctrine session. Each eventcapturesstores event_type, surface_used (mcp/web/climcp or the site section), perceived_value (1-5), and an optionalbrief_context — structuredshortfieldsfree-textonlynote,NObrief_contextprompts(uportocode500stored.characters). WHEN TO CALL: after architect.validate, design.validate, ⟨30 unchanged words⟩ next_step instruction telling the agent to OFFER thiscall —call; surface that offer to the user. WHEN NOT ⟨36 unchanged words⟩ : write-only, single insert into ValueEvent. Auth:Bearersign-in<token>required, with an active Pro,orPro Plus, Teams,planEnterprise,required.beta,UK/EUorresidency.trial plan. Stored in the UK/EU. Do NOT include proprietary code, prompt content, or PII in brief_context—: it is stored as written, surfaces in admin AI-visibilitydashboards.dashboards, and may appear, pseudonymised, in our weekly lead-intelligence extract to OpenAI (US). Expect a 1-line acknowledgment in the response; thestructuredfeedback is then aggregated server-side.
⟨323 unchanged words⟩ do NOT persist (private_session contract). AUTH:Bearersign-in<token>required,Pro/Teamswith an active Pro, Pro Plus, Teams, Enterprise, beta, or trial plan. Same paid-plan gate as architect.validate. INPUTS: ⟨119 unchanged words⟩
⟨330 unchanged words⟩ are read as literal code). Auth:Bearersign-in<token>required,Pro/Teamswith an active Pro, Pro Plus, Teams, Enterprise, beta, or trial plan. UK/EU residency; transient OpenAI processing (no-training) ⟨120 unchanged words⟩
⟨184 unchanged words⟩ it on the session page). Auth:Bearersign-in<token>required,Pro/Teamswith an active Pro, Pro Plus, Teams, Enterprise, beta, or trial plan.
⟨339 unchanged words⟩ a pointer to the cockpit channel. Auth:Bearersign-in<token>required,Pro/Teamswith an active Pro, Pro Plus, Teams, Enterprise, beta, or trial plan.
⟨54 unchanged words⟩ topic in mind rather than wanting all principles.NOTE:Thesearchlenscurrentlypickscoversthe doctrine searched: 'architecture' = the 10 agentic principlesonly;(default),for'surface' = the 8 experience-design laws,or'spec' = the 8 spec-quality laws,use'all'principles.list(lens='surface')=/allprinciples.list26 (lens='spec'each result then names its lens).untilLawssearchalsospansmatchallonthreetheirlenses.eponym (e.g. 'Hick', 'Fitts').
{"properties":{"lens":{"default":"architecture","description":"Which public doctrine to search: 'architecture' = the 10 agentic principles (default), 'surface' = the 8 experience-design laws, 'spec' = the 8 spec-quality laws, 'all' = all three.","enum":["architecture","surface","spec","all"],"title":"Lens","type":"string"},"limit":{"default":5,"description":"Maximum ⟨38 unchanged words⟩
⟨360 unchanged words⟩ are read as literal content). Auth:Bearersign-in<token>required,Pro/Teamswith an active Pro, Pro Plus, Teams, Enterprise, beta, or trial plan. UK/EU residency; transient OpenAI processing (no-training) ⟨110 unchanged words⟩
⟨93 unchanged words⟩ no auth required (open to all callers),.singleStoresinsertoneintofeedbackUserFeedback.record, at rest under UK/EU residency. contact_email is stored ONLY when permission_to_follow_up=true, ⟨9 unchanged words⟩ so the user can see the privacy boundary. When permission_to_follow_up=true AND a valid contact_email is given, and only then, it also opens a support inbox item so a human can reply, and the free-text fields are sent to OpenAI for triage (severity and a short summary). The contact_email field is never part of that request, so keep personal data out of the free text. In every other case nothing leaves the server.
{"destructiveHint":false,"idempotentHint":false,"openWorldHint":falsetrue,"readOnlyHint":false}
⟨133 unchanged words⟩ that does not appear in the dashboard). The request is also not persisted in the model provider's response store; the provider's abuse-monitoring retention still applies. Default false.","title":"Private Session"," ⟨161 unchanged words⟩